Fixed MTP to work with TWRP

drivers/block/Kconfig

@@ -0,0 +1,560 @@
+#
+# Block device driver configuration
+#
+
+menuconfig BLK_DEV
+	bool "Block devices"
+	depends on BLOCK
+	default y
+	---help---
+	  Say Y here to get to see options for various different block device
+	  drivers. This option alone does not add any kernel code.
+
+	  If you say N, all options in this submenu will be skipped and disabled;
+	  only do this if you know what you are doing.
+
+if BLK_DEV
+
+config BLK_DEV_NULL_BLK
+	tristate "Null test block driver"
+
+config BLK_DEV_FD
+	tristate "Normal floppy disk support"
+	depends on ARCH_MAY_HAVE_PC_FDC
+	---help---
+	  If you want to use the floppy disk drive(s) of your PC under Linux,
+	  say Y. Information about this driver, especially important for IBM
+	  Thinkpad users, is contained in
+	  <file:Documentation/blockdev/floppy.txt>.
+	  That file also contains the location of the Floppy driver FAQ as
+	  well as location of the fdutils package used to configure additional
+	  parameters of the driver at run time.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called floppy.
+
+config AMIGA_FLOPPY
+	tristate "Amiga floppy support"
+	depends on AMIGA
+
+config ATARI_FLOPPY
+	tristate "Atari floppy support"
+	depends on ATARI
+
+config MAC_FLOPPY
+	tristate "Support for PowerMac floppy"
+	depends on PPC_PMAC && !PPC_PMAC64
+	help
+	  If you have a SWIM-3 (Super Woz Integrated Machine 3; from Apple)
+	  floppy controller, say Y here. Most commonly found in PowerMacs.
+
+config BLK_DEV_SWIM
+	tristate "Support for SWIM Macintosh floppy"
+	depends on M68K && MAC
+	help
+	  You should select this option if you want floppy support
+	  and you don't have a II, IIfx, Q900, Q950 or AV series.
+
+config AMIGA_Z2RAM
+	tristate "Amiga Zorro II ramdisk support"
+	depends on ZORRO
+	help
+	  This enables support for using Chip RAM and Zorro II RAM as a
+	  ramdisk or as a swap partition. Say Y if you want to include this
+	  driver in the kernel.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called z2ram.
+
+config GDROM
+	tristate "SEGA Dreamcast GD-ROM drive"
+	depends on SH_DREAMCAST
+	help
+	  A standard SEGA Dreamcast comes with a modified CD ROM drive called a
+	  "GD-ROM" by SEGA to signify it is capable of reading special disks
+	  with up to 1 GB of data. This drive will also read standard CD ROM
+	  disks. Select this option to access any disks in your GD ROM drive.
+	  Most users will want to say "Y" here.
+	  You can also build this as a module which will be called gdrom.
+
+config PARIDE
+	tristate "Parallel port IDE device support"
+	depends on PARPORT_PC
+	---help---
+	  There are many external CD-ROM and disk devices that connect through
+	  your computer's parallel port. Most of them are actually IDE devices
+	  using a parallel port IDE adapter. This option enables the PARIDE
+	  subsystem which contains drivers for many of these external drives.
+	  Read <file:Documentation/blockdev/paride.txt> for more information.
+
+	  If you have said Y to the "Parallel-port support" configuration
+	  option, you may share a single port between your printer and other
+	  parallel port devices. Answer Y to build PARIDE support into your
+	  kernel, or M if you would like to build it as a loadable module. If
+	  your parallel port support is in a loadable module, you must build
+	  PARIDE as a module. If you built PARIDE support into your kernel,
+	  you may still build the individual protocol modules and high-level
+	  drivers as loadable modules. If you build this support as a module,
+	  it will be called paride.
+
+	  To use the PARIDE support, you must say Y or M here and also to at
+	  least one high-level driver (e.g. "Parallel port IDE disks",
+	  "Parallel port ATAPI CD-ROMs", "Parallel port ATAPI disks" etc.) and
+	  to at least one protocol driver (e.g. "ATEN EH-100 protocol",
+	  "MicroSolutions backpack protocol", "DataStor Commuter protocol"
+	  etc.).
+
+source "drivers/block/paride/Kconfig"
+
+source "drivers/block/mtip32xx/Kconfig"
+
+source "drivers/block/zram/Kconfig"
+
+config BLK_CPQ_DA
+	tristate "Compaq SMART2 support"
+	depends on PCI && VIRT_TO_BUS && 0
+	help
+	  This is the driver for Compaq Smart Array controllers.  Everyone
+	  using these boards should say Y here.  See the file
+	  <file:Documentation/blockdev/cpqarray.txt> for the current list of
+	  boards supported by this driver, and for further information on the
+	  use of this driver.
+
+config BLK_CPQ_CISS_DA
+	tristate "Compaq Smart Array 5xxx support"
+	depends on PCI
+	select CHECK_SIGNATURE
+	help
+	  This is the driver for Compaq Smart Array 5xxx controllers.
+	  Everyone using these boards should say Y here.
+	  See <file:Documentation/blockdev/cciss.txt> for the current list of
+	  boards supported by this driver, and for further information
+	  on the use of this driver.
+
+config CISS_SCSI_TAPE
+	bool "SCSI tape drive support for Smart Array 5xxx"
+	depends on BLK_CPQ_CISS_DA && PROC_FS
+	depends on SCSI=y || SCSI=BLK_CPQ_CISS_DA
+	help
+	  When enabled (Y), this option allows SCSI tape drives and SCSI medium
+	  changers (tape robots) to be accessed via a Compaq 5xxx array 
+	  controller.  (See <file:Documentation/blockdev/cciss.txt> for more details.)
+
+	  "SCSI support" and "SCSI tape support" must also be enabled for this 
+	  option to work.
+
+	  When this option is disabled (N), the SCSI portion of the driver 
+	  is not compiled.
+
+config BLK_DEV_DAC960
+	tristate "Mylex DAC960/DAC1100 PCI RAID Controller support"
+	depends on PCI
+	help
+	  This driver adds support for the Mylex DAC960, AcceleRAID, and
+	  eXtremeRAID PCI RAID controllers.  See the file
+	  <file:Documentation/blockdev/README.DAC960> for further information
+	  about this driver.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called DAC960.
+
+config BLK_DEV_UMEM
+	tristate "Micro Memory MM5415 Battery Backed RAM support"
+	depends on PCI
+	---help---
+	  Saying Y here will include support for the MM5415 family of
+	  battery backed (Non-volatile) RAM cards.
+	  <http://www.umem.com/>
+
+	  The cards appear as block devices that can be partitioned into
+	  as many as 15 partitions.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called umem.
+
+	  The umem driver has not yet been allocated a MAJOR number, so
+	  one is chosen dynamically.
+
+config BLK_DEV_UBD
+	bool "Virtual block device"
+	depends on UML
+	---help---
+          The User-Mode Linux port includes a driver called UBD which will let
+          you access arbitrary files on the host computer as block devices.
+          Unless you know that you do not need such virtual block devices say
+          Y here.
+
+config BLK_DEV_UBD_SYNC
+	bool "Always do synchronous disk IO for UBD"
+	depends on BLK_DEV_UBD
+	---help---
+	  Writes to the virtual block device are not immediately written to the
+	  host's disk; this may cause problems if, for example, the User-Mode
+	  Linux 'Virtual Machine' uses a journalling filesystem and the host
+	  computer crashes.
+
+          Synchronous operation (i.e. always writing data to the host's disk
+          immediately) is configurable on a per-UBD basis by using a special
+          kernel command line option.  Alternatively, you can say Y here to
+          turn on synchronous operation by default for all block devices.
+
+          If you're running a journalling file system (like reiserfs, for
+          example) in your virtual machine, you will want to say Y here.  If
+          you care for the safety of the data in your virtual machine, Y is a
+          wise choice too.  In all other cases (for example, if you're just
+          playing around with User-Mode Linux) you can choose N.
+
+config BLK_DEV_COW_COMMON
+	bool
+	default BLK_DEV_UBD
+
+config BLK_DEV_LOOP
+	tristate "Loopback device support"
+	---help---
+	  Saying Y here will allow you to use a regular file as a block
+	  device; you can then create a file system on that block device and
+	  mount it just as you would mount other block devices such as hard
+	  drive partitions, CD-ROM drives or floppy drives. The loop devices
+	  are block special device files with major number 7 and typically
+	  called /dev/loop0, /dev/loop1 etc.
+
+	  This is useful if you want to check an ISO 9660 file system before
+	  burning the CD, or if you want to use floppy images without first
+	  writing them to floppy. Furthermore, some Linux distributions avoid
+	  the need for a dedicated Linux partition by keeping their complete
+	  root file system inside a DOS FAT file using this loop device
+	  driver.
+
+	  To use the loop device, you need the losetup utility, found in the
+	  util-linux package, see
+	  <ftp://ftp.kernel.org/pub/linux/utils/util-linux/>.
+
+	  The loop device driver can also be used to "hide" a file system in
+	  a disk partition, floppy, or regular file, either using encryption
+	  (scrambling the data) or steganography (hiding the data in the low
+	  bits of, say, a sound file). This is also safe if the file resides
+	  on a remote file server.
+
+	  There are several ways of encrypting disks. Some of these require
+	  kernel patches. The vanilla kernel offers the cryptoloop option
+	  and a Device Mapper target (which is superior, as it supports all
+	  file systems). If you want to use the cryptoloop, say Y to both
+	  LOOP and CRYPTOLOOP, and make sure you have a recent (version 2.12
+	  or later) version of util-linux. Additionally, be aware that
+	  the cryptoloop is not safe for storing journaled filesystems.
+
+	  Note that this loop device has nothing to do with the loopback
+	  device used for network connections from the machine to itself.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called loop.
+
+	  Most users will answer N here.
+
+config BLK_DEV_LOOP_MIN_COUNT
+	int "Number of loop devices to pre-create at init time"
+	depends on BLK_DEV_LOOP
+	default 8
+	help
+	  Static number of loop devices to be unconditionally pre-created
+	  at init time.
+
+	  This default value can be overwritten on the kernel command
+	  line or with module-parameter loop.max_loop.
+
+	  The historic default is 8. If a late 2011 version of losetup(8)
+	  is used, it can be set to 0, since needed loop devices can be
+	  dynamically allocated with the /dev/loop-control interface.
+
+config BLK_DEV_CRYPTOLOOP
+	tristate "Cryptoloop Support"
+	select CRYPTO
+	select CRYPTO_CBC
+	depends on BLK_DEV_LOOP
+	---help---
+	  Say Y here if you want to be able to use the ciphers that are 
+	  provided by the CryptoAPI as loop transformation. This might be
+	  used as hard disk encryption.
+
+	  WARNING: This device is not safe for journaled file systems like
+	  ext3 or Reiserfs. Please use the Device Mapper crypto module
+	  instead, which can be configured to be on-disk compatible with the
+	  cryptoloop device.
+
+source "drivers/block/drbd/Kconfig"
+
+config BLK_DEV_NBD
+	tristate "Network block device support"
+	depends on NET
+	---help---
+	  Saying Y here will allow your computer to be a client for network
+	  block devices, i.e. it will be able to use block devices exported by
+	  servers (mount file systems on them etc.). Communication between
+	  client and server works over TCP/IP networking, but to the client
+	  program this is hidden: it looks like a regular local file access to
+	  a block device special file such as /dev/nd0.
+
+	  Network block devices also allows you to run a block-device in
+	  userland (making server and client physically the same computer,
+	  communicating using the loopback network device).
+
+	  Read <file:Documentation/blockdev/nbd.txt> for more information,
+	  especially about where to find the server code, which runs in user
+	  space and does not need special kernel support.
+
+	  Note that this has nothing to do with the network file systems NFS
+	  or Coda; you can say N here even if you intend to use NFS or Coda.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called nbd.
+
+	  If unsure, say N.
+
+config BLK_DEV_NVME
+	tristate "NVM Express block device"
+	depends on PCI
+	---help---
+	  The NVM Express driver is for solid state drives directly
+	  connected to the PCI or PCI Express bus.  If you know you
+	  don't have one of these, it is safe to answer N.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called nvme.
+
+config BLK_DEV_SKD
+	tristate "STEC S1120 Block Driver"
+	depends on PCI
+	depends on 64BIT
+	---help---
+	Saying Y or M here will enable support for the
+	STEC, Inc. S1120 PCIe SSD.
+
+	Use device /dev/skd$N amd /dev/skd$Np$M.
+
+config BLK_DEV_OSD
+	tristate "OSD object-as-blkdev support"
+	depends on SCSI_OSD_ULD
+	---help---
+	  Saying Y or M here will allow the exporting of a single SCSI
+	  OSD (object-based storage) object as a Linux block device.
+
+	  For example, if you create a 2G object on an OSD device,
+	  you can then use this module to present that 2G object as
+	  a Linux block device.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called osdblk.
+
+	  If unsure, say N.
+
+config BLK_DEV_SX8
+	tristate "Promise SATA SX8 support"
+	depends on PCI
+	---help---
+	  Saying Y or M here will enable support for the 
+	  Promise SATA SX8 controllers.
+
+	  Use devices /dev/sx8/$N and /dev/sx8/$Np$M.
+
+config BLK_DEV_RAM
+	tristate "RAM block device support"
+	---help---
+	  Saying Y here will allow you to use a portion of your RAM memory as
+	  a block device, so that you can make file systems on it, read and
+	  write to it and do all the other things that you can do with normal
+	  block devices (such as hard drives). It is usually used to load and
+	  store a copy of a minimal root file system off of a floppy into RAM
+	  during the initial install of Linux.
+
+	  Note that the kernel command line option "ramdisk=XX" is now obsolete.
+	  For details, read <file:Documentation/blockdev/ramdisk.txt>.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called brd. An alias "rd" has been defined
+	  for historical reasons.
+
+	  Most normal users won't need the RAM disk functionality, and can
+	  thus say N here.
+
+config BLK_DEV_RAM_COUNT
+	int "Default number of RAM disks"
+	default "16"
+	depends on BLK_DEV_RAM
+	help
+	  The default value is 16 RAM disks. Change this if you know what you
+	  are doing. If you boot from a filesystem that needs to be extracted
+	  in memory, you will need at least one RAM disk (e.g. root on cramfs).
+
+config BLK_DEV_RAM_SIZE
+	int "Default RAM disk size (kbytes)"
+	depends on BLK_DEV_RAM
+	default "4096"
+	help
+	  The default value is 4096 kilobytes. Only change this if you know
+	  what you are doing.
+
+config BLK_DEV_XIP
+	bool "Support XIP filesystems on RAM block device"
+	depends on BLK_DEV_RAM
+	default n
+	help
+	  Support XIP filesystems (such as ext2 with XIP support on) on
+	  top of block ram device. This will slightly enlarge the kernel, and
+	  will prevent RAM block device backing store memory from being
+	  allocated from highmem (only a problem for highmem systems).
+
+config CDROM_PKTCDVD
+	tristate "Packet writing on CD/DVD media"
+	depends on !UML
+	help
+	  If you have a CDROM/DVD drive that supports packet writing, say
+	  Y to include support. It should work with any MMC/Mt Fuji
+	  compliant ATAPI or SCSI drive, which is just about any newer
+	  DVD/CD writer.
+
+	  Currently only writing to CD-RW, DVD-RW, DVD+RW and DVDRAM discs
+	  is possible.
+	  DVD-RW disks must be in restricted overwrite mode.
+
+	  See the file <file:Documentation/cdrom/packet-writing.txt>
+	  for further information on the use of this driver.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called pktcdvd.
+
+config CDROM_PKTCDVD_BUFFERS
+	int "Free buffers for data gathering"
+	depends on CDROM_PKTCDVD
+	default "8"
+	help
+	  This controls the maximum number of active concurrent packets. More
+	  concurrent packets can increase write performance, but also require
+	  more memory. Each concurrent packet will require approximately 64Kb
+	  of non-swappable kernel memory, memory which will be allocated when
+	  a disc is opened for writing.
+
+config CDROM_PKTCDVD_WCACHE
+	bool "Enable write caching"
+	depends on CDROM_PKTCDVD
+	help
+	  If enabled, write caching will be set for the CD-R/W device. For now
+	  this option is dangerous unless the CD-RW media is known good, as we
+	  don't do deferred write error handling yet.
+
+config ATA_OVER_ETH
+	tristate "ATA over Ethernet support"
+	depends on NET
+	help
+	This driver provides Support for ATA over Ethernet block
+	devices like the Coraid EtherDrive (R) Storage Blade.
+
+config MG_DISK
+	tristate "mGine mflash, gflash support"
+	depends on ARM && GPIOLIB
+	help
+	  mGine mFlash(gFlash) block device driver
+
+config MG_DISK_RES
+	int "Size of reserved area before MBR"
+	depends on MG_DISK
+	default 0
+	help
+	  Define size of reserved area that usually used for boot. Unit is KB.
+	  All of the block device operation will be taken this value as start
+	  offset
+	  Examples:
+			1024 => 1 MB
+
+config SUNVDC
+	tristate "Sun Virtual Disk Client support"
+	depends on SUN_LDOMS
+	help
+	  Support for virtual disk devices as a client under Sun
+	  Logical Domains.
+
+source "drivers/s390/block/Kconfig"
+
+config XILINX_SYSACE
+	tristate "Xilinx SystemACE support"
+	depends on 4xx || MICROBLAZE
+	help
+	  Include support for the Xilinx SystemACE CompactFlash interface
+
+config XEN_BLKDEV_FRONTEND
+	tristate "Xen virtual block device support"
+	depends on XEN
+	default y
+	select XEN_XENBUS_FRONTEND
+	help
+	  This driver implements the front-end of the Xen virtual
+	  block device driver.  It communicates with a back-end driver
+	  in another domain which drives the actual block device.
+
+config XEN_BLKDEV_BACKEND
+	tristate "Xen block-device backend driver"
+	depends on XEN_BACKEND
+	help
+	  The block-device backend driver allows the kernel to export its
+	  block devices to other guests via a high-performance shared-memory
+	  interface.
+
+	  The corresponding Linux frontend driver is enabled by the
+	  CONFIG_XEN_BLKDEV_FRONTEND configuration option.
+
+	  The backend driver attaches itself to a any block device specified
+	  in the XenBus configuration. There are no limits to what the block
+	  device as long as it has a major and minor.
+
+	  If you are compiling a kernel to run in a Xen block backend driver
+	  domain (often this is domain 0) you should say Y here. To
+	  compile this driver as a module, chose M here: the module
+	  will be called xen-blkback.
+
+
+config VIRTIO_BLK
+	tristate "Virtio block driver"
+	depends on VIRTIO
+	---help---
+	  This is the virtual block driver for virtio.  It can be used with
+          lguest or QEMU based VMMs (like KVM or Xen).  Say Y or M.
+
+config BLK_DEV_HD
+	bool "Very old hard disk (MFM/RLL/IDE) driver"
+	depends on HAVE_IDE
+	depends on !ARM || ARCH_RPC || BROKEN
+	help
+	  This is a very old hard disk driver that lacks the enhanced
+	  functionality of the newer ones.
+
+	  It is required for systems with ancient MFM/RLL/ESDI drives.
+
+	  If unsure, say N.
+
+config BLK_DEV_RBD
+	tristate "Rados block device (RBD)"
+	depends on INET && BLOCK
+	select CEPH_LIB
+	select LIBCRC32C
+	select CRYPTO_AES
+	select CRYPTO
+	default n
+	help
+	  Say Y here if you want include the Rados block device, which stripes
+	  a block device over objects stored in the Ceph distributed object
+	  store.
+
+	  More information at http://ceph.newdream.net/.
+
+	  If unsure, say N.
+
+config BLK_DEV_RSXX
+	tristate "IBM Flash Adapter 900GB Full Height PCIe Device Driver"
+	depends on PCI
+	help
+	  Device driver for IBM's high speed PCIe SSD
+	  storage device: Flash Adapter 900GB Full Height.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called rsxx.
+
+endif # BLK_DEV

drivers/block/Makefile

@@ -0,0 +1,49 @@
+#
+# Makefile for the kernel block device drivers.
+#
+# 12 June 2000, Christoph Hellwig <hch@infradead.org>
+# Rewritten to use lists instead of if-statements.
+# 
+
+obj-$(CONFIG_MAC_FLOPPY)	+= swim3.o
+obj-$(CONFIG_BLK_DEV_SWIM)	+= swim_mod.o
+obj-$(CONFIG_BLK_DEV_FD)	+= floppy.o
+obj-$(CONFIG_AMIGA_FLOPPY)	+= amiflop.o
+obj-$(CONFIG_PS3_DISK)		+= ps3disk.o
+obj-$(CONFIG_PS3_VRAM)		+= ps3vram.o
+obj-$(CONFIG_ATARI_FLOPPY)	+= ataflop.o
+obj-$(CONFIG_AMIGA_Z2RAM)	+= z2ram.o
+obj-$(CONFIG_BLK_DEV_RAM)	+= brd.o
+obj-$(CONFIG_BLK_DEV_LOOP)	+= loop.o
+obj-$(CONFIG_BLK_CPQ_DA)	+= cpqarray.o
+obj-$(CONFIG_BLK_CPQ_CISS_DA)  += cciss.o
+obj-$(CONFIG_BLK_DEV_DAC960)	+= DAC960.o
+obj-$(CONFIG_XILINX_SYSACE)	+= xsysace.o
+obj-$(CONFIG_CDROM_PKTCDVD)	+= pktcdvd.o
+obj-$(CONFIG_MG_DISK)		+= mg_disk.o
+obj-$(CONFIG_SUNVDC)		+= sunvdc.o
+obj-$(CONFIG_BLK_DEV_NVME)	+= nvme.o
+obj-$(CONFIG_BLK_DEV_SKD)	+= skd.o
+obj-$(CONFIG_BLK_DEV_OSD)	+= osdblk.o
+
+obj-$(CONFIG_BLK_DEV_UMEM)	+= umem.o
+obj-$(CONFIG_BLK_DEV_NBD)	+= nbd.o
+obj-$(CONFIG_BLK_DEV_CRYPTOLOOP) += cryptoloop.o
+obj-$(CONFIG_VIRTIO_BLK)	+= virtio_blk.o
+
+obj-$(CONFIG_BLK_DEV_SX8)	+= sx8.o
+obj-$(CONFIG_BLK_DEV_HD)	+= hd.o
+
+obj-$(CONFIG_XEN_BLKDEV_FRONTEND)	+= xen-blkfront.o
+obj-$(CONFIG_XEN_BLKDEV_BACKEND)	+= xen-blkback/
+obj-$(CONFIG_BLK_DEV_DRBD)     += drbd/
+obj-$(CONFIG_BLK_DEV_RBD)     += rbd.o
+obj-$(CONFIG_BLK_DEV_PCIESSD_MTIP32XX)	+= mtip32xx/
+
+obj-$(CONFIG_BLK_DEV_RSXX) += rsxx/
+obj-$(CONFIG_BLK_DEV_NULL_BLK)	+= null_blk.o
+obj-$(CONFIG_ZRAM) += zram/
+
+nvme-y		:= nvme-core.o nvme-scsi.o
+skd-y		:= skd_main.o
+swim_mod-y	:= swim.o swim_asm.o

drivers/block/aoe/Makefile

@@ -0,0 +1,6 @@
+#
+# Makefile for ATA over Ethernet
+#
+
+obj-$(CONFIG_ATA_OVER_ETH)	+= aoe.o
+aoe-y := aoeblk.o aoechr.o aoecmd.o aoedev.o aoemain.o aoenet.o

drivers/block/aoe/aoe.h

@@ -0,0 +1,240 @@
+/* Copyright (c) 2013 Coraid, Inc.  See COPYING for GPL terms. */
+#define VERSION "85"
+#define AOE_MAJOR 152
+#define DEVICE_NAME "aoe"
+
+/* set AOE_PARTITIONS to 1 to use whole-disks only
+ * default is 16, which is 15 partitions plus the whole disk
+ */
+#ifndef AOE_PARTITIONS
+#define AOE_PARTITIONS (16)
+#endif
+
+#define WHITESPACE " \t\v\f\n,"
+
+enum {
+	AOECMD_ATA,
+	AOECMD_CFG,
+	AOECMD_VEND_MIN = 0xf0,
+
+	AOEFL_RSP = (1<<3),
+	AOEFL_ERR = (1<<2),
+
+	AOEAFL_EXT = (1<<6),
+	AOEAFL_DEV = (1<<4),
+	AOEAFL_ASYNC = (1<<1),
+	AOEAFL_WRITE = (1<<0),
+
+	AOECCMD_READ = 0,
+	AOECCMD_TEST,
+	AOECCMD_PTEST,
+	AOECCMD_SET,
+	AOECCMD_FSET,
+
+	AOE_HVER = 0x10,
+};
+
+struct aoe_hdr {
+	unsigned char dst[6];
+	unsigned char src[6];
+	__be16 type;
+	unsigned char verfl;
+	unsigned char err;
+	__be16 major;
+	unsigned char minor;
+	unsigned char cmd;
+	__be32 tag;
+};
+
+struct aoe_atahdr {
+	unsigned char aflags;
+	unsigned char errfeat;
+	unsigned char scnt;
+	unsigned char cmdstat;
+	unsigned char lba0;
+	unsigned char lba1;
+	unsigned char lba2;
+	unsigned char lba3;
+	unsigned char lba4;
+	unsigned char lba5;
+	unsigned char res[2];
+};
+
+struct aoe_cfghdr {
+	__be16 bufcnt;
+	__be16 fwver;
+	unsigned char scnt;
+	unsigned char aoeccmd;
+	unsigned char cslen[2];
+};
+
+enum {
+	DEVFL_UP = 1,	/* device is installed in system and ready for AoE->ATA commands */
+	DEVFL_TKILL = (1<<1),	/* flag for timer to know when to kill self */
+	DEVFL_EXT = (1<<2),	/* device accepts lba48 commands */
+	DEVFL_GDALLOC = (1<<3),	/* need to alloc gendisk */
+	DEVFL_GD_NOW = (1<<4),	/* allocating gendisk */
+	DEVFL_KICKME = (1<<5),	/* slow polling network card catch */
+	DEVFL_NEWSIZE = (1<<6),	/* need to update dev size in block layer */
+	DEVFL_FREEING = (1<<7),	/* set when device is being cleaned up */
+	DEVFL_FREED = (1<<8),	/* device has been cleaned up */
+};
+
+enum {
+	DEFAULTBCNT = 2 * 512,	/* 2 sectors */
+	MIN_BUFS = 16,
+	NTARGETS = 4,
+	NAOEIFS = 8,
+	NSKBPOOLMAX = 256,
+	NFACTIVE = 61,
+
+	TIMERTICK = HZ / 10,
+	RTTSCALE = 8,
+	RTTDSCALE = 3,
+	RTTAVG_INIT = USEC_PER_SEC / 4 << RTTSCALE,
+	RTTDEV_INIT = RTTAVG_INIT / 4,
+
+	HARD_SCORN_SECS = 10,	/* try another remote port after this */
+	MAX_TAINT = 1000,	/* cap on aoetgt taint */
+};
+
+struct buf {
+	ulong nframesout;
+	struct bio *bio;
+	struct bvec_iter iter;
+	struct request *rq;
+};
+
+enum frame_flags {
+	FFL_PROBE = 1,
+};
+
+struct frame {
+	struct list_head head;
+	u32 tag;
+	struct timeval sent;	/* high-res time packet was sent */
+	u32 sent_jiffs;		/* low-res jiffies-based sent time */
+	ulong waited;
+	ulong waited_total;
+	struct aoetgt *t;		/* parent target I belong to */
+	struct sk_buff *skb;		/* command skb freed on module exit */
+	struct sk_buff *r_skb;		/* response skb for async processing */
+	struct buf *buf;
+	struct bvec_iter iter;
+	char flags;
+};
+
+struct aoeif {
+	struct net_device *nd;
+	ulong lost;
+	int bcnt;
+};
+
+struct aoetgt {
+	unsigned char addr[6];
+	ushort nframes;		/* cap on frames to use */
+	struct aoedev *d;			/* parent device I belong to */
+	struct list_head ffree;			/* list of free frames */
+	struct aoeif ifs[NAOEIFS];
+	struct aoeif *ifp;	/* current aoeif in use */
+	ushort nout;		/* number of AoE commands outstanding */
+	ushort maxout;		/* current value for max outstanding */
+	ushort next_cwnd;	/* incr maxout after decrementing to zero */
+	ushort ssthresh;	/* slow start threshold */
+	ulong falloc;		/* number of allocated frames */
+	int taint;		/* how much we want to avoid this aoetgt */
+	int minbcnt;
+	int wpkts, rpkts;
+	char nout_probes;
+};
+
+struct aoedev {
+	struct aoedev *next;
+	ulong sysminor;
+	ulong aoemajor;
+	u32 rttavg;		/* scaled AoE round trip time average */
+	u32 rttdev;		/* scaled round trip time mean deviation */
+	u16 aoeminor;
+	u16 flags;
+	u16 nopen;		/* (bd_openers isn't available without sleeping) */
+	u16 fw_ver;		/* version of blade's firmware */
+	u16 lasttag;		/* last tag sent */
+	u16 useme;
+	ulong ref;
+	struct work_struct work;/* disk create work struct */
+	struct gendisk *gd;
+	struct dentry *debugfs;
+	struct request_queue *blkq;
+	struct hd_geometry geo;
+	sector_t ssize;
+	struct timer_list timer;
+	spinlock_t lock;
+	struct sk_buff_head skbpool;
+	mempool_t *bufpool;	/* for deadlock-free Buf allocation */
+	struct {		/* pointers to work in progress */
+		struct buf *buf;
+		struct bio *nxbio;
+		struct request *rq;
+	} ip;
+	ulong maxbcnt;
+	struct list_head factive[NFACTIVE];	/* hash of active frames */
+	struct list_head rexmitq; /* deferred retransmissions */
+	struct aoetgt **targets;
+	ulong ntargets;		/* number of allocated aoetgt pointers */
+	struct aoetgt **tgt;	/* target in use when working */
+	ulong kicked;
+	char ident[512];
+};
+
+/* kthread tracking */
+struct ktstate {
+	struct completion rendez;
+	struct task_struct *task;
+	wait_queue_head_t *waitq;
+	int (*fn) (int);
+	char name[12];
+	spinlock_t *lock;
+	int id;
+	int active;
+};
+
+int aoeblk_init(void);
+void aoeblk_exit(void);
+void aoeblk_gdalloc(void *);
+void aoedisk_rm_debugfs(struct aoedev *d);
+void aoedisk_rm_sysfs(struct aoedev *d);
+
+int aoechr_init(void);
+void aoechr_exit(void);
+void aoechr_error(char *);
+
+void aoecmd_work(struct aoedev *d);
+void aoecmd_cfg(ushort aoemajor, unsigned char aoeminor);
+struct sk_buff *aoecmd_ata_rsp(struct sk_buff *);
+void aoecmd_cfg_rsp(struct sk_buff *);
+void aoecmd_sleepwork(struct work_struct *);
+void aoecmd_wreset(struct aoetgt *t);
+void aoecmd_cleanslate(struct aoedev *);
+void aoecmd_exit(void);
+int aoecmd_init(void);
+struct sk_buff *aoecmd_ata_id(struct aoedev *);
+void aoe_freetframe(struct frame *);
+void aoe_flush_iocq(void);
+void aoe_flush_iocq_by_index(int);
+void aoe_end_request(struct aoedev *, struct request *, int);
+int aoe_ktstart(struct ktstate *k);
+void aoe_ktstop(struct ktstate *k);
+
+int aoedev_init(void);
+void aoedev_exit(void);
+struct aoedev *aoedev_by_aoeaddr(ulong maj, int min, int do_alloc);
+void aoedev_downdev(struct aoedev *d);
+int aoedev_flush(const char __user *str, size_t size);
+void aoe_failbuf(struct aoedev *, struct buf *);
+void aoedev_put(struct aoedev *);
+
+int aoenet_init(void);
+void aoenet_exit(void);
+void aoenet_xmit(struct sk_buff_head *);
+int is_aoe_netif(struct net_device *ifp);
+int set_aoe_iflist(const char __user *str, size_t size);

drivers/block/aoe/aoeblk.c

@@ -0,0 +1,464 @@
+/* Copyright (c) 2013 Coraid, Inc.  See COPYING for GPL terms. */
+/*
+ * aoeblk.c
+ * block device routines
+ */
+
+#include <linux/kernel.h>
+#include <linux/hdreg.h>
+#include <linux/blkdev.h>
+#include <linux/backing-dev.h>
+#include <linux/fs.h>
+#include <linux/ioctl.h>
+#include <linux/slab.h>
+#include <linux/ratelimit.h>
+#include <linux/genhd.h>
+#include <linux/netdevice.h>
+#include <linux/mutex.h>
+#include <linux/export.h>
+#include <linux/moduleparam.h>
+#include <linux/debugfs.h>
+#include <scsi/sg.h>
+#include "aoe.h"
+
+static DEFINE_MUTEX(aoeblk_mutex);
+static struct kmem_cache *buf_pool_cache;
+static struct dentry *aoe_debugfs_dir;
+
+/* GPFS needs a larger value than the default. */
+static int aoe_maxsectors;
+module_param(aoe_maxsectors, int, 0644);
+MODULE_PARM_DESC(aoe_maxsectors,
+	"When nonzero, set the maximum number of sectors per I/O request");
+
+static ssize_t aoedisk_show_state(struct device *dev,
+				  struct device_attribute *attr, char *page)
+{
+	struct gendisk *disk = dev_to_disk(dev);
+	struct aoedev *d = disk->private_data;
+
+	return snprintf(page, PAGE_SIZE,
+			"%s%s\n",
+			(d->flags & DEVFL_UP) ? "up" : "down",
+			(d->flags & DEVFL_KICKME) ? ",kickme" :
+			(d->nopen && !(d->flags & DEVFL_UP)) ? ",closewait" : "");
+	/* I'd rather see nopen exported so we can ditch closewait */
+}
+static ssize_t aoedisk_show_mac(struct device *dev,
+				struct device_attribute *attr, char *page)
+{
+	struct gendisk *disk = dev_to_disk(dev);
+	struct aoedev *d = disk->private_data;
+	struct aoetgt *t = d->targets[0];
+
+	if (t == NULL)
+		return snprintf(page, PAGE_SIZE, "none\n");
+	return snprintf(page, PAGE_SIZE, "%pm\n", t->addr);
+}
+static ssize_t aoedisk_show_netif(struct device *dev,
+				  struct device_attribute *attr, char *page)
+{
+	struct gendisk *disk = dev_to_disk(dev);
+	struct aoedev *d = disk->private_data;
+	struct net_device *nds[8], **nd, **nnd, **ne;
+	struct aoetgt **t, **te;
+	struct aoeif *ifp, *e;
+	char *p;
+
+	memset(nds, 0, sizeof nds);
+	nd = nds;
+	ne = nd + ARRAY_SIZE(nds);
+	t = d->targets;
+	te = t + d->ntargets;
+	for (; t < te && *t; t++) {
+		ifp = (*t)->ifs;
+		e = ifp + NAOEIFS;
+		for (; ifp < e && ifp->nd; ifp++) {
+			for (nnd = nds; nnd < nd; nnd++)
+				if (*nnd == ifp->nd)
+					break;
+			if (nnd == nd && nd != ne)
+				*nd++ = ifp->nd;
+		}
+	}
+
+	ne = nd;
+	nd = nds;
+	if (*nd == NULL)
+		return snprintf(page, PAGE_SIZE, "none\n");
+	for (p = page; nd < ne; nd++)
+		p += snprintf(p, PAGE_SIZE - (p-page), "%s%s",
+			p == page ? "" : ",", (*nd)->name);
+	p += snprintf(p, PAGE_SIZE - (p-page), "\n");
+	return p-page;
+}
+/* firmware version */
+static ssize_t aoedisk_show_fwver(struct device *dev,
+				  struct device_attribute *attr, char *page)
+{
+	struct gendisk *disk = dev_to_disk(dev);
+	struct aoedev *d = disk->private_data;
+
+	return snprintf(page, PAGE_SIZE, "0x%04x\n", (unsigned int) d->fw_ver);
+}
+static ssize_t aoedisk_show_payload(struct device *dev,
+				    struct device_attribute *attr, char *page)
+{
+	struct gendisk *disk = dev_to_disk(dev);
+	struct aoedev *d = disk->private_data;
+
+	return snprintf(page, PAGE_SIZE, "%lu\n", d->maxbcnt);
+}
+
+static int aoedisk_debugfs_show(struct seq_file *s, void *ignored)
+{
+	struct aoedev *d;
+	struct aoetgt **t, **te;
+	struct aoeif *ifp, *ife;
+	unsigned long flags;
+	char c;
+
+	d = s->private;
+	seq_printf(s, "rttavg: %d rttdev: %d\n",
+		d->rttavg >> RTTSCALE,
+		d->rttdev >> RTTDSCALE);
+	seq_printf(s, "nskbpool: %d\n", skb_queue_len(&d->skbpool));
+	seq_printf(s, "kicked: %ld\n", d->kicked);
+	seq_printf(s, "maxbcnt: %ld\n", d->maxbcnt);
+	seq_printf(s, "ref: %ld\n", d->ref);
+
+	spin_lock_irqsave(&d->lock, flags);
+	t = d->targets;
+	te = t + d->ntargets;
+	for (; t < te && *t; t++) {
+		c = '\t';
+		seq_printf(s, "falloc: %ld\n", (*t)->falloc);
+		seq_printf(s, "ffree: %p\n",
+			list_empty(&(*t)->ffree) ? NULL : (*t)->ffree.next);
+		seq_printf(s, "%pm:%d:%d:%d\n", (*t)->addr, (*t)->nout,
+			(*t)->maxout, (*t)->nframes);
+		seq_printf(s, "\tssthresh:%d\n", (*t)->ssthresh);
+		seq_printf(s, "\ttaint:%d\n", (*t)->taint);
+		seq_printf(s, "\tr:%d\n", (*t)->rpkts);
+		seq_printf(s, "\tw:%d\n", (*t)->wpkts);
+		ifp = (*t)->ifs;
+		ife = ifp + ARRAY_SIZE((*t)->ifs);
+		for (; ifp->nd && ifp < ife; ifp++) {
+			seq_printf(s, "%c%s", c, ifp->nd->name);
+			c = ',';
+		}
+		seq_puts(s, "\n");
+	}
+	spin_unlock_irqrestore(&d->lock, flags);
+
+	return 0;
+}
+
+static int aoe_debugfs_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, aoedisk_debugfs_show, inode->i_private);
+}
+
+static DEVICE_ATTR(state, S_IRUGO, aoedisk_show_state, NULL);
+static DEVICE_ATTR(mac, S_IRUGO, aoedisk_show_mac, NULL);
+static DEVICE_ATTR(netif, S_IRUGO, aoedisk_show_netif, NULL);
+static struct device_attribute dev_attr_firmware_version = {
+	.attr = { .name = "firmware-version", .mode = S_IRUGO },
+	.show = aoedisk_show_fwver,
+};
+static DEVICE_ATTR(payload, S_IRUGO, aoedisk_show_payload, NULL);
+
+static struct attribute *aoe_attrs[] = {
+	&dev_attr_state.attr,
+	&dev_attr_mac.attr,
+	&dev_attr_netif.attr,
+	&dev_attr_firmware_version.attr,
+	&dev_attr_payload.attr,
+	NULL,
+};
+
+static const struct attribute_group attr_group = {
+	.attrs = aoe_attrs,
+};
+
+static const struct file_operations aoe_debugfs_fops = {
+	.open = aoe_debugfs_open,
+	.read = seq_read,
+	.llseek = seq_lseek,
+	.release = single_release,
+};
+
+static void
+aoedisk_add_debugfs(struct aoedev *d)
+{
+	struct dentry *entry;
+	char *p;
+
+	if (aoe_debugfs_dir == NULL)
+		return;
+	p = strchr(d->gd->disk_name, '/');
+	if (p == NULL)
+		p = d->gd->disk_name;
+	else
+		p++;
+	BUG_ON(*p == '\0');
+	entry = debugfs_create_file(p, 0444, aoe_debugfs_dir, d,
+				    &aoe_debugfs_fops);
+	if (IS_ERR_OR_NULL(entry)) {
+		pr_info("aoe: cannot create debugfs file for %s\n",
+			d->gd->disk_name);
+		return;
+	}
+	BUG_ON(d->debugfs);
+	d->debugfs = entry;
+}
+void
+aoedisk_rm_debugfs(struct aoedev *d)
+{
+	debugfs_remove(d->debugfs);
+	d->debugfs = NULL;
+}
+
+static int
+aoedisk_add_sysfs(struct aoedev *d)
+{
+	return sysfs_create_group(&disk_to_dev(d->gd)->kobj, &attr_group);
+}
+void
+aoedisk_rm_sysfs(struct aoedev *d)
+{
+	sysfs_remove_group(&disk_to_dev(d->gd)->kobj, &attr_group);
+}
+
+static int
+aoeblk_open(struct block_device *bdev, fmode_t mode)
+{
+	struct aoedev *d = bdev->bd_disk->private_data;
+	ulong flags;
+
+	if (!virt_addr_valid(d)) {
+		pr_crit("aoe: invalid device pointer in %s\n",
+			__func__);
+		WARN_ON(1);
+		return -ENODEV;
+	}
+	if (!(d->flags & DEVFL_UP) || d->flags & DEVFL_TKILL)
+		return -ENODEV;
+
+	mutex_lock(&aoeblk_mutex);
+	spin_lock_irqsave(&d->lock, flags);
+	if (d->flags & DEVFL_UP && !(d->flags & DEVFL_TKILL)) {
+		d->nopen++;
+		spin_unlock_irqrestore(&d->lock, flags);
+		mutex_unlock(&aoeblk_mutex);
+		return 0;
+	}
+	spin_unlock_irqrestore(&d->lock, flags);
+	mutex_unlock(&aoeblk_mutex);
+	return -ENODEV;
+}
+
+static void
+aoeblk_release(struct gendisk *disk, fmode_t mode)
+{
+	struct aoedev *d = disk->private_data;
+	ulong flags;
+
+	spin_lock_irqsave(&d->lock, flags);
+
+	if (--d->nopen == 0) {
+		spin_unlock_irqrestore(&d->lock, flags);
+		aoecmd_cfg(d->aoemajor, d->aoeminor);
+		return;
+	}
+	spin_unlock_irqrestore(&d->lock, flags);
+}
+
+static void
+aoeblk_request(struct request_queue *q)
+{
+	struct aoedev *d;
+	struct request *rq;
+
+	d = q->queuedata;
+	if ((d->flags & DEVFL_UP) == 0) {
+		pr_info_ratelimited("aoe: device %ld.%d is not up\n",
+			d->aoemajor, d->aoeminor);
+		while ((rq = blk_peek_request(q))) {
+			blk_start_request(rq);
+			aoe_end_request(d, rq, 1);
+		}
+		return;
+	}
+	aoecmd_work(d);
+}
+
+static int
+aoeblk_getgeo(struct block_device *bdev, struct hd_geometry *geo)
+{
+	struct aoedev *d = bdev->bd_disk->private_data;
+
+	if ((d->flags & DEVFL_UP) == 0) {
+		printk(KERN_ERR "aoe: disk not up\n");
+		return -ENODEV;
+	}
+
+	geo->cylinders = d->geo.cylinders;
+	geo->heads = d->geo.heads;
+	geo->sectors = d->geo.sectors;
+	return 0;
+}
+
+static int
+aoeblk_ioctl(struct block_device *bdev, fmode_t mode, uint cmd, ulong arg)
+{
+	struct aoedev *d;
+
+	if (!arg)
+		return -EINVAL;
+
+	d = bdev->bd_disk->private_data;
+	if ((d->flags & DEVFL_UP) == 0) {
+		pr_err("aoe: disk not up\n");
+		return -ENODEV;
+	}
+
+	if (cmd == HDIO_GET_IDENTITY) {
+		if (!copy_to_user((void __user *) arg, &d->ident,
+			sizeof(d->ident)))
+			return 0;
+		return -EFAULT;
+	}
+
+	/* udev calls scsi_id, which uses SG_IO, resulting in noise */
+	if (cmd != SG_IO)
+		pr_info("aoe: unknown ioctl 0x%x\n", cmd);
+
+	return -ENOTTY;
+}
+
+static const struct block_device_operations aoe_bdops = {
+	.open = aoeblk_open,
+	.release = aoeblk_release,
+	.ioctl = aoeblk_ioctl,
+	.getgeo = aoeblk_getgeo,
+	.owner = THIS_MODULE,
+};
+
+/* alloc_disk and add_disk can sleep */
+void
+aoeblk_gdalloc(void *vp)
+{
+	struct aoedev *d = vp;
+	struct gendisk *gd;
+	mempool_t *mp;
+	struct request_queue *q;
+	enum { KB = 1024, MB = KB * KB, READ_AHEAD = 2 * MB, };
+	ulong flags;
+	int late = 0;
+
+	spin_lock_irqsave(&d->lock, flags);
+	if (d->flags & DEVFL_GDALLOC
+	&& !(d->flags & DEVFL_TKILL)
+	&& !(d->flags & DEVFL_GD_NOW))
+		d->flags |= DEVFL_GD_NOW;
+	else
+		late = 1;
+	spin_unlock_irqrestore(&d->lock, flags);
+	if (late)
+		return;
+
+	gd = alloc_disk(AOE_PARTITIONS);
+	if (gd == NULL) {
+		pr_err("aoe: cannot allocate disk structure for %ld.%d\n",
+			d->aoemajor, d->aoeminor);
+		goto err;
+	}
+
+	mp = mempool_create(MIN_BUFS, mempool_alloc_slab, mempool_free_slab,
+		buf_pool_cache);
+	if (mp == NULL) {
+		printk(KERN_ERR "aoe: cannot allocate bufpool for %ld.%d\n",
+			d->aoemajor, d->aoeminor);
+		goto err_disk;
+	}
+	q = blk_init_queue(aoeblk_request, &d->lock);
+	if (q == NULL) {
+		pr_err("aoe: cannot allocate block queue for %ld.%d\n",
+			d->aoemajor, d->aoeminor);
+		goto err_mempool;
+	}
+
+	spin_lock_irqsave(&d->lock, flags);
+	WARN_ON(!(d->flags & DEVFL_GD_NOW));
+	WARN_ON(!(d->flags & DEVFL_GDALLOC));
+	WARN_ON(d->flags & DEVFL_TKILL);
+	WARN_ON(d->gd);
+	WARN_ON(d->flags & DEVFL_UP);
+	blk_queue_max_hw_sectors(q, BLK_DEF_MAX_SECTORS);
+	q->backing_dev_info.name = "aoe";
+	q->backing_dev_info.ra_pages = READ_AHEAD / PAGE_CACHE_SIZE;
+	d->bufpool = mp;
+	d->blkq = gd->queue = q;
+	q->queuedata = d;
+	d->gd = gd;
+	if (aoe_maxsectors)
+		blk_queue_max_hw_sectors(q, aoe_maxsectors);
+	gd->major = AOE_MAJOR;
+	gd->first_minor = d->sysminor;
+	gd->fops = &aoe_bdops;
+	gd->private_data = d;
+	set_capacity(gd, d->ssize);
+	snprintf(gd->disk_name, sizeof gd->disk_name, "etherd/e%ld.%d",
+		d->aoemajor, d->aoeminor);
+
+	d->flags &= ~DEVFL_GDALLOC;
+	d->flags |= DEVFL_UP;
+
+	spin_unlock_irqrestore(&d->lock, flags);
+
+	add_disk(gd);
+	aoedisk_add_sysfs(d);
+	aoedisk_add_debugfs(d);
+
+	spin_lock_irqsave(&d->lock, flags);
+	WARN_ON(!(d->flags & DEVFL_GD_NOW));
+	d->flags &= ~DEVFL_GD_NOW;
+	spin_unlock_irqrestore(&d->lock, flags);
+	return;
+
+err_mempool:
+	mempool_destroy(mp);
+err_disk:
+	put_disk(gd);
+err:
+	spin_lock_irqsave(&d->lock, flags);
+	d->flags &= ~DEVFL_GD_NOW;
+	schedule_work(&d->work);
+	spin_unlock_irqrestore(&d->lock, flags);
+}
+
+void
+aoeblk_exit(void)
+{
+	debugfs_remove_recursive(aoe_debugfs_dir);
+	aoe_debugfs_dir = NULL;
+	kmem_cache_destroy(buf_pool_cache);
+}
+
+int __init
+aoeblk_init(void)
+{
+	buf_pool_cache = kmem_cache_create("aoe_bufs",
+					   sizeof(struct buf),
+					   0, 0, NULL);
+	if (buf_pool_cache == NULL)
+		return -ENOMEM;
+	aoe_debugfs_dir = debugfs_create_dir("aoe", NULL);
+	if (IS_ERR_OR_NULL(aoe_debugfs_dir)) {
+		pr_info("aoe: cannot create debugfs directory\n");
+		aoe_debugfs_dir = NULL;
+	}
+	return 0;
+}
+

drivers/block/aoe/aoechr.c

@@ -0,0 +1,320 @@
+/* Copyright (c) 2012 Coraid, Inc.  See COPYING for GPL terms. */
+/*
+ * aoechr.c
+ * AoE character device driver
+ */
+
+#include <linux/hdreg.h>
+#include <linux/blkdev.h>
+#include <linux/completion.h>
+#include <linux/delay.h>
+#include <linux/slab.h>
+#include <linux/mutex.h>
+#include <linux/skbuff.h>
+#include <linux/export.h>
+#include "aoe.h"
+
+enum {
+	//MINOR_STAT = 1, (moved to sysfs)
+	MINOR_ERR = 2,
+	MINOR_DISCOVER,
+	MINOR_INTERFACES,
+	MINOR_REVALIDATE,
+	MINOR_FLUSH,
+	MSGSZ = 2048,
+	NMSG = 100,		/* message backlog to retain */
+};
+
+struct aoe_chardev {
+	ulong minor;
+	char name[32];
+};
+
+enum { EMFL_VALID = 1 };
+
+struct ErrMsg {
+	short flags;
+	short len;
+	char *msg;
+};
+
+static DEFINE_MUTEX(aoechr_mutex);
+
+/* A ring buffer of error messages, to be read through
+ * "/dev/etherd/err".  When no messages are present,
+ * readers will block waiting for messages to appear.
+ */
+static struct ErrMsg emsgs[NMSG];
+static int emsgs_head_idx, emsgs_tail_idx;
+static struct completion emsgs_comp;
+static spinlock_t emsgs_lock;
+static int nblocked_emsgs_readers;
+static struct class *aoe_class;
+static struct aoe_chardev chardevs[] = {
+	{ MINOR_ERR, "err" },
+	{ MINOR_DISCOVER, "discover" },
+	{ MINOR_INTERFACES, "interfaces" },
+	{ MINOR_REVALIDATE, "revalidate" },
+	{ MINOR_FLUSH, "flush" },
+};
+
+static int
+discover(void)
+{
+	aoecmd_cfg(0xffff, 0xff);
+	return 0;
+}
+
+static int
+interfaces(const char __user *str, size_t size)
+{
+	if (set_aoe_iflist(str, size)) {
+		printk(KERN_ERR
+			"aoe: could not set interface list: too many interfaces\n");
+		return -EINVAL;
+	}
+	return 0;
+}
+
+static int
+revalidate(const char __user *str, size_t size)
+{
+	int major, minor, n;
+	ulong flags;
+	struct aoedev *d;
+	struct sk_buff *skb;
+	char buf[16];
+
+	if (size >= sizeof buf)
+		return -EINVAL;
+	buf[sizeof buf - 1] = '\0';
+	if (copy_from_user(buf, str, size))
+		return -EFAULT;
+
+	n = sscanf(buf, "e%d.%d", &major, &minor);
+	if (n != 2) {
+		pr_err("aoe: invalid device specification %s\n", buf);
+		return -EINVAL;
+	}
+	d = aoedev_by_aoeaddr(major, minor, 0);
+	if (!d)
+		return -EINVAL;
+	spin_lock_irqsave(&d->lock, flags);
+	aoecmd_cleanslate(d);
+	aoecmd_cfg(major, minor);
+loop:
+	skb = aoecmd_ata_id(d);
+	spin_unlock_irqrestore(&d->lock, flags);
+	/* try again if we are able to sleep a bit,
+	 * otherwise give up this revalidation
+	 */
+	if (!skb && !msleep_interruptible(250)) {
+		spin_lock_irqsave(&d->lock, flags);
+		goto loop;
+	}
+	aoedev_put(d);
+	if (skb) {
+		struct sk_buff_head queue;
+		__skb_queue_head_init(&queue);
+		__skb_queue_tail(&queue, skb);
+		aoenet_xmit(&queue);
+	}
+	return 0;
+}
+
+void
+aoechr_error(char *msg)
+{
+	struct ErrMsg *em;
+	char *mp;
+	ulong flags, n;
+
+	n = strlen(msg);
+
+	spin_lock_irqsave(&emsgs_lock, flags);
+
+	em = emsgs + emsgs_tail_idx;
+	if ((em->flags & EMFL_VALID)) {
+bail:		spin_unlock_irqrestore(&emsgs_lock, flags);
+		return;
+	}
+
+	mp = kmemdup(msg, n, GFP_ATOMIC);
+	if (mp == NULL) {
+		printk(KERN_ERR "aoe: allocation failure, len=%ld\n", n);
+		goto bail;
+	}
+
+	em->msg = mp;
+	em->flags |= EMFL_VALID;
+	em->len = n;
+
+	emsgs_tail_idx++;
+	emsgs_tail_idx %= ARRAY_SIZE(emsgs);
+
+	spin_unlock_irqrestore(&emsgs_lock, flags);
+
+	if (nblocked_emsgs_readers)
+		complete(&emsgs_comp);
+}
+
+static ssize_t
+aoechr_write(struct file *filp, const char __user *buf, size_t cnt, loff_t *offp)
+{
+	int ret = -EINVAL;
+
+	switch ((unsigned long) filp->private_data) {
+	default:
+		printk(KERN_INFO "aoe: can't write to that file.\n");
+		break;
+	case MINOR_DISCOVER:
+		ret = discover();
+		break;
+	case MINOR_INTERFACES:
+		ret = interfaces(buf, cnt);
+		break;
+	case MINOR_REVALIDATE:
+		ret = revalidate(buf, cnt);
+		break;
+	case MINOR_FLUSH:
+		ret = aoedev_flush(buf, cnt);
+		break;
+	}
+	if (ret == 0)
+		ret = cnt;
+	return ret;
+}
+
+static int
+aoechr_open(struct inode *inode, struct file *filp)
+{
+	int n, i;
+
+	mutex_lock(&aoechr_mutex);
+	n = iminor(inode);
+	filp->private_data = (void *) (unsigned long) n;
+
+	for (i = 0; i < ARRAY_SIZE(chardevs); ++i)
+		if (chardevs[i].minor == n) {
+			mutex_unlock(&aoechr_mutex);
+			return 0;
+		}
+	mutex_unlock(&aoechr_mutex);
+	return -EINVAL;
+}
+
+static int
+aoechr_rel(struct inode *inode, struct file *filp)
+{
+	return 0;
+}
+
+static ssize_t
+aoechr_read(struct file *filp, char __user *buf, size_t cnt, loff_t *off)
+{
+	unsigned long n;
+	char *mp;
+	struct ErrMsg *em;
+	ssize_t len;
+	ulong flags;
+
+	n = (unsigned long) filp->private_data;
+	if (n != MINOR_ERR)
+		return -EFAULT;
+
+	spin_lock_irqsave(&emsgs_lock, flags);
+
+	for (;;) {
+		em = emsgs + emsgs_head_idx;
+		if ((em->flags & EMFL_VALID) != 0)
+			break;
+		if (filp->f_flags & O_NDELAY) {
+			spin_unlock_irqrestore(&emsgs_lock, flags);
+			return -EAGAIN;
+		}
+		nblocked_emsgs_readers++;
+
+		spin_unlock_irqrestore(&emsgs_lock, flags);
+
+		n = wait_for_completion_interruptible(&emsgs_comp);
+
+		spin_lock_irqsave(&emsgs_lock, flags);
+
+		nblocked_emsgs_readers--;
+
+		if (n) {
+			spin_unlock_irqrestore(&emsgs_lock, flags);
+			return -ERESTARTSYS;
+		}
+	}
+	if (em->len > cnt) {
+		spin_unlock_irqrestore(&emsgs_lock, flags);
+		return -EAGAIN;
+	}
+	mp = em->msg;
+	len = em->len;
+	em->msg = NULL;
+	em->flags &= ~EMFL_VALID;
+
+	emsgs_head_idx++;
+	emsgs_head_idx %= ARRAY_SIZE(emsgs);
+
+	spin_unlock_irqrestore(&emsgs_lock, flags);
+
+	n = copy_to_user(buf, mp, len);
+	kfree(mp);
+	return n == 0 ? len : -EFAULT;
+}
+
+static const struct file_operations aoe_fops = {
+	.write = aoechr_write,
+	.read = aoechr_read,
+	.open = aoechr_open,
+	.release = aoechr_rel,
+	.owner = THIS_MODULE,
+	.llseek = noop_llseek,
+};
+
+static char *aoe_devnode(struct device *dev, umode_t *mode)
+{
+	return kasprintf(GFP_KERNEL, "etherd/%s", dev_name(dev));
+}
+
+int __init
+aoechr_init(void)
+{
+	int n, i;
+
+	n = register_chrdev(AOE_MAJOR, "aoechr", &aoe_fops);
+	if (n < 0) {
+		printk(KERN_ERR "aoe: can't register char device\n");
+		return n;
+	}
+	init_completion(&emsgs_comp);
+	spin_lock_init(&emsgs_lock);
+	aoe_class = class_create(THIS_MODULE, "aoe");
+	if (IS_ERR(aoe_class)) {
+		unregister_chrdev(AOE_MAJOR, "aoechr");
+		return PTR_ERR(aoe_class);
+	}
+	aoe_class->devnode = aoe_devnode;
+
+	for (i = 0; i < ARRAY_SIZE(chardevs); ++i)
+		device_create(aoe_class, NULL,
+			      MKDEV(AOE_MAJOR, chardevs[i].minor), NULL,
+			      chardevs[i].name);
+
+	return 0;
+}
+
+void
+aoechr_exit(void)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(chardevs); ++i)
+		device_destroy(aoe_class, MKDEV(AOE_MAJOR, chardevs[i].minor));
+	class_destroy(aoe_class);
+	unregister_chrdev(AOE_MAJOR, "aoechr");
+}
+

drivers/block/aoe/aoedev.c

@@ -0,0 +1,526 @@
+/* Copyright (c) 2013 Coraid, Inc.  See COPYING for GPL terms. */
+/*
+ * aoedev.c
+ * AoE device utility functions; maintains device list.
+ */
+
+#include <linux/hdreg.h>
+#include <linux/blkdev.h>
+#include <linux/netdevice.h>
+#include <linux/delay.h>
+#include <linux/slab.h>
+#include <linux/bitmap.h>
+#include <linux/kdev_t.h>
+#include <linux/moduleparam.h>
+#include <linux/string.h>
+#include "aoe.h"
+
+static void dummy_timer(ulong);
+static void freetgt(struct aoedev *d, struct aoetgt *t);
+static void skbpoolfree(struct aoedev *d);
+
+static int aoe_dyndevs = 1;
+module_param(aoe_dyndevs, int, 0644);
+MODULE_PARM_DESC(aoe_dyndevs, "Use dynamic minor numbers for devices.");
+
+static struct aoedev *devlist;
+static DEFINE_SPINLOCK(devlist_lock);
+
+/* Because some systems will have one, many, or no
+ *   - partitions,
+ *   - slots per shelf,
+ *   - or shelves,
+ * we need some flexibility in the way the minor numbers
+ * are allocated.  So they are dynamic.
+ */
+#define N_DEVS ((1U<<MINORBITS)/AOE_PARTITIONS)
+
+static DEFINE_SPINLOCK(used_minors_lock);
+static DECLARE_BITMAP(used_minors, N_DEVS);
+
+static int
+minor_get_dyn(ulong *sysminor)
+{
+	ulong flags;
+	ulong n;
+	int error = 0;
+
+	spin_lock_irqsave(&used_minors_lock, flags);
+	n = find_first_zero_bit(used_minors, N_DEVS);
+	if (n < N_DEVS)
+		set_bit(n, used_minors);
+	else
+		error = -1;
+	spin_unlock_irqrestore(&used_minors_lock, flags);
+
+	*sysminor = n * AOE_PARTITIONS;
+	return error;
+}
+
+static int
+minor_get_static(ulong *sysminor, ulong aoemaj, int aoemin)
+{
+	ulong flags;
+	ulong n;
+	int error = 0;
+	enum {
+		/* for backwards compatibility when !aoe_dyndevs,
+		 * a static number of supported slots per shelf */
+		NPERSHELF = 16,
+	};
+
+	if (aoemin >= NPERSHELF) {
+		pr_err("aoe: %s %d slots per shelf\n",
+			"static minor device numbers support only",
+			NPERSHELF);
+		error = -1;
+		goto out;
+	}
+
+	n = aoemaj * NPERSHELF + aoemin;
+	if (n >= N_DEVS) {
+		pr_err("aoe: %s with e%ld.%d\n",
+			"cannot use static minor device numbers",
+			aoemaj, aoemin);
+		error = -1;
+		goto out;
+	}
+
+	spin_lock_irqsave(&used_minors_lock, flags);
+	if (test_bit(n, used_minors)) {
+		pr_err("aoe: %s %lu\n",
+			"existing device already has static minor number",
+			n);
+		error = -1;
+	} else
+		set_bit(n, used_minors);
+	spin_unlock_irqrestore(&used_minors_lock, flags);
+	*sysminor = n * AOE_PARTITIONS;
+out:
+	return error;
+}
+
+static int
+minor_get(ulong *sysminor, ulong aoemaj, int aoemin)
+{
+	if (aoe_dyndevs)
+		return minor_get_dyn(sysminor);
+	else
+		return minor_get_static(sysminor, aoemaj, aoemin);
+}
+
+static void
+minor_free(ulong minor)
+{
+	ulong flags;
+
+	minor /= AOE_PARTITIONS;
+	BUG_ON(minor >= N_DEVS);
+
+	spin_lock_irqsave(&used_minors_lock, flags);
+	BUG_ON(!test_bit(minor, used_minors));
+	clear_bit(minor, used_minors);
+	spin_unlock_irqrestore(&used_minors_lock, flags);
+}
+
+/*
+ * Users who grab a pointer to the device with aoedev_by_aoeaddr
+ * automatically get a reference count and must be responsible
+ * for performing a aoedev_put.  With the addition of async
+ * kthread processing I'm no longer confident that we can
+ * guarantee consistency in the face of device flushes.
+ *
+ * For the time being, we only bother to add extra references for
+ * frames sitting on the iocq.  When the kthreads finish processing
+ * these frames, they will aoedev_put the device.
+ */
+
+void
+aoedev_put(struct aoedev *d)
+{
+	ulong flags;
+
+	spin_lock_irqsave(&devlist_lock, flags);
+	d->ref--;
+	spin_unlock_irqrestore(&devlist_lock, flags);
+}
+
+static void
+dummy_timer(ulong vp)
+{
+	struct aoedev *d;
+
+	d = (struct aoedev *)vp;
+	if (d->flags & DEVFL_TKILL)
+		return;
+	d->timer.expires = jiffies + HZ;
+	add_timer(&d->timer);
+}
+
+static void
+aoe_failip(struct aoedev *d)
+{
+	struct request *rq;
+	struct bio *bio;
+	unsigned long n;
+
+	aoe_failbuf(d, d->ip.buf);
+
+	rq = d->ip.rq;
+	if (rq == NULL)
+		return;
+	while ((bio = d->ip.nxbio)) {
+		clear_bit(BIO_UPTODATE, &bio->bi_flags);
+		d->ip.nxbio = bio->bi_next;
+		n = (unsigned long) rq->special;
+		rq->special = (void *) --n;
+	}
+	if ((unsigned long) rq->special == 0)
+		aoe_end_request(d, rq, 0);
+}
+
+static void
+downdev_frame(struct list_head *pos)
+{
+	struct frame *f;
+
+	f = list_entry(pos, struct frame, head);
+	list_del(pos);
+	if (f->buf) {
+		f->buf->nframesout--;
+		aoe_failbuf(f->t->d, f->buf);
+	}
+	aoe_freetframe(f);
+}
+
+void
+aoedev_downdev(struct aoedev *d)
+{
+	struct aoetgt *t, **tt, **te;
+	struct list_head *head, *pos, *nx;
+	struct request *rq;
+	int i;
+
+	d->flags &= ~DEVFL_UP;
+
+	/* clean out active and to-be-retransmitted buffers */
+	for (i = 0; i < NFACTIVE; i++) {
+		head = &d->factive[i];
+		list_for_each_safe(pos, nx, head)
+			downdev_frame(pos);
+	}
+	head = &d->rexmitq;
+	list_for_each_safe(pos, nx, head)
+		downdev_frame(pos);
+
+	/* reset window dressings */
+	tt = d->targets;
+	te = tt + d->ntargets;
+	for (; tt < te && (t = *tt); tt++) {
+		aoecmd_wreset(t);
+		t->nout = 0;
+	}
+
+	/* clean out the in-process request (if any) */
+	aoe_failip(d);
+
+	/* fast fail all pending I/O */
+	if (d->blkq) {
+		while ((rq = blk_peek_request(d->blkq))) {
+			blk_start_request(rq);
+			aoe_end_request(d, rq, 1);
+		}
+	}
+
+	if (d->gd)
+		set_capacity(d->gd, 0);
+}
+
+/* return whether the user asked for this particular
+ * device to be flushed
+ */
+static int
+user_req(char *s, size_t slen, struct aoedev *d)
+{
+	const char *p;
+	size_t lim;
+
+	if (!d->gd)
+		return 0;
+	p = kbasename(d->gd->disk_name);
+	lim = sizeof(d->gd->disk_name);
+	lim -= p - d->gd->disk_name;
+	if (slen < lim)
+		lim = slen;
+
+	return !strncmp(s, p, lim);
+}
+
+static void
+freedev(struct aoedev *d)
+{
+	struct aoetgt **t, **e;
+	int freeing = 0;
+	unsigned long flags;
+
+	spin_lock_irqsave(&d->lock, flags);
+	if (d->flags & DEVFL_TKILL
+	&& !(d->flags & DEVFL_FREEING)) {
+		d->flags |= DEVFL_FREEING;
+		freeing = 1;
+	}
+	spin_unlock_irqrestore(&d->lock, flags);
+	if (!freeing)
+		return;
+
+	del_timer_sync(&d->timer);
+	if (d->gd) {
+		aoedisk_rm_debugfs(d);
+		aoedisk_rm_sysfs(d);
+		del_gendisk(d->gd);
+		put_disk(d->gd);
+		blk_cleanup_queue(d->blkq);
+	}
+	t = d->targets;
+	e = t + d->ntargets;
+	for (; t < e && *t; t++)
+		freetgt(d, *t);
+	if (d->bufpool)
+		mempool_destroy(d->bufpool);
+	skbpoolfree(d);
+	minor_free(d->sysminor);
+
+	spin_lock_irqsave(&d->lock, flags);
+	d->flags |= DEVFL_FREED;
+	spin_unlock_irqrestore(&d->lock, flags);
+}
+
+enum flush_parms {
+	NOT_EXITING = 0,
+	EXITING = 1,
+};
+
+static int
+flush(const char __user *str, size_t cnt, int exiting)
+{
+	ulong flags;
+	struct aoedev *d, **dd;
+	char buf[16];
+	int all = 0;
+	int specified = 0;	/* flush a specific device */
+	unsigned int skipflags;
+
+	skipflags = DEVFL_GDALLOC | DEVFL_NEWSIZE | DEVFL_TKILL;
+
+	if (!exiting && cnt >= 3) {
+		if (cnt > sizeof buf)
+			cnt = sizeof buf;
+		if (copy_from_user(buf, str, cnt))
+			return -EFAULT;
+		all = !strncmp(buf, "all", 3);
+		if (!all)
+			specified = 1;
+	}
+
+	flush_scheduled_work();
+	/* pass one: without sleeping, do aoedev_downdev */
+	spin_lock_irqsave(&devlist_lock, flags);
+	for (d = devlist; d; d = d->next) {
+		spin_lock(&d->lock);
+		if (exiting) {
+			/* unconditionally take each device down */
+		} else if (specified) {
+			if (!user_req(buf, cnt, d))
+				goto cont;
+		} else if ((!all && (d->flags & DEVFL_UP))
+		|| d->flags & skipflags
+		|| d->nopen
+		|| d->ref)
+			goto cont;
+
+		aoedev_downdev(d);
+		d->flags |= DEVFL_TKILL;
+cont:
+		spin_unlock(&d->lock);
+	}
+	spin_unlock_irqrestore(&devlist_lock, flags);
+
+	/* pass two: call freedev, which might sleep,
+	 * for aoedevs marked with DEVFL_TKILL
+	 */
+restart:
+	spin_lock_irqsave(&devlist_lock, flags);
+	for (d = devlist; d; d = d->next) {
+		spin_lock(&d->lock);
+		if (d->flags & DEVFL_TKILL
+		&& !(d->flags & DEVFL_FREEING)) {
+			spin_unlock(&d->lock);
+			spin_unlock_irqrestore(&devlist_lock, flags);
+			freedev(d);
+			goto restart;
+		}
+		spin_unlock(&d->lock);
+	}
+
+	/* pass three: remove aoedevs marked with DEVFL_FREED */
+	for (dd = &devlist, d = *dd; d; d = *dd) {
+		struct aoedev *doomed = NULL;
+
+		spin_lock(&d->lock);
+		if (d->flags & DEVFL_FREED) {
+			*dd = d->next;
+			doomed = d;
+		} else {
+			dd = &d->next;
+		}
+		spin_unlock(&d->lock);
+		if (doomed)
+			kfree(doomed->targets);
+		kfree(doomed);
+	}
+	spin_unlock_irqrestore(&devlist_lock, flags);
+
+	return 0;
+}
+
+int
+aoedev_flush(const char __user *str, size_t cnt)
+{
+	return flush(str, cnt, NOT_EXITING);
+}
+
+/* This has been confirmed to occur once with Tms=3*1000 due to the
+ * driver changing link and not processing its transmit ring.  The
+ * problem is hard enough to solve by returning an error that I'm
+ * still punting on "solving" this.
+ */
+static void
+skbfree(struct sk_buff *skb)
+{
+	enum { Sms = 250, Tms = 30 * 1000};
+	int i = Tms / Sms;
+
+	if (skb == NULL)
+		return;
+	while (atomic_read(&skb_shinfo(skb)->dataref) != 1 && i-- > 0)
+		msleep(Sms);
+	if (i < 0) {
+		printk(KERN_ERR
+			"aoe: %s holds ref: %s\n",
+			skb->dev ? skb->dev->name : "netif",
+			"cannot free skb -- memory leaked.");
+		return;
+	}
+	skb->truesize -= skb->data_len;
+	skb_shinfo(skb)->nr_frags = skb->data_len = 0;
+	skb_trim(skb, 0);
+	dev_kfree_skb(skb);
+}
+
+static void
+skbpoolfree(struct aoedev *d)
+{
+	struct sk_buff *skb, *tmp;
+
+	skb_queue_walk_safe(&d->skbpool, skb, tmp)
+		skbfree(skb);
+
+	__skb_queue_head_init(&d->skbpool);
+}
+
+/* find it or allocate it */
+struct aoedev *
+aoedev_by_aoeaddr(ulong maj, int min, int do_alloc)
+{
+	struct aoedev *d;
+	int i;
+	ulong flags;
+	ulong sysminor = 0;
+
+	spin_lock_irqsave(&devlist_lock, flags);
+
+	for (d=devlist; d; d=d->next)
+		if (d->aoemajor == maj && d->aoeminor == min) {
+			spin_lock(&d->lock);
+			if (d->flags & DEVFL_TKILL) {
+				spin_unlock(&d->lock);
+				d = NULL;
+				goto out;
+			}
+			d->ref++;
+			spin_unlock(&d->lock);
+			break;
+		}
+	if (d || !do_alloc || minor_get(&sysminor, maj, min) < 0)
+		goto out;
+	d = kcalloc(1, sizeof *d, GFP_ATOMIC);
+	if (!d)
+		goto out;
+	d->targets = kcalloc(NTARGETS, sizeof(*d->targets), GFP_ATOMIC);
+	if (!d->targets) {
+		kfree(d);
+		d = NULL;
+		goto out;
+	}
+	d->ntargets = NTARGETS;
+	INIT_WORK(&d->work, aoecmd_sleepwork);
+	spin_lock_init(&d->lock);
+	skb_queue_head_init(&d->skbpool);
+	init_timer(&d->timer);
+	d->timer.data = (ulong) d;
+	d->timer.function = dummy_timer;
+	d->timer.expires = jiffies + HZ;
+	add_timer(&d->timer);
+	d->bufpool = NULL;	/* defer to aoeblk_gdalloc */
+	d->tgt = d->targets;
+	d->ref = 1;
+	for (i = 0; i < NFACTIVE; i++)
+		INIT_LIST_HEAD(&d->factive[i]);
+	INIT_LIST_HEAD(&d->rexmitq);
+	d->sysminor = sysminor;
+	d->aoemajor = maj;
+	d->aoeminor = min;
+	d->rttavg = RTTAVG_INIT;
+	d->rttdev = RTTDEV_INIT;
+	d->next = devlist;
+	devlist = d;
+ out:
+	spin_unlock_irqrestore(&devlist_lock, flags);
+	return d;
+}
+
+static void
+freetgt(struct aoedev *d, struct aoetgt *t)
+{
+	struct frame *f;
+	struct list_head *pos, *nx, *head;
+	struct aoeif *ifp;
+
+	for (ifp = t->ifs; ifp < &t->ifs[NAOEIFS]; ++ifp) {
+		if (!ifp->nd)
+			break;
+		dev_put(ifp->nd);
+	}
+
+	head = &t->ffree;
+	list_for_each_safe(pos, nx, head) {
+		list_del(pos);
+		f = list_entry(pos, struct frame, head);
+		skbfree(f->skb);
+		kfree(f);
+	}
+	kfree(t);
+}
+
+void
+aoedev_exit(void)
+{
+	flush_scheduled_work();
+	flush(NULL, 0, EXITING);
+}
+
+int __init
+aoedev_init(void)
+{
+	return 0;
+}

drivers/block/aoe/aoemain.c

@@ -0,0 +1,115 @@
+/* Copyright (c) 2012 Coraid, Inc.  See COPYING for GPL terms. */
+/*
+ * aoemain.c
+ * Module initialization routines, discover timer
+ */
+
+#include <linux/hdreg.h>
+#include <linux/blkdev.h>
+#include <linux/module.h>
+#include <linux/skbuff.h>
+#include "aoe.h"
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Sam Hopkins <sah@coraid.com>");
+MODULE_DESCRIPTION("AoE block/char driver for 2.6.2 and newer 2.6 kernels");
+MODULE_VERSION(VERSION);
+
+enum { TINIT, TRUN, TKILL };
+
+static void
+discover_timer(ulong vp)
+{
+	static struct timer_list t;
+	static volatile ulong die;
+	static spinlock_t lock;
+	ulong flags;
+	enum { DTIMERTICK = HZ * 60 }; /* one minute */
+
+	switch (vp) {
+	case TINIT:
+		init_timer(&t);
+		spin_lock_init(&lock);
+		t.data = TRUN;
+		t.function = discover_timer;
+		die = 0;
+	case TRUN:
+		spin_lock_irqsave(&lock, flags);
+		if (!die) {
+			t.expires = jiffies + DTIMERTICK;
+			add_timer(&t);
+		}
+		spin_unlock_irqrestore(&lock, flags);
+
+		aoecmd_cfg(0xffff, 0xff);
+		return;
+	case TKILL:
+		spin_lock_irqsave(&lock, flags);
+		die = 1;
+		spin_unlock_irqrestore(&lock, flags);
+
+		del_timer_sync(&t);
+	default:
+		return;
+	}
+}
+
+static void
+aoe_exit(void)
+{
+	discover_timer(TKILL);
+
+	aoenet_exit();
+	unregister_blkdev(AOE_MAJOR, DEVICE_NAME);
+	aoecmd_exit();
+	aoechr_exit();
+	aoedev_exit();
+	aoeblk_exit();		/* free cache after de-allocating bufs */
+}
+
+static int __init
+aoe_init(void)
+{
+	int ret;
+
+	ret = aoedev_init();
+	if (ret)
+		return ret;
+	ret = aoechr_init();
+	if (ret)
+		goto chr_fail;
+	ret = aoeblk_init();
+	if (ret)
+		goto blk_fail;
+	ret = aoenet_init();
+	if (ret)
+		goto net_fail;
+	ret = aoecmd_init();
+	if (ret)
+		goto cmd_fail;
+	ret = register_blkdev(AOE_MAJOR, DEVICE_NAME);
+	if (ret < 0) {
+		printk(KERN_ERR "aoe: can't register major\n");
+		goto blkreg_fail;
+	}
+	printk(KERN_INFO "aoe: AoE v%s initialised.\n", VERSION);
+	discover_timer(TINIT);
+	return 0;
+ blkreg_fail:
+	aoecmd_exit();
+ cmd_fail:
+	aoenet_exit();
+ net_fail:
+	aoeblk_exit();
+ blk_fail:
+	aoechr_exit();
+ chr_fail:
+	aoedev_exit();
+
+	printk(KERN_INFO "aoe: initialisation failure.\n");
+	return ret;
+}
+
+module_init(aoe_init);
+module_exit(aoe_exit);
+

drivers/block/aoe/aoenet.c

@@ -0,0 +1,223 @@
+/* Copyright (c) 2013 Coraid, Inc.  See COPYING for GPL terms. */
+/*
+ * aoenet.c
+ * Ethernet portion of AoE driver
+ */
+
+#include <linux/gfp.h>
+#include <linux/hdreg.h>
+#include <linux/blkdev.h>
+#include <linux/netdevice.h>
+#include <linux/moduleparam.h>
+#include <net/net_namespace.h>
+#include <asm/unaligned.h>
+#include "aoe.h"
+
+#define NECODES 5
+
+static char *aoe_errlist[] =
+{
+	"no such error",
+	"unrecognized command code",
+	"bad argument parameter",
+	"device unavailable",
+	"config string present",
+	"unsupported version"
+};
+
+enum {
+	IFLISTSZ = 1024,
+};
+
+static char aoe_iflist[IFLISTSZ];
+module_param_string(aoe_iflist, aoe_iflist, IFLISTSZ, 0600);
+MODULE_PARM_DESC(aoe_iflist, "aoe_iflist=dev1[,dev2...]");
+
+static wait_queue_head_t txwq;
+static struct ktstate kts;
+
+#ifndef MODULE
+static int __init aoe_iflist_setup(char *str)
+{
+	strncpy(aoe_iflist, str, IFLISTSZ);
+	aoe_iflist[IFLISTSZ - 1] = '\0';
+	return 1;
+}
+
+__setup("aoe_iflist=", aoe_iflist_setup);
+#endif
+
+static spinlock_t txlock;
+static struct sk_buff_head skbtxq;
+
+/* enters with txlock held */
+static int
+tx(int id) __must_hold(&txlock)
+{
+	struct sk_buff *skb;
+	struct net_device *ifp;
+
+	while ((skb = skb_dequeue(&skbtxq))) {
+		spin_unlock_irq(&txlock);
+		ifp = skb->dev;
+		if (dev_queue_xmit(skb) == NET_XMIT_DROP && net_ratelimit())
+			pr_warn("aoe: packet could not be sent on %s.  %s\n",
+				ifp ? ifp->name : "netif",
+				"consider increasing tx_queue_len");
+		spin_lock_irq(&txlock);
+	}
+	return 0;
+}
+
+int
+is_aoe_netif(struct net_device *ifp)
+{
+	register char *p, *q;
+	register int len;
+
+	if (aoe_iflist[0] == '\0')
+		return 1;
+
+	p = aoe_iflist + strspn(aoe_iflist, WHITESPACE);
+	for (; *p; p = q + strspn(q, WHITESPACE)) {
+		q = p + strcspn(p, WHITESPACE);
+		if (q != p)
+			len = q - p;
+		else
+			len = strlen(p); /* last token in aoe_iflist */
+
+		if (strlen(ifp->name) == len && !strncmp(ifp->name, p, len))
+			return 1;
+		if (q == p)
+			break;
+	}
+
+	return 0;
+}
+
+int
+set_aoe_iflist(const char __user *user_str, size_t size)
+{
+	if (size >= IFLISTSZ)
+		return -EINVAL;
+
+	if (copy_from_user(aoe_iflist, user_str, size)) {
+		printk(KERN_INFO "aoe: copy from user failed\n");
+		return -EFAULT;
+	}
+	aoe_iflist[size] = 0x00;
+	return 0;
+}
+
+void
+aoenet_xmit(struct sk_buff_head *queue)
+{
+	struct sk_buff *skb, *tmp;
+	ulong flags;
+
+	skb_queue_walk_safe(queue, skb, tmp) {
+		__skb_unlink(skb, queue);
+		spin_lock_irqsave(&txlock, flags);
+		skb_queue_tail(&skbtxq, skb);
+		spin_unlock_irqrestore(&txlock, flags);
+		wake_up(&txwq);
+	}
+}
+
+/*
+ * (1) len doesn't include the header by default.  I want this.
+ */
+static int
+aoenet_rcv(struct sk_buff *skb, struct net_device *ifp, struct packet_type *pt, struct net_device *orig_dev)
+{
+	struct aoe_hdr *h;
+	struct aoe_atahdr *ah;
+	u32 n;
+	int sn;
+
+	if (dev_net(ifp) != &init_net)
+		goto exit;
+
+	skb = skb_share_check(skb, GFP_ATOMIC);
+	if (skb == NULL)
+		return 0;
+	if (!is_aoe_netif(ifp))
+		goto exit;
+	skb_push(skb, ETH_HLEN);	/* (1) */
+	sn = sizeof(*h) + sizeof(*ah);
+	if (skb->len >= sn) {
+		sn -= skb_headlen(skb);
+		if (sn > 0 && !__pskb_pull_tail(skb, sn))
+			goto exit;
+	}
+	h = (struct aoe_hdr *) skb->data;
+	n = get_unaligned_be32(&h->tag);
+	if ((h->verfl & AOEFL_RSP) == 0 || (n & 1<<31))
+		goto exit;
+
+	if (h->verfl & AOEFL_ERR) {
+		n = h->err;
+		if (n > NECODES)
+			n = 0;
+		if (net_ratelimit())
+			printk(KERN_ERR
+				"%s%d.%d@%s; ecode=%d '%s'\n",
+				"aoe: error packet from ",
+				get_unaligned_be16(&h->major),
+				h->minor, skb->dev->name,
+				h->err, aoe_errlist[n]);
+		goto exit;
+	}
+
+	switch (h->cmd) {
+	case AOECMD_ATA:
+		/* ata_rsp may keep skb for later processing or give it back */
+		skb = aoecmd_ata_rsp(skb);
+		break;
+	case AOECMD_CFG:
+		aoecmd_cfg_rsp(skb);
+		break;
+	default:
+		if (h->cmd >= AOECMD_VEND_MIN)
+			break;	/* don't complain about vendor commands */
+		pr_info("aoe: unknown AoE command type 0x%02x\n", h->cmd);
+		break;
+	}
+
+	if (!skb)
+		return 0;
+exit:
+	dev_kfree_skb(skb);
+	return 0;
+}
+
+static struct packet_type aoe_pt __read_mostly = {
+	.type = __constant_htons(ETH_P_AOE),
+	.func = aoenet_rcv,
+};
+
+int __init
+aoenet_init(void)
+{
+	skb_queue_head_init(&skbtxq);
+	init_waitqueue_head(&txwq);
+	spin_lock_init(&txlock);
+	kts.lock = &txlock;
+	kts.fn = tx;
+	kts.waitq = &txwq;
+	kts.id = 0;
+	snprintf(kts.name, sizeof(kts.name), "aoe_tx%d", kts.id);
+	if (aoe_ktstart(&kts))
+		return -EAGAIN;
+	dev_add_pack(&aoe_pt);
+	return 0;
+}
+
+void
+aoenet_exit(void)
+{
+	aoe_ktstop(&kts);
+	skb_queue_purge(&skbtxq);
+	dev_remove_pack(&aoe_pt);
+}
+

drivers/block/brd.c

@@ -0,0 +1,666 @@
+/*
+ * Ram backed block device driver.
+ *
+ * Copyright (C) 2007 Nick Piggin
+ * Copyright (C) 2007 Novell Inc.
+ *
+ * Parts derived from drivers/block/rd.c, and drivers/block/loop.c, copyright
+ * of their respective owners.
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/major.h>
+#include <linux/blkdev.h>
+#include <linux/bio.h>
+#include <linux/highmem.h>
+#include <linux/mutex.h>
+#include <linux/radix-tree.h>
+#include <linux/fs.h>
+#include <linux/slab.h>
+
+#include <asm/uaccess.h>
+
+#define SECTOR_SHIFT		9
+#define PAGE_SECTORS_SHIFT	(PAGE_SHIFT - SECTOR_SHIFT)
+#define PAGE_SECTORS		(1 << PAGE_SECTORS_SHIFT)
+
+/*
+ * Each block ramdisk device has a radix_tree brd_pages of pages that stores
+ * the pages containing the block device's contents. A brd page's ->index is
+ * its offset in PAGE_SIZE units. This is similar to, but in no way connected
+ * with, the kernel's pagecache or buffer cache (which sit above our block
+ * device).
+ */
+struct brd_device {
+	int		brd_number;
+
+	struct request_queue	*brd_queue;
+	struct gendisk		*brd_disk;
+	struct list_head	brd_list;
+
+	/*
+	 * Backing store of pages and lock to protect it. This is the contents
+	 * of the block device.
+	 */
+	spinlock_t		brd_lock;
+	struct radix_tree_root	brd_pages;
+};
+
+/*
+ * Look up and return a brd's page for a given sector.
+ */
+static DEFINE_MUTEX(brd_mutex);
+static struct page *brd_lookup_page(struct brd_device *brd, sector_t sector)
+{
+	pgoff_t idx;
+	struct page *page;
+
+	/*
+	 * The page lifetime is protected by the fact that we have opened the
+	 * device node -- brd pages will never be deleted under us, so we
+	 * don't need any further locking or refcounting.
+	 *
+	 * This is strictly true for the radix-tree nodes as well (ie. we
+	 * don't actually need the rcu_read_lock()), however that is not a
+	 * documented feature of the radix-tree API so it is better to be
+	 * safe here (we don't have total exclusion from radix tree updates
+	 * here, only deletes).
+	 */
+	rcu_read_lock();
+	idx = sector >> PAGE_SECTORS_SHIFT; /* sector to page index */
+	page = radix_tree_lookup(&brd->brd_pages, idx);
+	rcu_read_unlock();
+
+	BUG_ON(page && page->index != idx);
+
+	return page;
+}
+
+/*
+ * Look up and return a brd's page for a given sector.
+ * If one does not exist, allocate an empty page, and insert that. Then
+ * return it.
+ */
+static struct page *brd_insert_page(struct brd_device *brd, sector_t sector)
+{
+	pgoff_t idx;
+	struct page *page;
+	gfp_t gfp_flags;
+
+	page = brd_lookup_page(brd, sector);
+	if (page)
+		return page;
+
+	/*
+	 * Must use NOIO because we don't want to recurse back into the
+	 * block or filesystem layers from page reclaim.
+	 *
+	 * Cannot support XIP and highmem, because our ->direct_access
+	 * routine for XIP must return memory that is always addressable.
+	 * If XIP was reworked to use pfns and kmap throughout, this
+	 * restriction might be able to be lifted.
+	 */
+	gfp_flags = GFP_NOIO | __GFP_ZERO;
+#ifndef CONFIG_BLK_DEV_XIP
+	gfp_flags |= __GFP_HIGHMEM;
+#endif
+	page = alloc_page(gfp_flags);
+	if (!page)
+		return NULL;
+
+	if (radix_tree_preload(GFP_NOIO)) {
+		__free_page(page);
+		return NULL;
+	}
+
+	spin_lock(&brd->brd_lock);
+	idx = sector >> PAGE_SECTORS_SHIFT;
+	page->index = idx;
+	if (radix_tree_insert(&brd->brd_pages, idx, page)) {
+		__free_page(page);
+		page = radix_tree_lookup(&brd->brd_pages, idx);
+		BUG_ON(!page);
+		BUG_ON(page->index != idx);
+	}
+	spin_unlock(&brd->brd_lock);
+
+	radix_tree_preload_end();
+
+	return page;
+}
+
+static void brd_free_page(struct brd_device *brd, sector_t sector)
+{
+	struct page *page;
+	pgoff_t idx;
+
+	spin_lock(&brd->brd_lock);
+	idx = sector >> PAGE_SECTORS_SHIFT;
+	page = radix_tree_delete(&brd->brd_pages, idx);
+	spin_unlock(&brd->brd_lock);
+	if (page)
+		__free_page(page);
+}
+
+static void brd_zero_page(struct brd_device *brd, sector_t sector)
+{
+	struct page *page;
+
+	page = brd_lookup_page(brd, sector);
+	if (page)
+		clear_highpage(page);
+}
+
+/*
+ * Free all backing store pages and radix tree. This must only be called when
+ * there are no other users of the device.
+ */
+#define FREE_BATCH 16
+static void brd_free_pages(struct brd_device *brd)
+{
+	unsigned long pos = 0;
+	struct page *pages[FREE_BATCH];
+	int nr_pages;
+
+	do {
+		int i;
+
+		nr_pages = radix_tree_gang_lookup(&brd->brd_pages,
+				(void **)pages, pos, FREE_BATCH);
+
+		for (i = 0; i < nr_pages; i++) {
+			void *ret;
+
+			BUG_ON(pages[i]->index < pos);
+			pos = pages[i]->index;
+			ret = radix_tree_delete(&brd->brd_pages, pos);
+			BUG_ON(!ret || ret != pages[i]);
+			__free_page(pages[i]);
+		}
+
+		pos++;
+
+		/*
+		 * This assumes radix_tree_gang_lookup always returns as
+		 * many pages as possible. If the radix-tree code changes,
+		 * so will this have to.
+		 */
+	} while (nr_pages == FREE_BATCH);
+}
+
+/*
+ * copy_to_brd_setup must be called before copy_to_brd. It may sleep.
+ */
+static int copy_to_brd_setup(struct brd_device *brd, sector_t sector, size_t n)
+{
+	unsigned int offset = (sector & (PAGE_SECTORS-1)) << SECTOR_SHIFT;
+	size_t copy;
+
+	copy = min_t(size_t, n, PAGE_SIZE - offset);
+	if (!brd_insert_page(brd, sector))
+		return -ENOSPC;
+	if (copy < n) {
+		sector += copy >> SECTOR_SHIFT;
+		if (!brd_insert_page(brd, sector))
+			return -ENOSPC;
+	}
+	return 0;
+}
+
+static void discard_from_brd(struct brd_device *brd,
+			sector_t sector, size_t n)
+{
+	while (n >= PAGE_SIZE) {
+		/*
+		 * Don't want to actually discard pages here because
+		 * re-allocating the pages can result in writeback
+		 * deadlocks under heavy load.
+		 */
+		if (0)
+			brd_free_page(brd, sector);
+		else
+			brd_zero_page(brd, sector);
+		sector += PAGE_SIZE >> SECTOR_SHIFT;
+		n -= PAGE_SIZE;
+	}
+}
+
+/*
+ * Copy n bytes from src to the brd starting at sector. Does not sleep.
+ */
+static void copy_to_brd(struct brd_device *brd, const void *src,
+			sector_t sector, size_t n)
+{
+	struct page *page;
+	void *dst;
+	unsigned int offset = (sector & (PAGE_SECTORS-1)) << SECTOR_SHIFT;
+	size_t copy;
+
+	copy = min_t(size_t, n, PAGE_SIZE - offset);
+	page = brd_lookup_page(brd, sector);
+	BUG_ON(!page);
+
+	dst = kmap_atomic(page);
+	memcpy(dst + offset, src, copy);
+	kunmap_atomic(dst);
+
+	if (copy < n) {
+		src += copy;
+		sector += copy >> SECTOR_SHIFT;
+		copy = n - copy;
+		page = brd_lookup_page(brd, sector);
+		BUG_ON(!page);
+
+		dst = kmap_atomic(page);
+		memcpy(dst, src, copy);
+		kunmap_atomic(dst);
+	}
+}
+
+/*
+ * Copy n bytes to dst from the brd starting at sector. Does not sleep.
+ */
+static void copy_from_brd(void *dst, struct brd_device *brd,
+			sector_t sector, size_t n)
+{
+	struct page *page;
+	void *src;
+	unsigned int offset = (sector & (PAGE_SECTORS-1)) << SECTOR_SHIFT;
+	size_t copy;
+
+	copy = min_t(size_t, n, PAGE_SIZE - offset);
+	page = brd_lookup_page(brd, sector);
+	if (page) {
+		src = kmap_atomic(page);
+		memcpy(dst, src + offset, copy);
+		kunmap_atomic(src);
+	} else
+		memset(dst, 0, copy);
+
+	if (copy < n) {
+		dst += copy;
+		sector += copy >> SECTOR_SHIFT;
+		copy = n - copy;
+		page = brd_lookup_page(brd, sector);
+		if (page) {
+			src = kmap_atomic(page);
+			memcpy(dst, src, copy);
+			kunmap_atomic(src);
+		} else
+			memset(dst, 0, copy);
+	}
+}
+
+/*
+ * Process a single bvec of a bio.
+ */
+static int brd_do_bvec(struct brd_device *brd, struct page *page,
+			unsigned int len, unsigned int off, int rw,
+			sector_t sector)
+{
+	void *mem;
+	int err = 0;
+
+	if (rw != READ) {
+		err = copy_to_brd_setup(brd, sector, len);
+		if (err)
+			goto out;
+	}
+
+	mem = kmap_atomic(page);
+	if (rw == READ) {
+		copy_from_brd(mem + off, brd, sector, len);
+		flush_dcache_page(page);
+	} else {
+		flush_dcache_page(page);
+		copy_to_brd(brd, mem + off, sector, len);
+	}
+	kunmap_atomic(mem);
+
+out:
+	return err;
+}
+
+static void brd_make_request(struct request_queue *q, struct bio *bio)
+{
+	struct block_device *bdev = bio->bi_bdev;
+	struct brd_device *brd = bdev->bd_disk->private_data;
+	int rw;
+	struct bio_vec bvec;
+	sector_t sector;
+	struct bvec_iter iter;
+	int err = -EIO;
+
+	sector = bio->bi_iter.bi_sector;
+	if (bio_end_sector(bio) > get_capacity(bdev->bd_disk))
+		goto out;
+
+	if (unlikely(bio->bi_rw & REQ_DISCARD)) {
+		err = 0;
+		discard_from_brd(brd, sector, bio->bi_iter.bi_size);
+		goto out;
+	}
+
+	rw = bio_rw(bio);
+	if (rw == READA)
+		rw = READ;
+
+	bio_for_each_segment(bvec, bio, iter) {
+		unsigned int len = bvec.bv_len;
+		err = brd_do_bvec(brd, bvec.bv_page, len,
+					bvec.bv_offset, rw, sector);
+		if (err)
+			break;
+		sector += len >> SECTOR_SHIFT;
+	}
+
+out:
+	bio_endio(bio, err);
+}
+
+static int brd_rw_page(struct block_device *bdev, sector_t sector,
+		       struct page *page, int rw)
+{
+	struct brd_device *brd = bdev->bd_disk->private_data;
+	int err = brd_do_bvec(brd, page, PAGE_CACHE_SIZE, 0, rw, sector);
+	page_endio(page, rw & WRITE, err);
+	return err;
+}
+
+#ifdef CONFIG_BLK_DEV_XIP
+static int brd_direct_access(struct block_device *bdev, sector_t sector,
+			void **kaddr, unsigned long *pfn)
+{
+	struct brd_device *brd = bdev->bd_disk->private_data;
+	struct page *page;
+
+	if (!brd)
+		return -ENODEV;
+	if (sector & (PAGE_SECTORS-1))
+		return -EINVAL;
+	if (sector + PAGE_SECTORS > get_capacity(bdev->bd_disk))
+		return -ERANGE;
+	page = brd_insert_page(brd, sector);
+	if (!page)
+		return -ENOSPC;
+	*kaddr = page_address(page);
+	*pfn = page_to_pfn(page);
+
+	return 0;
+}
+#endif
+
+static int brd_ioctl(struct block_device *bdev, fmode_t mode,
+			unsigned int cmd, unsigned long arg)
+{
+	int error;
+	struct brd_device *brd = bdev->bd_disk->private_data;
+
+	if (cmd != BLKFLSBUF)
+		return -ENOTTY;
+
+	/*
+	 * ram device BLKFLSBUF has special semantics, we want to actually
+	 * release and destroy the ramdisk data.
+	 */
+	mutex_lock(&brd_mutex);
+	mutex_lock(&bdev->bd_mutex);
+	error = -EBUSY;
+	if (bdev->bd_openers <= 1) {
+		/*
+		 * Kill the cache first, so it isn't written back to the
+		 * device.
+		 *
+		 * Another thread might instantiate more buffercache here,
+		 * but there is not much we can do to close that race.
+		 */
+		kill_bdev(bdev);
+		brd_free_pages(brd);
+		error = 0;
+	}
+	mutex_unlock(&bdev->bd_mutex);
+	mutex_unlock(&brd_mutex);
+
+	return error;
+}
+
+static const struct block_device_operations brd_fops = {
+	.owner =		THIS_MODULE,
+	.rw_page =		brd_rw_page,
+	.ioctl =		brd_ioctl,
+#ifdef CONFIG_BLK_DEV_XIP
+	.direct_access =	brd_direct_access,
+#endif
+};
+
+/*
+ * And now the modules code and kernel interface.
+ */
+static int rd_nr;
+int rd_size = CONFIG_BLK_DEV_RAM_SIZE;
+static int max_part;
+static int part_shift;
+static int part_show = 0;
+module_param(rd_nr, int, S_IRUGO);
+MODULE_PARM_DESC(rd_nr, "Maximum number of brd devices");
+module_param(rd_size, int, S_IRUGO);
+MODULE_PARM_DESC(rd_size, "Size of each RAM disk in kbytes.");
+module_param(max_part, int, S_IRUGO);
+MODULE_PARM_DESC(max_part, "Maximum number of partitions per RAM disk");
+module_param(part_show, int, S_IRUGO);
+MODULE_PARM_DESC(part_show, "Control RAM disk visibility in /proc/partitions");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS_BLOCKDEV_MAJOR(RAMDISK_MAJOR);
+MODULE_ALIAS("rd");
+
+#ifndef MODULE
+/* Legacy boot options - nonmodular */
+static int __init ramdisk_size(char *str)
+{
+	rd_size = simple_strtol(str, NULL, 0);
+	return 1;
+}
+__setup("ramdisk_size=", ramdisk_size);
+#endif
+
+/*
+ * The device scheme is derived from loop.c. Keep them in synch where possible
+ * (should share code eventually).
+ */
+static LIST_HEAD(brd_devices);
+static DEFINE_MUTEX(brd_devices_mutex);
+
+static struct brd_device *brd_alloc(int i)
+{
+	struct brd_device *brd;
+	struct gendisk *disk;
+
+	brd = kzalloc(sizeof(*brd), GFP_KERNEL);
+	if (!brd)
+		goto out;
+	brd->brd_number		= i;
+	spin_lock_init(&brd->brd_lock);
+	INIT_RADIX_TREE(&brd->brd_pages, GFP_ATOMIC);
+
+	brd->brd_queue = blk_alloc_queue(GFP_KERNEL);
+	if (!brd->brd_queue)
+		goto out_free_dev;
+	blk_queue_make_request(brd->brd_queue, brd_make_request);
+	blk_queue_max_hw_sectors(brd->brd_queue, 1024);
+	blk_queue_bounce_limit(brd->brd_queue, BLK_BOUNCE_ANY);
+
+	brd->brd_queue->limits.discard_granularity = PAGE_SIZE;
+	brd->brd_queue->limits.max_discard_sectors = UINT_MAX;
+	brd->brd_queue->limits.discard_zeroes_data = 1;
+	queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, brd->brd_queue);
+
+	disk = brd->brd_disk = alloc_disk(1 << part_shift);
+	if (!disk)
+		goto out_free_queue;
+	disk->major		= RAMDISK_MAJOR;
+	disk->first_minor	= i << part_shift;
+	disk->fops		= &brd_fops;
+	disk->private_data	= brd;
+	disk->queue		= brd->brd_queue;
+	if (!part_show)
+		disk->flags |= GENHD_FL_SUPPRESS_PARTITION_INFO;
+	sprintf(disk->disk_name, "ram%d", i);
+	set_capacity(disk, rd_size * 2);
+
+	return brd;
+
+out_free_queue:
+	blk_cleanup_queue(brd->brd_queue);
+out_free_dev:
+	kfree(brd);
+out:
+	return NULL;
+}
+
+static void brd_free(struct brd_device *brd)
+{
+	put_disk(brd->brd_disk);
+	blk_cleanup_queue(brd->brd_queue);
+	brd_free_pages(brd);
+	kfree(brd);
+}
+
+static struct brd_device *brd_init_one(int i)
+{
+	struct brd_device *brd;
+
+	list_for_each_entry(brd, &brd_devices, brd_list) {
+		if (brd->brd_number == i)
+			goto out;
+	}
+
+	brd = brd_alloc(i);
+	if (brd) {
+		add_disk(brd->brd_disk);
+		list_add_tail(&brd->brd_list, &brd_devices);
+	}
+out:
+	return brd;
+}
+
+static void brd_del_one(struct brd_device *brd)
+{
+	list_del(&brd->brd_list);
+	del_gendisk(brd->brd_disk);
+	brd_free(brd);
+}
+
+static struct kobject *brd_probe(dev_t dev, int *part, void *data)
+{
+	struct brd_device *brd;
+	struct kobject *kobj;
+
+	mutex_lock(&brd_devices_mutex);
+	brd = brd_init_one(MINOR(dev) >> part_shift);
+	kobj = brd ? get_disk(brd->brd_disk) : NULL;
+	mutex_unlock(&brd_devices_mutex);
+
+	*part = 0;
+	return kobj;
+}
+
+static int __init brd_init(void)
+{
+	int i, nr;
+	unsigned long range;
+	struct brd_device *brd, *next;
+
+	/*
+	 * brd module now has a feature to instantiate underlying device
+	 * structure on-demand, provided that there is an access dev node.
+	 * However, this will not work well with user space tool that doesn't
+	 * know about such "feature".  In order to not break any existing
+	 * tool, we do the following:
+	 *
+	 * (1) if rd_nr is specified, create that many upfront, and this
+	 *     also becomes a hard limit.
+	 * (2) if rd_nr is not specified, create CONFIG_BLK_DEV_RAM_COUNT
+	 *     (default 16) rd device on module load, user can further
+	 *     extend brd device by create dev node themselves and have
+	 *     kernel automatically instantiate actual device on-demand.
+	 */
+
+	part_shift = 0;
+	if (max_part > 0) {
+		part_shift = fls(max_part);
+
+		/*
+		 * Adjust max_part according to part_shift as it is exported
+		 * to user space so that user can decide correct minor number
+		 * if [s]he want to create more devices.
+		 *
+		 * Note that -1 is required because partition 0 is reserved
+		 * for the whole disk.
+		 */
+		max_part = (1UL << part_shift) - 1;
+	}
+
+	if ((1UL << part_shift) > DISK_MAX_PARTS)
+		return -EINVAL;
+
+	if (rd_nr > 1UL << (MINORBITS - part_shift))
+		return -EINVAL;
+
+	if (rd_nr) {
+		nr = rd_nr;
+		range = rd_nr << part_shift;
+	} else {
+		nr = CONFIG_BLK_DEV_RAM_COUNT;
+		range = 1UL << MINORBITS;
+	}
+
+	if (register_blkdev(RAMDISK_MAJOR, "ramdisk"))
+		return -EIO;
+
+	for (i = 0; i < nr; i++) {
+		brd = brd_alloc(i);
+		if (!brd)
+			goto out_free;
+		list_add_tail(&brd->brd_list, &brd_devices);
+	}
+
+	/* point of no return */
+
+	list_for_each_entry(brd, &brd_devices, brd_list)
+		add_disk(brd->brd_disk);
+
+	blk_register_region(MKDEV(RAMDISK_MAJOR, 0), range,
+				  THIS_MODULE, brd_probe, NULL, NULL);
+
+	printk(KERN_INFO "brd: module loaded\n");
+	return 0;
+
+out_free:
+	list_for_each_entry_safe(brd, next, &brd_devices, brd_list) {
+		list_del(&brd->brd_list);
+		brd_free(brd);
+	}
+	unregister_blkdev(RAMDISK_MAJOR, "ramdisk");
+
+	return -ENOMEM;
+}
+
+static void __exit brd_exit(void)
+{
+	unsigned long range;
+	struct brd_device *brd, *next;
+
+	range = rd_nr ? rd_nr << part_shift : 1UL << MINORBITS;
+
+	list_for_each_entry_safe(brd, next, &brd_devices, brd_list)
+		brd_del_one(brd);
+
+	blk_unregister_region(MKDEV(RAMDISK_MAJOR, 0), range);
+	unregister_blkdev(RAMDISK_MAJOR, "ramdisk");
+}
+
+module_init(brd_init);
+module_exit(brd_exit);
+

drivers/block/cciss.h

@@ -0,0 +1,435 @@
+#ifndef CCISS_H
+#define CCISS_H
+
+#include <linux/genhd.h>
+#include <linux/mutex.h>
+
+#include "cciss_cmd.h"
+
+
+#define NWD_SHIFT	4
+#define MAX_PART	(1 << NWD_SHIFT)
+
+#define IO_OK		0
+#define IO_ERROR	1
+#define IO_NEEDS_RETRY  3
+
+#define VENDOR_LEN	8
+#define MODEL_LEN	16
+#define REV_LEN		4
+
+struct ctlr_info;
+typedef struct ctlr_info ctlr_info_t;
+
+struct access_method {
+	void (*submit_command)(ctlr_info_t *h, CommandList_struct *c);
+	void (*set_intr_mask)(ctlr_info_t *h, unsigned long val);
+	unsigned long (*fifo_full)(ctlr_info_t *h);
+	bool (*intr_pending)(ctlr_info_t *h);
+	unsigned long (*command_completed)(ctlr_info_t *h);
+};
+typedef struct _drive_info_struct
+{
+	unsigned char LunID[8];
+	int 	usage_count;
+	struct request_queue *queue;
+	sector_t nr_blocks;
+	int	block_size;
+	int 	heads;
+	int	sectors;
+	int 	cylinders;
+	int	raid_level; /* set to -1 to indicate that
+			     * the drive is not in use/configured
+			     */
+	int	busy_configuring; /* This is set when a drive is being removed
+				   * to prevent it from being opened or it's
+				   * queue from being started.
+				   */
+	struct	device dev;
+	__u8 serial_no[16]; /* from inquiry page 0x83,
+			     * not necc. null terminated.
+			     */
+	char vendor[VENDOR_LEN + 1]; /* SCSI vendor string */
+	char model[MODEL_LEN + 1];   /* SCSI model string */
+	char rev[REV_LEN + 1];       /* SCSI revision string */
+	char device_initialized;     /* indicates whether dev is initialized */
+} drive_info_struct;
+
+struct ctlr_info
+{
+	int	ctlr;
+	char	devname[8];
+	char    *product_name;
+	char	firm_ver[4]; /* Firmware version */
+	struct pci_dev *pdev;
+	__u32	board_id;
+	void __iomem *vaddr;
+	unsigned long paddr;
+	int 	nr_cmds; /* Number of commands allowed on this controller */
+	CfgTable_struct __iomem *cfgtable;
+	int	interrupts_enabled;
+	int	major;
+	int 	max_commands;
+	int	commands_outstanding;
+	int 	max_outstanding; /* Debug */ 
+	int	num_luns;
+	int 	highest_lun;
+	int	usage_count;  /* number of opens all all minor devices */
+	/* Need space for temp sg list
+	 * number of scatter/gathers supported
+	 * number of scatter/gathers in chained block
+	 */
+	struct	scatterlist **scatter_list;
+	int	maxsgentries;
+	int	chainsize;
+	int	max_cmd_sgentries;
+	SGDescriptor_struct **cmd_sg_list;
+
+#	define PERF_MODE_INT	0
+#	define DOORBELL_INT	1
+#	define SIMPLE_MODE_INT	2
+#	define MEMQ_MODE_INT	3
+	unsigned int intr[4];
+	unsigned int msix_vector;
+	unsigned int msi_vector;
+	int	intr_mode;
+	int 	cciss_max_sectors;
+	BYTE	cciss_read;
+	BYTE	cciss_write;
+	BYTE	cciss_read_capacity;
+
+	/* information about each logical volume */
+	drive_info_struct *drv[CISS_MAX_LUN];
+
+	struct access_method access;
+
+	/* queue and queue Info */ 
+	struct list_head reqQ;
+	struct list_head cmpQ;
+	unsigned int Qdepth;
+	unsigned int maxQsinceinit;
+	unsigned int maxSG;
+	spinlock_t lock;
+
+	/* pointers to command and error info pool */
+	CommandList_struct 	*cmd_pool;
+	dma_addr_t		cmd_pool_dhandle; 
+	ErrorInfo_struct 	*errinfo_pool;
+	dma_addr_t		errinfo_pool_dhandle; 
+        unsigned long  		*cmd_pool_bits;
+	int			nr_allocs;
+	int			nr_frees; 
+	int			busy_configuring;
+	int			busy_initializing;
+	int			busy_scanning;
+	struct mutex		busy_shutting_down;
+
+	/* This element holds the zero based queue number of the last
+	 * queue to be started.  It is used for fairness.
+	*/
+	int			next_to_run;
+
+	/* Disk structures we need to pass back */
+	struct gendisk   *gendisk[CISS_MAX_LUN];
+#ifdef CONFIG_CISS_SCSI_TAPE
+	struct cciss_scsi_adapter_data_t *scsi_ctlr;
+#endif
+	unsigned char alive;
+	struct list_head scan_list;
+	struct completion scan_wait;
+	struct device dev;
+	/*
+	 * Performant mode tables.
+	 */
+	u32 trans_support;
+	u32 trans_offset;
+	struct TransTable_struct *transtable;
+	unsigned long transMethod;
+
+	/*
+	 * Performant mode completion buffer
+	 */
+	u64 *reply_pool;
+	dma_addr_t reply_pool_dhandle;
+	u64 *reply_pool_head;
+	size_t reply_pool_size;
+	unsigned char reply_pool_wraparound;
+	u32 *blockFetchTable;
+};
+
+/*  Defining the diffent access_methods
+ *
+ * Memory mapped FIFO interface (SMART 53xx cards)
+ */
+#define SA5_DOORBELL	0x20
+#define SA5_REQUEST_PORT_OFFSET	0x40
+#define SA5_REPLY_INTR_MASK_OFFSET	0x34
+#define SA5_REPLY_PORT_OFFSET		0x44
+#define SA5_INTR_STATUS		0x30
+#define SA5_SCRATCHPAD_OFFSET	0xB0
+
+#define SA5_CTCFG_OFFSET	0xB4
+#define SA5_CTMEM_OFFSET	0xB8
+
+#define SA5_INTR_OFF		0x08
+#define SA5B_INTR_OFF		0x04
+#define SA5_INTR_PENDING	0x08
+#define SA5B_INTR_PENDING	0x04
+#define FIFO_EMPTY		0xffffffff	
+#define CCISS_FIRMWARE_READY	0xffff0000 /* value in scratchpad register */
+/* Perf. mode flags */
+#define SA5_PERF_INTR_PENDING	0x04
+#define SA5_PERF_INTR_OFF	0x05
+#define SA5_OUTDB_STATUS_PERF_BIT	0x01
+#define SA5_OUTDB_CLEAR_PERF_BIT	0x01
+#define SA5_OUTDB_CLEAR         0xA0
+#define SA5_OUTDB_CLEAR_PERF_BIT        0x01
+#define SA5_OUTDB_STATUS        0x9C
+
+
+#define  CISS_ERROR_BIT		0x02
+
+#define CCISS_INTR_ON 	1 
+#define CCISS_INTR_OFF	0
+
+
+/* CCISS_BOARD_READY_WAIT_SECS is how long to wait for a board
+ * to become ready, in seconds, before giving up on it.
+ * CCISS_BOARD_READY_POLL_INTERVAL_MSECS * is how long to wait
+ * between polling the board to see if it is ready, in
+ * milliseconds.  CCISS_BOARD_READY_ITERATIONS is derived
+ * the above.
+ */
+#define CCISS_BOARD_READY_WAIT_SECS (120)
+#define CCISS_BOARD_NOT_READY_WAIT_SECS (100)
+#define CCISS_BOARD_READY_POLL_INTERVAL_MSECS (100)
+#define CCISS_BOARD_READY_ITERATIONS \
+	((CCISS_BOARD_READY_WAIT_SECS * 1000) / \
+		CCISS_BOARD_READY_POLL_INTERVAL_MSECS)
+#define CCISS_BOARD_NOT_READY_ITERATIONS \
+	((CCISS_BOARD_NOT_READY_WAIT_SECS * 1000) / \
+		CCISS_BOARD_READY_POLL_INTERVAL_MSECS)
+#define CCISS_POST_RESET_PAUSE_MSECS (3000)
+#define CCISS_POST_RESET_NOOP_INTERVAL_MSECS (4000)
+#define CCISS_POST_RESET_NOOP_RETRIES (12)
+#define CCISS_POST_RESET_NOOP_TIMEOUT_MSECS (10000)
+
+/* 
+	Send the command to the hardware 
+*/
+static void SA5_submit_command( ctlr_info_t *h, CommandList_struct *c) 
+{
+#ifdef CCISS_DEBUG
+	printk(KERN_WARNING "cciss%d: Sending %08x - down to controller\n",
+			h->ctlr, c->busaddr);
+#endif /* CCISS_DEBUG */
+         writel(c->busaddr, h->vaddr + SA5_REQUEST_PORT_OFFSET);
+	readl(h->vaddr + SA5_SCRATCHPAD_OFFSET);
+	 h->commands_outstanding++;
+	 if ( h->commands_outstanding > h->max_outstanding)
+		h->max_outstanding = h->commands_outstanding;
+}
+
+/*  
+ *  This card is the opposite of the other cards.  
+ *   0 turns interrupts on... 
+ *   0x08 turns them off... 
+ */
+static void SA5_intr_mask(ctlr_info_t *h, unsigned long val)
+{
+	if (val) 
+	{ /* Turn interrupts on */
+		h->interrupts_enabled = 1;
+		writel(0, h->vaddr + SA5_REPLY_INTR_MASK_OFFSET);
+		(void) readl(h->vaddr + SA5_REPLY_INTR_MASK_OFFSET);
+	} else /* Turn them off */
+	{
+		h->interrupts_enabled = 0;
+        	writel( SA5_INTR_OFF, 
+			h->vaddr + SA5_REPLY_INTR_MASK_OFFSET);
+		(void) readl(h->vaddr + SA5_REPLY_INTR_MASK_OFFSET);
+	}
+}
+/*
+ *  This card is the opposite of the other cards.
+ *   0 turns interrupts on...
+ *   0x04 turns them off...
+ */
+static void SA5B_intr_mask(ctlr_info_t *h, unsigned long val)
+{
+        if (val)
+        { /* Turn interrupts on */
+		h->interrupts_enabled = 1;
+                writel(0, h->vaddr + SA5_REPLY_INTR_MASK_OFFSET);
+		(void) readl(h->vaddr + SA5_REPLY_INTR_MASK_OFFSET);
+        } else /* Turn them off */
+        {
+		h->interrupts_enabled = 0;
+                writel( SA5B_INTR_OFF,
+                        h->vaddr + SA5_REPLY_INTR_MASK_OFFSET);
+		(void) readl(h->vaddr + SA5_REPLY_INTR_MASK_OFFSET);
+        }
+}
+
+/* Performant mode intr_mask */
+static void SA5_performant_intr_mask(ctlr_info_t *h, unsigned long val)
+{
+	if (val) { /* turn on interrupts */
+		h->interrupts_enabled = 1;
+		writel(0, h->vaddr + SA5_REPLY_INTR_MASK_OFFSET);
+		(void) readl(h->vaddr + SA5_REPLY_INTR_MASK_OFFSET);
+	} else {
+		h->interrupts_enabled = 0;
+		writel(SA5_PERF_INTR_OFF,
+				h->vaddr + SA5_REPLY_INTR_MASK_OFFSET);
+		(void) readl(h->vaddr + SA5_REPLY_INTR_MASK_OFFSET);
+	}
+}
+
+/*
+ *  Returns true if fifo is full.  
+ * 
+ */ 
+static unsigned long SA5_fifo_full(ctlr_info_t *h)
+{
+	if( h->commands_outstanding >= h->max_commands)
+		return(1);
+	else 
+		return(0);
+
+}
+/* 
+ *   returns value read from hardware. 
+ *     returns FIFO_EMPTY if there is nothing to read 
+ */ 
+static unsigned long SA5_completed(ctlr_info_t *h)
+{
+	unsigned long register_value 
+		= readl(h->vaddr + SA5_REPLY_PORT_OFFSET);
+	if(register_value != FIFO_EMPTY)
+	{
+		h->commands_outstanding--;
+#ifdef CCISS_DEBUG
+		printk("cciss:  Read %lx back from board\n", register_value);
+#endif /* CCISS_DEBUG */ 
+	} 
+#ifdef CCISS_DEBUG
+	else
+	{
+		printk("cciss:  FIFO Empty read\n");
+	}
+#endif 
+	return ( register_value); 
+
+}
+
+/* Performant mode command completed */
+static unsigned long SA5_performant_completed(ctlr_info_t *h)
+{
+	unsigned long register_value = FIFO_EMPTY;
+
+	/* flush the controller write of the reply queue by reading
+	 * outbound doorbell status register.
+	 */
+	register_value = readl(h->vaddr + SA5_OUTDB_STATUS);
+	/* msi auto clears the interrupt pending bit. */
+	if (!(h->msi_vector || h->msix_vector)) {
+		writel(SA5_OUTDB_CLEAR_PERF_BIT, h->vaddr + SA5_OUTDB_CLEAR);
+		/* Do a read in order to flush the write to the controller
+		 * (as per spec.)
+		 */
+		register_value = readl(h->vaddr + SA5_OUTDB_STATUS);
+	}
+
+	if ((*(h->reply_pool_head) & 1) == (h->reply_pool_wraparound)) {
+		register_value = *(h->reply_pool_head);
+		(h->reply_pool_head)++;
+		h->commands_outstanding--;
+	} else {
+		register_value = FIFO_EMPTY;
+	}
+	/* Check for wraparound */
+	if (h->reply_pool_head == (h->reply_pool + h->max_commands)) {
+		h->reply_pool_head = h->reply_pool;
+		h->reply_pool_wraparound ^= 1;
+	}
+
+	return register_value;
+}
+/*
+ *	Returns true if an interrupt is pending.. 
+ */
+static bool SA5_intr_pending(ctlr_info_t *h)
+{
+	unsigned long register_value  = 
+		readl(h->vaddr + SA5_INTR_STATUS);
+#ifdef CCISS_DEBUG
+	printk("cciss: intr_pending %lx\n", register_value);
+#endif  /* CCISS_DEBUG */
+	if( register_value &  SA5_INTR_PENDING) 
+		return  1;	
+	return 0 ;
+}
+
+/*
+ *      Returns true if an interrupt is pending..
+ */
+static bool SA5B_intr_pending(ctlr_info_t *h)
+{
+        unsigned long register_value  =
+                readl(h->vaddr + SA5_INTR_STATUS);
+#ifdef CCISS_DEBUG
+        printk("cciss: intr_pending %lx\n", register_value);
+#endif  /* CCISS_DEBUG */
+        if( register_value &  SA5B_INTR_PENDING)
+                return  1;
+        return 0 ;
+}
+
+static bool SA5_performant_intr_pending(ctlr_info_t *h)
+{
+	unsigned long register_value = readl(h->vaddr + SA5_INTR_STATUS);
+
+	if (!register_value)
+		return false;
+
+	if (h->msi_vector || h->msix_vector)
+		return true;
+
+	/* Read outbound doorbell to flush */
+	register_value = readl(h->vaddr + SA5_OUTDB_STATUS);
+	return register_value & SA5_OUTDB_STATUS_PERF_BIT;
+}
+
+static struct access_method SA5_access = {
+	SA5_submit_command,
+	SA5_intr_mask,
+	SA5_fifo_full,
+	SA5_intr_pending,
+	SA5_completed,
+};
+
+static struct access_method SA5B_access = {
+        SA5_submit_command,
+        SA5B_intr_mask,
+        SA5_fifo_full,
+        SA5B_intr_pending,
+        SA5_completed,
+};
+
+static struct access_method SA5_performant_access = {
+	SA5_submit_command,
+	SA5_performant_intr_mask,
+	SA5_fifo_full,
+	SA5_performant_intr_pending,
+	SA5_performant_completed,
+};
+
+struct board_type {
+	__u32	board_id;
+	char	*product_name;
+	struct access_method *access;
+	int nr_cmds; /* Max cmds this kind of ctlr can handle. */
+};
+
+#endif /* CCISS_H */

drivers/block/cciss_cmd.h

@@ -0,0 +1,269 @@
+#ifndef CCISS_CMD_H
+#define CCISS_CMD_H
+
+#include <linux/cciss_defs.h>
+
+/* DEFINES */
+#define CISS_VERSION "1.00"
+
+/* general boundary definitions */
+#define MAXSGENTRIES            32
+#define CCISS_SG_CHAIN          0x80000000
+#define MAXREPLYQS              256
+
+/* Unit Attentions ASC's as defined for the MSA2012sa */
+#define POWER_OR_RESET			0x29
+#define STATE_CHANGED			0x2a
+#define UNIT_ATTENTION_CLEARED		0x2f
+#define LUN_FAILED			0x3e
+#define REPORT_LUNS_CHANGED		0x3f
+
+/* Unit Attentions ASCQ's as defined for the MSA2012sa */
+
+	/* These ASCQ's defined for ASC = POWER_OR_RESET */
+#define POWER_ON_RESET			0x00
+#define POWER_ON_REBOOT			0x01
+#define SCSI_BUS_RESET			0x02
+#define MSA_TARGET_RESET		0x03
+#define CONTROLLER_FAILOVER		0x04
+#define TRANSCEIVER_SE			0x05
+#define TRANSCEIVER_LVD			0x06
+
+	/* These ASCQ's defined for ASC = STATE_CHANGED */
+#define RESERVATION_PREEMPTED		0x03
+#define ASYM_ACCESS_CHANGED		0x06
+#define LUN_CAPACITY_CHANGED		0x09
+
+/* config space register offsets */
+#define CFG_VENDORID            0x00
+#define CFG_DEVICEID            0x02
+#define CFG_I2OBAR              0x10
+#define CFG_MEM1BAR             0x14
+
+/* i2o space register offsets */
+#define I2O_IBDB_SET            0x20
+#define I2O_IBDB_CLEAR          0x70
+#define I2O_INT_STATUS          0x30
+#define I2O_INT_MASK            0x34
+#define I2O_IBPOST_Q            0x40
+#define I2O_OBPOST_Q            0x44
+#define I2O_DMA1_CFG		0x214
+
+/* Configuration Table */
+#define CFGTBL_ChangeReq        0x00000001l
+#define CFGTBL_AccCmds          0x00000001l
+#define DOORBELL_CTLR_RESET     0x00000004l
+#define DOORBELL_CTLR_RESET2    0x00000020l
+
+#define CFGTBL_Trans_Simple     0x00000002l
+#define CFGTBL_Trans_Performant 0x00000004l
+#define CFGTBL_Trans_use_short_tags 0x20000000l
+
+#define CFGTBL_BusType_Ultra2   0x00000001l
+#define CFGTBL_BusType_Ultra3   0x00000002l
+#define CFGTBL_BusType_Fibre1G  0x00000100l
+#define CFGTBL_BusType_Fibre2G  0x00000200l
+typedef struct _vals32
+{
+        __u32   lower;
+        __u32   upper;
+} vals32;
+
+typedef union _u64bit
+{
+   vals32	val32;
+   __u64	val;
+} u64bit;
+
+/* Type defs used in the following structs */
+#define QWORD vals32 
+
+/* STRUCTURES */
+#define CISS_MAX_PHYS_LUN	1024
+/* SCSI-3 Cmmands */
+
+#pragma pack(1)	
+
+#define CISS_INQUIRY 0x12
+/* Date returned */
+typedef struct _InquiryData_struct
+{
+  BYTE data_byte[36];
+} InquiryData_struct;
+
+#define CISS_REPORT_LOG 0xc2    /* Report Logical LUNs */
+#define CISS_REPORT_PHYS 0xc3   /* Report Physical LUNs */
+/* Data returned */
+typedef struct _ReportLUNdata_struct
+{
+  BYTE LUNListLength[4];
+  DWORD reserved;
+  BYTE LUN[CISS_MAX_LUN][8];
+} ReportLunData_struct;
+
+#define CCISS_READ_CAPACITY 0x25 /* Read Capacity */ 
+typedef struct _ReadCapdata_struct
+{
+  BYTE total_size[4];	/* Total size in blocks */
+  BYTE block_size[4];	/* Size of blocks in bytes */
+} ReadCapdata_struct;
+
+#define CCISS_READ_CAPACITY_16 0x9e /* Read Capacity 16 */
+
+/* service action to differentiate a 16 byte read capacity from
+   other commands that use the 0x9e SCSI op code */
+
+#define CCISS_READ_CAPACITY_16_SERVICE_ACT 0x10
+
+typedef struct _ReadCapdata_struct_16
+{
+	BYTE total_size[8];   /* Total size in blocks */
+	BYTE block_size[4];   /* Size of blocks in bytes */
+	BYTE prot_en:1;       /* protection enable bit */
+	BYTE rto_en:1;        /* reference tag own enable bit */
+	BYTE reserved:6;      /* reserved bits */
+	BYTE reserved2[18];   /* reserved bytes per spec */
+} ReadCapdata_struct_16;
+
+/* Define the supported read/write commands for cciss based controllers */
+
+#define CCISS_READ_10   0x28    /* Read(10)  */
+#define CCISS_WRITE_10  0x2a    /* Write(10) */
+#define CCISS_READ_16   0x88    /* Read(16)  */
+#define CCISS_WRITE_16  0x8a    /* Write(16) */
+
+/* Define the CDB lengths supported by cciss based controllers */
+
+#define CDB_LEN10	10
+#define CDB_LEN16	16
+
+/* BMIC commands */
+#define BMIC_READ 0x26
+#define BMIC_WRITE 0x27
+#define BMIC_CACHE_FLUSH 0xc2
+#define CCISS_CACHE_FLUSH 0x01	/* C2 was already being used by CCISS */
+
+#define CCISS_ABORT_MSG 0x00
+#define CCISS_RESET_MSG 0x01
+#define CCISS_RESET_TYPE_CONTROLLER 0x00
+#define CCISS_RESET_TYPE_BUS 0x01
+#define CCISS_RESET_TYPE_TARGET 0x03
+#define CCISS_RESET_TYPE_LUN 0x04
+#define CCISS_NOOP_MSG 0x03
+
+/* Command List Structure */
+#define CTLR_LUNID "\0\0\0\0\0\0\0\0"
+
+typedef struct _CommandListHeader_struct {
+  BYTE              ReplyQueue;
+  BYTE              SGList;
+  HWORD             SGTotal;
+  QWORD             Tag;
+  LUNAddr_struct    LUN;
+} CommandListHeader_struct;
+typedef struct _ErrDescriptor_struct {
+  QWORD  Addr;
+  DWORD  Len;
+} ErrDescriptor_struct;
+typedef struct _SGDescriptor_struct {
+  QWORD  Addr;
+  DWORD  Len;
+  DWORD  Ext;
+} SGDescriptor_struct;
+
+/* Command types */
+#define CMD_RWREQ       0x00
+#define CMD_IOCTL_PEND  0x01
+#define CMD_SCSI	0x03
+#define CMD_MSG_DONE	0x04
+#define CMD_MSG_TIMEOUT 0x05
+#define CMD_MSG_STALE	0xff
+
+/* This structure needs to be divisible by COMMANDLIST_ALIGNMENT
+ * because low bits of the address are used to to indicate that
+ * whether the tag contains an index or an address.  PAD_32 and
+ * PAD_64 can be adjusted independently as needed for 32-bit
+ * and 64-bits systems.
+ */
+#define COMMANDLIST_ALIGNMENT (32)
+#define IS_64_BIT ((sizeof(long) - 4)/4)
+#define IS_32_BIT (!IS_64_BIT)
+#define PAD_32 (0)
+#define PAD_64 (4)
+#define PADSIZE (IS_32_BIT * PAD_32 + IS_64_BIT * PAD_64)
+#define DIRECT_LOOKUP_BIT 0x10
+#define DIRECT_LOOKUP_SHIFT 5
+
+typedef struct _CommandList_struct {
+  CommandListHeader_struct Header;
+  RequestBlock_struct      Request;
+  ErrDescriptor_struct     ErrDesc;
+  SGDescriptor_struct      SG[MAXSGENTRIES];
+	/* information associated with the command */ 
+  __u32			   busaddr; /* physical address of this record */
+  ErrorInfo_struct * 	   err_info; /* pointer to the allocated mem */ 
+  int			   ctlr;
+  int			   cmd_type; 
+  long			   cmdindex;
+  struct list_head list;
+  struct request *	   rq;
+  struct completion *waiting;
+  int	 retry_count;
+  void * scsi_cmd;
+  char pad[PADSIZE];
+} CommandList_struct;
+
+/* Configuration Table Structure */
+typedef struct _HostWrite_struct {
+  DWORD TransportRequest;
+  DWORD Reserved;
+  DWORD CoalIntDelay;
+  DWORD CoalIntCount;
+} HostWrite_struct;
+
+typedef struct _CfgTable_struct {
+  BYTE             Signature[4];
+  DWORD            SpecValence;
+#define SIMPLE_MODE	0x02
+#define PERFORMANT_MODE	0x04
+#define MEMQ_MODE	0x08
+  DWORD            TransportSupport;
+  DWORD            TransportActive;
+  HostWrite_struct HostWrite;
+  DWORD            CmdsOutMax;
+  DWORD            BusTypes;
+  DWORD            TransMethodOffset;
+  BYTE             ServerName[16];
+  DWORD            HeartBeat;
+  DWORD            SCSI_Prefetch;
+  DWORD            MaxSGElements;
+  DWORD            MaxLogicalUnits;
+  DWORD            MaxPhysicalDrives;
+  DWORD            MaxPhysicalDrivesPerLogicalUnit;
+  DWORD            MaxPerformantModeCommands;
+  u8		   reserved[0x78 - 0x58];
+  u32		   misc_fw_support; /* offset 0x78 */
+#define MISC_FW_DOORBELL_RESET (0x02)
+#define MISC_FW_DOORBELL_RESET2 (0x10)
+	u8	   driver_version[32];
+} CfgTable_struct;
+
+struct TransTable_struct {
+  u32 BlockFetch0;
+  u32 BlockFetch1;
+  u32 BlockFetch2;
+  u32 BlockFetch3;
+  u32 BlockFetch4;
+  u32 BlockFetch5;
+  u32 BlockFetch6;
+  u32 BlockFetch7;
+  u32 RepQSize;
+  u32 RepQCount;
+  u32 RepQCtrAddrLow32;
+  u32 RepQCtrAddrHigh32;
+  u32 RepQAddr0Low32;
+  u32 RepQAddr0High32;
+};
+
+#pragma pack()	 
+#endif /* CCISS_CMD_H */

drivers/block/cciss_scsi.h

@@ -0,0 +1,79 @@
+/*
+ *    Disk Array driver for HP Smart Array controllers, SCSI Tape module.
+ *    (C) Copyright 2001, 2007 Hewlett-Packard Development Company, L.P.
+ *
+ *    This program is free software; you can redistribute it and/or modify
+ *    it under the terms of the GNU General Public License as published by
+ *    the Free Software Foundation; version 2 of the License.
+ *
+ *    This program is distributed in the hope that it will be useful,
+ *    but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *    MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ *    General Public License for more details.
+ *
+ *    You should have received a copy of the GNU General Public License
+ *    along with this program; if not, write to the Free Software
+ *    Foundation, Inc., 59 Temple Place, Suite 300, Boston, MA
+ *    02111-1307, USA.
+ *
+ *    Questions/Comments/Bugfixes to iss_storagedev@hp.com
+ *
+ */
+#ifdef CONFIG_CISS_SCSI_TAPE
+#ifndef _CCISS_SCSI_H_
+#define _CCISS_SCSI_H_
+
+#include <scsi/scsicam.h> /* possibly irrelevant, since we don't show disks */
+
+		/* the scsi id of the adapter... */
+#define SELF_SCSI_ID 15
+		/* 15 is somewhat arbitrary, since the scsi-2 bus
+		   that's presented by the driver to the OS is
+		   fabricated.  The "real" scsi-3 bus the
+		   hardware presents is fabricated too.
+		   The actual, honest-to-goodness physical
+		   bus that the devices are attached to is not
+		   addressible natively, and may in fact turn
+		   out to be not scsi at all. */
+
+
+/* 
+
+If the upper scsi layer tries to track how many commands we have 
+outstanding, it will be operating under the misapprehension that it is
+the only one sending us requests.  We also have the block interface,
+which is where most requests must surely come from, so the upper layer's
+notion of how many requests we have outstanding will be wrong most or
+all of the time. 
+
+Note, the normal SCSI mid-layer error handling doesn't work well
+for this driver because 1) it takes the io_request_lock before
+calling error handlers and uses a local variable to store flags,
+so the io_request_lock cannot be released and interrupts enabled
+inside the error handlers, and, the error handlers cannot poll
+for command completion because they might get commands from the
+block half of the driver completing, and not know what to do
+with them.  That's what we get for making a hybrid scsi/block
+driver, I suppose.
+
+*/
+
+struct cciss_scsi_dev_t {
+	int devtype;
+	int bus, target, lun;		/* as presented to the OS */
+	unsigned char scsi3addr[8];	/* as presented to the HW */
+	unsigned char device_id[16];	/* from inquiry pg. 0x83 */
+	unsigned char vendor[8];	/* bytes 8-15 of inquiry data */
+	unsigned char model[16];	/* bytes 16-31 of inquiry data */
+	unsigned char revision[4];	/* bytes 32-35 of inquiry data */
+};
+
+struct cciss_scsi_hba_t {
+	char *name;
+	int ndevices;
+#define CCISS_MAX_SCSI_DEVS_PER_HBA 16
+	struct cciss_scsi_dev_t dev[CCISS_MAX_SCSI_DEVS_PER_HBA];
+};
+
+#endif /* _CCISS_SCSI_H_ */
+#endif /* CONFIG_CISS_SCSI_TAPE */

drivers/block/cpqarray.h

@@ -0,0 +1,126 @@
+/*
+ *    Disk Array driver for Compaq SMART2 Controllers
+ *    Copyright 1998 Compaq Computer Corporation
+ *
+ *    This program is free software; you can redistribute it and/or modify
+ *    it under the terms of the GNU General Public License as published by
+ *    the Free Software Foundation; either version 2 of the License, or
+ *    (at your option) any later version.
+ *
+ *    This program is distributed in the hope that it will be useful,
+ *    but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *    MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
+ *    NON INFRINGEMENT.  See the GNU General Public License for more details.
+ *
+ *    You should have received a copy of the GNU General Public License
+ *    along with this program; if not, write to the Free Software
+ *    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ *
+ *    Questions/Comments/Bugfixes to iss_storagedev@hp.com
+ *
+ *    If you want to make changes, improve or add functionality to this
+ *    driver, you'll probably need the Compaq Array Controller Interface
+ *    Specificiation (Document number ECG086/1198)
+ */
+#ifndef CPQARRAY_H
+#define CPQARRAY_H
+
+#ifdef __KERNEL__
+#include <linux/blkdev.h>
+#include <linux/slab.h>
+#include <linux/proc_fs.h>
+#include <linux/timer.h>
+#endif
+
+#include "ida_cmd.h"
+
+#define IO_OK		0
+#define IO_ERROR	1
+#define NWD		16
+#define NWD_SHIFT	4
+
+#define IDA_TIMER	(5*HZ)
+#define IDA_TIMEOUT	(10*HZ)
+
+#define MISC_NONFATAL_WARN	0x01
+
+typedef struct {
+	unsigned blk_size;
+	unsigned nr_blks;
+	unsigned cylinders;
+	unsigned heads;
+	unsigned sectors;
+	int usage_count;
+} drv_info_t;
+
+#ifdef __KERNEL__
+
+struct ctlr_info;
+typedef struct ctlr_info ctlr_info_t;
+
+struct access_method {
+	void (*submit_command)(ctlr_info_t *h, cmdlist_t *c);
+	void (*set_intr_mask)(ctlr_info_t *h, unsigned long val);
+	unsigned long (*fifo_full)(ctlr_info_t *h);
+	unsigned long (*intr_pending)(ctlr_info_t *h);
+	unsigned long (*command_completed)(ctlr_info_t *h);
+};
+
+struct board_type {
+	__u32	board_id;
+	char	*product_name;
+	struct access_method *access;
+};
+
+struct ctlr_info {
+	int	ctlr;
+	char	devname[8];
+	__u32	log_drv_map;
+	__u32	drv_assign_map;
+	__u32	drv_spare_map;
+	__u32	mp_failed_drv_map;
+
+	char	firm_rev[4];
+	int	ctlr_sig;
+
+	int	log_drives;
+	int	phys_drives;
+
+	struct pci_dev *pci_dev;    /* NULL if EISA */
+	__u32	board_id;
+	char	*product_name;	
+
+	void __iomem *vaddr;
+	unsigned long paddr;
+	unsigned long io_mem_addr;
+	unsigned long io_mem_length;
+	int	intr;
+	int	usage_count;
+	drv_info_t	drv[NWD];
+	struct proc_dir_entry *proc;
+
+	struct access_method access;
+
+	cmdlist_t *reqQ;
+	cmdlist_t *cmpQ;
+	cmdlist_t *cmd_pool;
+	dma_addr_t cmd_pool_dhandle;
+	unsigned long *cmd_pool_bits;
+	struct request_queue *queue;
+	spinlock_t lock;
+
+	unsigned int Qdepth;
+	unsigned int maxQsinceinit;
+
+	unsigned int nr_requests;
+	unsigned int nr_allocs;
+	unsigned int nr_frees;
+	struct timer_list timer;
+	unsigned int misc_tflags;
+};
+
+#define IDA_LOCK(i)	(&hba[i]->lock)
+
+#endif
+
+#endif /* CPQARRAY_H */

drivers/block/cryptoloop.c

@@ -0,0 +1,216 @@
+/*
+   Linux loop encryption enabling module
+
+   Copyright (C)  2002 Herbert Valerio Riedel <hvr@gnu.org>
+   Copyright (C)  2003 Fruhwirth Clemens <clemens@endorphin.org>
+
+   This module is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation; either version 2 of the License, or
+   (at your option) any later version.
+
+   This module is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this module; if not, write to the Free Software
+   Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/module.h>
+
+#include <linux/init.h>
+#include <linux/string.h>
+#include <linux/crypto.h>
+#include <linux/blkdev.h>
+#include <linux/scatterlist.h>
+#include <asm/uaccess.h>
+#include "loop.h"
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("loop blockdevice transferfunction adaptor / CryptoAPI");
+MODULE_AUTHOR("Herbert Valerio Riedel <hvr@gnu.org>");
+
+#define LOOP_IV_SECTOR_BITS 9
+#define LOOP_IV_SECTOR_SIZE (1 << LOOP_IV_SECTOR_BITS)
+
+static int
+cryptoloop_init(struct loop_device *lo, const struct loop_info64 *info)
+{
+	int err = -EINVAL;
+	int cipher_len;
+	int mode_len;
+	char cms[LO_NAME_SIZE];			/* cipher-mode string */
+	char *cipher;
+	char *mode;
+	char *cmsp = cms;			/* c-m string pointer */
+	struct crypto_blkcipher *tfm;
+
+	/* encryption breaks for non sector aligned offsets */
+
+	if (info->lo_offset % LOOP_IV_SECTOR_SIZE)
+		goto out;
+
+	strncpy(cms, info->lo_crypt_name, LO_NAME_SIZE);
+	cms[LO_NAME_SIZE - 1] = 0;
+
+	cipher = cmsp;
+	cipher_len = strcspn(cmsp, "-");
+
+	mode = cmsp + cipher_len;
+	mode_len = 0;
+	if (*mode) {
+		mode++;
+		mode_len = strcspn(mode, "-");
+	}
+
+	if (!mode_len) {
+		mode = "cbc";
+		mode_len = 3;
+	}
+
+	if (cipher_len + mode_len + 3 > LO_NAME_SIZE)
+		return -EINVAL;
+
+	memmove(cms, mode, mode_len);
+	cmsp = cms + mode_len;
+	*cmsp++ = '(';
+	memcpy(cmsp, info->lo_crypt_name, cipher_len);
+	cmsp += cipher_len;
+	*cmsp++ = ')';
+	*cmsp = 0;
+
+	tfm = crypto_alloc_blkcipher(cms, 0, CRYPTO_ALG_ASYNC);
+	if (IS_ERR(tfm))
+		return PTR_ERR(tfm);
+
+	err = crypto_blkcipher_setkey(tfm, info->lo_encrypt_key,
+				      info->lo_encrypt_key_size);
+	
+	if (err != 0)
+		goto out_free_tfm;
+
+	lo->key_data = tfm;
+	return 0;
+
+ out_free_tfm:
+	crypto_free_blkcipher(tfm);
+
+ out:
+	return err;
+}
+
+
+typedef int (*encdec_cbc_t)(struct blkcipher_desc *desc,
+			struct scatterlist *sg_out,
+			struct scatterlist *sg_in,
+			unsigned int nsg);
+
+static int
+cryptoloop_transfer(struct loop_device *lo, int cmd,
+		    struct page *raw_page, unsigned raw_off,
+		    struct page *loop_page, unsigned loop_off,
+		    int size, sector_t IV)
+{
+	struct crypto_blkcipher *tfm = lo->key_data;
+	struct blkcipher_desc desc = {
+		.tfm = tfm,
+		.flags = CRYPTO_TFM_REQ_MAY_SLEEP,
+	};
+	struct scatterlist sg_out;
+	struct scatterlist sg_in;
+
+	encdec_cbc_t encdecfunc;
+	struct page *in_page, *out_page;
+	unsigned in_offs, out_offs;
+	int err;
+
+	sg_init_table(&sg_out, 1);
+	sg_init_table(&sg_in, 1);
+
+	if (cmd == READ) {
+		in_page = raw_page;
+		in_offs = raw_off;
+		out_page = loop_page;
+		out_offs = loop_off;
+		encdecfunc = crypto_blkcipher_crt(tfm)->decrypt;
+	} else {
+		in_page = loop_page;
+		in_offs = loop_off;
+		out_page = raw_page;
+		out_offs = raw_off;
+		encdecfunc = crypto_blkcipher_crt(tfm)->encrypt;
+	}
+
+	while (size > 0) {
+		const int sz = min(size, LOOP_IV_SECTOR_SIZE);
+		u32 iv[4] = { 0, };
+		iv[0] = cpu_to_le32(IV & 0xffffffff);
+
+		sg_set_page(&sg_in, in_page, sz, in_offs);
+		sg_set_page(&sg_out, out_page, sz, out_offs);
+
+		desc.info = iv;
+		err = encdecfunc(&desc, &sg_out, &sg_in, sz);
+		if (err)
+			return err;
+
+		IV++;
+		size -= sz;
+		in_offs += sz;
+		out_offs += sz;
+	}
+
+	return 0;
+}
+
+static int
+cryptoloop_ioctl(struct loop_device *lo, int cmd, unsigned long arg)
+{
+	return -EINVAL;
+}
+
+static int
+cryptoloop_release(struct loop_device *lo)
+{
+	struct crypto_blkcipher *tfm = lo->key_data;
+	if (tfm != NULL) {
+		crypto_free_blkcipher(tfm);
+		lo->key_data = NULL;
+		return 0;
+	}
+	printk(KERN_ERR "cryptoloop_release(): tfm == NULL?\n");
+	return -EINVAL;
+}
+
+static struct loop_func_table cryptoloop_funcs = {
+	.number = LO_CRYPT_CRYPTOAPI,
+	.init = cryptoloop_init,
+	.ioctl = cryptoloop_ioctl,
+	.transfer = cryptoloop_transfer,
+	.release = cryptoloop_release,
+	.owner = THIS_MODULE
+};
+
+static int __init
+init_cryptoloop(void)
+{
+	int rc = loop_register_transfer(&cryptoloop_funcs);
+
+	if (rc)
+		printk(KERN_ERR "cryptoloop: loop_register_transfer failed\n");
+	return rc;
+}
+
+static void __exit
+cleanup_cryptoloop(void)
+{
+	if (loop_unregister_transfer(LO_CRYPT_CRYPTOAPI))
+		printk(KERN_ERR
+			"cryptoloop: loop_unregister_transfer failed\n");
+}
+
+module_init(init_cryptoloop);
+module_exit(cleanup_cryptoloop);

drivers/block/drbd/Kconfig

@@ -0,0 +1,73 @@
+#
+# DRBD device driver configuration
+#
+
+comment "DRBD disabled because PROC_FS or INET not selected"
+	depends on PROC_FS='n' || INET='n'
+
+config BLK_DEV_DRBD
+	tristate "DRBD Distributed Replicated Block Device support"
+	depends on PROC_FS && INET
+	select LRU_CACHE
+	select LIBCRC32C
+	default n
+	help
+
+	  NOTE: In order to authenticate connections you have to select
+	  CRYPTO_HMAC and a hash function as well.
+
+	  DRBD is a shared-nothing, synchronously replicated block device. It
+	  is designed to serve as a building block for high availability
+	  clusters and in this context, is a "drop-in" replacement for shared
+	  storage. Simplistically, you could see it as a network RAID 1.
+
+	  Each minor device has a role, which can be 'primary' or 'secondary'.
+	  On the node with the primary device the application is supposed to
+	  run and to access the device (/dev/drbdX). Every write is sent to
+	  the local 'lower level block device' and, across the network, to the
+	  node with the device in 'secondary' state.  The secondary device
+	  simply writes the data to its lower level block device.
+
+	  DRBD can also be used in dual-Primary mode (device writable on both
+	  nodes), which means it can exhibit shared disk semantics in a
+	  shared-nothing cluster.  Needless to say, on top of dual-Primary
+	  DRBD utilizing a cluster file system is necessary to maintain for
+	  cache coherency.
+
+	  For automatic failover you need a cluster manager (e.g. heartbeat).
+	  See also: http://www.drbd.org/, http://www.linux-ha.org
+
+	  If unsure, say N.
+
+config DRBD_FAULT_INJECTION
+	bool "DRBD fault injection"
+	depends on BLK_DEV_DRBD
+	help
+
+	  Say Y here if you want to simulate IO errors, in order to test DRBD's
+	  behavior.
+
+	  The actual simulation of IO errors is done by writing 3 values to
+	  /sys/module/drbd/parameters/
+
+	  enable_faults: bitmask of...
+	  1	meta data write
+	  2               read
+	  4	resync data write
+	  8	            read
+	  16	data write
+	  32	data read
+	  64	read ahead
+	  128	kmalloc of bitmap
+	  256	allocation of peer_requests
+	  512	insert data corruption on receiving side
+
+	  fault_devs: bitmask of minor numbers
+	  fault_rate: frequency in percent
+
+	  Example: Simulate data write errors on /dev/drbd0 with a probability of 5%.
+		echo 16 > /sys/module/drbd/parameters/enable_faults
+		echo 1 > /sys/module/drbd/parameters/fault_devs
+		echo 5 > /sys/module/drbd/parameters/fault_rate
+
+	  If unsure, say N.

drivers/block/drbd/Makefile

@@ -0,0 +1,8 @@
+drbd-y := drbd_bitmap.o drbd_proc.o
+drbd-y += drbd_worker.o drbd_receiver.o drbd_req.o drbd_actlog.o
+drbd-y += drbd_main.o drbd_strings.o drbd_nl.o
+drbd-y += drbd_interval.o drbd_state.o
+drbd-y += drbd_nla.o
+drbd-$(CONFIG_DEBUG_FS) += drbd_debugfs.o
+
+obj-$(CONFIG_BLK_DEV_DRBD)     += drbd.o

drivers/block/drbd/drbd_debugfs.c

@@ -0,0 +1,958 @@
+#define pr_fmt(fmt) "drbd debugfs: " fmt
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/debugfs.h>
+#include <linux/seq_file.h>
+#include <linux/stat.h>
+#include <linux/jiffies.h>
+#include <linux/list.h>
+
+#include "drbd_int.h"
+#include "drbd_req.h"
+#include "drbd_debugfs.h"
+
+
+/**********************************************************************
+ * Whenever you change the file format, remember to bump the version. *
+ **********************************************************************/
+
+static struct dentry *drbd_debugfs_root;
+static struct dentry *drbd_debugfs_version;
+static struct dentry *drbd_debugfs_resources;
+static struct dentry *drbd_debugfs_minors;
+
+static void seq_print_age_or_dash(struct seq_file *m, bool valid, unsigned long dt)
+{
+	if (valid)
+		seq_printf(m, "\t%d", jiffies_to_msecs(dt));
+	else
+		seq_printf(m, "\t-");
+}
+
+static void __seq_print_rq_state_bit(struct seq_file *m,
+	bool is_set, char *sep, const char *set_name, const char *unset_name)
+{
+	if (is_set && set_name) {
+		seq_putc(m, *sep);
+		seq_puts(m, set_name);
+		*sep = '|';
+	} else if (!is_set && unset_name) {
+		seq_putc(m, *sep);
+		seq_puts(m, unset_name);
+		*sep = '|';
+	}
+}
+
+static void seq_print_rq_state_bit(struct seq_file *m,
+	bool is_set, char *sep, const char *set_name)
+{
+	__seq_print_rq_state_bit(m, is_set, sep, set_name, NULL);
+}
+
+/* pretty print enum drbd_req_state_bits req->rq_state */
+static void seq_print_request_state(struct seq_file *m, struct drbd_request *req)
+{
+	unsigned int s = req->rq_state;
+	char sep = ' ';
+	seq_printf(m, "\t0x%08x", s);
+	seq_printf(m, "\tmaster: %s", req->master_bio ? "pending" : "completed");
+
+	/* RQ_WRITE ignored, already reported */
+	seq_puts(m, "\tlocal:");
+	seq_print_rq_state_bit(m, s & RQ_IN_ACT_LOG, &sep, "in-AL");
+	seq_print_rq_state_bit(m, s & RQ_POSTPONED, &sep, "postponed");
+	seq_print_rq_state_bit(m, s & RQ_COMPLETION_SUSP, &sep, "suspended");
+	sep = ' ';
+	seq_print_rq_state_bit(m, s & RQ_LOCAL_PENDING, &sep, "pending");
+	seq_print_rq_state_bit(m, s & RQ_LOCAL_COMPLETED, &sep, "completed");
+	seq_print_rq_state_bit(m, s & RQ_LOCAL_ABORTED, &sep, "aborted");
+	seq_print_rq_state_bit(m, s & RQ_LOCAL_OK, &sep, "ok");
+	if (sep == ' ')
+		seq_puts(m, " -");
+
+	/* for_each_connection ... */
+	seq_printf(m, "\tnet:");
+	sep = ' ';
+	seq_print_rq_state_bit(m, s & RQ_NET_PENDING, &sep, "pending");
+	seq_print_rq_state_bit(m, s & RQ_NET_QUEUED, &sep, "queued");
+	seq_print_rq_state_bit(m, s & RQ_NET_SENT, &sep, "sent");
+	seq_print_rq_state_bit(m, s & RQ_NET_DONE, &sep, "done");
+	seq_print_rq_state_bit(m, s & RQ_NET_SIS, &sep, "sis");
+	seq_print_rq_state_bit(m, s & RQ_NET_OK, &sep, "ok");
+	if (sep == ' ')
+		seq_puts(m, " -");
+
+	seq_printf(m, " :");
+	sep = ' ';
+	seq_print_rq_state_bit(m, s & RQ_EXP_RECEIVE_ACK, &sep, "B");
+	seq_print_rq_state_bit(m, s & RQ_EXP_WRITE_ACK, &sep, "C");
+	seq_print_rq_state_bit(m, s & RQ_EXP_BARR_ACK, &sep, "barr");
+	if (sep == ' ')
+		seq_puts(m, " -");
+	seq_printf(m, "\n");
+}
+
+static void seq_print_one_request(struct seq_file *m, struct drbd_request *req, unsigned long now)
+{
+	/* change anything here, fixup header below! */
+	unsigned int s = req->rq_state;
+
+#define RQ_HDR_1 "epoch\tsector\tsize\trw"
+	seq_printf(m, "0x%x\t%llu\t%u\t%s",
+		req->epoch,
+		(unsigned long long)req->i.sector, req->i.size >> 9,
+		(s & RQ_WRITE) ? "W" : "R");
+
+#define RQ_HDR_2 "\tstart\tin AL\tsubmit"
+	seq_printf(m, "\t%d", jiffies_to_msecs(now - req->start_jif));
+	seq_print_age_or_dash(m, s & RQ_IN_ACT_LOG, now - req->in_actlog_jif);
+	seq_print_age_or_dash(m, s & RQ_LOCAL_PENDING, now - req->pre_submit_jif);
+
+#define RQ_HDR_3 "\tsent\tacked\tdone"
+	seq_print_age_or_dash(m, s & RQ_NET_SENT, now - req->pre_send_jif);
+	seq_print_age_or_dash(m, (s & RQ_NET_SENT) && !(s & RQ_NET_PENDING), now - req->acked_jif);
+	seq_print_age_or_dash(m, s & RQ_NET_DONE, now - req->net_done_jif);
+
+#define RQ_HDR_4 "\tstate\n"
+	seq_print_request_state(m, req);
+}
+#define RQ_HDR RQ_HDR_1 RQ_HDR_2 RQ_HDR_3 RQ_HDR_4
+
+static void seq_print_minor_vnr_req(struct seq_file *m, struct drbd_request *req, unsigned long now)
+{
+	seq_printf(m, "%u\t%u\t", req->device->minor, req->device->vnr);
+	seq_print_one_request(m, req, now);
+}
+
+static void seq_print_resource_pending_meta_io(struct seq_file *m, struct drbd_resource *resource, unsigned long now)
+{
+	struct drbd_device *device;
+	unsigned int i;
+
+	seq_puts(m, "minor\tvnr\tstart\tsubmit\tintent\n");
+	rcu_read_lock();
+	idr_for_each_entry(&resource->devices, device, i) {
+		struct drbd_md_io tmp;
+		/* In theory this is racy,
+		 * in the sense that there could have been a
+		 * drbd_md_put_buffer(); drbd_md_get_buffer();
+		 * between accessing these members here.  */
+		tmp = device->md_io;
+		if (atomic_read(&tmp.in_use)) {
+			seq_printf(m, "%u\t%u\t%d\t",
+				device->minor, device->vnr,
+				jiffies_to_msecs(now - tmp.start_jif));
+			if (time_before(tmp.submit_jif, tmp.start_jif))
+				seq_puts(m, "-\t");
+			else
+				seq_printf(m, "%d\t", jiffies_to_msecs(now - tmp.submit_jif));
+			seq_printf(m, "%s\n", tmp.current_use);
+		}
+	}
+	rcu_read_unlock();
+}
+
+static void seq_print_waiting_for_AL(struct seq_file *m, struct drbd_resource *resource, unsigned long now)
+{
+	struct drbd_device *device;
+	unsigned int i;
+
+	seq_puts(m, "minor\tvnr\tage\t#waiting\n");
+	rcu_read_lock();
+	idr_for_each_entry(&resource->devices, device, i) {
+		unsigned long jif;
+		struct drbd_request *req;
+		int n = atomic_read(&device->ap_actlog_cnt);
+		if (n) {
+			spin_lock_irq(&device->resource->req_lock);
+			req = list_first_entry_or_null(&device->pending_master_completion[1],
+				struct drbd_request, req_pending_master_completion);
+			/* if the oldest request does not wait for the activity log
+			 * it is not interesting for us here */
+			if (req && !(req->rq_state & RQ_IN_ACT_LOG))
+				jif = req->start_jif;
+			else
+				req = NULL;
+			spin_unlock_irq(&device->resource->req_lock);
+		}
+		if (n) {
+			seq_printf(m, "%u\t%u\t", device->minor, device->vnr);
+			if (req)
+				seq_printf(m, "%u\t", jiffies_to_msecs(now - jif));
+			else
+				seq_puts(m, "-\t");
+			seq_printf(m, "%u\n", n);
+		}
+	}
+	rcu_read_unlock();
+}
+
+static void seq_print_device_bitmap_io(struct seq_file *m, struct drbd_device *device, unsigned long now)
+{
+	struct drbd_bm_aio_ctx *ctx;
+	unsigned long start_jif;
+	unsigned int in_flight;
+	unsigned int flags;
+	spin_lock_irq(&device->resource->req_lock);
+	ctx = list_first_entry_or_null(&device->pending_bitmap_io, struct drbd_bm_aio_ctx, list);
+	if (ctx && ctx->done)
+		ctx = NULL;
+	if (ctx) {
+		start_jif = ctx->start_jif;
+		in_flight = atomic_read(&ctx->in_flight);
+		flags = ctx->flags;
+	}
+	spin_unlock_irq(&device->resource->req_lock);
+	if (ctx) {
+		seq_printf(m, "%u\t%u\t%c\t%u\t%u\n",
+			device->minor, device->vnr,
+			(flags & BM_AIO_READ) ? 'R' : 'W',
+			jiffies_to_msecs(now - start_jif),
+			in_flight);
+	}
+}
+
+static void seq_print_resource_pending_bitmap_io(struct seq_file *m, struct drbd_resource *resource, unsigned long now)
+{
+	struct drbd_device *device;
+	unsigned int i;
+
+	seq_puts(m, "minor\tvnr\trw\tage\t#in-flight\n");
+	rcu_read_lock();
+	idr_for_each_entry(&resource->devices, device, i) {
+		seq_print_device_bitmap_io(m, device, now);
+	}
+	rcu_read_unlock();
+}
+
+/* pretty print enum peer_req->flags */
+static void seq_print_peer_request_flags(struct seq_file *m, struct drbd_peer_request *peer_req)
+{
+	unsigned long f = peer_req->flags;
+	char sep = ' ';
+
+	__seq_print_rq_state_bit(m, f & EE_SUBMITTED, &sep, "submitted", "preparing");
+	__seq_print_rq_state_bit(m, f & EE_APPLICATION, &sep, "application", "internal");
+	seq_print_rq_state_bit(m, f & EE_CALL_AL_COMPLETE_IO, &sep, "in-AL");
+	seq_print_rq_state_bit(m, f & EE_SEND_WRITE_ACK, &sep, "C");
+	seq_print_rq_state_bit(m, f & EE_MAY_SET_IN_SYNC, &sep, "set-in-sync");
+
+	if (f & EE_IS_TRIM) {
+		seq_putc(m, sep);
+		sep = '|';
+		if (f & EE_IS_TRIM_USE_ZEROOUT)
+			seq_puts(m, "zero-out");
+		else
+			seq_puts(m, "trim");
+	}
+	seq_putc(m, '\n');
+}
+
+static void seq_print_peer_request(struct seq_file *m,
+	struct drbd_device *device, struct list_head *lh,
+	unsigned long now)
+{
+	bool reported_preparing = false;
+	struct drbd_peer_request *peer_req;
+	list_for_each_entry(peer_req, lh, w.list) {
+		if (reported_preparing && !(peer_req->flags & EE_SUBMITTED))
+			continue;
+
+		if (device)
+			seq_printf(m, "%u\t%u\t", device->minor, device->vnr);
+
+		seq_printf(m, "%llu\t%u\t%c\t%u\t",
+			(unsigned long long)peer_req->i.sector, peer_req->i.size >> 9,
+			(peer_req->flags & EE_WRITE) ? 'W' : 'R',
+			jiffies_to_msecs(now - peer_req->submit_jif));
+		seq_print_peer_request_flags(m, peer_req);
+		if (peer_req->flags & EE_SUBMITTED)
+			break;
+		else
+			reported_preparing = true;
+	}
+}
+
+static void seq_print_device_peer_requests(struct seq_file *m,
+	struct drbd_device *device, unsigned long now)
+{
+	seq_puts(m, "minor\tvnr\tsector\tsize\trw\tage\tflags\n");
+	spin_lock_irq(&device->resource->req_lock);
+	seq_print_peer_request(m, device, &device->active_ee, now);
+	seq_print_peer_request(m, device, &device->read_ee, now);
+	seq_print_peer_request(m, device, &device->sync_ee, now);
+	spin_unlock_irq(&device->resource->req_lock);
+	if (test_bit(FLUSH_PENDING, &device->flags)) {
+		seq_printf(m, "%u\t%u\t-\t-\tF\t%u\tflush\n",
+			device->minor, device->vnr,
+			jiffies_to_msecs(now - device->flush_jif));
+	}
+}
+
+static void seq_print_resource_pending_peer_requests(struct seq_file *m,
+	struct drbd_resource *resource, unsigned long now)
+{
+	struct drbd_device *device;
+	unsigned int i;
+
+	rcu_read_lock();
+	idr_for_each_entry(&resource->devices, device, i) {
+		seq_print_device_peer_requests(m, device, now);
+	}
+	rcu_read_unlock();
+}
+
+static void seq_print_resource_transfer_log_summary(struct seq_file *m,
+	struct drbd_resource *resource,
+	struct drbd_connection *connection,
+	unsigned long now)
+{
+	struct drbd_request *req;
+	unsigned int count = 0;
+	unsigned int show_state = 0;
+
+	seq_puts(m, "n\tdevice\tvnr\t" RQ_HDR);
+	spin_lock_irq(&resource->req_lock);
+	list_for_each_entry(req, &connection->transfer_log, tl_requests) {
+		unsigned int tmp = 0;
+		unsigned int s;
+		++count;
+
+		/* don't disable irq "forever" */
+		if (!(count & 0x1ff)) {
+			struct drbd_request *req_next;
+			kref_get(&req->kref);
+			spin_unlock_irq(&resource->req_lock);
+			cond_resched();
+			spin_lock_irq(&resource->req_lock);
+			req_next = list_next_entry(req, tl_requests);
+			if (kref_put(&req->kref, drbd_req_destroy))
+				req = req_next;
+			if (&req->tl_requests == &connection->transfer_log)
+				break;
+		}
+
+		s = req->rq_state;
+
+		/* This is meant to summarize timing issues, to be able to tell
+		 * local disk problems from network problems.
+		 * Skip requests, if we have shown an even older request with
+		 * similar aspects already.  */
+		if (req->master_bio == NULL)
+			tmp |= 1;
+		if ((s & RQ_LOCAL_MASK) && (s & RQ_LOCAL_PENDING))
+			tmp |= 2;
+		if (s & RQ_NET_MASK) {
+			if (!(s & RQ_NET_SENT))
+				tmp |= 4;
+			if (s & RQ_NET_PENDING)
+				tmp |= 8;
+			if (!(s & RQ_NET_DONE))
+				tmp |= 16;
+		}
+		if ((tmp & show_state) == tmp)
+			continue;
+		show_state |= tmp;
+		seq_printf(m, "%u\t", count);
+		seq_print_minor_vnr_req(m, req, now);
+		if (show_state == 0x1f)
+			break;
+	}
+	spin_unlock_irq(&resource->req_lock);
+}
+
+/* TODO: transfer_log and friends should be moved to resource */
+static int in_flight_summary_show(struct seq_file *m, void *pos)
+{
+	struct drbd_resource *resource = m->private;
+	struct drbd_connection *connection;
+	unsigned long jif = jiffies;
+
+	connection = first_connection(resource);
+	/* This does not happen, actually.
+	 * But be robust and prepare for future code changes. */
+	if (!connection || !kref_get_unless_zero(&connection->kref))
+		return -ESTALE;
+
+	/* BUMP me if you change the file format/content/presentation */
+	seq_printf(m, "v: %u\n\n", 0);
+
+	seq_puts(m, "oldest bitmap IO\n");
+	seq_print_resource_pending_bitmap_io(m, resource, jif);
+	seq_putc(m, '\n');
+
+	seq_puts(m, "meta data IO\n");
+	seq_print_resource_pending_meta_io(m, resource, jif);
+	seq_putc(m, '\n');
+
+	seq_puts(m, "socket buffer stats\n");
+	/* for each connection ... once we have more than one */
+	rcu_read_lock();
+	if (connection->data.socket) {
+		/* open coded SIOCINQ, the "relevant" part */
+		struct tcp_sock *tp = tcp_sk(connection->data.socket->sk);
+		int answ = tp->rcv_nxt - tp->copied_seq;
+		seq_printf(m, "unread receive buffer: %u Byte\n", answ);
+		/* open coded SIOCOUTQ, the "relevant" part */
+		answ = tp->write_seq - tp->snd_una;
+		seq_printf(m, "unacked send buffer: %u Byte\n", answ);
+	}
+	rcu_read_unlock();
+	seq_putc(m, '\n');
+
+	seq_puts(m, "oldest peer requests\n");
+	seq_print_resource_pending_peer_requests(m, resource, jif);
+	seq_putc(m, '\n');
+
+	seq_puts(m, "application requests waiting for activity log\n");
+	seq_print_waiting_for_AL(m, resource, jif);
+	seq_putc(m, '\n');
+
+	seq_puts(m, "oldest application requests\n");
+	seq_print_resource_transfer_log_summary(m, resource, connection, jif);
+	seq_putc(m, '\n');
+
+	jif = jiffies - jif;
+	if (jif)
+		seq_printf(m, "generated in %d ms\n", jiffies_to_msecs(jif));
+	kref_put(&connection->kref, drbd_destroy_connection);
+	return 0;
+}
+
+/* simple_positive(file->f_dentry) respectively debugfs_positive(),
+ * but neither is "reachable" from here.
+ * So we have our own inline version of it above.  :-( */
+static inline int debugfs_positive(struct dentry *dentry)
+{
+        return dentry->d_inode && !d_unhashed(dentry);
+}
+
+/* make sure at *open* time that the respective object won't go away. */
+static int drbd_single_open(struct file *file, int (*show)(struct seq_file *, void *),
+		                void *data, struct kref *kref,
+				void (*release)(struct kref *))
+{
+	struct dentry *parent;
+	int ret = -ESTALE;
+
+	/* Are we still linked,
+	 * or has debugfs_remove() already been called? */
+	parent = file->f_dentry->d_parent;
+	/* not sure if this can happen: */
+	if (!parent || !parent->d_inode)
+		goto out;
+	/* serialize with d_delete() */
+	mutex_lock(&parent->d_inode->i_mutex);
+	/* Make sure the object is still alive */
+	if (debugfs_positive(file->f_dentry)
+	&& kref_get_unless_zero(kref))
+		ret = 0;
+	mutex_unlock(&parent->d_inode->i_mutex);
+	if (!ret) {
+		ret = single_open(file, show, data);
+		if (ret)
+			kref_put(kref, release);
+	}
+out:
+	return ret;
+}
+
+static int in_flight_summary_open(struct inode *inode, struct file *file)
+{
+	struct drbd_resource *resource = inode->i_private;
+	return drbd_single_open(file, in_flight_summary_show, resource,
+				&resource->kref, drbd_destroy_resource);
+}
+
+static int in_flight_summary_release(struct inode *inode, struct file *file)
+{
+	struct drbd_resource *resource = inode->i_private;
+	kref_put(&resource->kref, drbd_destroy_resource);
+	return single_release(inode, file);
+}
+
+static const struct file_operations in_flight_summary_fops = {
+	.owner		= THIS_MODULE,
+	.open		= in_flight_summary_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= in_flight_summary_release,
+};
+
+void drbd_debugfs_resource_add(struct drbd_resource *resource)
+{
+	struct dentry *dentry;
+	if (!drbd_debugfs_resources)
+		return;
+
+	dentry = debugfs_create_dir(resource->name, drbd_debugfs_resources);
+	if (IS_ERR_OR_NULL(dentry))
+		goto fail;
+	resource->debugfs_res = dentry;
+
+	dentry = debugfs_create_dir("volumes", resource->debugfs_res);
+	if (IS_ERR_OR_NULL(dentry))
+		goto fail;
+	resource->debugfs_res_volumes = dentry;
+
+	dentry = debugfs_create_dir("connections", resource->debugfs_res);
+	if (IS_ERR_OR_NULL(dentry))
+		goto fail;
+	resource->debugfs_res_connections = dentry;
+
+	dentry = debugfs_create_file("in_flight_summary", S_IRUSR|S_IRGRP,
+			resource->debugfs_res, resource,
+			&in_flight_summary_fops);
+	if (IS_ERR_OR_NULL(dentry))
+		goto fail;
+	resource->debugfs_res_in_flight_summary = dentry;
+	return;
+
+fail:
+	drbd_debugfs_resource_cleanup(resource);
+	drbd_err(resource, "failed to create debugfs dentry\n");
+}
+
+static void drbd_debugfs_remove(struct dentry **dp)
+{
+	debugfs_remove(*dp);
+	*dp = NULL;
+}
+
+void drbd_debugfs_resource_cleanup(struct drbd_resource *resource)
+{
+	/* it is ok to call debugfs_remove(NULL) */
+	drbd_debugfs_remove(&resource->debugfs_res_in_flight_summary);
+	drbd_debugfs_remove(&resource->debugfs_res_connections);
+	drbd_debugfs_remove(&resource->debugfs_res_volumes);
+	drbd_debugfs_remove(&resource->debugfs_res);
+}
+
+static void seq_print_one_timing_detail(struct seq_file *m,
+	const struct drbd_thread_timing_details *tdp,
+	unsigned long now)
+{
+	struct drbd_thread_timing_details td;
+	/* No locking...
+	 * use temporary assignment to get at consistent data. */
+	do {
+		td = *tdp;
+	} while (td.cb_nr != tdp->cb_nr);
+	if (!td.cb_addr)
+		return;
+	seq_printf(m, "%u\t%d\t%s:%u\t%ps\n",
+			td.cb_nr,
+			jiffies_to_msecs(now - td.start_jif),
+			td.caller_fn, td.line,
+			td.cb_addr);
+}
+
+static void seq_print_timing_details(struct seq_file *m,
+		const char *title,
+		unsigned int cb_nr, struct drbd_thread_timing_details *tdp, unsigned long now)
+{
+	unsigned int start_idx;
+	unsigned int i;
+
+	seq_printf(m, "%s\n", title);
+	/* If not much is going on, this will result in natural ordering.
+	 * If it is very busy, we will possibly skip events, or even see wrap
+	 * arounds, which could only be avoided with locking.
+	 */
+	start_idx = cb_nr % DRBD_THREAD_DETAILS_HIST;
+	for (i = start_idx; i < DRBD_THREAD_DETAILS_HIST; i++)
+		seq_print_one_timing_detail(m, tdp+i, now);
+	for (i = 0; i < start_idx; i++)
+		seq_print_one_timing_detail(m, tdp+i, now);
+}
+
+static int callback_history_show(struct seq_file *m, void *ignored)
+{
+	struct drbd_connection *connection = m->private;
+	unsigned long jif = jiffies;
+
+	/* BUMP me if you change the file format/content/presentation */
+	seq_printf(m, "v: %u\n\n", 0);
+
+	seq_puts(m, "n\tage\tcallsite\tfn\n");
+	seq_print_timing_details(m, "worker", connection->w_cb_nr, connection->w_timing_details, jif);
+	seq_print_timing_details(m, "receiver", connection->r_cb_nr, connection->r_timing_details, jif);
+	return 0;
+}
+
+static int callback_history_open(struct inode *inode, struct file *file)
+{
+	struct drbd_connection *connection = inode->i_private;
+	return drbd_single_open(file, callback_history_show, connection,
+				&connection->kref, drbd_destroy_connection);
+}
+
+static int callback_history_release(struct inode *inode, struct file *file)
+{
+	struct drbd_connection *connection = inode->i_private;
+	kref_put(&connection->kref, drbd_destroy_connection);
+	return single_release(inode, file);
+}
+
+static const struct file_operations connection_callback_history_fops = {
+	.owner		= THIS_MODULE,
+	.open		= callback_history_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= callback_history_release,
+};
+
+static int connection_oldest_requests_show(struct seq_file *m, void *ignored)
+{
+	struct drbd_connection *connection = m->private;
+	unsigned long now = jiffies;
+	struct drbd_request *r1, *r2;
+
+	/* BUMP me if you change the file format/content/presentation */
+	seq_printf(m, "v: %u\n\n", 0);
+
+	spin_lock_irq(&connection->resource->req_lock);
+	r1 = connection->req_next;
+	if (r1)
+		seq_print_minor_vnr_req(m, r1, now);
+	r2 = connection->req_ack_pending;
+	if (r2 && r2 != r1) {
+		r1 = r2;
+		seq_print_minor_vnr_req(m, r1, now);
+	}
+	r2 = connection->req_not_net_done;
+	if (r2 && r2 != r1)
+		seq_print_minor_vnr_req(m, r2, now);
+	spin_unlock_irq(&connection->resource->req_lock);
+	return 0;
+}
+
+static int connection_oldest_requests_open(struct inode *inode, struct file *file)
+{
+	struct drbd_connection *connection = inode->i_private;
+	return drbd_single_open(file, connection_oldest_requests_show, connection,
+				&connection->kref, drbd_destroy_connection);
+}
+
+static int connection_oldest_requests_release(struct inode *inode, struct file *file)
+{
+	struct drbd_connection *connection = inode->i_private;
+	kref_put(&connection->kref, drbd_destroy_connection);
+	return single_release(inode, file);
+}
+
+static const struct file_operations connection_oldest_requests_fops = {
+	.owner		= THIS_MODULE,
+	.open		= connection_oldest_requests_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= connection_oldest_requests_release,
+};
+
+void drbd_debugfs_connection_add(struct drbd_connection *connection)
+{
+	struct dentry *conns_dir = connection->resource->debugfs_res_connections;
+	struct dentry *dentry;
+	if (!conns_dir)
+		return;
+
+	/* Once we enable mutliple peers,
+	 * these connections will have descriptive names.
+	 * For now, it is just the one connection to the (only) "peer". */
+	dentry = debugfs_create_dir("peer", conns_dir);
+	if (IS_ERR_OR_NULL(dentry))
+		goto fail;
+	connection->debugfs_conn = dentry;
+
+	dentry = debugfs_create_file("callback_history", S_IRUSR|S_IRGRP,
+			connection->debugfs_conn, connection,
+			&connection_callback_history_fops);
+	if (IS_ERR_OR_NULL(dentry))
+		goto fail;
+	connection->debugfs_conn_callback_history = dentry;
+
+	dentry = debugfs_create_file("oldest_requests", S_IRUSR|S_IRGRP,
+			connection->debugfs_conn, connection,
+			&connection_oldest_requests_fops);
+	if (IS_ERR_OR_NULL(dentry))
+		goto fail;
+	connection->debugfs_conn_oldest_requests = dentry;
+	return;
+
+fail:
+	drbd_debugfs_connection_cleanup(connection);
+	drbd_err(connection, "failed to create debugfs dentry\n");
+}
+
+void drbd_debugfs_connection_cleanup(struct drbd_connection *connection)
+{
+	drbd_debugfs_remove(&connection->debugfs_conn_callback_history);
+	drbd_debugfs_remove(&connection->debugfs_conn_oldest_requests);
+	drbd_debugfs_remove(&connection->debugfs_conn);
+}
+
+static void resync_dump_detail(struct seq_file *m, struct lc_element *e)
+{
+       struct bm_extent *bme = lc_entry(e, struct bm_extent, lce);
+
+       seq_printf(m, "%5d %s %s %s", bme->rs_left,
+		  test_bit(BME_NO_WRITES, &bme->flags) ? "NO_WRITES" : "---------",
+		  test_bit(BME_LOCKED, &bme->flags) ? "LOCKED" : "------",
+		  test_bit(BME_PRIORITY, &bme->flags) ? "PRIORITY" : "--------"
+		  );
+}
+
+static int device_resync_extents_show(struct seq_file *m, void *ignored)
+{
+	struct drbd_device *device = m->private;
+
+	/* BUMP me if you change the file format/content/presentation */
+	seq_printf(m, "v: %u\n\n", 0);
+
+	if (get_ldev_if_state(device, D_FAILED)) {
+		lc_seq_printf_stats(m, device->resync);
+		lc_seq_dump_details(m, device->resync, "rs_left flags", resync_dump_detail);
+		put_ldev(device);
+	}
+	return 0;
+}
+
+static int device_act_log_extents_show(struct seq_file *m, void *ignored)
+{
+	struct drbd_device *device = m->private;
+
+	/* BUMP me if you change the file format/content/presentation */
+	seq_printf(m, "v: %u\n\n", 0);
+
+	if (get_ldev_if_state(device, D_FAILED)) {
+		lc_seq_printf_stats(m, device->act_log);
+		lc_seq_dump_details(m, device->act_log, "", NULL);
+		put_ldev(device);
+	}
+	return 0;
+}
+
+static int device_oldest_requests_show(struct seq_file *m, void *ignored)
+{
+	struct drbd_device *device = m->private;
+	struct drbd_resource *resource = device->resource;
+	unsigned long now = jiffies;
+	struct drbd_request *r1, *r2;
+	int i;
+
+	/* BUMP me if you change the file format/content/presentation */
+	seq_printf(m, "v: %u\n\n", 0);
+
+	seq_puts(m, RQ_HDR);
+	spin_lock_irq(&resource->req_lock);
+	/* WRITE, then READ */
+	for (i = 1; i >= 0; --i) {
+		r1 = list_first_entry_or_null(&device->pending_master_completion[i],
+			struct drbd_request, req_pending_master_completion);
+		r2 = list_first_entry_or_null(&device->pending_completion[i],
+			struct drbd_request, req_pending_local);
+		if (r1)
+			seq_print_one_request(m, r1, now);
+		if (r2 && r2 != r1)
+			seq_print_one_request(m, r2, now);
+	}
+	spin_unlock_irq(&resource->req_lock);
+	return 0;
+}
+
+static int device_data_gen_id_show(struct seq_file *m, void *ignored)
+{
+	struct drbd_device *device = m->private;
+	struct drbd_md *md;
+	enum drbd_uuid_index idx;
+
+	if (!get_ldev_if_state(device, D_FAILED))
+		return -ENODEV;
+
+	md = &device->ldev->md;
+	spin_lock_irq(&md->uuid_lock);
+	for (idx = UI_CURRENT; idx <= UI_HISTORY_END; idx++) {
+		seq_printf(m, "0x%016llX\n", md->uuid[idx]);
+	}
+	spin_unlock_irq(&md->uuid_lock);
+	put_ldev(device);
+	return 0;
+}
+
+#define drbd_debugfs_device_attr(name)						\
+static int device_ ## name ## _open(struct inode *inode, struct file *file)	\
+{										\
+	struct drbd_device *device = inode->i_private;				\
+	return drbd_single_open(file, device_ ## name ## _show, device,		\
+				&device->kref, drbd_destroy_device);		\
+}										\
+static int device_ ## name ## _release(struct inode *inode, struct file *file)	\
+{										\
+	struct drbd_device *device = inode->i_private;				\
+	kref_put(&device->kref, drbd_destroy_device);				\
+	return single_release(inode, file);					\
+}										\
+static const struct file_operations device_ ## name ## _fops = {		\
+	.owner		= THIS_MODULE,						\
+	.open		= device_ ## name ## _open,				\
+	.read		= seq_read,						\
+	.llseek		= seq_lseek,						\
+	.release	= device_ ## name ## _release,				\
+};
+
+drbd_debugfs_device_attr(oldest_requests)
+drbd_debugfs_device_attr(act_log_extents)
+drbd_debugfs_device_attr(resync_extents)
+drbd_debugfs_device_attr(data_gen_id)
+
+void drbd_debugfs_device_add(struct drbd_device *device)
+{
+	struct dentry *vols_dir = device->resource->debugfs_res_volumes;
+	char minor_buf[8]; /* MINORMASK, MINORBITS == 20; */
+	char vnr_buf[8];   /* volume number vnr is even 16 bit only; */
+	char *slink_name = NULL;
+
+	struct dentry *dentry;
+	if (!vols_dir || !drbd_debugfs_minors)
+		return;
+
+	snprintf(vnr_buf, sizeof(vnr_buf), "%u", device->vnr);
+	dentry = debugfs_create_dir(vnr_buf, vols_dir);
+	if (IS_ERR_OR_NULL(dentry))
+		goto fail;
+	device->debugfs_vol = dentry;
+
+	snprintf(minor_buf, sizeof(minor_buf), "%u", device->minor);
+	slink_name = kasprintf(GFP_KERNEL, "../resources/%s/volumes/%u",
+			device->resource->name, device->vnr);
+	if (!slink_name)
+		goto fail;
+	dentry = debugfs_create_symlink(minor_buf, drbd_debugfs_minors, slink_name);
+	kfree(slink_name);
+	slink_name = NULL;
+	if (IS_ERR_OR_NULL(dentry))
+		goto fail;
+	device->debugfs_minor = dentry;
+
+#define DCF(name)	do {					\
+	dentry = debugfs_create_file(#name, S_IRUSR|S_IRGRP,	\
+			device->debugfs_vol, device,		\
+			&device_ ## name ## _fops);		\
+	if (IS_ERR_OR_NULL(dentry))				\
+		goto fail;					\
+	device->debugfs_vol_ ## name = dentry;			\
+	} while (0)
+
+	DCF(oldest_requests);
+	DCF(act_log_extents);
+	DCF(resync_extents);
+	DCF(data_gen_id);
+#undef DCF
+	return;
+
+fail:
+	drbd_debugfs_device_cleanup(device);
+	drbd_err(device, "failed to create debugfs entries\n");
+}
+
+void drbd_debugfs_device_cleanup(struct drbd_device *device)
+{
+	drbd_debugfs_remove(&device->debugfs_minor);
+	drbd_debugfs_remove(&device->debugfs_vol_oldest_requests);
+	drbd_debugfs_remove(&device->debugfs_vol_act_log_extents);
+	drbd_debugfs_remove(&device->debugfs_vol_resync_extents);
+	drbd_debugfs_remove(&device->debugfs_vol_data_gen_id);
+	drbd_debugfs_remove(&device->debugfs_vol);
+}
+
+void drbd_debugfs_peer_device_add(struct drbd_peer_device *peer_device)
+{
+	struct dentry *conn_dir = peer_device->connection->debugfs_conn;
+	struct dentry *dentry;
+	char vnr_buf[8];
+
+	if (!conn_dir)
+		return;
+
+	snprintf(vnr_buf, sizeof(vnr_buf), "%u", peer_device->device->vnr);
+	dentry = debugfs_create_dir(vnr_buf, conn_dir);
+	if (IS_ERR_OR_NULL(dentry))
+		goto fail;
+	peer_device->debugfs_peer_dev = dentry;
+	return;
+
+fail:
+	drbd_debugfs_peer_device_cleanup(peer_device);
+	drbd_err(peer_device, "failed to create debugfs entries\n");
+}
+
+void drbd_debugfs_peer_device_cleanup(struct drbd_peer_device *peer_device)
+{
+	drbd_debugfs_remove(&peer_device->debugfs_peer_dev);
+}
+
+static int drbd_version_show(struct seq_file *m, void *ignored)
+{
+	seq_printf(m, "# %s\n", drbd_buildtag());
+	seq_printf(m, "VERSION=%s\n", REL_VERSION);
+	seq_printf(m, "API_VERSION=%u\n", API_VERSION);
+	seq_printf(m, "PRO_VERSION_MIN=%u\n", PRO_VERSION_MIN);
+	seq_printf(m, "PRO_VERSION_MAX=%u\n", PRO_VERSION_MAX);
+	return 0;
+}
+
+static int drbd_version_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, drbd_version_show, NULL);
+}
+
+static struct file_operations drbd_version_fops = {
+	.owner = THIS_MODULE,
+	.open = drbd_version_open,
+	.llseek = seq_lseek,
+	.read = seq_read,
+	.release = single_release,
+};
+
+/* not __exit, may be indirectly called
+ * from the module-load-failure path as well. */
+void drbd_debugfs_cleanup(void)
+{
+	drbd_debugfs_remove(&drbd_debugfs_resources);
+	drbd_debugfs_remove(&drbd_debugfs_minors);
+	drbd_debugfs_remove(&drbd_debugfs_version);
+	drbd_debugfs_remove(&drbd_debugfs_root);
+}
+
+int __init drbd_debugfs_init(void)
+{
+	struct dentry *dentry;
+
+	dentry = debugfs_create_dir("drbd", NULL);
+	if (IS_ERR_OR_NULL(dentry))
+		goto fail;
+	drbd_debugfs_root = dentry;
+
+	dentry = debugfs_create_file("version", 0444, drbd_debugfs_root, NULL, &drbd_version_fops);
+	if (IS_ERR_OR_NULL(dentry))
+		goto fail;
+	drbd_debugfs_version = dentry;
+
+	dentry = debugfs_create_dir("resources", drbd_debugfs_root);
+	if (IS_ERR_OR_NULL(dentry))
+		goto fail;
+	drbd_debugfs_resources = dentry;
+
+	dentry = debugfs_create_dir("minors", drbd_debugfs_root);
+	if (IS_ERR_OR_NULL(dentry))
+		goto fail;
+	drbd_debugfs_minors = dentry;
+	return 0;
+
+fail:
+	drbd_debugfs_cleanup();
+	if (dentry)
+		return PTR_ERR(dentry);
+	else
+		return -EINVAL;
+}

drivers/block/drbd/drbd_debugfs.h

@@ -0,0 +1,39 @@
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/debugfs.h>
+
+#include "drbd_int.h"
+
+#ifdef CONFIG_DEBUG_FS
+int __init drbd_debugfs_init(void);
+void drbd_debugfs_cleanup(void);
+
+void drbd_debugfs_resource_add(struct drbd_resource *resource);
+void drbd_debugfs_resource_cleanup(struct drbd_resource *resource);
+
+void drbd_debugfs_connection_add(struct drbd_connection *connection);
+void drbd_debugfs_connection_cleanup(struct drbd_connection *connection);
+
+void drbd_debugfs_device_add(struct drbd_device *device);
+void drbd_debugfs_device_cleanup(struct drbd_device *device);
+
+void drbd_debugfs_peer_device_add(struct drbd_peer_device *peer_device);
+void drbd_debugfs_peer_device_cleanup(struct drbd_peer_device *peer_device);
+#else
+
+static inline int __init drbd_debugfs_init(void) { return -ENODEV; }
+static inline void drbd_debugfs_cleanup(void) { }
+
+static inline void drbd_debugfs_resource_add(struct drbd_resource *resource) { }
+static inline void drbd_debugfs_resource_cleanup(struct drbd_resource *resource) { }
+
+static inline void drbd_debugfs_connection_add(struct drbd_connection *connection) { }
+static inline void drbd_debugfs_connection_cleanup(struct drbd_connection *connection) { }
+
+static inline void drbd_debugfs_device_add(struct drbd_device *device) { }
+static inline void drbd_debugfs_device_cleanup(struct drbd_device *device) { }
+
+static inline void drbd_debugfs_peer_device_add(struct drbd_peer_device *peer_device) { }
+static inline void drbd_debugfs_peer_device_cleanup(struct drbd_peer_device *peer_device) { }
+
+#endif

drivers/block/drbd/drbd_interval.c

@@ -0,0 +1,179 @@
+#include <asm/bug.h>
+#include <linux/rbtree_augmented.h>
+#include "drbd_interval.h"
+
+/**
+ * interval_end  -  return end of @node
+ */
+static inline
+sector_t interval_end(struct rb_node *node)
+{
+	struct drbd_interval *this = rb_entry(node, struct drbd_interval, rb);
+	return this->end;
+}
+
+/**
+ * compute_subtree_last  -  compute end of @node
+ *
+ * The end of an interval is the highest (start + (size >> 9)) value of this
+ * node and of its children.  Called for @node and its parents whenever the end
+ * may have changed.
+ */
+static inline sector_t
+compute_subtree_last(struct drbd_interval *node)
+{
+	sector_t max = node->sector + (node->size >> 9);
+
+	if (node->rb.rb_left) {
+		sector_t left = interval_end(node->rb.rb_left);
+		if (left > max)
+			max = left;
+	}
+	if (node->rb.rb_right) {
+		sector_t right = interval_end(node->rb.rb_right);
+		if (right > max)
+			max = right;
+	}
+	return max;
+}
+
+RB_DECLARE_CALLBACKS(static, augment_callbacks, struct drbd_interval, rb,
+		     sector_t, end, compute_subtree_last);
+
+/**
+ * drbd_insert_interval  -  insert a new interval into a tree
+ */
+bool
+drbd_insert_interval(struct rb_root *root, struct drbd_interval *this)
+{
+	struct rb_node **new = &root->rb_node, *parent = NULL;
+	sector_t this_end = this->sector + (this->size >> 9);
+
+	BUG_ON(!IS_ALIGNED(this->size, 512));
+
+	while (*new) {
+		struct drbd_interval *here =
+			rb_entry(*new, struct drbd_interval, rb);
+
+		parent = *new;
+		if (here->end < this_end)
+			here->end = this_end;
+		if (this->sector < here->sector)
+			new = &(*new)->rb_left;
+		else if (this->sector > here->sector)
+			new = &(*new)->rb_right;
+		else if (this < here)
+			new = &(*new)->rb_left;
+		else if (this > here)
+			new = &(*new)->rb_right;
+		else
+			return false;
+	}
+
+	this->end = this_end;
+	rb_link_node(&this->rb, parent, new);
+	rb_insert_augmented(&this->rb, root, &augment_callbacks);
+	return true;
+}
+
+/**
+ * drbd_contains_interval  -  check if a tree contains a given interval
+ * @sector:	start sector of @interval
+ * @interval:	may not be a valid pointer
+ *
+ * Returns if the tree contains the node @interval with start sector @start.
+ * Does not dereference @interval until @interval is known to be a valid object
+ * in @tree.  Returns %false if @interval is in the tree but with a different
+ * sector number.
+ */
+bool
+drbd_contains_interval(struct rb_root *root, sector_t sector,
+		       struct drbd_interval *interval)
+{
+	struct rb_node *node = root->rb_node;
+
+	while (node) {
+		struct drbd_interval *here =
+			rb_entry(node, struct drbd_interval, rb);
+
+		if (sector < here->sector)
+			node = node->rb_left;
+		else if (sector > here->sector)
+			node = node->rb_right;
+		else if (interval < here)
+			node = node->rb_left;
+		else if (interval > here)
+			node = node->rb_right;
+		else
+			return true;
+	}
+	return false;
+}
+
+/**
+ * drbd_remove_interval  -  remove an interval from a tree
+ */
+void
+drbd_remove_interval(struct rb_root *root, struct drbd_interval *this)
+{
+	rb_erase_augmented(&this->rb, root, &augment_callbacks);
+}
+
+/**
+ * drbd_find_overlap  - search for an interval overlapping with [sector, sector + size)
+ * @sector:	start sector
+ * @size:	size, aligned to 512 bytes
+ *
+ * Returns an interval overlapping with [sector, sector + size), or NULL if
+ * there is none.  When there is more than one overlapping interval in the
+ * tree, the interval with the lowest start sector is returned, and all other
+ * overlapping intervals will be on the right side of the tree, reachable with
+ * rb_next().
+ */
+struct drbd_interval *
+drbd_find_overlap(struct rb_root *root, sector_t sector, unsigned int size)
+{
+	struct rb_node *node = root->rb_node;
+	struct drbd_interval *overlap = NULL;
+	sector_t end = sector + (size >> 9);
+
+	BUG_ON(!IS_ALIGNED(size, 512));
+
+	while (node) {
+		struct drbd_interval *here =
+			rb_entry(node, struct drbd_interval, rb);
+
+		if (node->rb_left &&
+		    sector < interval_end(node->rb_left)) {
+			/* Overlap if any must be on left side */
+			node = node->rb_left;
+		} else if (here->sector < end &&
+			   sector < here->sector + (here->size >> 9)) {
+			overlap = here;
+			break;
+		} else if (sector >= here->sector) {
+			/* Overlap if any must be on right side */
+			node = node->rb_right;
+		} else
+			break;
+	}
+	return overlap;
+}
+
+struct drbd_interval *
+drbd_next_overlap(struct drbd_interval *i, sector_t sector, unsigned int size)
+{
+	sector_t end = sector + (size >> 9);
+	struct rb_node *node;
+
+	for (;;) {
+		node = rb_next(&i->rb);
+		if (!node)
+			return NULL;
+		i = rb_entry(node, struct drbd_interval, rb);
+		if (i->sector >= end)
+			return NULL;
+		if (sector < i->sector + (i->size >> 9))
+			return i;
+	}
+}

drivers/block/drbd/drbd_interval.h

@@ -0,0 +1,42 @@
+#ifndef __DRBD_INTERVAL_H
+#define __DRBD_INTERVAL_H
+
+#include <linux/types.h>
+#include <linux/rbtree.h>
+
+struct drbd_interval {
+	struct rb_node rb;
+	sector_t sector;	/* start sector of the interval */
+	unsigned int size;	/* size in bytes */
+	sector_t end;		/* highest interval end in subtree */
+	int local:1		/* local or remote request? */;
+	int waiting:1;		/* someone is waiting for this to complete */
+	int completed:1;	/* this has been completed already;
+				 * ignore for conflict detection */
+};
+
+static inline void drbd_clear_interval(struct drbd_interval *i)
+{
+	RB_CLEAR_NODE(&i->rb);
+}
+
+static inline bool drbd_interval_empty(struct drbd_interval *i)
+{
+	return RB_EMPTY_NODE(&i->rb);
+}
+
+extern bool drbd_insert_interval(struct rb_root *, struct drbd_interval *);
+extern bool drbd_contains_interval(struct rb_root *, sector_t,
+				   struct drbd_interval *);
+extern void drbd_remove_interval(struct rb_root *, struct drbd_interval *);
+extern struct drbd_interval *drbd_find_overlap(struct rb_root *, sector_t,
+					unsigned int);
+extern struct drbd_interval *drbd_next_overlap(struct drbd_interval *, sector_t,
+					unsigned int);
+
+#define drbd_for_each_overlap(i, root, sector, size)		\
+	for (i = drbd_find_overlap(root, sector, size);		\
+	     i;							\
+	     i = drbd_next_overlap(i, sector, size))
+
+#endif  /* __DRBD_INTERVAL_H */

drivers/block/drbd/drbd_nla.c

@@ -0,0 +1,54 @@
+#include <linux/kernel.h>
+#include <net/netlink.h>
+#include <linux/drbd_genl_api.h>
+#include "drbd_nla.h"
+
+static int drbd_nla_check_mandatory(int maxtype, struct nlattr *nla)
+{
+	struct nlattr *head = nla_data(nla);
+	int len = nla_len(nla);
+	int rem;
+
+	/*
+	 * validate_nla (called from nla_parse_nested) ignores attributes
+	 * beyond maxtype, and does not understand the DRBD_GENLA_F_MANDATORY flag.
+	 * In order to have it validate attributes with the DRBD_GENLA_F_MANDATORY
+	 * flag set also, check and remove that flag before calling
+	 * nla_parse_nested.
+	 */
+
+	nla_for_each_attr(nla, head, len, rem) {
+		if (nla->nla_type & DRBD_GENLA_F_MANDATORY) {
+			nla->nla_type &= ~DRBD_GENLA_F_MANDATORY;
+			if (nla_type(nla) > maxtype)
+				return -EOPNOTSUPP;
+		}
+	}
+	return 0;
+}
+
+int drbd_nla_parse_nested(struct nlattr *tb[], int maxtype, struct nlattr *nla,
+			  const struct nla_policy *policy)
+{
+	int err;
+
+	err = drbd_nla_check_mandatory(maxtype, nla);
+	if (!err)
+		err = nla_parse_nested(tb, maxtype, nla, policy);
+
+	return err;
+}
+
+struct nlattr *drbd_nla_find_nested(int maxtype, struct nlattr *nla, int attrtype)
+{
+	int err;
+	/*
+	 * If any nested attribute has the DRBD_GENLA_F_MANDATORY flag set and
+	 * we don't know about that attribute, reject all the nested
+	 * attributes.
+	 */
+	err = drbd_nla_check_mandatory(maxtype, nla);
+	if (err)
+		return ERR_PTR(err);
+	return nla_find_nested(nla, attrtype);
+}

drivers/block/drbd/drbd_nla.h

@@ -0,0 +1,8 @@
+#ifndef __DRBD_NLA_H
+#define __DRBD_NLA_H
+
+extern int drbd_nla_parse_nested(struct nlattr *tb[], int maxtype, struct nlattr *nla,
+				 const struct nla_policy *policy);
+extern struct nlattr *drbd_nla_find_nested(int maxtype, struct nlattr *nla, int attrtype);
+
+#endif  /* __DRBD_NLA_H */

drivers/block/drbd/drbd_proc.c

@@ -0,0 +1,368 @@
+/*
+   drbd_proc.c
+
+   This file is part of DRBD by Philipp Reisner and Lars Ellenberg.
+
+   Copyright (C) 2001-2008, LINBIT Information Technologies GmbH.
+   Copyright (C) 1999-2008, Philipp Reisner <philipp.reisner@linbit.com>.
+   Copyright (C) 2002-2008, Lars Ellenberg <lars.ellenberg@linbit.com>.
+
+   drbd is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation; either version 2, or (at your option)
+   any later version.
+
+   drbd is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with drbd; see the file COPYING.  If not, write to
+   the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+
+ */
+
+#include <linux/module.h>
+
+#include <asm/uaccess.h>
+#include <linux/fs.h>
+#include <linux/file.h>
+#include <linux/proc_fs.h>
+#include <linux/seq_file.h>
+#include <linux/drbd.h>
+#include "drbd_int.h"
+
+static int drbd_proc_open(struct inode *inode, struct file *file);
+static int drbd_proc_release(struct inode *inode, struct file *file);
+
+
+struct proc_dir_entry *drbd_proc;
+const struct file_operations drbd_proc_fops = {
+	.owner		= THIS_MODULE,
+	.open		= drbd_proc_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= drbd_proc_release,
+};
+
+static void seq_printf_with_thousands_grouping(struct seq_file *seq, long v)
+{
+	/* v is in kB/sec. We don't expect TiByte/sec yet. */
+	if (unlikely(v >= 1000000)) {
+		/* cool: > GiByte/s */
+		seq_printf(seq, "%ld,", v / 1000000);
+		v %= 1000000;
+		seq_printf(seq, "%03ld,%03ld", v/1000, v % 1000);
+	} else if (likely(v >= 1000))
+		seq_printf(seq, "%ld,%03ld", v/1000, v % 1000);
+	else
+		seq_printf(seq, "%ld", v);
+}
+
+static void drbd_get_syncer_progress(struct drbd_device *device,
+		union drbd_dev_state state, unsigned long *rs_total,
+		unsigned long *bits_left, unsigned int *per_mil_done)
+{
+	/* this is to break it at compile time when we change that, in case we
+	 * want to support more than (1<<32) bits on a 32bit arch. */
+	typecheck(unsigned long, device->rs_total);
+	*rs_total = device->rs_total;
+
+	/* note: both rs_total and rs_left are in bits, i.e. in
+	 * units of BM_BLOCK_SIZE.
+	 * for the percentage, we don't care. */
+
+	if (state.conn == C_VERIFY_S || state.conn == C_VERIFY_T)
+		*bits_left = device->ov_left;
+	else
+		*bits_left = drbd_bm_total_weight(device) - device->rs_failed;
+	/* >> 10 to prevent overflow,
+	 * +1 to prevent division by zero */
+	if (*bits_left > *rs_total) {
+		/* D'oh. Maybe a logic bug somewhere.  More likely just a race
+		 * between state change and reset of rs_total.
+		 */
+		*bits_left = *rs_total;
+		*per_mil_done = *rs_total ? 0 : 1000;
+	} else {
+		/* Make sure the division happens in long context.
+		 * We allow up to one petabyte storage right now,
+		 * at a granularity of 4k per bit that is 2**38 bits.
+		 * After shift right and multiplication by 1000,
+		 * this should still fit easily into a 32bit long,
+		 * so we don't need a 64bit division on 32bit arch.
+		 * Note: currently we don't support such large bitmaps on 32bit
+		 * arch anyways, but no harm done to be prepared for it here.
+		 */
+		unsigned int shift = *rs_total > UINT_MAX ? 16 : 10;
+		unsigned long left = *bits_left >> shift;
+		unsigned long total = 1UL + (*rs_total >> shift);
+		unsigned long tmp = 1000UL - left * 1000UL/total;
+		*per_mil_done = tmp;
+	}
+}
+
+
+/*lge
+ * progress bars shamelessly adapted from driver/md/md.c
+ * output looks like
+ *	[=====>..............] 33.5% (23456/123456)
+ *	finish: 2:20:20 speed: 6,345 (6,456) K/sec
+ */
+static void drbd_syncer_progress(struct drbd_device *device, struct seq_file *seq,
+		union drbd_dev_state state)
+{
+	unsigned long db, dt, dbdt, rt, rs_total, rs_left;
+	unsigned int res;
+	int i, x, y;
+	int stalled = 0;
+
+	drbd_get_syncer_progress(device, state, &rs_total, &rs_left, &res);
+
+	x = res/50;
+	y = 20-x;
+	seq_printf(seq, "\t[");
+	for (i = 1; i < x; i++)
+		seq_printf(seq, "=");
+	seq_printf(seq, ">");
+	for (i = 0; i < y; i++)
+		seq_printf(seq, ".");
+	seq_printf(seq, "] ");
+
+	if (state.conn == C_VERIFY_S || state.conn == C_VERIFY_T)
+		seq_printf(seq, "verified:");
+	else
+		seq_printf(seq, "sync'ed:");
+	seq_printf(seq, "%3u.%u%% ", res / 10, res % 10);
+
+	/* if more than a few GB, display in MB */
+	if (rs_total > (4UL << (30 - BM_BLOCK_SHIFT)))
+		seq_printf(seq, "(%lu/%lu)M",
+			    (unsigned long) Bit2KB(rs_left >> 10),
+			    (unsigned long) Bit2KB(rs_total >> 10));
+	else
+		seq_printf(seq, "(%lu/%lu)K",
+			    (unsigned long) Bit2KB(rs_left),
+			    (unsigned long) Bit2KB(rs_total));
+
+	seq_printf(seq, "\n\t");
+
+	/* see drivers/md/md.c
+	 * We do not want to overflow, so the order of operands and
+	 * the * 100 / 100 trick are important. We do a +1 to be
+	 * safe against division by zero. We only estimate anyway.
+	 *
+	 * dt: time from mark until now
+	 * db: blocks written from mark until now
+	 * rt: remaining time
+	 */
+	/* Rolling marks. last_mark+1 may just now be modified.  last_mark+2 is
+	 * at least (DRBD_SYNC_MARKS-2)*DRBD_SYNC_MARK_STEP old, and has at
+	 * least DRBD_SYNC_MARK_STEP time before it will be modified. */
+	/* ------------------------ ~18s average ------------------------ */
+	i = (device->rs_last_mark + 2) % DRBD_SYNC_MARKS;
+	dt = (jiffies - device->rs_mark_time[i]) / HZ;
+	if (dt > 180)
+		stalled = 1;
+
+	if (!dt)
+		dt++;
+	db = device->rs_mark_left[i] - rs_left;
+	rt = (dt * (rs_left / (db/100+1)))/100; /* seconds */
+
+	seq_printf(seq, "finish: %lu:%02lu:%02lu",
+		rt / 3600, (rt % 3600) / 60, rt % 60);
+
+	dbdt = Bit2KB(db/dt);
+	seq_printf(seq, " speed: ");
+	seq_printf_with_thousands_grouping(seq, dbdt);
+	seq_printf(seq, " (");
+	/* ------------------------- ~3s average ------------------------ */
+	if (proc_details >= 1) {
+		/* this is what drbd_rs_should_slow_down() uses */
+		i = (device->rs_last_mark + DRBD_SYNC_MARKS-1) % DRBD_SYNC_MARKS;
+		dt = (jiffies - device->rs_mark_time[i]) / HZ;
+		if (!dt)
+			dt++;
+		db = device->rs_mark_left[i] - rs_left;
+		dbdt = Bit2KB(db/dt);
+		seq_printf_with_thousands_grouping(seq, dbdt);
+		seq_printf(seq, " -- ");
+	}
+
+	/* --------------------- long term average ---------------------- */
+	/* mean speed since syncer started
+	 * we do account for PausedSync periods */
+	dt = (jiffies - device->rs_start - device->rs_paused) / HZ;
+	if (dt == 0)
+		dt = 1;
+	db = rs_total - rs_left;
+	dbdt = Bit2KB(db/dt);
+	seq_printf_with_thousands_grouping(seq, dbdt);
+	seq_printf(seq, ")");
+
+	if (state.conn == C_SYNC_TARGET ||
+	    state.conn == C_VERIFY_S) {
+		seq_printf(seq, " want: ");
+		seq_printf_with_thousands_grouping(seq, device->c_sync_rate);
+	}
+	seq_printf(seq, " K/sec%s\n", stalled ? " (stalled)" : "");
+
+	if (proc_details >= 1) {
+		/* 64 bit:
+		 * we convert to sectors in the display below. */
+		unsigned long bm_bits = drbd_bm_bits(device);
+		unsigned long bit_pos;
+		unsigned long long stop_sector = 0;
+		if (state.conn == C_VERIFY_S ||
+		    state.conn == C_VERIFY_T) {
+			bit_pos = bm_bits - device->ov_left;
+			if (verify_can_do_stop_sector(device))
+				stop_sector = device->ov_stop_sector;
+		} else
+			bit_pos = device->bm_resync_fo;
+		/* Total sectors may be slightly off for oddly
+		 * sized devices. So what. */
+		seq_printf(seq,
+			"\t%3d%% sector pos: %llu/%llu",
+			(int)(bit_pos / (bm_bits/100+1)),
+			(unsigned long long)bit_pos * BM_SECT_PER_BIT,
+			(unsigned long long)bm_bits * BM_SECT_PER_BIT);
+		if (stop_sector != 0 && stop_sector != ULLONG_MAX)
+			seq_printf(seq, " stop sector: %llu", stop_sector);
+		seq_printf(seq, "\n");
+	}
+}
+
+static int drbd_seq_show(struct seq_file *seq, void *v)
+{
+	int i, prev_i = -1;
+	const char *sn;
+	struct drbd_device *device;
+	struct net_conf *nc;
+	union drbd_dev_state state;
+	char wp;
+
+	static char write_ordering_chars[] = {
+		[WO_none] = 'n',
+		[WO_drain_io] = 'd',
+		[WO_bdev_flush] = 'f',
+	};
+
+	seq_printf(seq, "version: " REL_VERSION " (api:%d/proto:%d-%d)\n%s\n",
+		   API_VERSION, PRO_VERSION_MIN, PRO_VERSION_MAX, drbd_buildtag());
+
+	/*
+	  cs .. connection state
+	  ro .. node role (local/remote)
+	  ds .. disk state (local/remote)
+	     protocol
+	     various flags
+	  ns .. network send
+	  nr .. network receive
+	  dw .. disk write
+	  dr .. disk read
+	  al .. activity log write count
+	  bm .. bitmap update write count
+	  pe .. pending (waiting for ack or data reply)
+	  ua .. unack'd (still need to send ack or data reply)
+	  ap .. application requests accepted, but not yet completed
+	  ep .. number of epochs currently "on the fly", P_BARRIER_ACK pending
+	  wo .. write ordering mode currently in use
+	 oos .. known out-of-sync kB
+	*/
+
+	rcu_read_lock();
+	idr_for_each_entry(&drbd_devices, device, i) {
+		if (prev_i != i - 1)
+			seq_printf(seq, "\n");
+		prev_i = i;
+
+		state = device->state;
+		sn = drbd_conn_str(state.conn);
+
+		if (state.conn == C_STANDALONE &&
+		    state.disk == D_DISKLESS &&
+		    state.role == R_SECONDARY) {
+			seq_printf(seq, "%2d: cs:Unconfigured\n", i);
+		} else {
+			/* reset device->congestion_reason */
+			bdi_rw_congested(&device->rq_queue->backing_dev_info);
+
+			nc = rcu_dereference(first_peer_device(device)->connection->net_conf);
+			wp = nc ? nc->wire_protocol - DRBD_PROT_A + 'A' : ' ';
+			seq_printf(seq,
+			   "%2d: cs:%s ro:%s/%s ds:%s/%s %c %c%c%c%c%c%c\n"
+			   "    ns:%u nr:%u dw:%u dr:%u al:%u bm:%u "
+			   "lo:%d pe:%d ua:%d ap:%d ep:%d wo:%c",
+			   i, sn,
+			   drbd_role_str(state.role),
+			   drbd_role_str(state.peer),
+			   drbd_disk_str(state.disk),
+			   drbd_disk_str(state.pdsk),
+			   wp,
+			   drbd_suspended(device) ? 's' : 'r',
+			   state.aftr_isp ? 'a' : '-',
+			   state.peer_isp ? 'p' : '-',
+			   state.user_isp ? 'u' : '-',
+			   device->congestion_reason ?: '-',
+			   test_bit(AL_SUSPENDED, &device->flags) ? 's' : '-',
+			   device->send_cnt/2,
+			   device->recv_cnt/2,
+			   device->writ_cnt/2,
+			   device->read_cnt/2,
+			   device->al_writ_cnt,
+			   device->bm_writ_cnt,
+			   atomic_read(&device->local_cnt),
+			   atomic_read(&device->ap_pending_cnt) +
+			   atomic_read(&device->rs_pending_cnt),
+			   atomic_read(&device->unacked_cnt),
+			   atomic_read(&device->ap_bio_cnt),
+			   first_peer_device(device)->connection->epochs,
+			   write_ordering_chars[device->resource->write_ordering]
+			);
+			seq_printf(seq, " oos:%llu\n",
+				   Bit2KB((unsigned long long)
+					   drbd_bm_total_weight(device)));
+		}
+		if (state.conn == C_SYNC_SOURCE ||
+		    state.conn == C_SYNC_TARGET ||
+		    state.conn == C_VERIFY_S ||
+		    state.conn == C_VERIFY_T)
+			drbd_syncer_progress(device, seq, state);
+
+		if (proc_details >= 1 && get_ldev_if_state(device, D_FAILED)) {
+			lc_seq_printf_stats(seq, device->resync);
+			lc_seq_printf_stats(seq, device->act_log);
+			put_ldev(device);
+		}
+
+		if (proc_details >= 2)
+			seq_printf(seq, "\tblocked on activity log: %d\n", atomic_read(&device->ap_actlog_cnt));
+	}
+	rcu_read_unlock();
+
+	return 0;
+}
+
+static int drbd_proc_open(struct inode *inode, struct file *file)
+{
+	int err;
+
+	if (try_module_get(THIS_MODULE)) {
+		err = single_open(file, drbd_seq_show, NULL);
+		if (err)
+			module_put(THIS_MODULE);
+		return err;
+	}
+	return -ENODEV;
+}
+
+static int drbd_proc_release(struct inode *inode, struct file *file)
+{
+	module_put(THIS_MODULE);
+	return single_release(inode, file);
+}
+
+/* PROC FS stuff end */

drivers/block/drbd/drbd_protocol.h

@@ -0,0 +1,307 @@
+#ifndef __DRBD_PROTOCOL_H
+#define __DRBD_PROTOCOL_H
+
+enum drbd_packet {
+	/* receiver (data socket) */
+	P_DATA		      = 0x00,
+	P_DATA_REPLY	      = 0x01, /* Response to P_DATA_REQUEST */
+	P_RS_DATA_REPLY	      = 0x02, /* Response to P_RS_DATA_REQUEST */
+	P_BARRIER	      = 0x03,
+	P_BITMAP	      = 0x04,
+	P_BECOME_SYNC_TARGET  = 0x05,
+	P_BECOME_SYNC_SOURCE  = 0x06,
+	P_UNPLUG_REMOTE	      = 0x07, /* Used at various times to hint the peer */
+	P_DATA_REQUEST	      = 0x08, /* Used to ask for a data block */
+	P_RS_DATA_REQUEST     = 0x09, /* Used to ask for a data block for resync */
+	P_SYNC_PARAM	      = 0x0a,
+	P_PROTOCOL	      = 0x0b,
+	P_UUIDS		      = 0x0c,
+	P_SIZES		      = 0x0d,
+	P_STATE		      = 0x0e,
+	P_SYNC_UUID	      = 0x0f,
+	P_AUTH_CHALLENGE      = 0x10,
+	P_AUTH_RESPONSE	      = 0x11,
+	P_STATE_CHG_REQ	      = 0x12,
+
+	/* asender (meta socket */
+	P_PING		      = 0x13,
+	P_PING_ACK	      = 0x14,
+	P_RECV_ACK	      = 0x15, /* Used in protocol B */
+	P_WRITE_ACK	      = 0x16, /* Used in protocol C */
+	P_RS_WRITE_ACK	      = 0x17, /* Is a P_WRITE_ACK, additionally call set_in_sync(). */
+	P_SUPERSEDED	      = 0x18, /* Used in proto C, two-primaries conflict detection */
+	P_NEG_ACK	      = 0x19, /* Sent if local disk is unusable */
+	P_NEG_DREPLY	      = 0x1a, /* Local disk is broken... */
+	P_NEG_RS_DREPLY	      = 0x1b, /* Local disk is broken... */
+	P_BARRIER_ACK	      = 0x1c,
+	P_STATE_CHG_REPLY     = 0x1d,
+
+	/* "new" commands, no longer fitting into the ordering scheme above */
+
+	P_OV_REQUEST	      = 0x1e, /* data socket */
+	P_OV_REPLY	      = 0x1f,
+	P_OV_RESULT	      = 0x20, /* meta socket */
+	P_CSUM_RS_REQUEST     = 0x21, /* data socket */
+	P_RS_IS_IN_SYNC	      = 0x22, /* meta socket */
+	P_SYNC_PARAM89	      = 0x23, /* data socket, protocol version 89 replacement for P_SYNC_PARAM */
+	P_COMPRESSED_BITMAP   = 0x24, /* compressed or otherwise encoded bitmap transfer */
+	/* P_CKPT_FENCE_REQ      = 0x25, * currently reserved for protocol D */
+	/* P_CKPT_DISABLE_REQ    = 0x26, * currently reserved for protocol D */
+	P_DELAY_PROBE         = 0x27, /* is used on BOTH sockets */
+	P_OUT_OF_SYNC         = 0x28, /* Mark as out of sync (Outrunning), data socket */
+	P_RS_CANCEL           = 0x29, /* meta: Used to cancel RS_DATA_REQUEST packet by SyncSource */
+	P_CONN_ST_CHG_REQ     = 0x2a, /* data sock: Connection wide state request */
+	P_CONN_ST_CHG_REPLY   = 0x2b, /* meta sock: Connection side state req reply */
+	P_RETRY_WRITE	      = 0x2c, /* Protocol C: retry conflicting write request */
+	P_PROTOCOL_UPDATE     = 0x2d, /* data sock: is used in established connections */
+        /* 0x2e to 0x30 reserved, used in drbd 9 */
+
+	/* REQ_DISCARD. We used "discard" in different contexts before,
+	 * which is why I chose TRIM here, to disambiguate. */
+	P_TRIM                = 0x31,
+
+	P_MAY_IGNORE	      = 0x100, /* Flag to test if (cmd > P_MAY_IGNORE) ... */
+	P_MAX_OPT_CMD	      = 0x101,
+
+	/* special command ids for handshake */
+
+	P_INITIAL_META	      = 0xfff1, /* First Packet on the MetaSock */
+	P_INITIAL_DATA	      = 0xfff2, /* First Packet on the Socket */
+
+	P_CONNECTION_FEATURES = 0xfffe	/* FIXED for the next century! */
+};
+
+#ifndef __packed
+#define __packed __attribute__((packed))
+#endif
+
+/* This is the layout for a packet on the wire.
+ * The byteorder is the network byte order.
+ *     (except block_id and barrier fields.
+ *	these are pointers to local structs
+ *	and have no relevance for the partner,
+ *	which just echoes them as received.)
+ *
+ * NOTE that the payload starts at a long aligned offset,
+ * regardless of 32 or 64 bit arch!
+ */
+struct p_header80 {
+	u32	  magic;
+	u16	  command;
+	u16	  length;	/* bytes of data after this header */
+} __packed;
+
+/* Header for big packets, Used for data packets exceeding 64kB */
+struct p_header95 {
+	u16	  magic;	/* use DRBD_MAGIC_BIG here */
+	u16	  command;
+	u32	  length;
+} __packed;
+
+struct p_header100 {
+	u32	  magic;
+	u16	  volume;
+	u16	  command;
+	u32	  length;
+	u32	  pad;
+} __packed;
+
+/* these defines must not be changed without changing the protocol version */
+#define DP_HARDBARRIER	      1 /* depricated */
+#define DP_RW_SYNC	      2 /* equals REQ_SYNC    */
+#define DP_MAY_SET_IN_SYNC    4
+#define DP_UNPLUG             8 /* not used anymore   */
+#define DP_FUA               16 /* equals REQ_FUA     */
+#define DP_FLUSH             32 /* equals REQ_FLUSH   */
+#define DP_DISCARD           64 /* equals REQ_DISCARD */
+#define DP_SEND_RECEIVE_ACK 128 /* This is a proto B write request */
+#define DP_SEND_WRITE_ACK   256 /* This is a proto C write request */
+
+struct p_data {
+	u64	    sector;    /* 64 bits sector number */
+	u64	    block_id;  /* to identify the request in protocol B&C */
+	u32	    seq_num;
+	u32	    dp_flags;
+} __packed;
+
+struct p_trim {
+	struct p_data p_data;
+	u32	    size;	/* == bio->bi_size */
+} __packed;
+
+/*
+ * commands which share a struct:
+ *  p_block_ack:
+ *   P_RECV_ACK (proto B), P_WRITE_ACK (proto C),
+ *   P_SUPERSEDED (proto C, two-primaries conflict detection)
+ *  p_block_req:
+ *   P_DATA_REQUEST, P_RS_DATA_REQUEST
+ */
+struct p_block_ack {
+	u64	    sector;
+	u64	    block_id;
+	u32	    blksize;
+	u32	    seq_num;
+} __packed;
+
+struct p_block_req {
+	u64 sector;
+	u64 block_id;
+	u32 blksize;
+	u32 pad;	/* to multiple of 8 Byte */
+} __packed;
+
+/*
+ * commands with their own struct for additional fields:
+ *   P_CONNECTION_FEATURES
+ *   P_BARRIER
+ *   P_BARRIER_ACK
+ *   P_SYNC_PARAM
+ *   ReportParams
+ */
+
+#define FF_TRIM      1
+
+struct p_connection_features {
+	u32 protocol_min;
+	u32 feature_flags;
+	u32 protocol_max;
+
+	/* should be more than enough for future enhancements
+	 * for now, feature_flags and the reserved array shall be zero.
+	 */
+
+	u32 _pad;
+	u64 reserved[7];
+} __packed;
+
+struct p_barrier {
+	u32 barrier;	/* barrier number _handle_ only */
+	u32 pad;	/* to multiple of 8 Byte */
+} __packed;
+
+struct p_barrier_ack {
+	u32 barrier;
+	u32 set_size;
+} __packed;
+
+struct p_rs_param {
+	u32 resync_rate;
+
+	      /* Since protocol version 88 and higher. */
+	char verify_alg[0];
+} __packed;
+
+struct p_rs_param_89 {
+	u32 resync_rate;
+	/* protocol version 89: */
+	char verify_alg[SHARED_SECRET_MAX];
+	char csums_alg[SHARED_SECRET_MAX];
+} __packed;
+
+struct p_rs_param_95 {
+	u32 resync_rate;
+	char verify_alg[SHARED_SECRET_MAX];
+	char csums_alg[SHARED_SECRET_MAX];
+	u32 c_plan_ahead;
+	u32 c_delay_target;
+	u32 c_fill_target;
+	u32 c_max_rate;
+} __packed;
+
+enum drbd_conn_flags {
+	CF_DISCARD_MY_DATA = 1,
+	CF_DRY_RUN = 2,
+};
+
+struct p_protocol {
+	u32 protocol;
+	u32 after_sb_0p;
+	u32 after_sb_1p;
+	u32 after_sb_2p;
+	u32 conn_flags;
+	u32 two_primaries;
+
+	/* Since protocol version 87 and higher. */
+	char integrity_alg[0];
+
+} __packed;
+
+struct p_uuids {
+	u64 uuid[UI_EXTENDED_SIZE];
+} __packed;
+
+struct p_rs_uuid {
+	u64	    uuid;
+} __packed;
+
+struct p_sizes {
+	u64	    d_size;  /* size of disk */
+	u64	    u_size;  /* user requested size */
+	u64	    c_size;  /* current exported size */
+	u32	    max_bio_size;  /* Maximal size of a BIO */
+	u16	    queue_order_type;  /* not yet implemented in DRBD*/
+	u16	    dds_flags; /* use enum dds_flags here. */
+} __packed;
+
+struct p_state {
+	u32	    state;
+} __packed;
+
+struct p_req_state {
+	u32	    mask;
+	u32	    val;
+} __packed;
+
+struct p_req_state_reply {
+	u32	    retcode;
+} __packed;
+
+struct p_drbd06_param {
+	u64	  size;
+	u32	  state;
+	u32	  blksize;
+	u32	  protocol;
+	u32	  version;
+	u32	  gen_cnt[5];
+	u32	  bit_map_gen[5];
+} __packed;
+
+struct p_block_desc {
+	u64 sector;
+	u32 blksize;
+	u32 pad;	/* to multiple of 8 Byte */
+} __packed;
+
+/* Valid values for the encoding field.
+ * Bump proto version when changing this. */
+enum drbd_bitmap_code {
+	/* RLE_VLI_Bytes = 0,
+	 * and other bit variants had been defined during
+	 * algorithm evaluation. */
+	RLE_VLI_Bits = 2,
+};
+
+struct p_compressed_bm {
+	/* (encoding & 0x0f): actual encoding, see enum drbd_bitmap_code
+	 * (encoding & 0x80): polarity (set/unset) of first runlength
+	 * ((encoding >> 4) & 0x07): pad_bits, number of trailing zero bits
+	 * used to pad up to head.length bytes
+	 */
+	u8 encoding;
+
+	u8 code[0];
+} __packed;
+
+struct p_delay_probe93 {
+	u32     seq_num; /* sequence number to match the two probe packets */
+	u32     offset;  /* usecs the probe got sent after the reference time point */
+} __packed;
+
+/*
+ * Bitmap packets need to fit within a single page on the sender and receiver,
+ * so we are limited to 4 KiB (and not to PAGE_SIZE, which can be bigger).
+ */
+#define DRBD_SOCKET_BUFFER_SIZE 4096
+
+#endif  /* __DRBD_PROTOCOL_H */

drivers/block/drbd/drbd_req.h

@@ -0,0 +1,351 @@
+/*
+   drbd_req.h
+
+   This file is part of DRBD by Philipp Reisner and Lars Ellenberg.
+
+   Copyright (C) 2006-2008, LINBIT Information Technologies GmbH.
+   Copyright (C) 2006-2008, Lars Ellenberg <lars.ellenberg@linbit.com>.
+   Copyright (C) 2006-2008, Philipp Reisner <philipp.reisner@linbit.com>.
+
+   DRBD is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation; either version 2, or (at your option)
+   any later version.
+
+   DRBD is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with drbd; see the file COPYING.  If not, write to
+   the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#ifndef _DRBD_REQ_H
+#define _DRBD_REQ_H
+
+#include <linux/module.h>
+
+#include <linux/slab.h>
+#include <linux/drbd.h>
+#include "drbd_int.h"
+
+/* The request callbacks will be called in irq context by the IDE drivers,
+   and in Softirqs/Tasklets/BH context by the SCSI drivers,
+   and by the receiver and worker in kernel-thread context.
+   Try to get the locking right :) */
+
+/*
+ * Objects of type struct drbd_request do only exist on a R_PRIMARY node, and are
+ * associated with IO requests originating from the block layer above us.
+ *
+ * There are quite a few things that may happen to a drbd request
+ * during its lifetime.
+ *
+ *  It will be created.
+ *  It will be marked with the intention to be
+ *    submitted to local disk and/or
+ *    send via the network.
+ *
+ *  It has to be placed on the transfer log and other housekeeping lists,
+ *  In case we have a network connection.
+ *
+ *  It may be identified as a concurrent (write) request
+ *    and be handled accordingly.
+ *
+ *  It may me handed over to the local disk subsystem.
+ *  It may be completed by the local disk subsystem,
+ *    either successfully or with io-error.
+ *  In case it is a READ request, and it failed locally,
+ *    it may be retried remotely.
+ *
+ *  It may be queued for sending.
+ *  It may be handed over to the network stack,
+ *    which may fail.
+ *  It may be acknowledged by the "peer" according to the wire_protocol in use.
+ *    this may be a negative ack.
+ *  It may receive a faked ack when the network connection is lost and the
+ *  transfer log is cleaned up.
+ *  Sending may be canceled due to network connection loss.
+ *  When it finally has outlived its time,
+ *    corresponding dirty bits in the resync-bitmap may be cleared or set,
+ *    it will be destroyed,
+ *    and completion will be signalled to the originator,
+ *      with or without "success".
+ */
+
+enum drbd_req_event {
+	CREATED,
+	TO_BE_SENT,
+	TO_BE_SUBMITTED,
+
+	/* XXX yes, now I am inconsistent...
+	 * these are not "events" but "actions"
+	 * oh, well... */
+	QUEUE_FOR_NET_WRITE,
+	QUEUE_FOR_NET_READ,
+	QUEUE_FOR_SEND_OOS,
+
+	/* An empty flush is queued as P_BARRIER,
+	 * which will cause it to complete "successfully",
+	 * even if the local disk flush failed.
+	 *
+	 * Just like "real" requests, empty flushes (blkdev_issue_flush()) will
+	 * only see an error if neither local nor remote data is reachable. */
+	QUEUE_AS_DRBD_BARRIER,
+
+	SEND_CANCELED,
+	SEND_FAILED,
+	HANDED_OVER_TO_NETWORK,
+	OOS_HANDED_TO_NETWORK,
+	CONNECTION_LOST_WHILE_PENDING,
+	READ_RETRY_REMOTE_CANCELED,
+	RECV_ACKED_BY_PEER,
+	WRITE_ACKED_BY_PEER,
+	WRITE_ACKED_BY_PEER_AND_SIS, /* and set_in_sync */
+	CONFLICT_RESOLVED,
+	POSTPONE_WRITE,
+	NEG_ACKED,
+	BARRIER_ACKED, /* in protocol A and B */
+	DATA_RECEIVED, /* (remote read) */
+
+	COMPLETED_OK,
+	READ_COMPLETED_WITH_ERROR,
+	READ_AHEAD_COMPLETED_WITH_ERROR,
+	WRITE_COMPLETED_WITH_ERROR,
+	DISCARD_COMPLETED_NOTSUPP,
+	DISCARD_COMPLETED_WITH_ERROR,
+
+	ABORT_DISK_IO,
+	RESEND,
+	FAIL_FROZEN_DISK_IO,
+	RESTART_FROZEN_DISK_IO,
+	NOTHING,
+};
+
+/* encoding of request states for now.  we don't actually need that many bits.
+ * we don't need to do atomic bit operations either, since most of the time we
+ * need to look at the connection state and/or manipulate some lists at the
+ * same time, so we should hold the request lock anyways.
+ */
+enum drbd_req_state_bits {
+	/* 3210
+	 * 0000: no local possible
+	 * 0001: to be submitted
+	 *    UNUSED, we could map: 011: submitted, completion still pending
+	 * 0110: completed ok
+	 * 0010: completed with error
+	 * 1001: Aborted (before completion)
+	 * 1x10: Aborted and completed -> free
+	 */
+	__RQ_LOCAL_PENDING,
+	__RQ_LOCAL_COMPLETED,
+	__RQ_LOCAL_OK,
+	__RQ_LOCAL_ABORTED,
+
+	/* 87654
+	 * 00000: no network possible
+	 * 00001: to be send
+	 * 00011: to be send, on worker queue
+	 * 00101: sent, expecting recv_ack (B) or write_ack (C)
+	 * 11101: sent,
+	 *        recv_ack (B) or implicit "ack" (A),
+	 *        still waiting for the barrier ack.
+	 *        master_bio may already be completed and invalidated.
+	 * 11100: write acked (C),
+	 *        data received (for remote read, any protocol)
+	 *        or finally the barrier ack has arrived (B,A)...
+	 *        request can be freed
+	 * 01100: neg-acked (write, protocol C)
+	 *        or neg-d-acked (read, any protocol)
+	 *        or killed from the transfer log
+	 *        during cleanup after connection loss
+	 *        request can be freed
+	 * 01000: canceled or send failed...
+	 *        request can be freed
+	 */
+
+	/* if "SENT" is not set, yet, this can still fail or be canceled.
+	 * if "SENT" is set already, we still wait for an Ack packet.
+	 * when cleared, the master_bio may be completed.
+	 * in (B,A) the request object may still linger on the transaction log
+	 * until the corresponding barrier ack comes in */
+	__RQ_NET_PENDING,
+
+	/* If it is QUEUED, and it is a WRITE, it is also registered in the
+	 * transfer log. Currently we need this flag to avoid conflicts between
+	 * worker canceling the request and tl_clear_barrier killing it from
+	 * transfer log.  We should restructure the code so this conflict does
+	 * no longer occur. */
+	__RQ_NET_QUEUED,
+
+	/* well, actually only "handed over to the network stack".
+	 *
+	 * TODO can potentially be dropped because of the similar meaning
+	 * of RQ_NET_SENT and ~RQ_NET_QUEUED.
+	 * however it is not exactly the same. before we drop it
+	 * we must ensure that we can tell a request with network part
+	 * from a request without, regardless of what happens to it. */
+	__RQ_NET_SENT,
+
+	/* when set, the request may be freed (if RQ_NET_QUEUED is clear).
+	 * basically this means the corresponding P_BARRIER_ACK was received */
+	__RQ_NET_DONE,
+
+	/* whether or not we know (C) or pretend (B,A) that the write
+	 * was successfully written on the peer.
+	 */
+	__RQ_NET_OK,
+
+	/* peer called drbd_set_in_sync() for this write */
+	__RQ_NET_SIS,
+
+	/* keep this last, its for the RQ_NET_MASK */
+	__RQ_NET_MAX,
+
+	/* Set when this is a write, clear for a read */
+	__RQ_WRITE,
+
+	/* Should call drbd_al_complete_io() for this request... */
+	__RQ_IN_ACT_LOG,
+
+	/* The peer has sent a retry ACK */
+	__RQ_POSTPONED,
+
+	/* would have been completed,
+	 * but was not, because of drbd_suspended() */
+	__RQ_COMPLETION_SUSP,
+
+	/* We expect a receive ACK (wire proto B) */
+	__RQ_EXP_RECEIVE_ACK,
+
+	/* We expect a write ACK (wite proto C) */
+	__RQ_EXP_WRITE_ACK,
+
+	/* waiting for a barrier ack, did an extra kref_get */
+	__RQ_EXP_BARR_ACK,
+};
+
+#define RQ_LOCAL_PENDING   (1UL << __RQ_LOCAL_PENDING)
+#define RQ_LOCAL_COMPLETED (1UL << __RQ_LOCAL_COMPLETED)
+#define RQ_LOCAL_OK        (1UL << __RQ_LOCAL_OK)
+#define RQ_LOCAL_ABORTED   (1UL << __RQ_LOCAL_ABORTED)
+
+#define RQ_LOCAL_MASK      ((RQ_LOCAL_ABORTED << 1)-1)
+
+#define RQ_NET_PENDING     (1UL << __RQ_NET_PENDING)
+#define RQ_NET_QUEUED      (1UL << __RQ_NET_QUEUED)
+#define RQ_NET_SENT        (1UL << __RQ_NET_SENT)
+#define RQ_NET_DONE        (1UL << __RQ_NET_DONE)
+#define RQ_NET_OK          (1UL << __RQ_NET_OK)
+#define RQ_NET_SIS         (1UL << __RQ_NET_SIS)
+
+/* 0x1f8 */
+#define RQ_NET_MASK        (((1UL << __RQ_NET_MAX)-1) & ~RQ_LOCAL_MASK)
+
+#define RQ_WRITE           (1UL << __RQ_WRITE)
+#define RQ_IN_ACT_LOG      (1UL << __RQ_IN_ACT_LOG)
+#define RQ_POSTPONED	   (1UL << __RQ_POSTPONED)
+#define RQ_COMPLETION_SUSP (1UL << __RQ_COMPLETION_SUSP)
+#define RQ_EXP_RECEIVE_ACK (1UL << __RQ_EXP_RECEIVE_ACK)
+#define RQ_EXP_WRITE_ACK   (1UL << __RQ_EXP_WRITE_ACK)
+#define RQ_EXP_BARR_ACK    (1UL << __RQ_EXP_BARR_ACK)
+
+/* For waking up the frozen transfer log mod_req() has to return if the request
+   should be counted in the epoch object*/
+#define MR_WRITE       1
+#define MR_READ        2
+
+static inline void drbd_req_make_private_bio(struct drbd_request *req, struct bio *bio_src)
+{
+	struct bio *bio;
+	bio = bio_clone(bio_src, GFP_NOIO); /* XXX cannot fail?? */
+
+	req->private_bio = bio;
+
+	bio->bi_private  = req;
+	bio->bi_end_io   = drbd_request_endio;
+	bio->bi_next     = NULL;
+}
+
+/* Short lived temporary struct on the stack.
+ * We could squirrel the error to be returned into
+ * bio->bi_iter.bi_size, or similar. But that would be too ugly. */
+struct bio_and_error {
+	struct bio *bio;
+	int error;
+};
+
+extern void start_new_tl_epoch(struct drbd_connection *connection);
+extern void drbd_req_destroy(struct kref *kref);
+extern void _req_may_be_done(struct drbd_request *req,
+		struct bio_and_error *m);
+extern int __req_mod(struct drbd_request *req, enum drbd_req_event what,
+		struct bio_and_error *m);
+extern void complete_master_bio(struct drbd_device *device,
+		struct bio_and_error *m);
+extern void request_timer_fn(unsigned long data);
+extern void tl_restart(struct drbd_connection *connection, enum drbd_req_event what);
+extern void _tl_restart(struct drbd_connection *connection, enum drbd_req_event what);
+extern void tl_abort_disk_io(struct drbd_device *device);
+
+/* this is in drbd_main.c */
+extern void drbd_restart_request(struct drbd_request *req);
+
+/* use this if you don't want to deal with calling complete_master_bio()
+ * outside the spinlock, e.g. when walking some list on cleanup. */
+static inline int _req_mod(struct drbd_request *req, enum drbd_req_event what)
+{
+	struct drbd_device *device = req->device;
+	struct bio_and_error m;
+	int rv;
+
+	/* __req_mod possibly frees req, do not touch req after that! */
+	rv = __req_mod(req, what, &m);
+	if (m.bio)
+		complete_master_bio(device, &m);
+
+	return rv;
+}
+
+/* completion of master bio is outside of our spinlock.
+ * We still may or may not be inside some irqs disabled section
+ * of the lower level driver completion callback, so we need to
+ * spin_lock_irqsave here. */
+static inline int req_mod(struct drbd_request *req,
+		enum drbd_req_event what)
+{
+	unsigned long flags;
+	struct drbd_device *device = req->device;
+	struct bio_and_error m;
+	int rv;
+
+	spin_lock_irqsave(&device->resource->req_lock, flags);
+	rv = __req_mod(req, what, &m);
+	spin_unlock_irqrestore(&device->resource->req_lock, flags);
+
+	if (m.bio)
+		complete_master_bio(device, &m);
+
+	return rv;
+}
+
+static inline bool drbd_should_do_remote(union drbd_dev_state s)
+{
+	return s.pdsk == D_UP_TO_DATE ||
+		(s.pdsk >= D_INCONSISTENT &&
+		 s.conn >= C_WF_BITMAP_T &&
+		 s.conn < C_AHEAD);
+	/* Before proto 96 that was >= CONNECTED instead of >= C_WF_BITMAP_T.
+	   That is equivalent since before 96 IO was frozen in the C_WF_BITMAP*
+	   states. */
+}
+static inline bool drbd_should_send_out_of_sync(union drbd_dev_state s)
+{
+	return s.conn == C_AHEAD || s.conn == C_WF_BITMAP_S;
+	/* pdsk = D_INCONSISTENT as a consequence. Protocol 96 check not necessary
+	   since we enter state C_AHEAD only if proto >= 96 */
+}
+
+#endif

drivers/block/drbd/drbd_state.h

@@ -0,0 +1,161 @@
+#ifndef DRBD_STATE_H
+#define DRBD_STATE_H
+
+struct drbd_device;
+struct drbd_connection;
+
+/**
+ * DOC: DRBD State macros
+ *
+ * These macros are used to express state changes in easily readable form.
+ *
+ * The NS macros expand to a mask and a value, that can be bit ored onto the
+ * current state as soon as the spinlock (req_lock) was taken.
+ *
+ * The _NS macros are used for state functions that get called with the
+ * spinlock. These macros expand directly to the new state value.
+ *
+ * Besides the basic forms NS() and _NS() additional _?NS[23] are defined
+ * to express state changes that affect more than one aspect of the state.
+ *
+ * E.g. NS2(conn, C_CONNECTED, peer, R_SECONDARY)
+ * Means that the network connection was established and that the peer
+ * is in secondary role.
+ */
+#define role_MASK R_MASK
+#define peer_MASK R_MASK
+#define disk_MASK D_MASK
+#define pdsk_MASK D_MASK
+#define conn_MASK C_MASK
+#define susp_MASK 1
+#define user_isp_MASK 1
+#define aftr_isp_MASK 1
+#define susp_nod_MASK 1
+#define susp_fen_MASK 1
+
+#define NS(T, S) \
+	({ union drbd_state mask; mask.i = 0; mask.T = T##_MASK; mask; }), \
+	({ union drbd_state val; val.i = 0; val.T = (S); val; })
+#define NS2(T1, S1, T2, S2) \
+	({ union drbd_state mask; mask.i = 0; mask.T1 = T1##_MASK; \
+	  mask.T2 = T2##_MASK; mask; }), \
+	({ union drbd_state val; val.i = 0; val.T1 = (S1); \
+	  val.T2 = (S2); val; })
+#define NS3(T1, S1, T2, S2, T3, S3) \
+	({ union drbd_state mask; mask.i = 0; mask.T1 = T1##_MASK; \
+	  mask.T2 = T2##_MASK; mask.T3 = T3##_MASK; mask; }), \
+	({ union drbd_state val;  val.i = 0; val.T1 = (S1); \
+	  val.T2 = (S2); val.T3 = (S3); val; })
+
+#define _NS(D, T, S) \
+	D, ({ union drbd_state __ns; __ns = drbd_read_state(D); __ns.T = (S); __ns; })
+#define _NS2(D, T1, S1, T2, S2) \
+	D, ({ union drbd_state __ns; __ns = drbd_read_state(D); __ns.T1 = (S1); \
+	__ns.T2 = (S2); __ns; })
+#define _NS3(D, T1, S1, T2, S2, T3, S3) \
+	D, ({ union drbd_state __ns; __ns = drbd_read_state(D); __ns.T1 = (S1); \
+	__ns.T2 = (S2); __ns.T3 = (S3); __ns; })
+
+enum chg_state_flags {
+	CS_HARD	         = 1 << 0,
+	CS_VERBOSE       = 1 << 1,
+	CS_WAIT_COMPLETE = 1 << 2,
+	CS_SERIALIZE     = 1 << 3,
+	CS_ORDERED       = CS_WAIT_COMPLETE + CS_SERIALIZE,
+	CS_LOCAL_ONLY    = 1 << 4, /* Do not consider a device pair wide state change */
+	CS_DC_ROLE       = 1 << 5, /* DC = display as connection state change */
+	CS_DC_PEER       = 1 << 6,
+	CS_DC_CONN       = 1 << 7,
+	CS_DC_DISK       = 1 << 8,
+	CS_DC_PDSK       = 1 << 9,
+	CS_DC_SUSP       = 1 << 10,
+	CS_DC_MASK       = CS_DC_ROLE + CS_DC_PEER + CS_DC_CONN + CS_DC_DISK + CS_DC_PDSK,
+	CS_IGN_OUTD_FAIL = 1 << 11,
+};
+
+/* drbd_dev_state and drbd_state are different types. This is to stress the
+   small difference. There is no suspended flag (.susp), and no suspended
+   while fence handler runs flas (susp_fen). */
+union drbd_dev_state {
+	struct {
+#if defined(__LITTLE_ENDIAN_BITFIELD)
+		unsigned role:2 ;   /* 3/4	 primary/secondary/unknown */
+		unsigned peer:2 ;   /* 3/4	 primary/secondary/unknown */
+		unsigned conn:5 ;   /* 17/32	 cstates */
+		unsigned disk:4 ;   /* 8/16	 from D_DISKLESS to D_UP_TO_DATE */
+		unsigned pdsk:4 ;   /* 8/16	 from D_DISKLESS to D_UP_TO_DATE */
+		unsigned _unused:1 ;
+		unsigned aftr_isp:1 ; /* isp .. imposed sync pause */
+		unsigned peer_isp:1 ;
+		unsigned user_isp:1 ;
+		unsigned _pad:11;   /* 0	 unused */
+#elif defined(__BIG_ENDIAN_BITFIELD)
+		unsigned _pad:11;
+		unsigned user_isp:1 ;
+		unsigned peer_isp:1 ;
+		unsigned aftr_isp:1 ; /* isp .. imposed sync pause */
+		unsigned _unused:1 ;
+		unsigned pdsk:4 ;   /* 8/16	 from D_DISKLESS to D_UP_TO_DATE */
+		unsigned disk:4 ;   /* 8/16	 from D_DISKLESS to D_UP_TO_DATE */
+		unsigned conn:5 ;   /* 17/32	 cstates */
+		unsigned peer:2 ;   /* 3/4	 primary/secondary/unknown */
+		unsigned role:2 ;   /* 3/4	 primary/secondary/unknown */
+#else
+# error "this endianess is not supported"
+#endif
+	};
+	unsigned int i;
+};
+
+extern enum drbd_state_rv drbd_change_state(struct drbd_device *device,
+					    enum chg_state_flags f,
+					    union drbd_state mask,
+					    union drbd_state val);
+extern void drbd_force_state(struct drbd_device *, union drbd_state,
+			union drbd_state);
+extern enum drbd_state_rv _drbd_request_state(struct drbd_device *,
+					      union drbd_state,
+					      union drbd_state,
+					      enum chg_state_flags);
+extern enum drbd_state_rv __drbd_set_state(struct drbd_device *, union drbd_state,
+					   enum chg_state_flags,
+					   struct completion *done);
+extern void print_st_err(struct drbd_device *, union drbd_state,
+			union drbd_state, int);
+
+enum drbd_state_rv
+_conn_request_state(struct drbd_connection *connection, union drbd_state mask, union drbd_state val,
+		    enum chg_state_flags flags);
+
+enum drbd_state_rv
+conn_request_state(struct drbd_connection *connection, union drbd_state mask, union drbd_state val,
+		   enum chg_state_flags flags);
+
+extern void drbd_resume_al(struct drbd_device *device);
+extern bool conn_all_vols_unconf(struct drbd_connection *connection);
+
+/**
+ * drbd_request_state() - Reqest a state change
+ * @device:	DRBD device.
+ * @mask:	mask of state bits to change.
+ * @val:	value of new state bits.
+ *
+ * This is the most graceful way of requesting a state change. It is verbose
+ * quite verbose in case the state change is not possible, and all those
+ * state changes are globally serialized.
+ */
+static inline int drbd_request_state(struct drbd_device *device,
+				     union drbd_state mask,
+				     union drbd_state val)
+{
+	return _drbd_request_state(device, mask, val, CS_VERBOSE + CS_ORDERED);
+}
+
+enum drbd_role conn_highest_role(struct drbd_connection *connection);
+enum drbd_role conn_highest_peer(struct drbd_connection *connection);
+enum drbd_disk_state conn_highest_disk(struct drbd_connection *connection);
+enum drbd_disk_state conn_lowest_disk(struct drbd_connection *connection);
+enum drbd_disk_state conn_highest_pdsk(struct drbd_connection *connection);
+enum drbd_conns conn_lowest_conn(struct drbd_connection *connection);
+
+#endif

drivers/block/drbd/drbd_strings.c

@@ -0,0 +1,118 @@
+/*
+  drbd.h
+
+  This file is part of DRBD by Philipp Reisner and Lars Ellenberg.
+
+  Copyright (C) 2003-2008, LINBIT Information Technologies GmbH.
+  Copyright (C) 2003-2008, Philipp Reisner <philipp.reisner@linbit.com>.
+  Copyright (C) 2003-2008, Lars Ellenberg <lars.ellenberg@linbit.com>.
+
+  drbd is free software; you can redistribute it and/or modify
+  it under the terms of the GNU General Public License as published by
+  the Free Software Foundation; either version 2, or (at your option)
+  any later version.
+
+  drbd is distributed in the hope that it will be useful,
+  but WITHOUT ANY WARRANTY; without even the implied warranty of
+  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+  GNU General Public License for more details.
+
+  You should have received a copy of the GNU General Public License
+  along with drbd; see the file COPYING.  If not, write to
+  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+
+*/
+
+#include <linux/drbd.h>
+#include "drbd_strings.h"
+
+static const char *drbd_conn_s_names[] = {
+	[C_STANDALONE]       = "StandAlone",
+	[C_DISCONNECTING]    = "Disconnecting",
+	[C_UNCONNECTED]      = "Unconnected",
+	[C_TIMEOUT]          = "Timeout",
+	[C_BROKEN_PIPE]      = "BrokenPipe",
+	[C_NETWORK_FAILURE]  = "NetworkFailure",
+	[C_PROTOCOL_ERROR]   = "ProtocolError",
+	[C_WF_CONNECTION]    = "WFConnection",
+	[C_WF_REPORT_PARAMS] = "WFReportParams",
+	[C_TEAR_DOWN]        = "TearDown",
+	[C_CONNECTED]        = "Connected",
+	[C_STARTING_SYNC_S]  = "StartingSyncS",
+	[C_STARTING_SYNC_T]  = "StartingSyncT",
+	[C_WF_BITMAP_S]      = "WFBitMapS",
+	[C_WF_BITMAP_T]      = "WFBitMapT",
+	[C_WF_SYNC_UUID]     = "WFSyncUUID",
+	[C_SYNC_SOURCE]      = "SyncSource",
+	[C_SYNC_TARGET]      = "SyncTarget",
+	[C_PAUSED_SYNC_S]    = "PausedSyncS",
+	[C_PAUSED_SYNC_T]    = "PausedSyncT",
+	[C_VERIFY_S]         = "VerifyS",
+	[C_VERIFY_T]         = "VerifyT",
+	[C_AHEAD]            = "Ahead",
+	[C_BEHIND]           = "Behind",
+};
+
+static const char *drbd_role_s_names[] = {
+	[R_PRIMARY]   = "Primary",
+	[R_SECONDARY] = "Secondary",
+	[R_UNKNOWN]   = "Unknown"
+};
+
+static const char *drbd_disk_s_names[] = {
+	[D_DISKLESS]     = "Diskless",
+	[D_ATTACHING]    = "Attaching",
+	[D_FAILED]       = "Failed",
+	[D_NEGOTIATING]  = "Negotiating",
+	[D_INCONSISTENT] = "Inconsistent",
+	[D_OUTDATED]     = "Outdated",
+	[D_UNKNOWN]      = "DUnknown",
+	[D_CONSISTENT]   = "Consistent",
+	[D_UP_TO_DATE]   = "UpToDate",
+};
+
+static const char *drbd_state_sw_errors[] = {
+	[-SS_TWO_PRIMARIES] = "Multiple primaries not allowed by config",
+	[-SS_NO_UP_TO_DATE_DISK] = "Need access to UpToDate data",
+	[-SS_NO_LOCAL_DISK] = "Can not resync without local disk",
+	[-SS_NO_REMOTE_DISK] = "Can not resync without remote disk",
+	[-SS_CONNECTED_OUTDATES] = "Refusing to be Outdated while Connected",
+	[-SS_PRIMARY_NOP] = "Refusing to be Primary while peer is not outdated",
+	[-SS_RESYNC_RUNNING] = "Can not start OV/resync since it is already active",
+	[-SS_ALREADY_STANDALONE] = "Can not disconnect a StandAlone device",
+	[-SS_CW_FAILED_BY_PEER] = "State change was refused by peer node",
+	[-SS_IS_DISKLESS] = "Device is diskless, the requested operation requires a disk",
+	[-SS_DEVICE_IN_USE] = "Device is held open by someone",
+	[-SS_NO_NET_CONFIG] = "Have no net/connection configuration",
+	[-SS_NO_VERIFY_ALG] = "Need a verify algorithm to start online verify",
+	[-SS_NEED_CONNECTION] = "Need a connection to start verify or resync",
+	[-SS_NOT_SUPPORTED] = "Peer does not support protocol",
+	[-SS_LOWER_THAN_OUTDATED] = "Disk state is lower than outdated",
+	[-SS_IN_TRANSIENT_STATE] = "In transient state, retry after next state change",
+	[-SS_CONCURRENT_ST_CHG] = "Concurrent state changes detected and aborted",
+	[-SS_OUTDATE_WO_CONN] = "Need a connection for a graceful disconnect/outdate peer",
+	[-SS_O_VOL_PEER_PRI] = "Other vol primary on peer not allowed by config",
+};
+
+const char *drbd_conn_str(enum drbd_conns s)
+{
+	/* enums are unsigned... */
+	return s > C_BEHIND ? "TOO_LARGE" : drbd_conn_s_names[s];
+}
+
+const char *drbd_role_str(enum drbd_role s)
+{
+	return s > R_SECONDARY   ? "TOO_LARGE" : drbd_role_s_names[s];
+}
+
+const char *drbd_disk_str(enum drbd_disk_state s)
+{
+	return s > D_UP_TO_DATE    ? "TOO_LARGE" : drbd_disk_s_names[s];
+}
+
+const char *drbd_set_st_err_str(enum drbd_state_rv err)
+{
+	return err <= SS_AFTER_LAST_ERROR ? "TOO_SMALL" :
+	       err > SS_TWO_PRIMARIES ? "TOO_LARGE"
+			: drbd_state_sw_errors[-err];
+}

drivers/block/drbd/drbd_strings.h

@@ -0,0 +1,9 @@
+#ifndef __DRBD_STRINGS_H
+#define __DRBD_STRINGS_H
+
+extern const char *drbd_conn_str(enum drbd_conns);
+extern const char *drbd_role_str(enum drbd_role);
+extern const char *drbd_disk_str(enum drbd_disk_state);
+extern const char *drbd_set_st_err_str(enum drbd_state_rv);
+
+#endif  /* __DRBD_STRINGS_H */

drivers/block/drbd/drbd_vli.h

@@ -0,0 +1,351 @@
+/*
+-*- linux-c -*-
+   drbd_receiver.c
+   This file is part of DRBD by Philipp Reisner and Lars Ellenberg.
+
+   Copyright (C) 2001-2008, LINBIT Information Technologies GmbH.
+   Copyright (C) 1999-2008, Philipp Reisner <philipp.reisner@linbit.com>.
+   Copyright (C) 2002-2008, Lars Ellenberg <lars.ellenberg@linbit.com>.
+
+   drbd is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation; either version 2, or (at your option)
+   any later version.
+
+   drbd is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with drbd; see the file COPYING.  If not, write to
+   the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+
+#ifndef _DRBD_VLI_H
+#define _DRBD_VLI_H
+
+/*
+ * At a granularity of 4KiB storage represented per bit,
+ * and stroage sizes of several TiB,
+ * and possibly small-bandwidth replication,
+ * the bitmap transfer time can take much too long,
+ * if transmitted in plain text.
+ *
+ * We try to reduce the transferred bitmap information
+ * by encoding runlengths of bit polarity.
+ *
+ * We never actually need to encode a "zero" (runlengths are positive).
+ * But then we have to store the value of the first bit.
+ * The first bit of information thus shall encode if the first runlength
+ * gives the number of set or unset bits.
+ *
+ * We assume that large areas are either completely set or unset,
+ * which gives good compression with any runlength method,
+ * even when encoding the runlength as fixed size 32bit/64bit integers.
+ *
+ * Still, there may be areas where the polarity flips every few bits,
+ * and encoding the runlength sequence of those areas with fix size
+ * integers would be much worse than plaintext.
+ *
+ * We want to encode small runlength values with minimum code length,
+ * while still being able to encode a Huge run of all zeros.
+ *
+ * Thus we need a Variable Length Integer encoding, VLI.
+ *
+ * For some cases, we produce more code bits than plaintext input.
+ * We need to send incompressible chunks as plaintext, skip over them
+ * and then see if the next chunk compresses better.
+ *
+ * We don't care too much about "excellent" compression ratio for large
+ * runlengths (all set/all clear): whether we achieve a factor of 100
+ * or 1000 is not that much of an issue.
+ * We do not want to waste too much on short runlengths in the "noisy"
+ * parts of the bitmap, though.
+ *
+ * There are endless variants of VLI, we experimented with:
+ *  * simple byte-based
+ *  * various bit based with different code word length.
+ *
+ * To avoid yet an other configuration parameter (choice of bitmap compression
+ * algorithm) which was difficult to explain and tune, we just chose the one
+ * variant that turned out best in all test cases.
+ * Based on real world usage patterns, with device sizes ranging from a few GiB
+ * to several TiB, file server/mailserver/webserver/mysql/postgress,
+ * mostly idle to really busy, the all time winner (though sometimes only
+ * marginally better) is:
+ */
+
+/*
+ * encoding is "visualised" as
+ * __little endian__ bitstream, least significant bit first (left most)
+ *
+ * this particular encoding is chosen so that the prefix code
+ * starts as unary encoding the level, then modified so that
+ * 10 levels can be described in 8bit, with minimal overhead
+ * for the smaller levels.
+ *
+ * Number of data bits follow fibonacci sequence, with the exception of the
+ * last level (+1 data bit, so it makes 64bit total).  The only worse code when
+ * encoding bit polarity runlength is 1 plain bits => 2 code bits.
+prefix    data bits                                    max val  Nº data bits
+0 x                                                         0x2            1
+10 x                                                        0x4            1
+110 xx                                                      0x8            2
+1110 xxx                                                   0x10            3
+11110 xxx xx                                               0x30            5
+111110 xx xxxxxx                                          0x130            8
+11111100  xxxxxxxx xxxxx                                 0x2130           13
+11111110  xxxxxxxx xxxxxxxx xxxxx                      0x202130           21
+11111101  xxxxxxxx xxxxxxxx xxxxxxxx  xxxxxxxx xx   0x400202130           34
+11111111  xxxxxxxx xxxxxxxx xxxxxxxx  xxxxxxxx xxxxxxxx xxxxxxxx xxxxxxxx 56
+ * maximum encodable value: 0x100000400202130 == 2**56 + some */
+
+/* compression "table":
+ transmitted   x                                0.29
+ as plaintext x                                  ........................
+             x                                   ........................
+            x                                    ........................
+           x    0.59                         0.21........................
+          x      ........................................................
+         x       .. c ...................................................
+        x    0.44.. o ...................................................
+       x .......... d ...................................................
+      x  .......... e ...................................................
+     X.............   ...................................................
+    x.............. b ...................................................
+2.0x............... i ...................................................
+ #X................ t ...................................................
+ #................. s ...........................  plain bits  ..........
+-+-----------------------------------------------------------------------
+ 1             16              32                              64
+*/
+
+/* LEVEL: (total bits, prefix bits, prefix value),
+ * sorted ascending by number of total bits.
+ * The rest of the code table is calculated at compiletime from this. */
+
+/* fibonacci data 1, 1, ... */
+#define VLI_L_1_1() do { \
+	LEVEL( 2, 1, 0x00); \
+	LEVEL( 3, 2, 0x01); \
+	LEVEL( 5, 3, 0x03); \
+	LEVEL( 7, 4, 0x07); \
+	LEVEL(10, 5, 0x0f); \
+	LEVEL(14, 6, 0x1f); \
+	LEVEL(21, 8, 0x3f); \
+	LEVEL(29, 8, 0x7f); \
+	LEVEL(42, 8, 0xbf); \
+	LEVEL(64, 8, 0xff); \
+	} while (0)
+
+/* finds a suitable level to decode the least significant part of in.
+ * returns number of bits consumed.
+ *
+ * BUG() for bad input, as that would mean a buggy code table. */
+static inline int vli_decode_bits(u64 *out, const u64 in)
+{
+	u64 adj = 1;
+
+#define LEVEL(t,b,v)					\
+	do {						\
+		if ((in & ((1 << b) -1)) == v) {	\
+			*out = ((in & ((~0ULL) >> (64-t))) >> b) + adj;	\
+			return t;			\
+		}					\
+		adj += 1ULL << (t - b);			\
+	} while (0)
+
+	VLI_L_1_1();
+
+	/* NOT REACHED, if VLI_LEVELS code table is defined properly */
+	BUG();
+#undef LEVEL
+}
+
+/* return number of code bits needed,
+ * or negative error number */
+static inline int __vli_encode_bits(u64 *out, const u64 in)
+{
+	u64 max = 0;
+	u64 adj = 1;
+
+	if (in == 0)
+		return -EINVAL;
+
+#define LEVEL(t,b,v) do {		\
+		max += 1ULL << (t - b);	\
+		if (in <= max) {	\
+			if (out)	\
+				*out = ((in - adj) << b) | v;	\
+			return t;	\
+		}			\
+		adj = max + 1;		\
+	} while (0)
+
+	VLI_L_1_1();
+
+	return -EOVERFLOW;
+#undef LEVEL
+}
+
+#undef VLI_L_1_1
+
+/* code from here down is independend of actually used bit code */
+
+/*
+ * Code length is determined by some unique (e.g. unary) prefix.
+ * This encodes arbitrary bit length, not whole bytes: we have a bit-stream,
+ * not a byte stream.
+ */
+
+/* for the bitstream, we need a cursor */
+struct bitstream_cursor {
+	/* the current byte */
+	u8 *b;
+	/* the current bit within *b, nomalized: 0..7 */
+	unsigned int bit;
+};
+
+/* initialize cursor to point to first bit of stream */
+static inline void bitstream_cursor_reset(struct bitstream_cursor *cur, void *s)
+{
+	cur->b = s;
+	cur->bit = 0;
+}
+
+/* advance cursor by that many bits; maximum expected input value: 64,
+ * but depending on VLI implementation, it may be more. */
+static inline void bitstream_cursor_advance(struct bitstream_cursor *cur, unsigned int bits)
+{
+	bits += cur->bit;
+	cur->b = cur->b + (bits >> 3);
+	cur->bit = bits & 7;
+}
+
+/* the bitstream itself knows its length */
+struct bitstream {
+	struct bitstream_cursor cur;
+	unsigned char *buf;
+	size_t buf_len;		/* in bytes */
+
+	/* for input stream:
+	 * number of trailing 0 bits for padding
+	 * total number of valid bits in stream: buf_len * 8 - pad_bits */
+	unsigned int pad_bits;
+};
+
+static inline void bitstream_init(struct bitstream *bs, void *s, size_t len, unsigned int pad_bits)
+{
+	bs->buf = s;
+	bs->buf_len = len;
+	bs->pad_bits = pad_bits;
+	bitstream_cursor_reset(&bs->cur, bs->buf);
+}
+
+static inline void bitstream_rewind(struct bitstream *bs)
+{
+	bitstream_cursor_reset(&bs->cur, bs->buf);
+	memset(bs->buf, 0, bs->buf_len);
+}
+
+/* Put (at most 64) least significant bits of val into bitstream, and advance cursor.
+ * Ignores "pad_bits".
+ * Returns zero if bits == 0 (nothing to do).
+ * Returns number of bits used if successful.
+ *
+ * If there is not enough room left in bitstream,
+ * leaves bitstream unchanged and returns -ENOBUFS.
+ */
+static inline int bitstream_put_bits(struct bitstream *bs, u64 val, const unsigned int bits)
+{
+	unsigned char *b = bs->cur.b;
+	unsigned int tmp;
+
+	if (bits == 0)
+		return 0;
+
+	if ((bs->cur.b + ((bs->cur.bit + bits -1) >> 3)) - bs->buf >= bs->buf_len)
+		return -ENOBUFS;
+
+	/* paranoia: strip off hi bits; they should not be set anyways. */
+	if (bits < 64)
+		val &= ~0ULL >> (64 - bits);
+
+	*b++ |= (val & 0xff) << bs->cur.bit;
+
+	for (tmp = 8 - bs->cur.bit; tmp < bits; tmp += 8)
+		*b++ |= (val >> tmp) & 0xff;
+
+	bitstream_cursor_advance(&bs->cur, bits);
+	return bits;
+}
+
+/* Fetch (at most 64) bits from bitstream into *out, and advance cursor.
+ *
+ * If more than 64 bits are requested, returns -EINVAL and leave *out unchanged.
+ *
+ * If there are less than the requested number of valid bits left in the
+ * bitstream, still fetches all available bits.
+ *
+ * Returns number of actually fetched bits.
+ */
+static inline int bitstream_get_bits(struct bitstream *bs, u64 *out, int bits)
+{
+	u64 val;
+	unsigned int n;
+
+	if (bits > 64)
+		return -EINVAL;
+
+	if (bs->cur.b + ((bs->cur.bit + bs->pad_bits + bits -1) >> 3) - bs->buf >= bs->buf_len)
+		bits = ((bs->buf_len - (bs->cur.b - bs->buf)) << 3)
+			- bs->cur.bit - bs->pad_bits;
+
+	if (bits == 0) {
+		*out = 0;
+		return 0;
+	}
+
+	/* get the high bits */
+	val = 0;
+	n = (bs->cur.bit + bits + 7) >> 3;
+	/* n may be at most 9, if cur.bit + bits > 64 */
+	/* which means this copies at most 8 byte */
+	if (n) {
+		memcpy(&val, bs->cur.b+1, n - 1);
+		val = le64_to_cpu(val) << (8 - bs->cur.bit);
+	}
+
+	/* we still need the low bits */
+	val |= bs->cur.b[0] >> bs->cur.bit;
+
+	/* and mask out bits we don't want */
+	val &= ~0ULL >> (64 - bits);
+
+	bitstream_cursor_advance(&bs->cur, bits);
+	*out = val;
+
+	return bits;
+}
+
+/* encodes @in as vli into @bs;
+
+ * return values
+ *  > 0: number of bits successfully stored in bitstream
+ * -ENOBUFS @bs is full
+ * -EINVAL input zero (invalid)
+ * -EOVERFLOW input too large for this vli code (invalid)
+ */
+static inline int vli_encode_bits(struct bitstream *bs, u64 in)
+{
+	u64 code = code;
+	int bits = __vli_encode_bits(&code, in);
+
+	if (bits <= 0)
+		return bits;
+
+	return bitstream_put_bits(bs, code, bits);
+}
+
+#endif

drivers/block/hd.c

@@ -0,0 +1,804 @@
+/*
+ *  Copyright (C) 1991, 1992  Linus Torvalds
+ *
+ * This is the low-level hd interrupt support. It traverses the
+ * request-list, using interrupts to jump between functions. As
+ * all the functions are called within interrupts, we may not
+ * sleep. Special care is recommended.
+ *
+ *  modified by Drew Eckhardt to check nr of hd's from the CMOS.
+ *
+ *  Thanks to Branko Lankester, lankeste@fwi.uva.nl, who found a bug
+ *  in the early extended-partition checks and added DM partitions
+ *
+ *  IRQ-unmask, drive-id, multiple-mode, support for ">16 heads",
+ *  and general streamlining by Mark Lord.
+ *
+ *  Removed 99% of above. Use Mark's ide driver for those options.
+ *  This is now a lightweight ST-506 driver. (Paul Gortmaker)
+ *
+ *  Modified 1995 Russell King for ARM processor.
+ *
+ *  Bugfix: max_sectors must be <= 255 or the wheels tend to come
+ *  off in a hurry once you queue things up - Paul G. 02/2001
+ */
+
+/* Uncomment the following if you want verbose error reports. */
+/* #define VERBOSE_ERRORS */
+
+#include <linux/blkdev.h>
+#include <linux/errno.h>
+#include <linux/signal.h>
+#include <linux/interrupt.h>
+#include <linux/timer.h>
+#include <linux/fs.h>
+#include <linux/kernel.h>
+#include <linux/genhd.h>
+#include <linux/string.h>
+#include <linux/ioport.h>
+#include <linux/init.h>
+#include <linux/blkpg.h>
+#include <linux/ata.h>
+#include <linux/hdreg.h>
+
+#define HD_IRQ 14
+
+#define REALLY_SLOW_IO
+#include <asm/io.h>
+#include <asm/uaccess.h>
+
+#ifdef __arm__
+#undef  HD_IRQ
+#endif
+#include <asm/irq.h>
+#ifdef __arm__
+#define HD_IRQ IRQ_HARDDISK
+#endif
+
+/* Hd controller regster ports */
+
+#define HD_DATA		0x1f0		/* _CTL when writing */
+#define HD_ERROR	0x1f1		/* see err-bits */
+#define HD_NSECTOR	0x1f2		/* nr of sectors to read/write */
+#define HD_SECTOR	0x1f3		/* starting sector */
+#define HD_LCYL		0x1f4		/* starting cylinder */
+#define HD_HCYL		0x1f5		/* high byte of starting cyl */
+#define HD_CURRENT	0x1f6		/* 101dhhhh , d=drive, hhhh=head */
+#define HD_STATUS	0x1f7		/* see status-bits */
+#define HD_FEATURE	HD_ERROR	/* same io address, read=error, write=feature */
+#define HD_PRECOMP	HD_FEATURE	/* obsolete use of this port - predates IDE */
+#define HD_COMMAND	HD_STATUS	/* same io address, read=status, write=cmd */
+
+#define HD_CMD		0x3f6		/* used for resets */
+#define HD_ALTSTATUS	0x3f6		/* same as HD_STATUS but doesn't clear irq */
+
+/* Bits of HD_STATUS */
+#define ERR_STAT		0x01
+#define INDEX_STAT		0x02
+#define ECC_STAT		0x04	/* Corrected error */
+#define DRQ_STAT		0x08
+#define SEEK_STAT		0x10
+#define SERVICE_STAT		SEEK_STAT
+#define WRERR_STAT		0x20
+#define READY_STAT		0x40
+#define BUSY_STAT		0x80
+
+/* Bits for HD_ERROR */
+#define MARK_ERR		0x01	/* Bad address mark */
+#define TRK0_ERR		0x02	/* couldn't find track 0 */
+#define ABRT_ERR		0x04	/* Command aborted */
+#define MCR_ERR			0x08	/* media change request */
+#define ID_ERR			0x10	/* ID field not found */
+#define MC_ERR			0x20	/* media changed */
+#define ECC_ERR			0x40	/* Uncorrectable ECC error */
+#define BBD_ERR			0x80	/* pre-EIDE meaning:  block marked bad */
+#define ICRC_ERR		0x80	/* new meaning:  CRC error during transfer */
+
+static DEFINE_SPINLOCK(hd_lock);
+static struct request_queue *hd_queue;
+static struct request *hd_req;
+
+#define TIMEOUT_VALUE	(6*HZ)
+#define	HD_DELAY	0
+
+#define MAX_ERRORS     16	/* Max read/write errors/sector */
+#define RESET_FREQ      8	/* Reset controller every 8th retry */
+#define RECAL_FREQ      4	/* Recalibrate every 4th retry */
+#define MAX_HD		2
+
+#define STAT_OK		(READY_STAT|SEEK_STAT)
+#define OK_STATUS(s)	(((s)&(STAT_OK|(BUSY_STAT|WRERR_STAT|ERR_STAT)))==STAT_OK)
+
+static void recal_intr(void);
+static void bad_rw_intr(void);
+
+static int reset;
+static int hd_error;
+
+/*
+ *  This struct defines the HD's and their types.
+ */
+struct hd_i_struct {
+	unsigned int head, sect, cyl, wpcom, lzone, ctl;
+	int unit;
+	int recalibrate;
+	int special_op;
+};
+
+#ifdef HD_TYPE
+static struct hd_i_struct hd_info[] = { HD_TYPE };
+static int NR_HD = ARRAY_SIZE(hd_info);
+#else
+static struct hd_i_struct hd_info[MAX_HD];
+static int NR_HD;
+#endif
+
+static struct gendisk *hd_gendisk[MAX_HD];
+
+static struct timer_list device_timer;
+
+#define TIMEOUT_VALUE (6*HZ)
+
+#define SET_TIMER							\
+	do {								\
+		mod_timer(&device_timer, jiffies + TIMEOUT_VALUE);	\
+	} while (0)
+
+static void (*do_hd)(void) = NULL;
+#define SET_HANDLER(x) \
+if ((do_hd = (x)) != NULL) \
+	SET_TIMER; \
+else \
+	del_timer(&device_timer);
+
+
+#if (HD_DELAY > 0)
+
+#include <linux/i8253.h>
+
+unsigned long last_req;
+
+unsigned long read_timer(void)
+{
+	unsigned long t, flags;
+	int i;
+
+	raw_spin_lock_irqsave(&i8253_lock, flags);
+	t = jiffies * 11932;
+	outb_p(0, 0x43);
+	i = inb_p(0x40);
+	i |= inb(0x40) << 8;
+	raw_spin_unlock_irqrestore(&i8253_lock, flags);
+	return(t - i);
+}
+#endif
+
+static void __init hd_setup(char *str, int *ints)
+{
+	int hdind = 0;
+
+	if (ints[0] != 3)
+		return;
+	if (hd_info[0].head != 0)
+		hdind = 1;
+	hd_info[hdind].head = ints[2];
+	hd_info[hdind].sect = ints[3];
+	hd_info[hdind].cyl = ints[1];
+	hd_info[hdind].wpcom = 0;
+	hd_info[hdind].lzone = ints[1];
+	hd_info[hdind].ctl = (ints[2] > 8 ? 8 : 0);
+	NR_HD = hdind+1;
+}
+
+static bool hd_end_request(int err, unsigned int bytes)
+{
+	if (__blk_end_request(hd_req, err, bytes))
+		return true;
+	hd_req = NULL;
+	return false;
+}
+
+static bool hd_end_request_cur(int err)
+{
+	return hd_end_request(err, blk_rq_cur_bytes(hd_req));
+}
+
+static void dump_status(const char *msg, unsigned int stat)
+{
+	char *name = "hd?";
+	if (hd_req)
+		name = hd_req->rq_disk->disk_name;
+
+#ifdef VERBOSE_ERRORS
+	printk("%s: %s: status=0x%02x { ", name, msg, stat & 0xff);
+	if (stat & BUSY_STAT)	printk("Busy ");
+	if (stat & READY_STAT)	printk("DriveReady ");
+	if (stat & WRERR_STAT)	printk("WriteFault ");
+	if (stat & SEEK_STAT)	printk("SeekComplete ");
+	if (stat & DRQ_STAT)	printk("DataRequest ");
+	if (stat & ECC_STAT)	printk("CorrectedError ");
+	if (stat & INDEX_STAT)	printk("Index ");
+	if (stat & ERR_STAT)	printk("Error ");
+	printk("}\n");
+	if ((stat & ERR_STAT) == 0) {
+		hd_error = 0;
+	} else {
+		hd_error = inb(HD_ERROR);
+		printk("%s: %s: error=0x%02x { ", name, msg, hd_error & 0xff);
+		if (hd_error & BBD_ERR)		printk("BadSector ");
+		if (hd_error & ECC_ERR)		printk("UncorrectableError ");
+		if (hd_error & ID_ERR)		printk("SectorIdNotFound ");
+		if (hd_error & ABRT_ERR)	printk("DriveStatusError ");
+		if (hd_error & TRK0_ERR)	printk("TrackZeroNotFound ");
+		if (hd_error & MARK_ERR)	printk("AddrMarkNotFound ");
+		printk("}");
+		if (hd_error & (BBD_ERR|ECC_ERR|ID_ERR|MARK_ERR)) {
+			printk(", CHS=%d/%d/%d", (inb(HD_HCYL)<<8) + inb(HD_LCYL),
+				inb(HD_CURRENT) & 0xf, inb(HD_SECTOR));
+			if (hd_req)
+				printk(", sector=%ld", blk_rq_pos(hd_req));
+		}
+		printk("\n");
+	}
+#else
+	printk("%s: %s: status=0x%02x.\n", name, msg, stat & 0xff);
+	if ((stat & ERR_STAT) == 0) {
+		hd_error = 0;
+	} else {
+		hd_error = inb(HD_ERROR);
+		printk("%s: %s: error=0x%02x.\n", name, msg, hd_error & 0xff);
+	}
+#endif
+}
+
+static void check_status(void)
+{
+	int i = inb_p(HD_STATUS);
+
+	if (!OK_STATUS(i)) {
+		dump_status("check_status", i);
+		bad_rw_intr();
+	}
+}
+
+static int controller_busy(void)
+{
+	int retries = 100000;
+	unsigned char status;
+
+	do {
+		status = inb_p(HD_STATUS);
+	} while ((status & BUSY_STAT) && --retries);
+	return status;
+}
+
+static int status_ok(void)
+{
+	unsigned char status = inb_p(HD_STATUS);
+
+	if (status & BUSY_STAT)
+		return 1;	/* Ancient, but does it make sense??? */
+	if (status & WRERR_STAT)
+		return 0;
+	if (!(status & READY_STAT))
+		return 0;
+	if (!(status & SEEK_STAT))
+		return 0;
+	return 1;
+}
+
+static int controller_ready(unsigned int drive, unsigned int head)
+{
+	int retry = 100;
+
+	do {
+		if (controller_busy() & BUSY_STAT)
+			return 0;
+		outb_p(0xA0 | (drive<<4) | head, HD_CURRENT);
+		if (status_ok())
+			return 1;
+	} while (--retry);
+	return 0;
+}
+
+static void hd_out(struct hd_i_struct *disk,
+		   unsigned int nsect,
+		   unsigned int sect,
+		   unsigned int head,
+		   unsigned int cyl,
+		   unsigned int cmd,
+		   void (*intr_addr)(void))
+{
+	unsigned short port;
+
+#if (HD_DELAY > 0)
+	while (read_timer() - last_req < HD_DELAY)
+		/* nothing */;
+#endif
+	if (reset)
+		return;
+	if (!controller_ready(disk->unit, head)) {
+		reset = 1;
+		return;
+	}
+	SET_HANDLER(intr_addr);
+	outb_p(disk->ctl, HD_CMD);
+	port = HD_DATA;
+	outb_p(disk->wpcom >> 2, ++port);
+	outb_p(nsect, ++port);
+	outb_p(sect, ++port);
+	outb_p(cyl, ++port);
+	outb_p(cyl >> 8, ++port);
+	outb_p(0xA0 | (disk->unit << 4) | head, ++port);
+	outb_p(cmd, ++port);
+}
+
+static void hd_request (void);
+
+static int drive_busy(void)
+{
+	unsigned int i;
+	unsigned char c;
+
+	for (i = 0; i < 500000 ; i++) {
+		c = inb_p(HD_STATUS);
+		if ((c & (BUSY_STAT | READY_STAT | SEEK_STAT)) == STAT_OK)
+			return 0;
+	}
+	dump_status("reset timed out", c);
+	return 1;
+}
+
+static void reset_controller(void)
+{
+	int	i;
+
+	outb_p(4, HD_CMD);
+	for (i = 0; i < 1000; i++) barrier();
+	outb_p(hd_info[0].ctl & 0x0f, HD_CMD);
+	for (i = 0; i < 1000; i++) barrier();
+	if (drive_busy())
+		printk("hd: controller still busy\n");
+	else if ((hd_error = inb(HD_ERROR)) != 1)
+		printk("hd: controller reset failed: %02x\n", hd_error);
+}
+
+static void reset_hd(void)
+{
+	static int i;
+
+repeat:
+	if (reset) {
+		reset = 0;
+		i = -1;
+		reset_controller();
+	} else {
+		check_status();
+		if (reset)
+			goto repeat;
+	}
+	if (++i < NR_HD) {
+		struct hd_i_struct *disk = &hd_info[i];
+		disk->special_op = disk->recalibrate = 1;
+		hd_out(disk, disk->sect, disk->sect, disk->head-1,
+			disk->cyl, ATA_CMD_INIT_DEV_PARAMS, &reset_hd);
+		if (reset)
+			goto repeat;
+	} else
+		hd_request();
+}
+
+/*
+ * Ok, don't know what to do with the unexpected interrupts: on some machines
+ * doing a reset and a retry seems to result in an eternal loop. Right now I
+ * ignore it, and just set the timeout.
+ *
+ * On laptops (and "green" PCs), an unexpected interrupt occurs whenever the
+ * drive enters "idle", "standby", or "sleep" mode, so if the status looks
+ * "good", we just ignore the interrupt completely.
+ */
+static void unexpected_hd_interrupt(void)
+{
+	unsigned int stat = inb_p(HD_STATUS);
+
+	if (stat & (BUSY_STAT|DRQ_STAT|ECC_STAT|ERR_STAT)) {
+		dump_status("unexpected interrupt", stat);
+		SET_TIMER;
+	}
+}
+
+/*
+ * bad_rw_intr() now tries to be a bit smarter and does things
+ * according to the error returned by the controller.
+ * -Mika Liljeberg (liljeber@cs.Helsinki.FI)
+ */
+static void bad_rw_intr(void)
+{
+	struct request *req = hd_req;
+
+	if (req != NULL) {
+		struct hd_i_struct *disk = req->rq_disk->private_data;
+		if (++req->errors >= MAX_ERRORS || (hd_error & BBD_ERR)) {
+			hd_end_request_cur(-EIO);
+			disk->special_op = disk->recalibrate = 1;
+		} else if (req->errors % RESET_FREQ == 0)
+			reset = 1;
+		else if ((hd_error & TRK0_ERR) || req->errors % RECAL_FREQ == 0)
+			disk->special_op = disk->recalibrate = 1;
+		/* Otherwise just retry */
+	}
+}
+
+static inline int wait_DRQ(void)
+{
+	int retries;
+	int stat;
+
+	for (retries = 0; retries < 100000; retries++) {
+		stat = inb_p(HD_STATUS);
+		if (stat & DRQ_STAT)
+			return 0;
+	}
+	dump_status("wait_DRQ", stat);
+	return -1;
+}
+
+static void read_intr(void)
+{
+	struct request *req;
+	int i, retries = 100000;
+
+	do {
+		i = (unsigned) inb_p(HD_STATUS);
+		if (i & BUSY_STAT)
+			continue;
+		if (!OK_STATUS(i))
+			break;
+		if (i & DRQ_STAT)
+			goto ok_to_read;
+	} while (--retries > 0);
+	dump_status("read_intr", i);
+	bad_rw_intr();
+	hd_request();
+	return;
+
+ok_to_read:
+	req = hd_req;
+	insw(HD_DATA, bio_data(req->bio), 256);
+#ifdef DEBUG
+	printk("%s: read: sector %ld, remaining = %u, buffer=%p\n",
+	       req->rq_disk->disk_name, blk_rq_pos(req) + 1,
+	       blk_rq_sectors(req) - 1, bio_data(req->bio)+512);
+#endif
+	if (hd_end_request(0, 512)) {
+		SET_HANDLER(&read_intr);
+		return;
+	}
+
+	(void) inb_p(HD_STATUS);
+#if (HD_DELAY > 0)
+	last_req = read_timer();
+#endif
+	hd_request();
+}
+
+static void write_intr(void)
+{
+	struct request *req = hd_req;
+	int i;
+	int retries = 100000;
+
+	do {
+		i = (unsigned) inb_p(HD_STATUS);
+		if (i & BUSY_STAT)
+			continue;
+		if (!OK_STATUS(i))
+			break;
+		if ((blk_rq_sectors(req) <= 1) || (i & DRQ_STAT))
+			goto ok_to_write;
+	} while (--retries > 0);
+	dump_status("write_intr", i);
+	bad_rw_intr();
+	hd_request();
+	return;
+
+ok_to_write:
+	if (hd_end_request(0, 512)) {
+		SET_HANDLER(&write_intr);
+		outsw(HD_DATA, bio_data(req->bio), 256);
+		return;
+	}
+
+#if (HD_DELAY > 0)
+	last_req = read_timer();
+#endif
+	hd_request();
+}
+
+static void recal_intr(void)
+{
+	check_status();
+#if (HD_DELAY > 0)
+	last_req = read_timer();
+#endif
+	hd_request();
+}
+
+/*
+ * This is another of the error-routines I don't know what to do with. The
+ * best idea seems to just set reset, and start all over again.
+ */
+static void hd_times_out(unsigned long dummy)
+{
+	char *name;
+
+	do_hd = NULL;
+
+	if (!hd_req)
+		return;
+
+	spin_lock_irq(hd_queue->queue_lock);
+	reset = 1;
+	name = hd_req->rq_disk->disk_name;
+	printk("%s: timeout\n", name);
+	if (++hd_req->errors >= MAX_ERRORS) {
+#ifdef DEBUG
+		printk("%s: too many errors\n", name);
+#endif
+		hd_end_request_cur(-EIO);
+	}
+	hd_request();
+	spin_unlock_irq(hd_queue->queue_lock);
+}
+
+static int do_special_op(struct hd_i_struct *disk, struct request *req)
+{
+	if (disk->recalibrate) {
+		disk->recalibrate = 0;
+		hd_out(disk, disk->sect, 0, 0, 0, ATA_CMD_RESTORE, &recal_intr);
+		return reset;
+	}
+	if (disk->head > 16) {
+		printk("%s: cannot handle device with more than 16 heads - giving up\n", req->rq_disk->disk_name);
+		hd_end_request_cur(-EIO);
+	}
+	disk->special_op = 0;
+	return 1;
+}
+
+/*
+ * The driver enables interrupts as much as possible.  In order to do this,
+ * (a) the device-interrupt is disabled before entering hd_request(),
+ * and (b) the timeout-interrupt is disabled before the sti().
+ *
+ * Interrupts are still masked (by default) whenever we are exchanging
+ * data/cmds with a drive, because some drives seem to have very poor
+ * tolerance for latency during I/O. The IDE driver has support to unmask
+ * interrupts for non-broken hardware, so use that driver if required.
+ */
+static void hd_request(void)
+{
+	unsigned int block, nsect, sec, track, head, cyl;
+	struct hd_i_struct *disk;
+	struct request *req;
+
+	if (do_hd)
+		return;
+repeat:
+	del_timer(&device_timer);
+
+	if (!hd_req) {
+		hd_req = blk_fetch_request(hd_queue);
+		if (!hd_req) {
+			do_hd = NULL;
+			return;
+		}
+	}
+	req = hd_req;
+
+	if (reset) {
+		reset_hd();
+		return;
+	}
+	disk = req->rq_disk->private_data;
+	block = blk_rq_pos(req);
+	nsect = blk_rq_sectors(req);
+	if (block >= get_capacity(req->rq_disk) ||
+	    ((block+nsect) > get_capacity(req->rq_disk))) {
+		printk("%s: bad access: block=%d, count=%d\n",
+			req->rq_disk->disk_name, block, nsect);
+		hd_end_request_cur(-EIO);
+		goto repeat;
+	}
+
+	if (disk->special_op) {
+		if (do_special_op(disk, req))
+			goto repeat;
+		return;
+	}
+	sec   = block % disk->sect + 1;
+	track = block / disk->sect;
+	head  = track % disk->head;
+	cyl   = track / disk->head;
+#ifdef DEBUG
+	printk("%s: %sing: CHS=%d/%d/%d, sectors=%d, buffer=%p\n",
+		req->rq_disk->disk_name,
+		req_data_dir(req) == READ ? "read" : "writ",
+		cyl, head, sec, nsect, bio_data(req->bio));
+#endif
+	if (req->cmd_type == REQ_TYPE_FS) {
+		switch (rq_data_dir(req)) {
+		case READ:
+			hd_out(disk, nsect, sec, head, cyl, ATA_CMD_PIO_READ,
+				&read_intr);
+			if (reset)
+				goto repeat;
+			break;
+		case WRITE:
+			hd_out(disk, nsect, sec, head, cyl, ATA_CMD_PIO_WRITE,
+				&write_intr);
+			if (reset)
+				goto repeat;
+			if (wait_DRQ()) {
+				bad_rw_intr();
+				goto repeat;
+			}
+			outsw(HD_DATA, bio_data(req->bio), 256);
+			break;
+		default:
+			printk("unknown hd-command\n");
+			hd_end_request_cur(-EIO);
+			break;
+		}
+	}
+}
+
+static void do_hd_request(struct request_queue *q)
+{
+	hd_request();
+}
+
+static int hd_getgeo(struct block_device *bdev, struct hd_geometry *geo)
+{
+	struct hd_i_struct *disk = bdev->bd_disk->private_data;
+
+	geo->heads = disk->head;
+	geo->sectors = disk->sect;
+	geo->cylinders = disk->cyl;
+	return 0;
+}
+
+/*
+ * Releasing a block device means we sync() it, so that it can safely
+ * be forgotten about...
+ */
+
+static irqreturn_t hd_interrupt(int irq, void *dev_id)
+{
+	void (*handler)(void) = do_hd;
+
+	spin_lock(hd_queue->queue_lock);
+
+	do_hd = NULL;
+	del_timer(&device_timer);
+	if (!handler)
+		handler = unexpected_hd_interrupt;
+	handler();
+
+	spin_unlock(hd_queue->queue_lock);
+
+	return IRQ_HANDLED;
+}
+
+static const struct block_device_operations hd_fops = {
+	.getgeo =	hd_getgeo,
+};
+
+static int __init hd_init(void)
+{
+	int drive;
+
+	if (register_blkdev(HD_MAJOR, "hd"))
+		return -1;
+
+	hd_queue = blk_init_queue(do_hd_request, &hd_lock);
+	if (!hd_queue) {
+		unregister_blkdev(HD_MAJOR, "hd");
+		return -ENOMEM;
+	}
+
+	blk_queue_max_hw_sectors(hd_queue, 255);
+	init_timer(&device_timer);
+	device_timer.function = hd_times_out;
+	blk_queue_logical_block_size(hd_queue, 512);
+
+	if (!NR_HD) {
+		/*
+		 * We don't know anything about the drive.  This means
+		 * that you *MUST* specify the drive parameters to the
+		 * kernel yourself.
+		 *
+		 * If we were on an i386, we used to read this info from
+		 * the BIOS or CMOS.  This doesn't work all that well,
+		 * since this assumes that this is a primary or secondary
+		 * drive, and if we're using this legacy driver, it's
+		 * probably an auxiliary controller added to recover
+		 * legacy data off an ST-506 drive.  Either way, it's
+		 * definitely safest to have the user explicitly specify
+		 * the information.
+		 */
+		printk("hd: no drives specified - use hd=cyl,head,sectors"
+			" on kernel command line\n");
+		goto out;
+	}
+
+	for (drive = 0 ; drive < NR_HD ; drive++) {
+		struct gendisk *disk = alloc_disk(64);
+		struct hd_i_struct *p = &hd_info[drive];
+		if (!disk)
+			goto Enomem;
+		disk->major = HD_MAJOR;
+		disk->first_minor = drive << 6;
+		disk->fops = &hd_fops;
+		sprintf(disk->disk_name, "hd%c", 'a'+drive);
+		disk->private_data = p;
+		set_capacity(disk, p->head * p->sect * p->cyl);
+		disk->queue = hd_queue;
+		p->unit = drive;
+		hd_gendisk[drive] = disk;
+		printk("%s: %luMB, CHS=%d/%d/%d\n",
+			disk->disk_name, (unsigned long)get_capacity(disk)/2048,
+			p->cyl, p->head, p->sect);
+	}
+
+	if (request_irq(HD_IRQ, hd_interrupt, 0, "hd", NULL)) {
+		printk("hd: unable to get IRQ%d for the hard disk driver\n",
+			HD_IRQ);
+		goto out1;
+	}
+	if (!request_region(HD_DATA, 8, "hd")) {
+		printk(KERN_WARNING "hd: port 0x%x busy\n", HD_DATA);
+		goto out2;
+	}
+	if (!request_region(HD_CMD, 1, "hd(cmd)")) {
+		printk(KERN_WARNING "hd: port 0x%x busy\n", HD_CMD);
+		goto out3;
+	}
+
+	/* Let them fly */
+	for (drive = 0; drive < NR_HD; drive++)
+		add_disk(hd_gendisk[drive]);
+
+	return 0;
+
+out3:
+	release_region(HD_DATA, 8);
+out2:
+	free_irq(HD_IRQ, NULL);
+out1:
+	for (drive = 0; drive < NR_HD; drive++)
+		put_disk(hd_gendisk[drive]);
+	NR_HD = 0;
+out:
+	del_timer(&device_timer);
+	unregister_blkdev(HD_MAJOR, "hd");
+	blk_cleanup_queue(hd_queue);
+	return -1;
+Enomem:
+	while (drive--)
+		put_disk(hd_gendisk[drive]);
+	goto out;
+}
+
+static int __init parse_hd_setup(char *line)
+{
+	int ints[6];
+
+	(void) get_options(line, ARRAY_SIZE(ints), ints);
+	hd_setup(NULL, ints);
+
+	return 1;
+}
+__setup("hd=", parse_hd_setup);
+
+late_initcall(hd_init);

drivers/block/ida_cmd.h

@@ -0,0 +1,349 @@
+/*
+ *    Disk Array driver for Compaq SMART2 Controllers
+ *    Copyright 1998 Compaq Computer Corporation
+ *
+ *    This program is free software; you can redistribute it and/or modify
+ *    it under the terms of the GNU General Public License as published by
+ *    the Free Software Foundation; either version 2 of the License, or
+ *    (at your option) any later version.
+ *
+ *    This program is distributed in the hope that it will be useful,
+ *    but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *    MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
+ *    NON INFRINGEMENT.  See the GNU General Public License for more details.
+ *
+ *    You should have received a copy of the GNU General Public License
+ *    along with this program; if not, write to the Free Software
+ *    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ *
+ *    Questions/Comments/Bugfixes to iss_storagedev@hp.com
+ *
+ */
+#ifndef ARRAYCMD_H
+#define ARRAYCMD_H
+
+#include <asm/types.h>
+#if 0
+#include <linux/blkdev.h>
+#endif
+
+/* for the Smart Array 42XX cards */
+#define S42XX_REQUEST_PORT_OFFSET	0x40
+#define S42XX_REPLY_INTR_MASK_OFFSET	0x34
+#define S42XX_REPLY_PORT_OFFSET		0x44
+#define S42XX_INTR_STATUS		0x30
+
+#define S42XX_INTR_OFF		0x08
+#define S42XX_INTR_PENDING	0x08
+
+#define COMMAND_FIFO		0x04
+#define COMMAND_COMPLETE_FIFO	0x08
+#define INTR_MASK		0x0C
+#define INTR_STATUS		0x10
+#define INTR_PENDING		0x14
+
+#define FIFO_NOT_EMPTY		0x01
+#define FIFO_NOT_FULL		0x02
+
+#define BIG_PROBLEM		0x40
+#define LOG_NOT_CONF		2
+
+#pragma pack(1)
+typedef struct {
+	__u32	size;
+	__u32	addr;
+} sg_t;
+
+#define RCODE_NONFATAL	0x02
+#define RCODE_FATAL	0x04
+#define RCODE_INVREQ	0x10
+typedef struct {
+	__u16	next;
+	__u8	cmd;
+	__u8	rcode;
+	__u32	blk;
+	__u16	blk_cnt;
+	__u8	sg_cnt;
+	__u8	reserved;
+} rhdr_t;
+
+#define SG_MAX			32
+typedef struct {
+	rhdr_t	hdr;
+	sg_t	sg[SG_MAX];
+	__u32	bp;
+} rblk_t;
+
+typedef struct {
+	__u8	unit;
+	__u8	prio;
+	__u16	size;
+} chdr_t;
+
+#define CMD_RWREQ	0x00
+#define CMD_IOCTL_PEND	0x01
+#define CMD_IOCTL_DONE	0x02
+
+typedef struct cmdlist {
+	chdr_t	hdr;
+	rblk_t	req;
+	__u32	size;
+	int	retry_cnt;
+	__u32	busaddr;
+	int	ctlr;
+	struct cmdlist *prev;
+	struct cmdlist *next;
+	struct request *rq;
+	int type;
+} cmdlist_t;
+	
+#define ID_CTLR		0x11
+typedef struct {
+	__u8	nr_drvs;
+	__u32	cfg_sig;
+	__u8	firm_rev[4];
+	__u8	rom_rev[4];
+	__u8	hw_rev;
+	__u32	bb_rev;
+	__u32	drv_present_map;
+	__u32	ext_drv_map;
+	__u32	board_id;
+	__u8	cfg_error;
+	__u32	non_disk_bits;
+	__u8	bad_ram_addr;
+	__u8	cpu_rev;
+	__u8	pdpi_rev;
+	__u8	epic_rev;
+	__u8	wcxc_rev;
+	__u8	marketing_rev;
+	__u8	ctlr_flags;
+	__u8	host_flags;
+	__u8	expand_dis;
+	__u8	scsi_chips;
+	__u32	max_req_blocks;
+	__u32	ctlr_clock;
+	__u8	drvs_per_bus;
+	__u16	big_drv_present_map[8];
+	__u16	big_ext_drv_map[8];
+	__u16	big_non_disk_map[8];
+	__u16	task_flags;
+	__u8	icl_bus;
+	__u8	red_modes;
+	__u8	cur_red_mode;
+	__u8	red_ctlr_stat;
+	__u8	red_fail_reason;
+	__u8	reserved[403];
+} id_ctlr_t;
+
+typedef struct {
+	__u16	cyl;
+	__u8	heads;
+	__u8	xsig;
+	__u8	psectors;
+	__u16	wpre;
+	__u8	maxecc;
+	__u8	drv_ctrl;
+	__u16	pcyls;
+	__u8	pheads;
+	__u16	landz;
+	__u8	sect_per_track;
+	__u8	cksum;
+} drv_param_t;
+
+#define ID_LOG_DRV	0x10
+typedef struct {
+	__u16	blk_size;
+	__u32	nr_blks;
+	drv_param_t drv;
+	__u8	fault_tol;
+	__u8	reserved;
+	__u8	bios_disable;
+} id_log_drv_t;
+
+#define ID_LOG_DRV_EXT	0x18
+typedef struct {
+	__u32	log_drv_id;
+	__u8	log_drv_label[64];
+	__u8	reserved[418];
+} id_log_drv_ext_t;
+
+#define SENSE_LOG_DRV_STAT	0x12
+typedef struct {
+	__u8	status;
+	__u32	fail_map;
+	__u16	read_err[32];
+	__u16	write_err[32];
+	__u8	drv_err_data[256];
+	__u8	drq_timeout[32];
+	__u32	blks_to_recover;
+	__u8	drv_recovering;
+	__u16	remap_cnt[32];
+	__u32	replace_drv_map;
+	__u32	act_spare_map;
+	__u8	spare_stat;
+	__u8	spare_repl_map[32];
+	__u32	repl_ok_map;
+	__u8	media_exch;
+	__u8	cache_fail;
+	__u8	expn_fail;
+	__u8	unit_flags;
+	__u16	big_fail_map[8];
+	__u16	big_remap_map[128];
+	__u16	big_repl_map[8];
+	__u16	big_act_spare_map[8];
+	__u8	big_spar_repl_map[128];
+	__u16	big_repl_ok_map[8];
+	__u8	big_drv_rebuild;
+	__u8	reserved[36];
+} sense_log_drv_stat_t;
+
+#define START_RECOVER		0x13
+
+#define ID_PHYS_DRV		0x15
+typedef struct {
+	__u8	scsi_bus;
+	__u8	scsi_id;
+	__u16	blk_size;
+	__u32	nr_blks;
+	__u32	rsvd_blks;
+	__u8	drv_model[40];
+	__u8	drv_sn[40];
+	__u8	drv_fw[8];
+	__u8	scsi_iq_bits;
+	__u8	compaq_drv_stmp;
+	__u8	last_fail;
+	__u8	phys_drv_flags;
+	__u8	phys_drv_flags1;
+	__u8	scsi_lun;
+	__u8	phys_drv_flags2;
+	__u8	reserved;
+	__u32	spi_speed_rules;
+	__u8	phys_connector[2];
+	__u8	phys_box_on_bus;
+	__u8	phys_bay_in_box;
+} id_phys_drv_t;
+
+#define BLINK_DRV_LEDS		0x16
+typedef struct {
+	__u32	blink_duration;
+	__u32	reserved;
+	__u8	blink[256];
+	__u8	reserved1[248];
+} blink_drv_leds_t;
+
+#define SENSE_BLINK_LEDS	0x17
+typedef struct {
+	__u32	blink_duration;
+	__u32	btime_elap;
+	__u8	blink[256];
+	__u8	reserved1[248];
+} sense_blink_leds_t;
+
+#define IDA_READ		0x20
+#define IDA_WRITE		0x30
+#define IDA_WRITE_MEDIA		0x31
+#define RESET_TO_DIAG		0x40
+#define DIAG_PASS_THRU		0x41
+
+#define SENSE_CONFIG		0x50
+#define SET_CONFIG		0x51
+typedef struct {
+	__u32	cfg_sig;
+	__u16	compat_port;
+	__u8	data_dist_mode;
+	__u8	surf_an_ctrl;
+	__u16	ctlr_phys_drv;
+	__u16	log_unit_phys_drv;
+	__u16	fault_tol_mode;
+	__u8	phys_drv_param[16];
+	drv_param_t drv;
+	__u32	drv_asgn_map;
+	__u16	dist_factor;
+	__u32	spare_asgn_map;
+	__u8	reserved[6];
+	__u16	os;
+	__u8	ctlr_order;
+	__u8	extra_info;
+	__u32	data_offs;
+	__u8	parity_backedout_write_drvs;
+	__u8	parity_dist_mode;
+	__u8	parity_shift_fact;
+	__u8	bios_disable_flag;
+	__u32	blks_on_vol;
+	__u32	blks_per_drv;
+	__u8	scratch[16];
+	__u16	big_drv_map[8];
+	__u16	big_spare_map[8];
+	__u8	ss_source_vol;
+	__u8	mix_drv_cap_range;
+	struct {
+		__u16	big_drv_map[8];
+		__u32	blks_per_drv;
+		__u16	fault_tol_mode;
+		__u16	dist_factor;
+	} MDC_range[4];
+	__u8	reserved1[248];
+} config_t;
+
+#define BYPASS_VOL_STATE	0x52
+#define SS_CREATE_VOL		0x53
+#define CHANGE_CONFIG		0x54
+#define SENSE_ORIG_CONF		0x55
+#define REORDER_LOG_DRV		0x56
+typedef struct {
+	__u8	old_units[32];
+} reorder_log_drv_t;
+
+#define LABEL_LOG_DRV		0x57
+typedef struct {
+	__u8	log_drv_label[64];
+} label_log_drv_t;
+
+#define SS_TO_VOL		0x58
+	
+#define SET_SURF_DELAY		0x60
+typedef struct {
+	__u16	delay;
+	__u8	reserved[510];
+} surf_delay_t;
+
+#define SET_OVERHEAT_DELAY	0x61
+typedef struct {
+	__u16	delay;
+} overhead_delay_t;
+ 
+#define SET_MP_DELAY
+typedef struct {
+	__u16	delay;
+	__u8	reserved[510];
+} mp_delay_t;
+
+#define PASSTHRU_A	0x91
+typedef struct {
+	__u8	target;
+	__u8	bus;
+	__u8	lun;
+	__u32	timeout;
+	__u32	flags;
+	__u8	status;
+	__u8	error;
+	__u8	cdb_len;
+	__u8	sense_error;
+	__u8	sense_key;
+	__u32	sense_info;
+	__u8	sense_code;
+	__u8	sense_qual;
+	__u32	residual;
+	__u8	reserved[4];
+	__u8	cdb[12];	
+} scsi_param_t;
+
+#define RESUME_BACKGROUND_ACTIVITY	0x99
+#define SENSE_CONTROLLER_PERFORMANCE	0xa8
+#define FLUSH_CACHE			0xc2
+#define COLLECT_BUFFER			0xd2
+#define READ_FLASH_ROM			0xf6
+#define WRITE_FLASH_ROM			0xf7
+#pragma pack()	
+
+#endif /* ARRAYCMD_H */

drivers/block/ida_ioctl.h

@@ -0,0 +1,87 @@
+/*
+ *    Disk Array driver for Compaq SMART2 Controllers
+ *    Copyright 1998 Compaq Computer Corporation
+ *
+ *    This program is free software; you can redistribute it and/or modify
+ *    it under the terms of the GNU General Public License as published by
+ *    the Free Software Foundation; either version 2 of the License, or
+ *    (at your option) any later version.
+ *
+ *    This program is distributed in the hope that it will be useful,
+ *    but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *    MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
+ *    NON INFRINGEMENT.  See the GNU General Public License for more details.
+ *
+ *    You should have received a copy of the GNU General Public License
+ *    along with this program; if not, write to the Free Software
+ *    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ *
+ *    Questions/Comments/Bugfixes to iss_storagedev@hp.com
+ *
+ */
+#ifndef IDA_IOCTL_H
+#define IDA_IOCTL_H
+
+#include "ida_cmd.h"
+#include "cpqarray.h"
+
+#define IDAGETDRVINFO		0x27272828
+#define IDAPASSTHRU		0x28282929
+#define IDAGETCTLRSIG		0x29293030
+#define IDAREVALIDATEVOLS	0x30303131
+#define IDADRIVERVERSION	0x31313232
+#define IDAGETPCIINFO		0x32323333
+
+typedef struct _ida_pci_info_struct
+{
+	unsigned char 	bus;
+	unsigned char 	dev_fn;
+	__u32 		board_id;
+} ida_pci_info_struct;
+/*
+ * Normally, the ioctl determines the logical unit for this command by
+ * the major,minor number of the fd passed to ioctl.  If you need to send
+ * a command to a different/nonexistant unit (such as during config), you
+ * can override the normal behavior by setting the unit valid bit. (Normally,
+ * it should be zero) The controller the command is sent to is still
+ * determined by the major number of the open device.
+ */
+
+#define UNITVALID	0x80
+typedef struct {
+	__u8	cmd;
+	__u8	rcode;
+	__u8	unit;
+	__u32	blk;
+	__u16	blk_cnt;
+
+/* currently, sg_cnt is assumed to be 1: only the 0th element of sg is used */
+	struct {
+		void	__user *addr;
+		size_t	size;
+	} sg[SG_MAX];
+	int	sg_cnt;
+
+	union ctlr_cmds {
+		drv_info_t		drv;
+		unsigned char		buf[1024];
+
+		id_ctlr_t		id_ctlr;
+		drv_param_t		drv_param;
+		id_log_drv_t		id_log_drv;
+		id_log_drv_ext_t	id_log_drv_ext;
+		sense_log_drv_stat_t	sense_log_drv_stat;
+		id_phys_drv_t		id_phys_drv;
+		blink_drv_leds_t	blink_drv_leds;
+		sense_blink_leds_t	sense_blink_leds;
+		config_t		config;
+		reorder_log_drv_t	reorder_log_drv;
+		label_log_drv_t		label_log_drv;
+		surf_delay_t		surf_delay;
+		overhead_delay_t	overhead_delay;
+		mp_delay_t		mp_delay;
+		scsi_param_t		scsi_param;
+	} c;
+} ida_ioctl_t;
+
+#endif /* IDA_IOCTL_H */

drivers/block/loop.h

@@ -0,0 +1,85 @@
+/*
+ * loop.h
+ *
+ * Written by Theodore Ts'o, 3/29/93.
+ *
+ * Copyright 1993 by Theodore Ts'o.  Redistribution of this file is
+ * permitted under the GNU General Public License.
+ */
+#ifndef _LINUX_LOOP_H
+#define _LINUX_LOOP_H
+
+#include <linux/bio.h>
+#include <linux/blkdev.h>
+#include <linux/spinlock.h>
+#include <linux/mutex.h>
+#include <uapi/linux/loop.h>
+
+/* Possible states of device */
+enum {
+	Lo_unbound,
+	Lo_bound,
+	Lo_rundown,
+};
+
+struct loop_func_table;
+
+struct loop_device {
+	int		lo_number;
+	int		lo_refcnt;
+	loff_t		lo_offset;
+	loff_t		lo_sizelimit;
+	int		lo_flags;
+	int		(*transfer)(struct loop_device *, int cmd,
+				    struct page *raw_page, unsigned raw_off,
+				    struct page *loop_page, unsigned loop_off,
+				    int size, sector_t real_block);
+	char		lo_file_name[LO_NAME_SIZE];
+	char		lo_crypt_name[LO_NAME_SIZE];
+	char		lo_encrypt_key[LO_KEY_SIZE];
+	int		lo_encrypt_key_size;
+	struct loop_func_table *lo_encryption;
+	__u32           lo_init[2];
+	kuid_t		lo_key_owner;	/* Who set the key */
+	int		(*ioctl)(struct loop_device *, int cmd, 
+				 unsigned long arg); 
+
+	struct file *	lo_backing_file;
+	struct block_device *lo_device;
+	unsigned	lo_blocksize;
+	void		*key_data; 
+
+	gfp_t		old_gfp_mask;
+
+	spinlock_t		lo_lock;
+	struct bio_list		lo_bio_list;
+	unsigned int		lo_bio_count;
+	int			lo_state;
+	struct mutex		lo_ctl_mutex;
+	struct task_struct	*lo_thread;
+	wait_queue_head_t	lo_event;
+	/* wait queue for incoming requests */
+	wait_queue_head_t	lo_req_wait;
+
+	struct request_queue	*lo_queue;
+	struct gendisk		*lo_disk;
+};
+
+/* Support for loadable transfer modules */
+struct loop_func_table {
+	int number;	/* filter type */ 
+	int (*transfer)(struct loop_device *lo, int cmd,
+			struct page *raw_page, unsigned raw_off,
+			struct page *loop_page, unsigned loop_off,
+			int size, sector_t real_block);
+	int (*init)(struct loop_device *, const struct loop_info64 *); 
+	/* release is called from loop_unregister_transfer or clr_fd */
+	int (*release)(struct loop_device *); 
+	int (*ioctl)(struct loop_device *, int cmd, unsigned long arg);
+	struct module *owner;
+}; 
+
+int loop_register_transfer(struct loop_func_table *funcs);
+int loop_unregister_transfer(int number); 
+
+#endif

drivers/block/mtip32xx/Kconfig

@@ -0,0 +1,9 @@
+#
+# mtip32xx device driver configuration
+#
+
+config BLK_DEV_PCIESSD_MTIP32XX
+	tristate "Block Device Driver for Micron PCIe SSDs"
+	depends on PCI
+	help
+          This enables the block driver for Micron PCIe SSDs.

drivers/block/mtip32xx/Makefile

@@ -0,0 +1,5 @@
+#
+# Makefile for  Block device driver for Micron PCIe SSD
+#
+
+obj-$(CONFIG_BLK_DEV_PCIESSD_MTIP32XX) += mtip32xx.o

drivers/block/mtip32xx/mtip32xx.h

@@ -0,0 +1,511 @@
+/*
+ * mtip32xx.h - Header file for the P320 SSD Block Driver
+ *   Copyright (C) 2011 Micron Technology, Inc.
+ *
+ * Portions of this code were derived from works subjected to the
+ * following copyright:
+ *    Copyright (C) 2009 Integrated Device Technology, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ */
+
+#ifndef __MTIP32XX_H__
+#define __MTIP32XX_H__
+
+#include <linux/spinlock.h>
+#include <linux/rwsem.h>
+#include <linux/ata.h>
+#include <linux/interrupt.h>
+#include <linux/genhd.h>
+
+/* Offset of Subsystem Device ID in pci confoguration space */
+#define PCI_SUBSYSTEM_DEVICEID	0x2E
+
+/* offset of Device Control register in PCIe extended capabilites space */
+#define PCIE_CONFIG_EXT_DEVICE_CONTROL_OFFSET	0x48
+
+/* check for erase mode support during secure erase */
+#define MTIP_SEC_ERASE_MODE     0x2
+
+/* # of times to retry timed out/failed IOs */
+#define MTIP_MAX_RETRIES	2
+
+/* Various timeout values in ms */
+#define MTIP_NCQ_CMD_TIMEOUT_MS      15000
+#define MTIP_IOCTL_CMD_TIMEOUT_MS    5000
+#define MTIP_INT_CMD_TIMEOUT_MS      5000
+#define MTIP_QUIESCE_IO_TIMEOUT_MS   (MTIP_NCQ_CMD_TIMEOUT_MS * \
+				     (MTIP_MAX_RETRIES + 1))
+
+/* check for timeouts every 500ms */
+#define MTIP_TIMEOUT_CHECK_PERIOD	500
+
+/* ftl rebuild */
+#define MTIP_FTL_REBUILD_OFFSET		142
+#define MTIP_FTL_REBUILD_MAGIC		0xED51
+#define MTIP_FTL_REBUILD_TIMEOUT_MS	2400000
+
+/* unaligned IO handling */
+#define MTIP_MAX_UNALIGNED_SLOTS	2
+
+/* Macro to extract the tag bit number from a tag value. */
+#define MTIP_TAG_BIT(tag)	(tag & 0x1F)
+
+/*
+ * Macro to extract the tag index from a tag value. The index
+ * is used to access the correct s_active/Command Issue register based
+ * on the tag value.
+ */
+#define MTIP_TAG_INDEX(tag)	(tag >> 5)
+
+/*
+ * Maximum number of scatter gather entries
+ * a single command may have.
+ */
+#define MTIP_MAX_SG		504
+
+/*
+ * Maximum number of slot groups (Command Issue & s_active registers)
+ * NOTE: This is the driver maximum; check dd->slot_groups for actual value.
+ */
+#define MTIP_MAX_SLOT_GROUPS	8
+
+/* Internal command tag. */
+#define MTIP_TAG_INTERNAL	0
+
+/* Micron Vendor ID & P320x SSD Device ID */
+#define PCI_VENDOR_ID_MICRON    0x1344
+#define P320H_DEVICE_ID		0x5150
+#define P320M_DEVICE_ID		0x5151
+#define P320S_DEVICE_ID		0x5152
+#define P325M_DEVICE_ID		0x5153
+#define P420H_DEVICE_ID		0x5160
+#define P420M_DEVICE_ID		0x5161
+#define P425M_DEVICE_ID		0x5163
+
+/* Driver name and version strings */
+#define MTIP_DRV_NAME		"mtip32xx"
+#define MTIP_DRV_VERSION	"1.3.1"
+
+/* Maximum number of minor device numbers per device. */
+#define MTIP_MAX_MINORS		16
+
+/* Maximum number of supported command slots. */
+#define MTIP_MAX_COMMAND_SLOTS	(MTIP_MAX_SLOT_GROUPS * 32)
+
+/*
+ * Per-tag bitfield size in longs.
+ * Linux bit manipulation functions
+ * (i.e. test_and_set_bit, find_next_zero_bit)
+ * manipulate memory in longs, so we try to make the math work.
+ * take the slot groups and find the number of longs, rounding up.
+ * Careful! i386 and x86_64 use different size longs!
+ */
+#define U32_PER_LONG	(sizeof(long) / sizeof(u32))
+#define SLOTBITS_IN_LONGS ((MTIP_MAX_SLOT_GROUPS + \
+					(U32_PER_LONG-1))/U32_PER_LONG)
+
+/* BAR number used to access the HBA registers. */
+#define MTIP_ABAR		5
+
+#ifdef DEBUG
+ #define dbg_printk(format, arg...)	\
+	printk(pr_fmt(format), ##arg);
+#else
+ #define dbg_printk(format, arg...)
+#endif
+
+#define MTIP_DFS_MAX_BUF_SIZE 1024
+
+#define __force_bit2int (unsigned int __force)
+
+enum {
+	/* below are bit numbers in 'flags' defined in mtip_port */
+	MTIP_PF_IC_ACTIVE_BIT       = 0, /* pio/ioctl */
+	MTIP_PF_EH_ACTIVE_BIT       = 1, /* error handling */
+	MTIP_PF_SE_ACTIVE_BIT       = 2, /* secure erase */
+	MTIP_PF_DM_ACTIVE_BIT       = 3, /* download microcde */
+	MTIP_PF_PAUSE_IO      =	((1 << MTIP_PF_IC_ACTIVE_BIT) |
+				(1 << MTIP_PF_EH_ACTIVE_BIT) |
+				(1 << MTIP_PF_SE_ACTIVE_BIT) |
+				(1 << MTIP_PF_DM_ACTIVE_BIT)),
+
+	MTIP_PF_SVC_THD_ACTIVE_BIT  = 4,
+	MTIP_PF_ISSUE_CMDS_BIT      = 5,
+	MTIP_PF_REBUILD_BIT         = 6,
+	MTIP_PF_SR_CLEANUP_BIT      = 7,
+	MTIP_PF_SVC_THD_STOP_BIT    = 8,
+
+	/* below are bit numbers in 'dd_flag' defined in driver_data */
+	MTIP_DDF_SEC_LOCK_BIT	    = 0,
+	MTIP_DDF_REMOVE_PENDING_BIT = 1,
+	MTIP_DDF_OVER_TEMP_BIT      = 2,
+	MTIP_DDF_WRITE_PROTECT_BIT  = 3,
+	MTIP_DDF_REMOVE_DONE_BIT    = 4,
+	MTIP_DDF_CLEANUP_BIT        = 5,
+	MTIP_DDF_RESUME_BIT         = 6,
+	MTIP_DDF_INIT_DONE_BIT      = 7,
+	MTIP_DDF_REBUILD_FAILED_BIT = 8,
+
+	MTIP_DDF_STOP_IO      = ((1 << MTIP_DDF_REMOVE_PENDING_BIT) |
+				(1 << MTIP_DDF_SEC_LOCK_BIT) |
+				(1 << MTIP_DDF_OVER_TEMP_BIT) |
+				(1 << MTIP_DDF_WRITE_PROTECT_BIT) |
+				(1 << MTIP_DDF_REBUILD_FAILED_BIT)),
+
+};
+
+struct smart_attr {
+	u8 attr_id;
+	u16 flags;
+	u8 cur;
+	u8 worst;
+	u32 data;
+	u8 res[3];
+} __packed;
+
+struct mtip_work {
+	struct work_struct work;
+	void *port;
+	int cpu_binding;
+	u32 completed;
+} ____cacheline_aligned_in_smp;
+
+#define DEFINE_HANDLER(group)                                  \
+	void mtip_workq_sdbf##group(struct work_struct *work)       \
+	{                                                      \
+		struct mtip_work *w = (struct mtip_work *) work;         \
+		mtip_workq_sdbfx(w->port, group, w->completed);     \
+	}
+
+#define MTIP_TRIM_TIMEOUT_MS		240000
+#define MTIP_MAX_TRIM_ENTRIES		8
+#define MTIP_MAX_TRIM_ENTRY_LEN		0xfff8
+
+struct mtip_trim_entry {
+	u32 lba;   /* starting lba of region */
+	u16 rsvd;  /* unused */
+	u16 range; /* # of 512b blocks to trim */
+} __packed;
+
+struct mtip_trim {
+	/* Array of regions to trim */
+	struct mtip_trim_entry entry[MTIP_MAX_TRIM_ENTRIES];
+} __packed;
+
+/* Register Frame Information Structure (FIS), host to device. */
+struct host_to_dev_fis {
+	/*
+	 * FIS type.
+	 * - 27h Register FIS, host to device.
+	 * - 34h Register FIS, device to host.
+	 * - 39h DMA Activate FIS, device to host.
+	 * - 41h DMA Setup FIS, bi-directional.
+	 * - 46h Data FIS, bi-directional.
+	 * - 58h BIST Activate FIS, bi-directional.
+	 * - 5Fh PIO Setup FIS, device to host.
+	 * - A1h Set Device Bits FIS, device to host.
+	 */
+	unsigned char type;
+	unsigned char opts;
+	unsigned char command;
+	unsigned char features;
+
+	union {
+		unsigned char lba_low;
+		unsigned char sector;
+	};
+	union {
+		unsigned char lba_mid;
+		unsigned char cyl_low;
+	};
+	union {
+		unsigned char lba_hi;
+		unsigned char cyl_hi;
+	};
+	union {
+		unsigned char device;
+		unsigned char head;
+	};
+
+	union {
+		unsigned char lba_low_ex;
+		unsigned char sector_ex;
+	};
+	union {
+		unsigned char lba_mid_ex;
+		unsigned char cyl_low_ex;
+	};
+	union {
+		unsigned char lba_hi_ex;
+		unsigned char cyl_hi_ex;
+	};
+	unsigned char features_ex;
+
+	unsigned char sect_count;
+	unsigned char sect_cnt_ex;
+	unsigned char res2;
+	unsigned char control;
+
+	unsigned int res3;
+};
+
+/* Command header structure. */
+struct mtip_cmd_hdr {
+	/*
+	 * Command options.
+	 * - Bits 31:16 Number of PRD entries.
+	 * - Bits 15:8 Unused in this implementation.
+	 * - Bit 7 Prefetch bit, informs the drive to prefetch PRD entries.
+	 * - Bit 6 Write bit, should be set when writing data to the device.
+	 * - Bit 5 Unused in this implementation.
+	 * - Bits 4:0 Length of the command FIS in DWords (DWord = 4 bytes).
+	 */
+	unsigned int opts;
+	/* This field is unsed when using NCQ. */
+	union {
+		unsigned int byte_count;
+		unsigned int status;
+	};
+	/*
+	 * Lower 32 bits of the command table address associated with this
+	 * header. The command table addresses must be 128 byte aligned.
+	 */
+	unsigned int ctba;
+	/*
+	 * If 64 bit addressing is used this field is the upper 32 bits
+	 * of the command table address associated with this command.
+	 */
+	unsigned int ctbau;
+	/* Reserved and unused. */
+	unsigned int res[4];
+};
+
+/* Command scatter gather structure (PRD). */
+struct mtip_cmd_sg {
+	/*
+	 * Low 32 bits of the data buffer address. For P320 this
+	 * address must be 8 byte aligned signified by bits 2:0 being
+	 * set to 0.
+	 */
+	unsigned int dba;
+	/*
+	 * When 64 bit addressing is used this field is the upper
+	 * 32 bits of the data buffer address.
+	 */
+	unsigned int dba_upper;
+	/* Unused. */
+	unsigned int reserved;
+	/*
+	 * Bit 31: interrupt when this data block has been transferred.
+	 * Bits 30..22: reserved
+	 * Bits 21..0: byte count (minus 1).  For P320 the byte count must be
+	 * 8 byte aligned signified by bits 2:0 being set to 1.
+	 */
+	unsigned int info;
+};
+struct mtip_port;
+
+/* Structure used to describe a command. */
+struct mtip_cmd {
+
+	struct mtip_cmd_hdr *command_header; /* ptr to command header entry */
+
+	dma_addr_t command_header_dma; /* corresponding physical address */
+
+	void *command; /* ptr to command table entry */
+
+	dma_addr_t command_dma; /* corresponding physical address */
+
+	void *comp_data; /* data passed to completion function comp_func() */
+	/*
+	 * Completion function called by the ISR upon completion of
+	 * a command.
+	 */
+	void (*comp_func)(struct mtip_port *port,
+				int tag,
+				struct mtip_cmd *cmd,
+				int status);
+
+	int scatter_ents; /* Number of scatter list entries used */
+
+	int unaligned; /* command is unaligned on 4k boundary */
+
+	struct scatterlist sg[MTIP_MAX_SG]; /* Scatter list entries */
+
+	int retries; /* The number of retries left for this command. */
+
+	int direction; /* Data transfer direction */
+};
+
+/* Structure used to describe a port. */
+struct mtip_port {
+	/* Pointer back to the driver data for this port. */
+	struct driver_data *dd;
+	/*
+	 * Used to determine if the data pointed to by the
+	 * identify field is valid.
+	 */
+	unsigned long identify_valid;
+	/* Base address of the memory mapped IO for the port. */
+	void __iomem *mmio;
+	/* Array of pointers to the memory mapped s_active registers. */
+	void __iomem *s_active[MTIP_MAX_SLOT_GROUPS];
+	/* Array of pointers to the memory mapped completed registers. */
+	void __iomem *completed[MTIP_MAX_SLOT_GROUPS];
+	/* Array of pointers to the memory mapped Command Issue registers. */
+	void __iomem *cmd_issue[MTIP_MAX_SLOT_GROUPS];
+	/*
+	 * Pointer to the beginning of the command header memory as used
+	 * by the driver.
+	 */
+	void *command_list;
+	/*
+	 * Pointer to the beginning of the command header memory as used
+	 * by the DMA.
+	 */
+	dma_addr_t command_list_dma;
+	/*
+	 * Pointer to the beginning of the RX FIS memory as used
+	 * by the driver.
+	 */
+	void *rxfis;
+	/*
+	 * Pointer to the beginning of the RX FIS memory as used
+	 * by the DMA.
+	 */
+	dma_addr_t rxfis_dma;
+	/*
+	 * Pointer to the DMA region for RX Fis, Identify, RLE10, and SMART
+	 */
+	void *block1;
+	/*
+	 * DMA address of region for RX Fis, Identify, RLE10, and SMART
+	 */
+	dma_addr_t block1_dma;
+	/*
+	 * Pointer to the beginning of the identify data memory as used
+	 * by the driver.
+	 */
+	u16 *identify;
+	/*
+	 * Pointer to the beginning of the identify data memory as used
+	 * by the DMA.
+	 */
+	dma_addr_t identify_dma;
+	/*
+	 * Pointer to the beginning of a sector buffer that is used
+	 * by the driver when issuing internal commands.
+	 */
+	u16 *sector_buffer;
+	/*
+	 * Pointer to the beginning of a sector buffer that is used
+	 * by the DMA when the driver issues internal commands.
+	 */
+	dma_addr_t sector_buffer_dma;
+	/*
+	 * Bit significant, used to determine if a command slot has
+	 * been allocated. i.e. the slot is in use.  Bits are cleared
+	 * when the command slot and all associated data structures
+	 * are no longer needed.
+	 */
+	u16 *log_buf;
+	dma_addr_t log_buf_dma;
+
+	u8 *smart_buf;
+	dma_addr_t smart_buf_dma;
+
+	unsigned long allocated[SLOTBITS_IN_LONGS];
+	/*
+	 * used to queue commands when an internal command is in progress
+	 * or error handling is active
+	 */
+	unsigned long cmds_to_issue[SLOTBITS_IN_LONGS];
+	/* Used by mtip_service_thread to wait for an event */
+	wait_queue_head_t svc_wait;
+	/*
+	 * indicates the state of the port. Also, helps the service thread
+	 * to determine its action on wake up.
+	 */
+	unsigned long flags;
+	/*
+	 * Timer used to complete commands that have been active for too long.
+	 */
+	unsigned long ic_pause_timer;
+
+	/* Semaphore to control queue depth of unaligned IOs */
+	struct semaphore cmd_slot_unal;
+
+	/* Spinlock for working around command-issue bug. */
+	spinlock_t cmd_issue_lock[MTIP_MAX_SLOT_GROUPS];
+};
+
+/*
+ * Driver private data structure.
+ *
+ * One structure is allocated per probed device.
+ */
+struct driver_data {
+	void __iomem *mmio; /* Base address of the HBA registers. */
+
+	int major; /* Major device number. */
+
+	int instance; /* Instance number. First device probed is 0, ... */
+
+	struct gendisk *disk; /* Pointer to our gendisk structure. */
+
+	struct pci_dev *pdev; /* Pointer to the PCI device structure. */
+
+	struct request_queue *queue; /* Our request queue. */
+
+	struct blk_mq_tag_set tags; /* blk_mq tags */
+
+	struct mtip_port *port; /* Pointer to the port data structure. */
+
+	unsigned product_type; /* magic value declaring the product type */
+
+	unsigned slot_groups; /* number of slot groups the product supports */
+
+	unsigned long index; /* Index to determine the disk name */
+
+	unsigned long dd_flag; /* NOTE: use atomic bit operations on this */
+
+	struct task_struct *mtip_svc_handler; /* task_struct of svc thd */
+
+	struct dentry *dfs_node;
+
+	bool trim_supp; /* flag indicating trim support */
+
+	bool sr;
+
+	int numa_node; /* NUMA support */
+
+	char workq_name[32];
+
+	struct workqueue_struct *isr_workq;
+
+	atomic_t irq_workers_active;
+
+	struct mtip_work work[MTIP_MAX_SLOT_GROUPS];
+
+	int isr_binding;
+
+	struct block_device *bdev;
+
+	struct list_head online_list; /* linkage for online list */
+
+	struct list_head remove_list; /* linkage for removing list */
+
+	int unal_qdepth; /* qdepth of unaligned IO queue */
+};
+
+#endif

drivers/block/nbd.c

@@ -0,0 +1,926 @@
+/*
+ * Network block device - make block devices work over TCP
+ *
+ * Note that you can not swap over this thing, yet. Seems to work but
+ * deadlocks sometimes - you can not swap over TCP in general.
+ * 
+ * Copyright 1997-2000, 2008 Pavel Machek <pavel@ucw.cz>
+ * Parts copyright 2001 Steven Whitehouse <steve@chygwyn.com>
+ *
+ * This file is released under GPLv2 or later.
+ *
+ * (part of code stolen from loop.c)
+ */
+
+#include <linux/major.h>
+
+#include <linux/blkdev.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/sched.h>
+#include <linux/fs.h>
+#include <linux/bio.h>
+#include <linux/stat.h>
+#include <linux/errno.h>
+#include <linux/file.h>
+#include <linux/ioctl.h>
+#include <linux/mutex.h>
+#include <linux/compiler.h>
+#include <linux/err.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <net/sock.h>
+#include <linux/net.h>
+#include <linux/kthread.h>
+
+#include <asm/uaccess.h>
+#include <asm/types.h>
+
+#include <linux/nbd.h>
+
+#define NBD_MAGIC 0x68797548
+
+#ifdef NDEBUG
+#define dprintk(flags, fmt...)
+#else /* NDEBUG */
+#define dprintk(flags, fmt...) do { \
+	if (debugflags & (flags)) printk(KERN_DEBUG fmt); \
+} while (0)
+#define DBG_IOCTL       0x0004
+#define DBG_INIT        0x0010
+#define DBG_EXIT        0x0020
+#define DBG_BLKDEV      0x0100
+#define DBG_RX          0x0200
+#define DBG_TX          0x0400
+static unsigned int debugflags;
+#endif /* NDEBUG */
+
+static unsigned int nbds_max = 16;
+static struct nbd_device *nbd_dev;
+static int max_part;
+
+/*
+ * Use just one lock (or at most 1 per NIC). Two arguments for this:
+ * 1. Each NIC is essentially a synchronization point for all servers
+ *    accessed through that NIC so there's no need to have more locks
+ *    than NICs anyway.
+ * 2. More locks lead to more "Dirty cache line bouncing" which will slow
+ *    down each lock to the point where they're actually slower than just
+ *    a single lock.
+ * Thanks go to Jens Axboe and Al Viro for their LKML emails explaining this!
+ */
+static DEFINE_SPINLOCK(nbd_lock);
+
+#ifndef NDEBUG
+static const char *ioctl_cmd_to_ascii(int cmd)
+{
+	switch (cmd) {
+	case NBD_SET_SOCK: return "set-sock";
+	case NBD_SET_BLKSIZE: return "set-blksize";
+	case NBD_SET_SIZE: return "set-size";
+	case NBD_SET_TIMEOUT: return "set-timeout";
+	case NBD_SET_FLAGS: return "set-flags";
+	case NBD_DO_IT: return "do-it";
+	case NBD_CLEAR_SOCK: return "clear-sock";
+	case NBD_CLEAR_QUE: return "clear-que";
+	case NBD_PRINT_DEBUG: return "print-debug";
+	case NBD_SET_SIZE_BLOCKS: return "set-size-blocks";
+	case NBD_DISCONNECT: return "disconnect";
+	case BLKROSET: return "set-read-only";
+	case BLKFLSBUF: return "flush-buffer-cache";
+	}
+	return "unknown";
+}
+
+static const char *nbdcmd_to_ascii(int cmd)
+{
+	switch (cmd) {
+	case  NBD_CMD_READ: return "read";
+	case NBD_CMD_WRITE: return "write";
+	case  NBD_CMD_DISC: return "disconnect";
+	case NBD_CMD_FLUSH: return "flush";
+	case  NBD_CMD_TRIM: return "trim/discard";
+	}
+	return "invalid";
+}
+#endif /* NDEBUG */
+
+static void nbd_end_request(struct request *req)
+{
+	int error = req->errors ? -EIO : 0;
+	struct request_queue *q = req->q;
+	unsigned long flags;
+
+	dprintk(DBG_BLKDEV, "%s: request %p: %s\n", req->rq_disk->disk_name,
+			req, error ? "failed" : "done");
+
+	spin_lock_irqsave(q->queue_lock, flags);
+	__blk_end_request_all(req, error);
+	spin_unlock_irqrestore(q->queue_lock, flags);
+}
+
+static void sock_shutdown(struct nbd_device *nbd, int lock)
+{
+	/* Forcibly shutdown the socket causing all listeners
+	 * to error
+	 *
+	 * FIXME: This code is duplicated from sys_shutdown, but
+	 * there should be a more generic interface rather than
+	 * calling socket ops directly here */
+	if (lock)
+		mutex_lock(&nbd->tx_lock);
+	if (nbd->sock) {
+		dev_warn(disk_to_dev(nbd->disk), "shutting down socket\n");
+		kernel_sock_shutdown(nbd->sock, SHUT_RDWR);
+		nbd->sock = NULL;
+	}
+	if (lock)
+		mutex_unlock(&nbd->tx_lock);
+}
+
+static void nbd_xmit_timeout(unsigned long arg)
+{
+	struct task_struct *task = (struct task_struct *)arg;
+
+	printk(KERN_WARNING "nbd: killing hung xmit (%s, pid: %d)\n",
+		task->comm, task->pid);
+	force_sig(SIGKILL, task);
+}
+
+/*
+ *  Send or receive packet.
+ */
+static int sock_xmit(struct nbd_device *nbd, int send, void *buf, int size,
+		int msg_flags)
+{
+	struct socket *sock = nbd->sock;
+	int result;
+	struct msghdr msg;
+	struct kvec iov;
+	sigset_t blocked, oldset;
+	unsigned long pflags = current->flags;
+
+	if (unlikely(!sock)) {
+		dev_err(disk_to_dev(nbd->disk),
+			"Attempted %s on closed socket in sock_xmit\n",
+			(send ? "send" : "recv"));
+		return -EINVAL;
+	}
+
+	/* Allow interception of SIGKILL only
+	 * Don't allow other signals to interrupt the transmission */
+	siginitsetinv(&blocked, sigmask(SIGKILL));
+	sigprocmask(SIG_SETMASK, &blocked, &oldset);
+
+	current->flags |= PF_MEMALLOC;
+	do {
+		sock->sk->sk_allocation = GFP_NOIO | __GFP_MEMALLOC;
+		iov.iov_base = buf;
+		iov.iov_len = size;
+		msg.msg_name = NULL;
+		msg.msg_namelen = 0;
+		msg.msg_control = NULL;
+		msg.msg_controllen = 0;
+		msg.msg_flags = msg_flags | MSG_NOSIGNAL;
+
+		if (send) {
+			struct timer_list ti;
+
+			if (nbd->xmit_timeout) {
+				init_timer(&ti);
+				ti.function = nbd_xmit_timeout;
+				ti.data = (unsigned long)current;
+				ti.expires = jiffies + nbd->xmit_timeout;
+				add_timer(&ti);
+			}
+			result = kernel_sendmsg(sock, &msg, &iov, 1, size);
+			if (nbd->xmit_timeout)
+				del_timer_sync(&ti);
+		} else
+			result = kernel_recvmsg(sock, &msg, &iov, 1, size,
+						msg.msg_flags);
+
+		if (signal_pending(current)) {
+			siginfo_t info;
+			printk(KERN_WARNING "nbd (pid %d: %s) got signal %d\n",
+				task_pid_nr(current), current->comm,
+				dequeue_signal_lock(current, &current->blocked, &info));
+			result = -EINTR;
+			sock_shutdown(nbd, !send);
+			break;
+		}
+
+		if (result <= 0) {
+			if (result == 0)
+				result = -EPIPE; /* short read */
+			break;
+		}
+		size -= result;
+		buf += result;
+	} while (size > 0);
+
+	sigprocmask(SIG_SETMASK, &oldset, NULL);
+	tsk_restore_flags(current, pflags, PF_MEMALLOC);
+
+	return result;
+}
+
+static inline int sock_send_bvec(struct nbd_device *nbd, struct bio_vec *bvec,
+		int flags)
+{
+	int result;
+	void *kaddr = kmap(bvec->bv_page);
+	result = sock_xmit(nbd, 1, kaddr + bvec->bv_offset,
+			   bvec->bv_len, flags);
+	kunmap(bvec->bv_page);
+	return result;
+}
+
+/* always call with the tx_lock held */
+static int nbd_send_req(struct nbd_device *nbd, struct request *req)
+{
+	int result, flags;
+	struct nbd_request request;
+	unsigned long size = blk_rq_bytes(req);
+
+	memset(&request, 0, sizeof(request));
+	request.magic = htonl(NBD_REQUEST_MAGIC);
+	request.type = htonl(nbd_cmd(req));
+
+	if (nbd_cmd(req) != NBD_CMD_FLUSH && nbd_cmd(req) != NBD_CMD_DISC) {
+		request.from = cpu_to_be64((u64)blk_rq_pos(req) << 9);
+		request.len = htonl(size);
+	}
+	memcpy(request.handle, &req, sizeof(req));
+
+	dprintk(DBG_TX, "%s: request %p: sending control (%s@%llu,%uB)\n",
+			nbd->disk->disk_name, req,
+			nbdcmd_to_ascii(nbd_cmd(req)),
+			(unsigned long long)blk_rq_pos(req) << 9,
+			blk_rq_bytes(req));
+	result = sock_xmit(nbd, 1, &request, sizeof(request),
+			(nbd_cmd(req) == NBD_CMD_WRITE) ? MSG_MORE : 0);
+	if (result <= 0) {
+		dev_err(disk_to_dev(nbd->disk),
+			"Send control failed (result %d)\n", result);
+		goto error_out;
+	}
+
+	if (nbd_cmd(req) == NBD_CMD_WRITE) {
+		struct req_iterator iter;
+		struct bio_vec bvec;
+		/*
+		 * we are really probing at internals to determine
+		 * whether to set MSG_MORE or not...
+		 */
+		rq_for_each_segment(bvec, req, iter) {
+			flags = 0;
+			if (!rq_iter_last(bvec, iter))
+				flags = MSG_MORE;
+			dprintk(DBG_TX, "%s: request %p: sending %d bytes data\n",
+					nbd->disk->disk_name, req, bvec.bv_len);
+			result = sock_send_bvec(nbd, &bvec, flags);
+			if (result <= 0) {
+				dev_err(disk_to_dev(nbd->disk),
+					"Send data failed (result %d)\n",
+					result);
+				goto error_out;
+			}
+		}
+	}
+	return 0;
+
+error_out:
+	return -EIO;
+}
+
+static struct request *nbd_find_request(struct nbd_device *nbd,
+					struct request *xreq)
+{
+	struct request *req, *tmp;
+	int err;
+
+	err = wait_event_interruptible(nbd->active_wq, nbd->active_req != xreq);
+	if (unlikely(err))
+		goto out;
+
+	spin_lock(&nbd->queue_lock);
+	list_for_each_entry_safe(req, tmp, &nbd->queue_head, queuelist) {
+		if (req != xreq)
+			continue;
+		list_del_init(&req->queuelist);
+		spin_unlock(&nbd->queue_lock);
+		return req;
+	}
+	spin_unlock(&nbd->queue_lock);
+
+	err = -ENOENT;
+
+out:
+	return ERR_PTR(err);
+}
+
+static inline int sock_recv_bvec(struct nbd_device *nbd, struct bio_vec *bvec)
+{
+	int result;
+	void *kaddr = kmap(bvec->bv_page);
+	result = sock_xmit(nbd, 0, kaddr + bvec->bv_offset, bvec->bv_len,
+			MSG_WAITALL);
+	kunmap(bvec->bv_page);
+	return result;
+}
+
+/* NULL returned = something went wrong, inform userspace */
+static struct request *nbd_read_stat(struct nbd_device *nbd)
+{
+	int result;
+	struct nbd_reply reply;
+	struct request *req;
+
+	reply.magic = 0;
+	result = sock_xmit(nbd, 0, &reply, sizeof(reply), MSG_WAITALL);
+	if (result <= 0) {
+		dev_err(disk_to_dev(nbd->disk),
+			"Receive control failed (result %d)\n", result);
+		goto harderror;
+	}
+
+	if (ntohl(reply.magic) != NBD_REPLY_MAGIC) {
+		dev_err(disk_to_dev(nbd->disk), "Wrong magic (0x%lx)\n",
+				(unsigned long)ntohl(reply.magic));
+		result = -EPROTO;
+		goto harderror;
+	}
+
+	req = nbd_find_request(nbd, *(struct request **)reply.handle);
+	if (IS_ERR(req)) {
+		result = PTR_ERR(req);
+		if (result != -ENOENT)
+			goto harderror;
+
+		dev_err(disk_to_dev(nbd->disk), "Unexpected reply (%p)\n",
+			reply.handle);
+		result = -EBADR;
+		goto harderror;
+	}
+
+	if (ntohl(reply.error)) {
+		dev_err(disk_to_dev(nbd->disk), "Other side returned error (%d)\n",
+			ntohl(reply.error));
+		req->errors++;
+		return req;
+	}
+
+	dprintk(DBG_RX, "%s: request %p: got reply\n",
+			nbd->disk->disk_name, req);
+	if (nbd_cmd(req) == NBD_CMD_READ) {
+		struct req_iterator iter;
+		struct bio_vec bvec;
+
+		rq_for_each_segment(bvec, req, iter) {
+			result = sock_recv_bvec(nbd, &bvec);
+			if (result <= 0) {
+				dev_err(disk_to_dev(nbd->disk), "Receive data failed (result %d)\n",
+					result);
+				req->errors++;
+				return req;
+			}
+			dprintk(DBG_RX, "%s: request %p: got %d bytes data\n",
+				nbd->disk->disk_name, req, bvec.bv_len);
+		}
+	}
+	return req;
+harderror:
+	nbd->harderror = result;
+	return NULL;
+}
+
+static ssize_t pid_show(struct device *dev,
+			struct device_attribute *attr, char *buf)
+{
+	struct gendisk *disk = dev_to_disk(dev);
+
+	return sprintf(buf, "%ld\n",
+		(long) ((struct nbd_device *)disk->private_data)->pid);
+}
+
+static struct device_attribute pid_attr = {
+	.attr = { .name = "pid", .mode = S_IRUGO},
+	.show = pid_show,
+};
+
+static int nbd_do_it(struct nbd_device *nbd)
+{
+	struct request *req;
+	int ret;
+
+	BUG_ON(nbd->magic != NBD_MAGIC);
+
+	sk_set_memalloc(nbd->sock->sk);
+	nbd->pid = task_pid_nr(current);
+	ret = device_create_file(disk_to_dev(nbd->disk), &pid_attr);
+	if (ret) {
+		dev_err(disk_to_dev(nbd->disk), "device_create_file failed!\n");
+		nbd->pid = 0;
+		return ret;
+	}
+
+	while ((req = nbd_read_stat(nbd)) != NULL)
+		nbd_end_request(req);
+
+	device_remove_file(disk_to_dev(nbd->disk), &pid_attr);
+	nbd->pid = 0;
+	return 0;
+}
+
+static void nbd_clear_que(struct nbd_device *nbd)
+{
+	struct request *req;
+
+	BUG_ON(nbd->magic != NBD_MAGIC);
+
+	/*
+	 * Because we have set nbd->sock to NULL under the tx_lock, all
+	 * modifications to the list must have completed by now.  For
+	 * the same reason, the active_req must be NULL.
+	 *
+	 * As a consequence, we don't need to take the spin lock while
+	 * purging the list here.
+	 */
+	BUG_ON(nbd->sock);
+	BUG_ON(nbd->active_req);
+
+	while (!list_empty(&nbd->queue_head)) {
+		req = list_entry(nbd->queue_head.next, struct request,
+				 queuelist);
+		list_del_init(&req->queuelist);
+		req->errors++;
+		nbd_end_request(req);
+	}
+
+	while (!list_empty(&nbd->waiting_queue)) {
+		req = list_entry(nbd->waiting_queue.next, struct request,
+				 queuelist);
+		list_del_init(&req->queuelist);
+		req->errors++;
+		nbd_end_request(req);
+	}
+}
+
+
+static void nbd_handle_req(struct nbd_device *nbd, struct request *req)
+{
+	if (req->cmd_type != REQ_TYPE_FS)
+		goto error_out;
+
+	nbd_cmd(req) = NBD_CMD_READ;
+	if (rq_data_dir(req) == WRITE) {
+		if ((req->cmd_flags & REQ_DISCARD)) {
+			WARN_ON(!(nbd->flags & NBD_FLAG_SEND_TRIM));
+			nbd_cmd(req) = NBD_CMD_TRIM;
+		} else
+			nbd_cmd(req) = NBD_CMD_WRITE;
+		if (nbd->flags & NBD_FLAG_READ_ONLY) {
+			dev_err(disk_to_dev(nbd->disk),
+				"Write on read-only\n");
+			goto error_out;
+		}
+	}
+
+	if (req->cmd_flags & REQ_FLUSH) {
+		BUG_ON(unlikely(blk_rq_sectors(req)));
+		nbd_cmd(req) = NBD_CMD_FLUSH;
+	}
+
+	req->errors = 0;
+
+	mutex_lock(&nbd->tx_lock);
+	if (unlikely(!nbd->sock)) {
+		mutex_unlock(&nbd->tx_lock);
+		dev_err(disk_to_dev(nbd->disk),
+			"Attempted send on closed socket\n");
+		goto error_out;
+	}
+
+	nbd->active_req = req;
+
+	if (nbd_send_req(nbd, req) != 0) {
+		dev_err(disk_to_dev(nbd->disk), "Request send failed\n");
+		req->errors++;
+		nbd_end_request(req);
+	} else {
+		spin_lock(&nbd->queue_lock);
+		list_add_tail(&req->queuelist, &nbd->queue_head);
+		spin_unlock(&nbd->queue_lock);
+	}
+
+	nbd->active_req = NULL;
+	mutex_unlock(&nbd->tx_lock);
+	wake_up_all(&nbd->active_wq);
+
+	return;
+
+error_out:
+	req->errors++;
+	nbd_end_request(req);
+}
+
+static int nbd_thread(void *data)
+{
+	struct nbd_device *nbd = data;
+	struct request *req;
+
+	set_user_nice(current, MIN_NICE);
+	while (!kthread_should_stop() || !list_empty(&nbd->waiting_queue)) {
+		/* wait for something to do */
+		wait_event_interruptible(nbd->waiting_wq,
+					 kthread_should_stop() ||
+					 !list_empty(&nbd->waiting_queue));
+
+		/* extract request */
+		if (list_empty(&nbd->waiting_queue))
+			continue;
+
+		spin_lock_irq(&nbd->queue_lock);
+		req = list_entry(nbd->waiting_queue.next, struct request,
+				 queuelist);
+		list_del_init(&req->queuelist);
+		spin_unlock_irq(&nbd->queue_lock);
+
+		/* handle request */
+		nbd_handle_req(nbd, req);
+	}
+	return 0;
+}
+
+/*
+ * We always wait for result of write, for now. It would be nice to make it optional
+ * in future
+ * if ((rq_data_dir(req) == WRITE) && (nbd->flags & NBD_WRITE_NOCHK))
+ *   { printk( "Warning: Ignoring result!\n"); nbd_end_request( req ); }
+ */
+
+static void do_nbd_request(struct request_queue *q)
+		__releases(q->queue_lock) __acquires(q->queue_lock)
+{
+	struct request *req;
+	
+	while ((req = blk_fetch_request(q)) != NULL) {
+		struct nbd_device *nbd;
+
+		spin_unlock_irq(q->queue_lock);
+
+		dprintk(DBG_BLKDEV, "%s: request %p: dequeued (flags=%x)\n",
+				req->rq_disk->disk_name, req, req->cmd_type);
+
+		nbd = req->rq_disk->private_data;
+
+		BUG_ON(nbd->magic != NBD_MAGIC);
+
+		if (unlikely(!nbd->sock)) {
+			dev_err(disk_to_dev(nbd->disk),
+				"Attempted send on closed socket\n");
+			req->errors++;
+			nbd_end_request(req);
+			spin_lock_irq(q->queue_lock);
+			continue;
+		}
+
+		spin_lock_irq(&nbd->queue_lock);
+		list_add_tail(&req->queuelist, &nbd->waiting_queue);
+		spin_unlock_irq(&nbd->queue_lock);
+
+		wake_up(&nbd->waiting_wq);
+
+		spin_lock_irq(q->queue_lock);
+	}
+}
+
+/* Must be called with tx_lock held */
+
+static int __nbd_ioctl(struct block_device *bdev, struct nbd_device *nbd,
+		       unsigned int cmd, unsigned long arg)
+{
+	switch (cmd) {
+	case NBD_DISCONNECT: {
+		struct request sreq;
+
+		dev_info(disk_to_dev(nbd->disk), "NBD_DISCONNECT\n");
+		if (!nbd->sock)
+			return -EINVAL;
+
+		mutex_unlock(&nbd->tx_lock);
+		fsync_bdev(bdev);
+		mutex_lock(&nbd->tx_lock);
+		blk_rq_init(NULL, &sreq);
+		sreq.cmd_type = REQ_TYPE_SPECIAL;
+		nbd_cmd(&sreq) = NBD_CMD_DISC;
+
+		/* Check again after getting mutex back.  */
+		if (!nbd->sock)
+			return -EINVAL;
+
+		nbd->disconnect = 1;
+
+		nbd_send_req(nbd, &sreq);
+		return 0;
+	}
+ 
+	case NBD_CLEAR_SOCK: {
+		struct socket *sock = nbd->sock;
+		nbd->sock = NULL;
+		nbd_clear_que(nbd);
+		BUG_ON(!list_empty(&nbd->queue_head));
+		BUG_ON(!list_empty(&nbd->waiting_queue));
+		kill_bdev(bdev);
+		if (sock)
+			sockfd_put(sock);
+		return 0;
+	}
+
+	case NBD_SET_SOCK: {
+		struct socket *sock;
+		int err;
+		if (nbd->sock)
+			return -EBUSY;
+		sock = sockfd_lookup(arg, &err);
+		if (sock) {
+			nbd->sock = sock;
+			if (max_part > 0)
+				bdev->bd_invalidated = 1;
+			nbd->disconnect = 0; /* we're connected now */
+			return 0;
+		}
+		return -EINVAL;
+	}
+
+	case NBD_SET_BLKSIZE:
+		nbd->blksize = arg;
+		nbd->bytesize &= ~(nbd->blksize-1);
+		bdev->bd_inode->i_size = nbd->bytesize;
+		set_blocksize(bdev, nbd->blksize);
+		set_capacity(nbd->disk, nbd->bytesize >> 9);
+		return 0;
+
+	case NBD_SET_SIZE:
+		nbd->bytesize = arg & ~(nbd->blksize-1);
+		bdev->bd_inode->i_size = nbd->bytesize;
+		set_blocksize(bdev, nbd->blksize);
+		set_capacity(nbd->disk, nbd->bytesize >> 9);
+		return 0;
+
+	case NBD_SET_TIMEOUT:
+		nbd->xmit_timeout = arg * HZ;
+		return 0;
+
+	case NBD_SET_FLAGS:
+		nbd->flags = arg;
+		return 0;
+
+	case NBD_SET_SIZE_BLOCKS:
+		nbd->bytesize = ((u64) arg) * nbd->blksize;
+		bdev->bd_inode->i_size = nbd->bytesize;
+		set_blocksize(bdev, nbd->blksize);
+		set_capacity(nbd->disk, nbd->bytesize >> 9);
+		return 0;
+
+	case NBD_DO_IT: {
+		struct task_struct *thread;
+		struct socket *sock;
+		int error;
+
+		if (nbd->pid)
+			return -EBUSY;
+		if (!nbd->sock)
+			return -EINVAL;
+
+		mutex_unlock(&nbd->tx_lock);
+
+		if (nbd->flags & NBD_FLAG_READ_ONLY)
+			set_device_ro(bdev, true);
+		if (nbd->flags & NBD_FLAG_SEND_TRIM)
+			queue_flag_set_unlocked(QUEUE_FLAG_DISCARD,
+				nbd->disk->queue);
+		if (nbd->flags & NBD_FLAG_SEND_FLUSH)
+			blk_queue_flush(nbd->disk->queue, REQ_FLUSH);
+		else
+			blk_queue_flush(nbd->disk->queue, 0);
+
+		thread = kthread_create(nbd_thread, nbd, "%s",
+					nbd->disk->disk_name);
+		if (IS_ERR(thread)) {
+			mutex_lock(&nbd->tx_lock);
+			return PTR_ERR(thread);
+		}
+		wake_up_process(thread);
+		error = nbd_do_it(nbd);
+		kthread_stop(thread);
+
+		mutex_lock(&nbd->tx_lock);
+		if (error)
+			return error;
+		sock_shutdown(nbd, 0);
+		sock = nbd->sock;
+		nbd->sock = NULL;
+		nbd_clear_que(nbd);
+		dev_warn(disk_to_dev(nbd->disk), "queue cleared\n");
+		kill_bdev(bdev);
+		queue_flag_clear_unlocked(QUEUE_FLAG_DISCARD, nbd->disk->queue);
+		set_device_ro(bdev, false);
+		if (sock)
+			sockfd_put(sock);
+		nbd->flags = 0;
+		nbd->bytesize = 0;
+		bdev->bd_inode->i_size = 0;
+		set_capacity(nbd->disk, 0);
+		if (max_part > 0)
+			ioctl_by_bdev(bdev, BLKRRPART, 0);
+		if (nbd->disconnect) /* user requested, ignore socket errors */
+			return 0;
+		return nbd->harderror;
+	}
+
+	case NBD_CLEAR_QUE:
+		/*
+		 * This is for compatibility only.  The queue is always cleared
+		 * by NBD_DO_IT or NBD_CLEAR_SOCK.
+		 */
+		return 0;
+
+	case NBD_PRINT_DEBUG:
+		dev_info(disk_to_dev(nbd->disk),
+			"next = %p, prev = %p, head = %p\n",
+			nbd->queue_head.next, nbd->queue_head.prev,
+			&nbd->queue_head);
+		return 0;
+	}
+	return -ENOTTY;
+}
+
+static int nbd_ioctl(struct block_device *bdev, fmode_t mode,
+		     unsigned int cmd, unsigned long arg)
+{
+	struct nbd_device *nbd = bdev->bd_disk->private_data;
+	int error;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	BUG_ON(nbd->magic != NBD_MAGIC);
+
+	/* Anyone capable of this syscall can do *real bad* things */
+	dprintk(DBG_IOCTL, "%s: nbd_ioctl cmd=%s(0x%x) arg=%lu\n",
+		nbd->disk->disk_name, ioctl_cmd_to_ascii(cmd), cmd, arg);
+
+	mutex_lock(&nbd->tx_lock);
+	error = __nbd_ioctl(bdev, nbd, cmd, arg);
+	mutex_unlock(&nbd->tx_lock);
+
+	return error;
+}
+
+static const struct block_device_operations nbd_fops =
+{
+	.owner =	THIS_MODULE,
+	.ioctl =	nbd_ioctl,
+};
+
+/*
+ * And here should be modules and kernel interface 
+ *  (Just smiley confuses emacs :-)
+ */
+
+static int __init nbd_init(void)
+{
+	int err = -ENOMEM;
+	int i;
+	int part_shift;
+
+	BUILD_BUG_ON(sizeof(struct nbd_request) != 28);
+
+	if (max_part < 0) {
+		printk(KERN_ERR "nbd: max_part must be >= 0\n");
+		return -EINVAL;
+	}
+
+	part_shift = 0;
+	if (max_part > 0) {
+		part_shift = fls(max_part);
+
+		/*
+		 * Adjust max_part according to part_shift as it is exported
+		 * to user space so that user can know the max number of
+		 * partition kernel should be able to manage.
+		 *
+		 * Note that -1 is required because partition 0 is reserved
+		 * for the whole disk.
+		 */
+		max_part = (1UL << part_shift) - 1;
+	}
+
+	if ((1UL << part_shift) > DISK_MAX_PARTS)
+		return -EINVAL;
+
+	if (nbds_max > 1UL << (MINORBITS - part_shift))
+		return -EINVAL;
+
+	nbd_dev = kcalloc(nbds_max, sizeof(*nbd_dev), GFP_KERNEL);
+	if (!nbd_dev)
+		return -ENOMEM;
+
+	for (i = 0; i < nbds_max; i++) {
+		struct gendisk *disk = alloc_disk(1 << part_shift);
+		if (!disk)
+			goto out;
+		nbd_dev[i].disk = disk;
+		/*
+		 * The new linux 2.5 block layer implementation requires
+		 * every gendisk to have its very own request_queue struct.
+		 * These structs are big so we dynamically allocate them.
+		 */
+		disk->queue = blk_init_queue(do_nbd_request, &nbd_lock);
+		if (!disk->queue) {
+			put_disk(disk);
+			goto out;
+		}
+		/*
+		 * Tell the block layer that we are not a rotational device
+		 */
+		queue_flag_set_unlocked(QUEUE_FLAG_NONROT, disk->queue);
+		queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, disk->queue);
+		disk->queue->limits.discard_granularity = 512;
+		disk->queue->limits.max_discard_sectors = UINT_MAX;
+		disk->queue->limits.discard_zeroes_data = 0;
+		blk_queue_max_hw_sectors(disk->queue, 65536);
+		disk->queue->limits.max_sectors = 256;
+	}
+
+	if (register_blkdev(NBD_MAJOR, "nbd")) {
+		err = -EIO;
+		goto out;
+	}
+
+	printk(KERN_INFO "nbd: registered device at major %d\n", NBD_MAJOR);
+	dprintk(DBG_INIT, "nbd: debugflags=0x%x\n", debugflags);
+
+	for (i = 0; i < nbds_max; i++) {
+		struct gendisk *disk = nbd_dev[i].disk;
+		nbd_dev[i].magic = NBD_MAGIC;
+		INIT_LIST_HEAD(&nbd_dev[i].waiting_queue);
+		spin_lock_init(&nbd_dev[i].queue_lock);
+		INIT_LIST_HEAD(&nbd_dev[i].queue_head);
+		mutex_init(&nbd_dev[i].tx_lock);
+		init_waitqueue_head(&nbd_dev[i].active_wq);
+		init_waitqueue_head(&nbd_dev[i].waiting_wq);
+		nbd_dev[i].blksize = 1024;
+		nbd_dev[i].bytesize = 0;
+		disk->major = NBD_MAJOR;
+		disk->first_minor = i << part_shift;
+		disk->fops = &nbd_fops;
+		disk->private_data = &nbd_dev[i];
+		sprintf(disk->disk_name, "nbd%d", i);
+		set_capacity(disk, 0);
+		add_disk(disk);
+	}
+
+	return 0;
+out:
+	while (i--) {
+		blk_cleanup_queue(nbd_dev[i].disk->queue);
+		put_disk(nbd_dev[i].disk);
+	}
+	kfree(nbd_dev);
+	return err;
+}
+
+static void __exit nbd_cleanup(void)
+{
+	int i;
+	for (i = 0; i < nbds_max; i++) {
+		struct gendisk *disk = nbd_dev[i].disk;
+		nbd_dev[i].magic = 0;
+		if (disk) {
+			del_gendisk(disk);
+			blk_cleanup_queue(disk->queue);
+			put_disk(disk);
+		}
+	}
+	unregister_blkdev(NBD_MAJOR, "nbd");
+	kfree(nbd_dev);
+	printk(KERN_INFO "nbd: unregistered device at major %d\n", NBD_MAJOR);
+}
+
+module_init(nbd_init);
+module_exit(nbd_cleanup);
+
+MODULE_DESCRIPTION("Network Block Device");
+MODULE_LICENSE("GPL");
+
+module_param(nbds_max, int, 0444);
+MODULE_PARM_DESC(nbds_max, "number of network block devices to initialize (default: 16)");
+module_param(max_part, int, 0444);
+MODULE_PARM_DESC(max_part, "number of partitions per device (default: 0)");
+#ifndef NDEBUG
+module_param(debugflags, int, 0644);
+MODULE_PARM_DESC(debugflags, "flags for controlling debug output");
+#endif

drivers/block/null_blk.c

@@ -0,0 +1,636 @@
+#include <linux/module.h>
+
+#include <linux/moduleparam.h>
+#include <linux/sched.h>
+#include <linux/fs.h>
+#include <linux/blkdev.h>
+#include <linux/init.h>
+#include <linux/slab.h>
+#include <linux/blk-mq.h>
+#include <linux/hrtimer.h>
+
+struct nullb_cmd {
+	struct list_head list;
+	struct llist_node ll_list;
+	struct call_single_data csd;
+	struct request *rq;
+	struct bio *bio;
+	unsigned int tag;
+	struct nullb_queue *nq;
+};
+
+struct nullb_queue {
+	unsigned long *tag_map;
+	wait_queue_head_t wait;
+	unsigned int queue_depth;
+
+	struct nullb_cmd *cmds;
+};
+
+struct nullb {
+	struct list_head list;
+	unsigned int index;
+	struct request_queue *q;
+	struct gendisk *disk;
+	struct blk_mq_tag_set tag_set;
+	struct hrtimer timer;
+	unsigned int queue_depth;
+	spinlock_t lock;
+
+	struct nullb_queue *queues;
+	unsigned int nr_queues;
+};
+
+static LIST_HEAD(nullb_list);
+static struct mutex lock;
+static int null_major;
+static int nullb_indexes;
+
+struct completion_queue {
+	struct llist_head list;
+	struct hrtimer timer;
+};
+
+/*
+ * These are per-cpu for now, they will need to be configured by the
+ * complete_queues parameter and appropriately mapped.
+ */
+static DEFINE_PER_CPU(struct completion_queue, completion_queues);
+
+enum {
+	NULL_IRQ_NONE		= 0,
+	NULL_IRQ_SOFTIRQ	= 1,
+	NULL_IRQ_TIMER		= 2,
+};
+
+enum {
+	NULL_Q_BIO		= 0,
+	NULL_Q_RQ		= 1,
+	NULL_Q_MQ		= 2,
+};
+
+static int submit_queues;
+module_param(submit_queues, int, S_IRUGO);
+MODULE_PARM_DESC(submit_queues, "Number of submission queues");
+
+static int home_node = NUMA_NO_NODE;
+module_param(home_node, int, S_IRUGO);
+MODULE_PARM_DESC(home_node, "Home node for the device");
+
+static int queue_mode = NULL_Q_MQ;
+module_param(queue_mode, int, S_IRUGO);
+MODULE_PARM_DESC(queue_mode, "Block interface to use (0=bio,1=rq,2=multiqueue)");
+
+static int gb = 250;
+module_param(gb, int, S_IRUGO);
+MODULE_PARM_DESC(gb, "Size in GB");
+
+static int bs = 512;
+module_param(bs, int, S_IRUGO);
+MODULE_PARM_DESC(bs, "Block size (in bytes)");
+
+static int nr_devices = 2;
+module_param(nr_devices, int, S_IRUGO);
+MODULE_PARM_DESC(nr_devices, "Number of devices to register");
+
+static int irqmode = NULL_IRQ_SOFTIRQ;
+module_param(irqmode, int, S_IRUGO);
+MODULE_PARM_DESC(irqmode, "IRQ completion handler. 0-none, 1-softirq, 2-timer");
+
+static int completion_nsec = 10000;
+module_param(completion_nsec, int, S_IRUGO);
+MODULE_PARM_DESC(completion_nsec, "Time in ns to complete a request in hardware. Default: 10,000ns");
+
+static int hw_queue_depth = 64;
+module_param(hw_queue_depth, int, S_IRUGO);
+MODULE_PARM_DESC(hw_queue_depth, "Queue depth for each hardware queue. Default: 64");
+
+static bool use_per_node_hctx = false;
+module_param(use_per_node_hctx, bool, S_IRUGO);
+MODULE_PARM_DESC(use_per_node_hctx, "Use per-node allocation for hardware context queues. Default: false");
+
+static void put_tag(struct nullb_queue *nq, unsigned int tag)
+{
+	clear_bit_unlock(tag, nq->tag_map);
+
+	if (waitqueue_active(&nq->wait))
+		wake_up(&nq->wait);
+}
+
+static unsigned int get_tag(struct nullb_queue *nq)
+{
+	unsigned int tag;
+
+	do {
+		tag = find_first_zero_bit(nq->tag_map, nq->queue_depth);
+		if (tag >= nq->queue_depth)
+			return -1U;
+	} while (test_and_set_bit_lock(tag, nq->tag_map));
+
+	return tag;
+}
+
+static void free_cmd(struct nullb_cmd *cmd)
+{
+	put_tag(cmd->nq, cmd->tag);
+}
+
+static struct nullb_cmd *__alloc_cmd(struct nullb_queue *nq)
+{
+	struct nullb_cmd *cmd;
+	unsigned int tag;
+
+	tag = get_tag(nq);
+	if (tag != -1U) {
+		cmd = &nq->cmds[tag];
+		cmd->tag = tag;
+		cmd->nq = nq;
+		return cmd;
+	}
+
+	return NULL;
+}
+
+static struct nullb_cmd *alloc_cmd(struct nullb_queue *nq, int can_wait)
+{
+	struct nullb_cmd *cmd;
+	DEFINE_WAIT(wait);
+
+	cmd = __alloc_cmd(nq);
+	if (cmd || !can_wait)
+		return cmd;
+
+	do {
+		prepare_to_wait(&nq->wait, &wait, TASK_UNINTERRUPTIBLE);
+		cmd = __alloc_cmd(nq);
+		if (cmd)
+			break;
+
+		io_schedule();
+	} while (1);
+
+	finish_wait(&nq->wait, &wait);
+	return cmd;
+}
+
+static void end_cmd(struct nullb_cmd *cmd)
+{
+	switch (queue_mode)  {
+	case NULL_Q_MQ:
+		blk_mq_end_request(cmd->rq, 0);
+		return;
+	case NULL_Q_RQ:
+		INIT_LIST_HEAD(&cmd->rq->queuelist);
+		blk_end_request_all(cmd->rq, 0);
+		break;
+	case NULL_Q_BIO:
+		bio_endio(cmd->bio, 0);
+		break;
+	}
+
+	free_cmd(cmd);
+}
+
+static enum hrtimer_restart null_cmd_timer_expired(struct hrtimer *timer)
+{
+	struct completion_queue *cq;
+	struct llist_node *entry;
+	struct nullb_cmd *cmd;
+
+	cq = &per_cpu(completion_queues, smp_processor_id());
+
+	while ((entry = llist_del_all(&cq->list)) != NULL) {
+		entry = llist_reverse_order(entry);
+		do {
+			cmd = container_of(entry, struct nullb_cmd, ll_list);
+			entry = entry->next;
+			end_cmd(cmd);
+		} while (entry);
+	}
+
+	return HRTIMER_NORESTART;
+}
+
+static void null_cmd_end_timer(struct nullb_cmd *cmd)
+{
+	struct completion_queue *cq = &per_cpu(completion_queues, get_cpu());
+
+	cmd->ll_list.next = NULL;
+	if (llist_add(&cmd->ll_list, &cq->list)) {
+		ktime_t kt = ktime_set(0, completion_nsec);
+
+		hrtimer_start(&cq->timer, kt, HRTIMER_MODE_REL);
+	}
+
+	put_cpu();
+}
+
+static void null_softirq_done_fn(struct request *rq)
+{
+	if (queue_mode == NULL_Q_MQ)
+		end_cmd(blk_mq_rq_to_pdu(rq));
+	else
+		end_cmd(rq->special);
+}
+
+static inline void null_handle_cmd(struct nullb_cmd *cmd)
+{
+	/* Complete IO by inline, softirq or timer */
+	switch (irqmode) {
+	case NULL_IRQ_SOFTIRQ:
+		switch (queue_mode)  {
+		case NULL_Q_MQ:
+			blk_mq_complete_request(cmd->rq);
+			break;
+		case NULL_Q_RQ:
+			blk_complete_request(cmd->rq);
+			break;
+		case NULL_Q_BIO:
+			/*
+			 * XXX: no proper submitting cpu information available.
+			 */
+			end_cmd(cmd);
+			break;
+		}
+		break;
+	case NULL_IRQ_NONE:
+		end_cmd(cmd);
+		break;
+	case NULL_IRQ_TIMER:
+		null_cmd_end_timer(cmd);
+		break;
+	}
+}
+
+static struct nullb_queue *nullb_to_queue(struct nullb *nullb)
+{
+	int index = 0;
+
+	if (nullb->nr_queues != 1)
+		index = raw_smp_processor_id() / ((nr_cpu_ids + nullb->nr_queues - 1) / nullb->nr_queues);
+
+	return &nullb->queues[index];
+}
+
+static void null_queue_bio(struct request_queue *q, struct bio *bio)
+{
+	struct nullb *nullb = q->queuedata;
+	struct nullb_queue *nq = nullb_to_queue(nullb);
+	struct nullb_cmd *cmd;
+
+	cmd = alloc_cmd(nq, 1);
+	cmd->bio = bio;
+
+	null_handle_cmd(cmd);
+}
+
+static int null_rq_prep_fn(struct request_queue *q, struct request *req)
+{
+	struct nullb *nullb = q->queuedata;
+	struct nullb_queue *nq = nullb_to_queue(nullb);
+	struct nullb_cmd *cmd;
+
+	cmd = alloc_cmd(nq, 0);
+	if (cmd) {
+		cmd->rq = req;
+		req->special = cmd;
+		return BLKPREP_OK;
+	}
+
+	return BLKPREP_DEFER;
+}
+
+static void null_request_fn(struct request_queue *q)
+{
+	struct request *rq;
+
+	while ((rq = blk_fetch_request(q)) != NULL) {
+		struct nullb_cmd *cmd = rq->special;
+
+		spin_unlock_irq(q->queue_lock);
+		null_handle_cmd(cmd);
+		spin_lock_irq(q->queue_lock);
+	}
+}
+
+static int null_queue_rq(struct blk_mq_hw_ctx *hctx, struct request *rq,
+		bool last)
+{
+	struct nullb_cmd *cmd = blk_mq_rq_to_pdu(rq);
+
+	cmd->rq = rq;
+	cmd->nq = hctx->driver_data;
+
+	blk_mq_start_request(rq);
+
+	null_handle_cmd(cmd);
+	return BLK_MQ_RQ_QUEUE_OK;
+}
+
+static void null_init_queue(struct nullb *nullb, struct nullb_queue *nq)
+{
+	BUG_ON(!nullb);
+	BUG_ON(!nq);
+
+	init_waitqueue_head(&nq->wait);
+	nq->queue_depth = nullb->queue_depth;
+}
+
+static int null_init_hctx(struct blk_mq_hw_ctx *hctx, void *data,
+			  unsigned int index)
+{
+	struct nullb *nullb = data;
+	struct nullb_queue *nq = &nullb->queues[index];
+
+	hctx->driver_data = nq;
+	null_init_queue(nullb, nq);
+	nullb->nr_queues++;
+
+	return 0;
+}
+
+static struct blk_mq_ops null_mq_ops = {
+	.queue_rq       = null_queue_rq,
+	.map_queue      = blk_mq_map_queue,
+	.init_hctx	= null_init_hctx,
+	.complete	= null_softirq_done_fn,
+};
+
+static void null_del_dev(struct nullb *nullb)
+{
+	list_del_init(&nullb->list);
+
+	del_gendisk(nullb->disk);
+	blk_cleanup_queue(nullb->q);
+	if (queue_mode == NULL_Q_MQ)
+		blk_mq_free_tag_set(&nullb->tag_set);
+	put_disk(nullb->disk);
+	kfree(nullb);
+}
+
+static int null_open(struct block_device *bdev, fmode_t mode)
+{
+	return 0;
+}
+
+static void null_release(struct gendisk *disk, fmode_t mode)
+{
+}
+
+static const struct block_device_operations null_fops = {
+	.owner =	THIS_MODULE,
+	.open =		null_open,
+	.release =	null_release,
+};
+
+static int setup_commands(struct nullb_queue *nq)
+{
+	struct nullb_cmd *cmd;
+	int i, tag_size;
+
+	nq->cmds = kzalloc(nq->queue_depth * sizeof(*cmd), GFP_KERNEL);
+	if (!nq->cmds)
+		return -ENOMEM;
+
+	tag_size = ALIGN(nq->queue_depth, BITS_PER_LONG) / BITS_PER_LONG;
+	nq->tag_map = kzalloc(tag_size * sizeof(unsigned long), GFP_KERNEL);
+	if (!nq->tag_map) {
+		kfree(nq->cmds);
+		return -ENOMEM;
+	}
+
+	for (i = 0; i < nq->queue_depth; i++) {
+		cmd = &nq->cmds[i];
+		INIT_LIST_HEAD(&cmd->list);
+		cmd->ll_list.next = NULL;
+		cmd->tag = -1U;
+	}
+
+	return 0;
+}
+
+static void cleanup_queue(struct nullb_queue *nq)
+{
+	kfree(nq->tag_map);
+	kfree(nq->cmds);
+}
+
+static void cleanup_queues(struct nullb *nullb)
+{
+	int i;
+
+	for (i = 0; i < nullb->nr_queues; i++)
+		cleanup_queue(&nullb->queues[i]);
+
+	kfree(nullb->queues);
+}
+
+static int setup_queues(struct nullb *nullb)
+{
+	nullb->queues = kzalloc(submit_queues * sizeof(struct nullb_queue),
+								GFP_KERNEL);
+	if (!nullb->queues)
+		return -ENOMEM;
+
+	nullb->nr_queues = 0;
+	nullb->queue_depth = hw_queue_depth;
+
+	return 0;
+}
+
+static int init_driver_queues(struct nullb *nullb)
+{
+	struct nullb_queue *nq;
+	int i, ret = 0;
+
+	for (i = 0; i < submit_queues; i++) {
+		nq = &nullb->queues[i];
+
+		null_init_queue(nullb, nq);
+
+		ret = setup_commands(nq);
+		if (ret)
+			return ret;
+		nullb->nr_queues++;
+	}
+	return 0;
+}
+
+static int null_add_dev(void)
+{
+	struct gendisk *disk;
+	struct nullb *nullb;
+	sector_t size;
+	int rv;
+
+	nullb = kzalloc_node(sizeof(*nullb), GFP_KERNEL, home_node);
+	if (!nullb) {
+		rv = -ENOMEM;
+		goto out;
+	}
+
+	spin_lock_init(&nullb->lock);
+
+	if (queue_mode == NULL_Q_MQ && use_per_node_hctx)
+		submit_queues = nr_online_nodes;
+
+	rv = setup_queues(nullb);
+	if (rv)
+		goto out_free_nullb;
+
+	if (queue_mode == NULL_Q_MQ) {
+		nullb->tag_set.ops = &null_mq_ops;
+		nullb->tag_set.nr_hw_queues = submit_queues;
+		nullb->tag_set.queue_depth = hw_queue_depth;
+		nullb->tag_set.numa_node = home_node;
+		nullb->tag_set.cmd_size	= sizeof(struct nullb_cmd);
+		nullb->tag_set.flags = BLK_MQ_F_SHOULD_MERGE;
+		nullb->tag_set.driver_data = nullb;
+
+		rv = blk_mq_alloc_tag_set(&nullb->tag_set);
+		if (rv)
+			goto out_cleanup_queues;
+
+		nullb->q = blk_mq_init_queue(&nullb->tag_set);
+		if (!nullb->q) {
+			rv = -ENOMEM;
+			goto out_cleanup_tags;
+		}
+	} else if (queue_mode == NULL_Q_BIO) {
+		nullb->q = blk_alloc_queue_node(GFP_KERNEL, home_node);
+		if (!nullb->q) {
+			rv = -ENOMEM;
+			goto out_cleanup_queues;
+		}
+		blk_queue_make_request(nullb->q, null_queue_bio);
+		rv = init_driver_queues(nullb);
+		if (rv)
+			goto out_cleanup_blk_queue;
+	} else {
+		nullb->q = blk_init_queue_node(null_request_fn, &nullb->lock, home_node);
+		if (!nullb->q) {
+			rv = -ENOMEM;
+			goto out_cleanup_queues;
+		}
+		blk_queue_prep_rq(nullb->q, null_rq_prep_fn);
+		blk_queue_softirq_done(nullb->q, null_softirq_done_fn);
+		rv = init_driver_queues(nullb);
+		if (rv)
+			goto out_cleanup_blk_queue;
+	}
+
+	nullb->q->queuedata = nullb;
+	queue_flag_set_unlocked(QUEUE_FLAG_NONROT, nullb->q);
+	queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, nullb->q);
+
+	disk = nullb->disk = alloc_disk_node(1, home_node);
+	if (!disk) {
+		rv = -ENOMEM;
+		goto out_cleanup_blk_queue;
+	}
+
+	mutex_lock(&lock);
+	list_add_tail(&nullb->list, &nullb_list);
+	nullb->index = nullb_indexes++;
+	mutex_unlock(&lock);
+
+	blk_queue_logical_block_size(nullb->q, bs);
+	blk_queue_physical_block_size(nullb->q, bs);
+
+	size = gb * 1024 * 1024 * 1024ULL;
+	sector_div(size, bs);
+	set_capacity(disk, size);
+
+	disk->flags |= GENHD_FL_EXT_DEVT;
+	disk->major		= null_major;
+	disk->first_minor	= nullb->index;
+	disk->fops		= &null_fops;
+	disk->private_data	= nullb;
+	disk->queue		= nullb->q;
+	sprintf(disk->disk_name, "nullb%d", nullb->index);
+	add_disk(disk);
+	return 0;
+
+out_cleanup_blk_queue:
+	blk_cleanup_queue(nullb->q);
+out_cleanup_tags:
+	if (queue_mode == NULL_Q_MQ)
+		blk_mq_free_tag_set(&nullb->tag_set);
+out_cleanup_queues:
+	cleanup_queues(nullb);
+out_free_nullb:
+	kfree(nullb);
+out:
+	return rv;
+}
+
+static int __init null_init(void)
+{
+	unsigned int i;
+
+	if (bs > PAGE_SIZE) {
+		pr_warn("null_blk: invalid block size\n");
+		pr_warn("null_blk: defaults block size to %lu\n", PAGE_SIZE);
+		bs = PAGE_SIZE;
+	}
+
+	if (queue_mode == NULL_Q_MQ && use_per_node_hctx) {
+		if (submit_queues < nr_online_nodes) {
+			pr_warn("null_blk: submit_queues param is set to %u.",
+							nr_online_nodes);
+			submit_queues = nr_online_nodes;
+		}
+	} else if (submit_queues > nr_cpu_ids)
+		submit_queues = nr_cpu_ids;
+	else if (!submit_queues)
+		submit_queues = 1;
+
+	mutex_init(&lock);
+
+	/* Initialize a separate list for each CPU for issuing softirqs */
+	for_each_possible_cpu(i) {
+		struct completion_queue *cq = &per_cpu(completion_queues, i);
+
+		init_llist_head(&cq->list);
+
+		if (irqmode != NULL_IRQ_TIMER)
+			continue;
+
+		hrtimer_init(&cq->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
+		cq->timer.function = null_cmd_timer_expired;
+	}
+
+	null_major = register_blkdev(0, "nullb");
+	if (null_major < 0)
+		return null_major;
+
+	for (i = 0; i < nr_devices; i++) {
+		if (null_add_dev()) {
+			unregister_blkdev(null_major, "nullb");
+			return -EINVAL;
+		}
+	}
+
+	pr_info("null: module loaded\n");
+	return 0;
+}
+
+static void __exit null_exit(void)
+{
+	struct nullb *nullb;
+
+	unregister_blkdev(null_major, "nullb");
+
+	mutex_lock(&lock);
+	while (!list_empty(&nullb_list)) {
+		nullb = list_entry(nullb_list.next, struct nullb, list);
+		null_del_dev(nullb);
+	}
+	mutex_unlock(&lock);
+}
+
+module_init(null_init);
+module_exit(null_exit);
+
+MODULE_AUTHOR("Jens Axboe <jaxboe@fusionio.com>");
+MODULE_LICENSE("GPL");

drivers/block/osdblk.c

@@ -0,0 +1,699 @@
+
+/*
+   osdblk.c -- Export a single SCSI OSD object as a Linux block device
+
+
+   Copyright 2009 Red Hat, Inc.
+
+   This program is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program; see the file COPYING.  If not, write to
+   the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+
+
+   Instructions for use
+   --------------------
+
+   1) Map a Linux block device to an existing OSD object.
+
+      In this example, we will use partition id 1234, object id 5678,
+      OSD device /dev/osd1.
+
+      $ echo "1234 5678 /dev/osd1" > /sys/class/osdblk/add
+
+
+   2) List all active blkdev<->object mappings.
+
+      In this example, we have performed step #1 twice, creating two blkdevs,
+      mapped to two separate OSD objects.
+
+      $ cat /sys/class/osdblk/list
+      0 174 1234 5678 /dev/osd1
+      1 179 1994 897123 /dev/osd0
+
+      The columns, in order, are:
+      - blkdev unique id
+      - blkdev assigned major
+      - OSD object partition id
+      - OSD object id
+      - OSD device
+
+
+   3) Remove an active blkdev<->object mapping.
+
+      In this example, we remove the mapping with blkdev unique id 1.
+
+      $ echo 1 > /sys/class/osdblk/remove
+
+
+   NOTE:  The actual creation and deletion of OSD objects is outside the scope
+   of this driver.
+
+ */
+
+#include <linux/kernel.h>
+#include <linux/device.h>
+#include <linux/module.h>
+#include <linux/fs.h>
+#include <linux/slab.h>
+#include <scsi/osd_initiator.h>
+#include <scsi/osd_attributes.h>
+#include <scsi/osd_sec.h>
+#include <scsi/scsi_device.h>
+
+#define DRV_NAME "osdblk"
+#define PFX DRV_NAME ": "
+
+/* #define _OSDBLK_DEBUG */
+#ifdef _OSDBLK_DEBUG
+#define OSDBLK_DEBUG(fmt, a...) \
+	printk(KERN_NOTICE "osdblk @%s:%d: " fmt, __func__, __LINE__, ##a)
+#else
+#define OSDBLK_DEBUG(fmt, a...) \
+	do { if (0) printk(fmt, ##a); } while (0)
+#endif
+
+MODULE_AUTHOR("Jeff Garzik <jeff@garzik.org>");
+MODULE_DESCRIPTION("block device inside an OSD object osdblk.ko");
+MODULE_LICENSE("GPL");
+
+struct osdblk_device;
+
+enum {
+	OSDBLK_MINORS_PER_MAJOR	= 256,		/* max minors per blkdev */
+	OSDBLK_MAX_REQ		= 32,		/* max parallel requests */
+	OSDBLK_OP_TIMEOUT	= 4 * 60,	/* sync OSD req timeout */
+};
+
+struct osdblk_request {
+	struct request		*rq;		/* blk layer request */
+	struct bio		*bio;		/* cloned bio */
+	struct osdblk_device	*osdev;		/* associated blkdev */
+};
+
+struct osdblk_device {
+	int			id;		/* blkdev unique id */
+
+	int			major;		/* blkdev assigned major */
+	struct gendisk		*disk;		/* blkdev's gendisk and rq */
+	struct request_queue	*q;
+
+	struct osd_dev		*osd;		/* associated OSD */
+
+	char			name[32];	/* blkdev name, e.g. osdblk34 */
+
+	spinlock_t		lock;		/* queue lock */
+
+	struct osd_obj_id	obj;		/* OSD partition, obj id */
+	uint8_t			obj_cred[OSD_CAP_LEN]; /* OSD cred */
+
+	struct osdblk_request	req[OSDBLK_MAX_REQ]; /* request table */
+
+	struct list_head	node;
+
+	char			osd_path[0];	/* OSD device path */
+};
+
+static struct class *class_osdblk;		/* /sys/class/osdblk */
+static DEFINE_MUTEX(ctl_mutex);	/* Serialize open/close/setup/teardown */
+static LIST_HEAD(osdblkdev_list);
+
+static const struct block_device_operations osdblk_bd_ops = {
+	.owner		= THIS_MODULE,
+};
+
+static const struct osd_attr g_attr_logical_length = ATTR_DEF(
+	OSD_APAGE_OBJECT_INFORMATION, OSD_ATTR_OI_LOGICAL_LENGTH, 8);
+
+static void osdblk_make_credential(u8 cred_a[OSD_CAP_LEN],
+				   const struct osd_obj_id *obj)
+{
+	osd_sec_init_nosec_doall_caps(cred_a, obj, false, true);
+}
+
+/* copied from exofs; move to libosd? */
+/*
+ * Perform a synchronous OSD operation.  copied from exofs; move to libosd?
+ */
+static int osd_sync_op(struct osd_request *or, int timeout, uint8_t *credential)
+{
+	int ret;
+
+	or->timeout = timeout;
+	ret = osd_finalize_request(or, 0, credential, NULL);
+	if (ret)
+		return ret;
+
+	ret = osd_execute_request(or);
+
+	/* osd_req_decode_sense(or, ret); */
+	return ret;
+}
+
+/*
+ * Perform an asynchronous OSD operation.  copied from exofs; move to libosd?
+ */
+static int osd_async_op(struct osd_request *or, osd_req_done_fn *async_done,
+		   void *caller_context, u8 *cred)
+{
+	int ret;
+
+	ret = osd_finalize_request(or, 0, cred, NULL);
+	if (ret)
+		return ret;
+
+	ret = osd_execute_request_async(or, async_done, caller_context);
+
+	return ret;
+}
+
+/* copied from exofs; move to libosd? */
+static int extract_attr_from_req(struct osd_request *or, struct osd_attr *attr)
+{
+	struct osd_attr cur_attr = {.attr_page = 0}; /* start with zeros */
+	void *iter = NULL;
+	int nelem;
+
+	do {
+		nelem = 1;
+		osd_req_decode_get_attr_list(or, &cur_attr, &nelem, &iter);
+		if ((cur_attr.attr_page == attr->attr_page) &&
+		    (cur_attr.attr_id == attr->attr_id)) {
+			attr->len = cur_attr.len;
+			attr->val_ptr = cur_attr.val_ptr;
+			return 0;
+		}
+	} while (iter);
+
+	return -EIO;
+}
+
+static int osdblk_get_obj_size(struct osdblk_device *osdev, u64 *size_out)
+{
+	struct osd_request *or;
+	struct osd_attr attr;
+	int ret;
+
+	/* start request */
+	or = osd_start_request(osdev->osd, GFP_KERNEL);
+	if (!or)
+		return -ENOMEM;
+
+	/* create a get-attributes(length) request */
+	osd_req_get_attributes(or, &osdev->obj);
+
+	osd_req_add_get_attr_list(or, &g_attr_logical_length, 1);
+
+	/* execute op synchronously */
+	ret = osd_sync_op(or, OSDBLK_OP_TIMEOUT, osdev->obj_cred);
+	if (ret)
+		goto out;
+
+	/* extract length from returned attribute info */
+	attr = g_attr_logical_length;
+	ret = extract_attr_from_req(or, &attr);
+	if (ret)
+		goto out;
+
+	*size_out = get_unaligned_be64(attr.val_ptr);
+
+out:
+	osd_end_request(or);
+	return ret;
+
+}
+
+static void osdblk_osd_complete(struct osd_request *or, void *private)
+{
+	struct osdblk_request *orq = private;
+	struct osd_sense_info osi;
+	int ret = osd_req_decode_sense(or, &osi);
+
+	if (ret) {
+		ret = -EIO;
+		OSDBLK_DEBUG("osdblk_osd_complete with err=%d\n", ret);
+	}
+
+	/* complete OSD request */
+	osd_end_request(or);
+
+	/* complete request passed to osdblk by block layer */
+	__blk_end_request_all(orq->rq, ret);
+}
+
+static void bio_chain_put(struct bio *chain)
+{
+	struct bio *tmp;
+
+	while (chain) {
+		tmp = chain;
+		chain = chain->bi_next;
+
+		bio_put(tmp);
+	}
+}
+
+static struct bio *bio_chain_clone(struct bio *old_chain, gfp_t gfpmask)
+{
+	struct bio *tmp, *new_chain = NULL, *tail = NULL;
+
+	while (old_chain) {
+		tmp = bio_clone_kmalloc(old_chain, gfpmask);
+		if (!tmp)
+			goto err_out;
+
+		tmp->bi_bdev = NULL;
+		gfpmask &= ~__GFP_WAIT;
+		tmp->bi_next = NULL;
+
+		if (!new_chain)
+			new_chain = tail = tmp;
+		else {
+			tail->bi_next = tmp;
+			tail = tmp;
+		}
+
+		old_chain = old_chain->bi_next;
+	}
+
+	return new_chain;
+
+err_out:
+	OSDBLK_DEBUG("bio_chain_clone with err\n");
+	bio_chain_put(new_chain);
+	return NULL;
+}
+
+static void osdblk_rq_fn(struct request_queue *q)
+{
+	struct osdblk_device *osdev = q->queuedata;
+
+	while (1) {
+		struct request *rq;
+		struct osdblk_request *orq;
+		struct osd_request *or;
+		struct bio *bio;
+		bool do_write, do_flush;
+
+		/* peek at request from block layer */
+		rq = blk_fetch_request(q);
+		if (!rq)
+			break;
+
+		/* filter out block requests we don't understand */
+		if (rq->cmd_type != REQ_TYPE_FS) {
+			blk_end_request_all(rq, 0);
+			continue;
+		}
+
+		/* deduce our operation (read, write, flush) */
+		/* I wish the block layer simplified cmd_type/cmd_flags/cmd[]
+		 * into a clearly defined set of RPC commands:
+		 * read, write, flush, scsi command, power mgmt req,
+		 * driver-specific, etc.
+		 */
+
+		do_flush = rq->cmd_flags & REQ_FLUSH;
+		do_write = (rq_data_dir(rq) == WRITE);
+
+		if (!do_flush) { /* osd_flush does not use a bio */
+			/* a bio clone to be passed down to OSD request */
+			bio = bio_chain_clone(rq->bio, GFP_ATOMIC);
+			if (!bio)
+				break;
+		} else
+			bio = NULL;
+
+		/* alloc internal OSD request, for OSD command execution */
+		or = osd_start_request(osdev->osd, GFP_ATOMIC);
+		if (!or) {
+			bio_chain_put(bio);
+			OSDBLK_DEBUG("osd_start_request with err\n");
+			break;
+		}
+
+		orq = &osdev->req[rq->tag];
+		orq->rq = rq;
+		orq->bio = bio;
+		orq->osdev = osdev;
+
+		/* init OSD command: flush, write or read */
+		if (do_flush)
+			osd_req_flush_object(or, &osdev->obj,
+					     OSD_CDB_FLUSH_ALL, 0, 0);
+		else if (do_write)
+			osd_req_write(or, &osdev->obj, blk_rq_pos(rq) * 512ULL,
+				      bio, blk_rq_bytes(rq));
+		else
+			osd_req_read(or, &osdev->obj, blk_rq_pos(rq) * 512ULL,
+				     bio, blk_rq_bytes(rq));
+
+		OSDBLK_DEBUG("%s 0x%x bytes at 0x%llx\n",
+			do_flush ? "flush" : do_write ?
+				"write" : "read", blk_rq_bytes(rq),
+			blk_rq_pos(rq) * 512ULL);
+
+		/* begin OSD command execution */
+		if (osd_async_op(or, osdblk_osd_complete, orq,
+				 osdev->obj_cred)) {
+			osd_end_request(or);
+			blk_requeue_request(q, rq);
+			bio_chain_put(bio);
+			OSDBLK_DEBUG("osd_execute_request_async with err\n");
+			break;
+		}
+
+		/* remove the special 'flush' marker, now that the command
+		 * is executing
+		 */
+		rq->special = NULL;
+	}
+}
+
+static void osdblk_free_disk(struct osdblk_device *osdev)
+{
+	struct gendisk *disk = osdev->disk;
+
+	if (!disk)
+		return;
+
+	if (disk->flags & GENHD_FL_UP)
+		del_gendisk(disk);
+	if (disk->queue)
+		blk_cleanup_queue(disk->queue);
+	put_disk(disk);
+}
+
+static int osdblk_init_disk(struct osdblk_device *osdev)
+{
+	struct gendisk *disk;
+	struct request_queue *q;
+	int rc;
+	u64 obj_size = 0;
+
+	/* contact OSD, request size info about the object being mapped */
+	rc = osdblk_get_obj_size(osdev, &obj_size);
+	if (rc)
+		return rc;
+
+	/* create gendisk info */
+	disk = alloc_disk(OSDBLK_MINORS_PER_MAJOR);
+	if (!disk)
+		return -ENOMEM;
+
+	sprintf(disk->disk_name, DRV_NAME "%d", osdev->id);
+	disk->major = osdev->major;
+	disk->first_minor = 0;
+	disk->fops = &osdblk_bd_ops;
+	disk->private_data = osdev;
+
+	/* init rq */
+	q = blk_init_queue(osdblk_rq_fn, &osdev->lock);
+	if (!q) {
+		put_disk(disk);
+		return -ENOMEM;
+	}
+
+	/* switch queue to TCQ mode; allocate tag map */
+	rc = blk_queue_init_tags(q, OSDBLK_MAX_REQ, NULL);
+	if (rc) {
+		blk_cleanup_queue(q);
+		put_disk(disk);
+		return rc;
+	}
+
+	/* Set our limits to the lower device limits, because osdblk cannot
+	 * sleep when allocating a lower-request and therefore cannot be
+	 * bouncing.
+	 */
+	blk_queue_stack_limits(q, osd_request_queue(osdev->osd));
+
+	blk_queue_prep_rq(q, blk_queue_start_tag);
+	blk_queue_flush(q, REQ_FLUSH);
+
+	disk->queue = q;
+
+	q->queuedata = osdev;
+
+	osdev->disk = disk;
+	osdev->q = q;
+
+	/* finally, announce the disk to the world */
+	set_capacity(disk, obj_size / 512ULL);
+	add_disk(disk);
+
+	printk(KERN_INFO "%s: Added of size 0x%llx\n",
+		disk->disk_name, (unsigned long long)obj_size);
+
+	return 0;
+}
+
+/********************************************************************
+ * /sys/class/osdblk/
+ *                   add	map OSD object to blkdev
+ *                   remove	unmap OSD object
+ *                   list	show mappings
+ *******************************************************************/
+
+static void class_osdblk_release(struct class *cls)
+{
+	kfree(cls);
+}
+
+static ssize_t class_osdblk_list(struct class *c,
+				struct class_attribute *attr,
+				char *data)
+{
+	int n = 0;
+	struct list_head *tmp;
+
+	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+
+	list_for_each(tmp, &osdblkdev_list) {
+		struct osdblk_device *osdev;
+
+		osdev = list_entry(tmp, struct osdblk_device, node);
+
+		n += sprintf(data+n, "%d %d %llu %llu %s\n",
+			osdev->id,
+			osdev->major,
+			osdev->obj.partition,
+			osdev->obj.id,
+			osdev->osd_path);
+	}
+
+	mutex_unlock(&ctl_mutex);
+	return n;
+}
+
+static ssize_t class_osdblk_add(struct class *c,
+				struct class_attribute *attr,
+				const char *buf, size_t count)
+{
+	struct osdblk_device *osdev;
+	ssize_t rc;
+	int irc, new_id = 0;
+	struct list_head *tmp;
+
+	if (!try_module_get(THIS_MODULE))
+		return -ENODEV;
+
+	/* new osdblk_device object */
+	osdev = kzalloc(sizeof(*osdev) + strlen(buf) + 1, GFP_KERNEL);
+	if (!osdev) {
+		rc = -ENOMEM;
+		goto err_out_mod;
+	}
+
+	/* static osdblk_device initialization */
+	spin_lock_init(&osdev->lock);
+	INIT_LIST_HEAD(&osdev->node);
+
+	/* generate unique id: find highest unique id, add one */
+
+	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+
+	list_for_each(tmp, &osdblkdev_list) {
+		struct osdblk_device *osdev;
+
+		osdev = list_entry(tmp, struct osdblk_device, node);
+		if (osdev->id > new_id)
+			new_id = osdev->id + 1;
+	}
+
+	osdev->id = new_id;
+
+	/* add to global list */
+	list_add_tail(&osdev->node, &osdblkdev_list);
+
+	mutex_unlock(&ctl_mutex);
+
+	/* parse add command */
+	if (sscanf(buf, "%llu %llu %s", &osdev->obj.partition, &osdev->obj.id,
+		   osdev->osd_path) != 3) {
+		rc = -EINVAL;
+		goto err_out_slot;
+	}
+
+	/* initialize rest of new object */
+	sprintf(osdev->name, DRV_NAME "%d", osdev->id);
+
+	/* contact requested OSD */
+	osdev->osd = osduld_path_lookup(osdev->osd_path);
+	if (IS_ERR(osdev->osd)) {
+		rc = PTR_ERR(osdev->osd);
+		goto err_out_slot;
+	}
+
+	/* build OSD credential */
+	osdblk_make_credential(osdev->obj_cred, &osdev->obj);
+
+	/* register our block device */
+	irc = register_blkdev(0, osdev->name);
+	if (irc < 0) {
+		rc = irc;
+		goto err_out_osd;
+	}
+
+	osdev->major = irc;
+
+	/* set up and announce blkdev mapping */
+	rc = osdblk_init_disk(osdev);
+	if (rc)
+		goto err_out_blkdev;
+
+	return count;
+
+err_out_blkdev:
+	unregister_blkdev(osdev->major, osdev->name);
+err_out_osd:
+	osduld_put_device(osdev->osd);
+err_out_slot:
+	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+	list_del_init(&osdev->node);
+	mutex_unlock(&ctl_mutex);
+
+	kfree(osdev);
+err_out_mod:
+	OSDBLK_DEBUG("Error adding device %s\n", buf);
+	module_put(THIS_MODULE);
+	return rc;
+}
+
+static ssize_t class_osdblk_remove(struct class *c,
+					struct class_attribute *attr,
+					const char *buf,
+					size_t count)
+{
+	struct osdblk_device *osdev = NULL;
+	int target_id, rc;
+	unsigned long ul;
+	struct list_head *tmp;
+
+	rc = kstrtoul(buf, 10, &ul);
+	if (rc)
+		return rc;
+
+	/* convert to int; abort if we lost anything in the conversion */
+	target_id = (int) ul;
+	if (target_id != ul)
+		return -EINVAL;
+
+	/* remove object from list immediately */
+	mutex_lock_nested(&ctl_mutex, SINGLE_DEPTH_NESTING);
+
+	list_for_each(tmp, &osdblkdev_list) {
+		osdev = list_entry(tmp, struct osdblk_device, node);
+		if (osdev->id == target_id) {
+			list_del_init(&osdev->node);
+			break;
+		}
+		osdev = NULL;
+	}
+
+	mutex_unlock(&ctl_mutex);
+
+	if (!osdev)
+		return -ENOENT;
+
+	/* clean up and free blkdev and associated OSD connection */
+	osdblk_free_disk(osdev);
+	unregister_blkdev(osdev->major, osdev->name);
+	osduld_put_device(osdev->osd);
+	kfree(osdev);
+
+	/* release module ref */
+	module_put(THIS_MODULE);
+
+	return count;
+}
+
+static struct class_attribute class_osdblk_attrs[] = {
+	__ATTR(add,	0200, NULL, class_osdblk_add),
+	__ATTR(remove,	0200, NULL, class_osdblk_remove),
+	__ATTR(list,	0444, class_osdblk_list, NULL),
+	__ATTR_NULL
+};
+
+static int osdblk_sysfs_init(void)
+{
+	int ret = 0;
+
+	/*
+	 * create control files in sysfs
+	 * /sys/class/osdblk/...
+	 */
+	class_osdblk = kzalloc(sizeof(*class_osdblk), GFP_KERNEL);
+	if (!class_osdblk)
+		return -ENOMEM;
+
+	class_osdblk->name = DRV_NAME;
+	class_osdblk->owner = THIS_MODULE;
+	class_osdblk->class_release = class_osdblk_release;
+	class_osdblk->class_attrs = class_osdblk_attrs;
+
+	ret = class_register(class_osdblk);
+	if (ret) {
+		kfree(class_osdblk);
+		class_osdblk = NULL;
+		printk(PFX "failed to create class osdblk\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static void osdblk_sysfs_cleanup(void)
+{
+	if (class_osdblk)
+		class_destroy(class_osdblk);
+	class_osdblk = NULL;
+}
+
+static int __init osdblk_init(void)
+{
+	int rc;
+
+	rc = osdblk_sysfs_init();
+	if (rc)
+		return rc;
+
+	return 0;
+}
+
+static void __exit osdblk_exit(void)
+{
+	osdblk_sysfs_cleanup();
+}
+
+module_init(osdblk_init);
+module_exit(osdblk_exit);
+

drivers/block/paride/Kconfig

@@ -0,0 +1,300 @@
+#
+# PARIDE configuration
+#
+# PARIDE doesn't need PARPORT, but if PARPORT is configured as a module,
+# PARIDE must also be a module.
+# PARIDE only supports PC style parports. Tough for USB or other parports...
+
+comment "Parallel IDE high-level drivers"
+	depends on PARIDE
+
+config PARIDE_PD
+	tristate "Parallel port IDE disks"
+	depends on PARIDE
+	help
+	  This option enables the high-level driver for IDE-type disk devices
+	  connected through a parallel port. If you chose to build PARIDE
+	  support into your kernel, you may answer Y here to build in the
+	  parallel port IDE driver, otherwise you should answer M to build
+	  it as a loadable module. The module will be called pd. You
+	  must also have at least one parallel port protocol driver in your
+	  system. Among the devices supported by this driver are the SyQuest
+	  EZ-135, EZ-230 and SparQ drives, the Avatar Shark and the backpack
+	  hard drives from MicroSolutions.
+
+config PARIDE_PCD
+	tristate "Parallel port ATAPI CD-ROMs"
+	depends on PARIDE
+	---help---
+	  This option enables the high-level driver for ATAPI CD-ROM devices
+	  connected through a parallel port. If you chose to build PARIDE
+	  support into your kernel, you may answer Y here to build in the
+	  parallel port ATAPI CD-ROM driver, otherwise you should answer M to
+	  build it as a loadable module. The module will be called pcd. You
+	  must also have at least one parallel port protocol driver in your
+	  system. Among the devices supported by this driver are the
+	  MicroSolutions backpack CD-ROM drives and the Freecom Power CD. If
+	  you have such a CD-ROM drive, you should also say Y or M to "ISO
+	  9660 CD-ROM file system support" below, because that's the file
+	  system used on CD-ROMs.
+
+config PARIDE_PF
+	tristate "Parallel port ATAPI disks"
+	depends on PARIDE
+	help
+	  This option enables the high-level driver for ATAPI disk devices
+	  connected through a parallel port. If you chose to build PARIDE
+	  support into your kernel, you may answer Y here to build in the
+	  parallel port ATAPI disk driver, otherwise you should answer M
+	  to build it as a loadable module. The module will be called pf.
+	  You must also have at least one parallel port protocol driver in
+	  your system. Among the devices supported by this driver are the
+	  MicroSolutions backpack PD/CD drive and the Imation Superdisk
+	  LS-120 drive.
+
+config PARIDE_PT
+	tristate "Parallel port ATAPI tapes"
+	depends on PARIDE
+	help
+	  This option enables the high-level driver for ATAPI tape devices
+	  connected through a parallel port. If you chose to build PARIDE
+	  support into your kernel, you may answer Y here to build in the
+	  parallel port ATAPI disk driver, otherwise you should answer M
+	  to build it as a loadable module. The module will be called pt.
+	  You must also have at least one parallel port protocol driver in
+	  your system. Among the devices supported by this driver is the
+	  parallel port version of the HP 5GB drive.
+
+config PARIDE_PG
+	tristate "Parallel port generic ATAPI devices"
+	depends on PARIDE
+	---help---
+	  This option enables a special high-level driver for generic ATAPI
+	  devices connected through a parallel port. The driver allows user
+	  programs, such as cdrtools, to send ATAPI commands directly to a
+	  device.
+
+	  If you chose to build PARIDE support into your kernel, you may
+	  answer Y here to build in the parallel port generic ATAPI driver,
+	  otherwise you should answer M to build it as a loadable module. The
+	  module will be called pg.
+
+	  You must also have at least one parallel port protocol driver in
+	  your system.
+
+	  This driver implements an API loosely related to the generic SCSI
+	  driver. See <file:include/linux/pg.h>. for details.
+
+	  You can obtain the most recent version of cdrtools from
+	  <ftp://ftp.berlios.de/pub/cdrecord/>. Versions 1.6.1a3 and
+	  later fully support this driver.
+
+comment "Parallel IDE protocol modules"
+	depends on PARIDE
+
+config PARIDE_ATEN
+	tristate "ATEN EH-100 protocol"
+	depends on PARIDE
+	help
+	  This option enables support for the ATEN EH-100 parallel port IDE
+	  protocol. This protocol is used in some inexpensive low performance
+	  parallel port kits made in Hong Kong. If you chose to build PARIDE
+	  support into your kernel, you may answer Y here to build in the
+	  protocol driver, otherwise you should answer M to build it as a
+	  loadable module. The module will be called aten. You must also
+	  have a high-level driver for the type of device that you want to
+	  support.
+
+config PARIDE_BPCK
+	tristate "MicroSolutions backpack (Series 5) protocol"
+	depends on PARIDE
+	---help---
+	  This option enables support for the Micro Solutions BACKPACK
+	  parallel port Series 5 IDE protocol.  (Most BACKPACK drives made
+	  before 1999 were Series 5) Series 5 drives will NOT always have the
+	  Series noted on the bottom of the drive. Series 6 drivers will.
+
+	  In other words, if your BACKPACK drive doesn't say "Series 6" on the
+	  bottom, enable this option.
+
+	  If you chose to build PARIDE support into your kernel, you may
+	  answer Y here to build in the protocol driver, otherwise you should
+	  answer M to build it as a loadable module.  The module will be
+	  called bpck.  You must also have a high-level driver for the type
+	  of device that you want to support.
+
+config PARIDE_BPCK6
+	tristate "MicroSolutions backpack (Series 6) protocol"
+	depends on PARIDE && !64BIT
+	---help---
+	  This option enables support for the Micro Solutions BACKPACK
+	  parallel port Series 6 IDE protocol.  (Most BACKPACK drives made
+	  after 1999 were Series 6) Series 6 drives will have the Series noted
+	  on the bottom of the drive.  Series 5 drivers don't always have it
+	  noted.
+
+	  In other words, if your BACKPACK drive says "Series 6" on the
+	  bottom, enable this option.
+
+	  If you chose to build PARIDE support into your kernel, you may
+	  answer Y here to build in the protocol driver, otherwise you should
+	  answer M to build it as a loadable module.  The module will be
+	  called bpck6.  You must also have a high-level driver for the type
+	  of device that you want to support.
+
+config PARIDE_COMM
+	tristate "DataStor Commuter protocol"
+	depends on PARIDE
+	help
+	  This option enables support for the Commuter parallel port IDE
+	  protocol from DataStor. If you chose to build PARIDE support
+	  into your kernel, you may answer Y here to build in the protocol
+	  driver, otherwise you should answer M to build it as a loadable
+	  module. The module will be called comm. You must also have
+	  a high-level driver for the type of device that you want to support.
+
+config PARIDE_DSTR
+	tristate "DataStor EP-2000 protocol"
+	depends on PARIDE
+	help
+	  This option enables support for the EP-2000 parallel port IDE
+	  protocol from DataStor. If you chose to build PARIDE support
+	  into your kernel, you may answer Y here to build in the protocol
+	  driver, otherwise you should answer M to build it as a loadable
+	  module. The module will be called dstr. You must also have
+	  a high-level driver for the type of device that you want to support.
+
+config PARIDE_FIT2
+	tristate "FIT TD-2000 protocol"
+	depends on PARIDE
+	help
+	  This option enables support for the TD-2000 parallel port IDE
+	  protocol from Fidelity International Technology. This is a simple
+	  (low speed) adapter that is used in some portable hard drives. If
+	  you chose to build PARIDE support into your kernel, you may answer Y
+	  here to build in the protocol driver, otherwise you should answer M
+	  to build it as a loadable module. The module will be called ktti.
+	  You must also have a high-level driver for the type of device that
+	  you want to support.
+
+config PARIDE_FIT3
+	tristate "FIT TD-3000 protocol"
+	depends on PARIDE
+	help
+	  This option enables support for the TD-3000 parallel port IDE
+	  protocol from Fidelity International Technology. This protocol is
+	  used in newer models of their portable disk, CD-ROM and PD/CD
+	  devices. If you chose to build PARIDE support into your kernel, you
+	  may answer Y here to build in the protocol driver, otherwise you
+	  should answer M to build it as a loadable module. The module will be
+	  called fit3. You must also have a high-level driver for the type
+	  of device that you want to support.
+
+config PARIDE_EPAT
+	tristate "Shuttle EPAT/EPEZ protocol"
+	depends on PARIDE
+	help
+	  This option enables support for the EPAT parallel port IDE protocol.
+	  EPAT is a parallel port IDE adapter manufactured by Shuttle
+	  Technology and widely used in devices from major vendors such as
+	  Hewlett-Packard, SyQuest, Imation and Avatar. If you chose to build
+	  PARIDE support into your kernel, you may answer Y here to build in
+	  the protocol driver, otherwise you should answer M to build it as a
+	  loadable module. The module will be called epat. You must also
+	  have a high-level driver for the type of device that you want to
+	  support.
+
+config PARIDE_EPATC8
+	bool "Support c7/c8 chips"
+	depends on PARIDE_EPAT
+	help
+	  This option enables support for the newer Shuttle EP1284 (aka c7 and
+	  c8) chip. You need this if you are using any recent Imation SuperDisk
+	  (LS-120) drive.
+
+config PARIDE_EPIA
+	tristate "Shuttle EPIA protocol"
+	depends on PARIDE
+	help
+	  This option enables support for the (obsolete) EPIA parallel port
+	  IDE protocol from Shuttle Technology. This adapter can still be
+	  found in some no-name kits. If you chose to build PARIDE support
+	  into your kernel, you may answer Y here to build in the protocol
+	  driver, otherwise you should answer M to build it as a loadable
+	  module. The module will be called epia. You must also have a
+	  high-level driver for the type of device that you want to support.
+
+config PARIDE_FRIQ
+	tristate "Freecom IQ ASIC-2 protocol"
+	depends on PARIDE
+	help
+	  This option enables support for version 2 of the Freecom IQ parallel
+	  port IDE adapter.  This adapter is used by the Maxell Superdisk
+	  drive.  If you chose to build PARIDE support into your kernel, you
+	  may answer Y here to build in the protocol driver, otherwise you
+	  should answer M to build it as a loadable module. The module will be
+	  called friq. You must also have a high-level driver for the type
+	  of device that you want to support.
+
+config PARIDE_FRPW
+	tristate "FreeCom power protocol"
+	depends on PARIDE
+	help
+	  This option enables support for the Freecom power parallel port IDE
+	  protocol. If you chose to build PARIDE support into your kernel, you
+	  may answer Y here to build in the protocol driver, otherwise you
+	  should answer M to build it as a loadable module. The module will be
+	  called frpw. You must also have a high-level driver for the type
+	  of device that you want to support.
+
+config PARIDE_KBIC
+	tristate "KingByte KBIC-951A/971A protocols"
+	depends on PARIDE
+	help
+	  This option enables support for the KBIC-951A and KBIC-971A parallel
+	  port IDE protocols from KingByte Information Corp. KingByte's
+	  adapters appear in many no-name portable disk and CD-ROM products,
+	  especially in Europe. If you chose to build PARIDE support into your
+	  kernel, you may answer Y here to build in the protocol driver,
+	  otherwise you should answer M to build it as a loadable module. The
+	  module will be called kbic. You must also have a high-level driver
+	  for the type of device that you want to support.
+
+config PARIDE_KTTI
+	tristate "KT PHd protocol"
+	depends on PARIDE
+	help
+	  This option enables support for the "PHd" parallel port IDE protocol
+	  from KT Technology. This is a simple (low speed) adapter that is
+	  used in some 2.5" portable hard drives. If you chose to build PARIDE
+	  support into your kernel, you may answer Y here to build in the
+	  protocol driver, otherwise you should answer M to build it as a
+	  loadable module. The module will be called ktti. You must also
+	  have a high-level driver for the type of device that you want to
+	  support.
+
+config PARIDE_ON20
+	tristate "OnSpec 90c20 protocol"
+	depends on PARIDE
+	help
+	  This option enables support for the (obsolete) 90c20 parallel port
+	  IDE protocol from OnSpec (often marketed under the ValuStore brand
+	  name). If you chose to build PARIDE support into your kernel, you
+	  may answer Y here to build in the protocol driver, otherwise you
+	  should answer M to build it as a loadable module. The module will
+	  be called on20. You must also have a high-level driver for the
+	  type of device that you want to support.
+
+config PARIDE_ON26
+	tristate "OnSpec 90c26 protocol"
+	depends on PARIDE
+	help
+	  This option enables support for the 90c26 parallel port IDE protocol
+	  from OnSpec Electronics (often marketed under the ValuStore brand
+	  name). If you chose to build PARIDE support into your kernel, you
+	  may answer Y here to build in the protocol driver, otherwise you
+	  should answer M to build it as a loadable module. The module will be
+	  called on26. You must also have a high-level driver for the type
+	  of device that you want to support.
+
+#

drivers/block/paride/Makefile

@@ -0,0 +1,28 @@
+#
+# Makefile for Parallel port IDE device drivers.
+#
+# 7 October 2000, Bartlomiej Zolnierkiewicz <bkz@linux-ide.org>
+# Rewritten to use lists instead of if-statements.
+#
+
+obj-$(CONFIG_PARIDE)		+= paride.o
+obj-$(CONFIG_PARIDE_ATEN)	+= aten.o
+obj-$(CONFIG_PARIDE_BPCK)	+= bpck.o
+obj-$(CONFIG_PARIDE_COMM)	+= comm.o
+obj-$(CONFIG_PARIDE_DSTR)	+= dstr.o
+obj-$(CONFIG_PARIDE_KBIC)	+= kbic.o
+obj-$(CONFIG_PARIDE_EPAT)	+= epat.o
+obj-$(CONFIG_PARIDE_EPIA)	+= epia.o
+obj-$(CONFIG_PARIDE_FRPW)	+= frpw.o
+obj-$(CONFIG_PARIDE_FRIQ)	+= friq.o
+obj-$(CONFIG_PARIDE_FIT2)	+= fit2.o
+obj-$(CONFIG_PARIDE_FIT3)	+= fit3.o
+obj-$(CONFIG_PARIDE_ON20)	+= on20.o
+obj-$(CONFIG_PARIDE_ON26)	+= on26.o
+obj-$(CONFIG_PARIDE_KTTI)	+= ktti.o
+obj-$(CONFIG_PARIDE_BPCK6)	+= bpck6.o
+obj-$(CONFIG_PARIDE_PD)		+= pd.o
+obj-$(CONFIG_PARIDE_PCD)	+= pcd.o
+obj-$(CONFIG_PARIDE_PF)		+= pf.o
+obj-$(CONFIG_PARIDE_PT)		+= pt.o
+obj-$(CONFIG_PARIDE_PG)		+= pg.o

drivers/block/paride/Transition-notes

@@ -0,0 +1,128 @@
+Lemma 1:
+	If ps_tq is scheduled, ps_tq_active is 1.  ps_tq_int() can be called
+	only when ps_tq_active is 1.
+Proof:	All assignments to ps_tq_active and all scheduling of ps_tq happen
+	under ps_spinlock.  There are three places where that can happen:
+	one in ps_set_intr() (A) and two in ps_tq_int() (B and C).
+	Consider the sequnce of these events.  A can not be preceded by
+	anything except B, since it is under if (!ps_tq_active) under
+	ps_spinlock.  C is always preceded by B, since we can't reach it
+	other than through B and we don't drop ps_spinlock between them.
+	IOW, the sequence is A?(BA|BC|B)*.  OTOH, number of B can not exceed
+	the sum of numbers of A and C, since each call of ps_tq_int() is
+	the result of ps_tq execution.  Therefore, the sequence starts with
+	A and each B is preceded by either A or C.  Moments when we enter
+	ps_tq_int() are sandwiched between {A,C} and B in that sequence,
+	since at any time number of B can not exceed the number of these
+	moments which, in turn, can not exceed the number of A and C.
+	In other words, the sequence of events is (A or C set ps_tq_active to
+	1 and schedule ps_tq, ps_tq is executed, ps_tq_int() is entered,
+	B resets ps_tq_active)*.
+
+
+consider the following area:
+	* in do_pd_request1(): to calls of pi_do_claimed() and return in
+	  case when pd_req is NULL.
+	* in next_request(): to call of do_pd_request1()
+	* in do_pd_read(): to call of ps_set_intr()
+	* in do_pd_read_start(): to calls of pi_do_claimed(), next_request()
+and ps_set_intr()
+	* in do_pd_read_drq(): to calls of pi_do_claimed() and next_request()
+	* in do_pd_write(): to call of ps_set_intr()
+	* in do_pd_write_start(): to calls of pi_do_claimed(), next_request()
+and ps_set_intr()
+	* in do_pd_write_done(): to calls of pi_do_claimed() and next_request()
+	* in ps_set_intr(): to check for ps_tq_active and to scheduling
+	  ps_tq if ps_tq_active was 0.
+	* in ps_tq_int(): from the moment when we get ps_spinlock() to the
+	  return, call of con() or scheduling ps_tq.
+	* in pi_schedule_claimed() when called from pi_do_claimed() called from
+	  pd.c, everything until returning 1 or setting or setting ->claim_cont
+	  on the path that returns 0
+	* in pi_do_claimed() when called from pd.c, everything until the call
+	  of pi_do_claimed() plus the everything until the call of cont() if
+	  pi_do_claimed() has returned 1.
+	* in pi_wake_up() called for PIA that belongs to pd.c, everything from
+	  the moment when pi_spinlock has been acquired.
+
+Lemma 2:
+	1) at any time at most one thread of execution can be in that area or
+	be preempted there.
+	2) When there is such a thread, pd_busy is set or pd_lock is held by
+	that thread.
+	3) When there is such a thread, ps_tq_active is 0 or ps_spinlock is
+	held by that thread.
+	4) When there is such a thread, all PIA belonging to pd.c have NULL
+	->claim_cont or pi_spinlock is held by thread in question.
+
+Proof:	consider the first moment when the above is not true.
+
+(1) can become not true if some thread enters that area while another is there.
+	a) do_pd_request1() can be called from next_request() or do_pd_request()
+	   In the first case the thread was already in the area.  In the second,
+	   the thread was holding pd_lock and found pd_busy not set, which would
+	   mean that (2) was already not true.
+	b) ps_set_intr() and pi_schedule_claimed() can be called only from the
+	   area.
+	c) pi_do_claimed() is called by pd.c only from the area.
+	d) ps_tq_int() can enter the area only when the thread is holding
+	   ps_spinlock and ps_tq_active is 1 (due to Lemma 1).  It means that
+	   (3) was already not true.
+	e) do_pd_{read,write}* could be called only from the area.  The only
+	   case that needs consideration is call from pi_wake_up() and there
+	   we would have to be called for the PIA that got ->claimed_cont
+	   from pd.c.  That could happen only if pi_do_claimed() had been
+	   called from pd.c for that PIA, which happens only for PIA belonging
+	   to pd.c.
+	f) pi_wake_up() can enter the area only when the thread is holding
+	   pi_spinlock and ->claimed_cont is non-NULL for PIA belonging to
+	   pd.c.  It means that (4) was already not true.
+
+(2) can become not true only when pd_lock is released by the thread in question.
+	Indeed, pd_busy is reset only in the area and thread that resets
+	it is holding pd_lock.	The only place within the area where we
+	release pd_lock is in pd_next_buf() (called from within the area).
+	But that code does not reset pd_busy, so pd_busy would have to be
+	0 when pd_next_buf() had acquired pd_lock.  If it become 0 while
+	we were acquiring the lock, (1) would be already false, since
+	the thread that had reset it would be in the area simulateously.
+	If it was 0 before we tried to acquire pd_lock, (2) would be
+	already false.
+
+For similar reasons, (3) can become not true only when ps_spinlock is released
+by the thread in question.  However, all such places within the area are right
+after resetting ps_tq_active to 0.
+
+(4) is done the same way - all places where we release pi_spinlock within
+the area are either after resetting ->claimed_cont to NULL while holding
+pi_spinlock, or after not tocuhing ->claimed_cont since acquiring pi_spinlock
+also in the area.  The only place where ->claimed_cont is made non-NULL is
+in the area, under pi_spinlock and we do not release it until after leaving
+the area.
+
+QED.
+
+
+Corollary 1: ps_tq_active can be killed.  Indeed, the only place where we
+check its value is in ps_set_intr() and if it had been non-zero at that
+point, we would have violated either (2.1) (if it was set while ps_set_intr()
+was acquiring ps_spinlock) or (2.3) (if it was set when we started to
+acquire ps_spinlock).
+
+Corollary 2: ps_spinlock can be killed.  Indeed, Lemma 1 and Lemma 2 show
+that the only possible contention is between scheduling ps_tq followed by
+immediate release of spinlock and beginning of execution of ps_tq on
+another CPU.
+
+Corollary 3: assignment to pd_busy in do_pd_read_start() and do_pd_write_start()
+can be killed.  Indeed, we are not holding pd_lock and thus pd_busy is already
+1 here.
+
+Corollary 4: in ps_tq_int() uses of con can be replaced with uses of
+ps_continuation, since the latter is changed only from the area.
+We don't need to reset it to NULL, since we are guaranteed that there
+will be a call of ps_set_intr() before we look at ps_continuation again.
+We can remove the check for ps_continuation being NULL for the same
+reason - the value is guaranteed to be set by the last ps_set_intr() and
+we never pass it NULL.  Assignements in the beginning of ps_set_intr()
+can be taken to callers as long as they remain within the area.

drivers/block/paride/aten.c

@@ -0,0 +1,162 @@
+/* 
+        aten.c  (c) 1997-8  Grant R. Guenther <grant@torque.net>
+                            Under the terms of the GNU General Public License.
+
+	aten.c is a low-level protocol driver for the ATEN EH-100
+	parallel port adapter.  The EH-100 supports 4-bit and 8-bit
+        modes only.  There is also an EH-132 which supports EPP mode
+        transfers.  The EH-132 is not yet supported.
+
+*/
+
+/* Changes:
+
+	1.01	GRG 1998.05.05	init_proto, release_proto
+
+*/
+
+#define ATEN_VERSION      "1.01"
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/wait.h>
+#include <linux/types.h>
+#include <asm/io.h>
+
+#include "paride.h"
+
+#define j44(a,b)                ((((a>>4)&0x0f)|(b&0xf0))^0x88)
+
+/* cont = 0 - access the IDE register file 
+   cont = 1 - access the IDE command set 
+*/
+
+static int  cont_map[2] = { 0x08, 0x20 };
+
+static void  aten_write_regr( PIA *pi, int cont, int regr, int val)
+
+{	int r;
+
+	r = regr + cont_map[cont] + 0x80;
+
+	w0(r); w2(0xe); w2(6); w0(val); w2(7); w2(6); w2(0xc);
+}
+
+static int aten_read_regr( PIA *pi, int cont, int regr )
+
+{	int  a, b, r;
+
+        r = regr + cont_map[cont] + 0x40;
+
+	switch (pi->mode) {
+
+        case 0: w0(r); w2(0xe); w2(6); 
+		w2(7); w2(6); w2(0);
+		a = r1(); w0(0x10); b = r1(); w2(0xc);
+		return j44(a,b);
+
+        case 1: r |= 0x10;
+		w0(r); w2(0xe); w2(6); w0(0xff); 
+		w2(0x27); w2(0x26); w2(0x20);
+		a = r0();
+		w2(0x26); w2(0xc);
+		return a;
+	}
+	return -1;
+}
+
+static void aten_read_block( PIA *pi, char * buf, int count )
+
+{	int  k, a, b, c, d;
+
+	switch (pi->mode) {
+
+	case 0:	w0(0x48); w2(0xe); w2(6);
+		for (k=0;k<count/2;k++) {
+			w2(7); w2(6); w2(2);
+			a = r1(); w0(0x58); b = r1();
+			w2(0); d = r1(); w0(0x48); c = r1();
+			buf[2*k] = j44(c,d);
+			buf[2*k+1] = j44(a,b);
+		}
+		w2(0xc);
+		break;
+
+	case 1: w0(0x58); w2(0xe); w2(6);
+		for (k=0;k<count/2;k++) {
+			w2(0x27); w2(0x26); w2(0x22);
+			a = r0(); w2(0x20); b = r0();
+			buf[2*k] = b; buf[2*k+1] = a;
+		}
+		w2(0x26); w2(0xc);
+		break;
+	}
+}
+
+static void aten_write_block( PIA *pi, char * buf, int count )
+
+{	int k;
+
+	w0(0x88); w2(0xe); w2(6);
+	for (k=0;k<count/2;k++) {
+		w0(buf[2*k+1]); w2(0xe); w2(6);
+		w0(buf[2*k]); w2(7); w2(6);
+	}
+	w2(0xc);
+}
+
+static void aten_connect ( PIA *pi  )
+
+{       pi->saved_r0 = r0();
+        pi->saved_r2 = r2();
+	w2(0xc);	
+}
+
+static void aten_disconnect ( PIA *pi )
+
+{       w0(pi->saved_r0);
+        w2(pi->saved_r2);
+} 
+
+static void aten_log_adapter( PIA *pi, char * scratch, int verbose )
+
+{       char    *mode_string[2] = {"4-bit","8-bit"};
+
+        printk("%s: aten %s, ATEN EH-100 at 0x%x, ",
+                pi->device,ATEN_VERSION,pi->port);
+        printk("mode %d (%s), delay %d\n",pi->mode,
+		mode_string[pi->mode],pi->delay);
+
+}
+
+static struct pi_protocol aten = {
+	.owner		= THIS_MODULE,
+	.name		= "aten",
+	.max_mode	= 2,
+	.epp_first	= 2,
+	.default_delay	= 1,
+	.max_units	= 1,
+	.write_regr	= aten_write_regr,
+	.read_regr	= aten_read_regr,
+	.write_block	= aten_write_block,
+	.read_block	= aten_read_block,
+	.connect	= aten_connect,
+	.disconnect	= aten_disconnect,
+	.log_adapter	= aten_log_adapter,
+};
+
+static int __init aten_init(void)
+{
+	return paride_register(&aten);
+}
+
+static void __exit aten_exit(void)
+{
+	paride_unregister( &aten );
+}
+
+MODULE_LICENSE("GPL");
+module_init(aten_init)
+module_exit(aten_exit)

drivers/block/paride/bpck.c

@@ -0,0 +1,477 @@
+/* 
+	bpck.c	(c) 1996-8  Grant R. Guenther <grant@torque.net>
+		            Under the terms of the GNU General Public License.
+
+	bpck.c is a low-level protocol driver for the MicroSolutions 
+	"backpack" parallel port IDE adapter.  
+
+*/
+
+/* Changes:
+
+	1.01	GRG 1998.05.05 init_proto, release_proto, pi->delay 
+	1.02    GRG 1998.08.15 default pi->delay returned to 4
+
+*/
+
+#define	BPCK_VERSION	"1.02" 
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/wait.h>
+#include <asm/io.h>
+
+#include "paride.h"
+
+#undef r2
+#undef w2
+
+#define PC			pi->private
+#define r2()			(PC=(in_p(2) & 0xff))
+#define w2(byte)  		{out_p(2,byte); PC = byte;}
+#define t2(pat)   		{PC ^= pat; out_p(2,PC);}
+#define e2()			{PC &= 0xfe; out_p(2,PC);}
+#define o2()			{PC |= 1; out_p(2,PC);}
+
+#define j44(l,h)     (((l>>3)&0x7)|((l>>4)&0x8)|((h<<1)&0x70)|(h&0x80))
+
+/* cont = 0 - access the IDE register file 
+   cont = 1 - access the IDE command set 
+   cont = 2 - use internal bpck register addressing
+*/
+
+static int  cont_map[3] = { 0x40, 0x48, 0 };
+
+static int bpck_read_regr( PIA *pi, int cont, int regr )
+
+{       int r, l, h;
+
+	r = regr + cont_map[cont];
+
+	switch (pi->mode) {
+
+	case 0: w0(r & 0xf); w0(r); t2(2); t2(4);
+	        l = r1();
+        	t2(4);
+        	h = r1();
+        	return j44(l,h);
+
+	case 1: w0(r & 0xf); w0(r); t2(2);
+	        e2(); t2(0x20);
+		t2(4); h = r0();
+	        t2(1); t2(0x20);
+	        return h;
+
+	case 2:
+	case 3:
+	case 4: w0(r); w2(9); w2(0); w2(0x20);
+		h = r4();
+		w2(0);
+		return h;
+
+	}
+	return -1;
+}	
+
+static void bpck_write_regr( PIA *pi, int cont, int regr, int val )
+
+{	int	r;
+
+        r = regr + cont_map[cont];
+
+	switch (pi->mode) {
+
+	case 0:
+	case 1: w0(r);
+		t2(2);
+		w0(val);
+		o2(); t2(4); t2(1);
+		break;
+
+	case 2:
+	case 3:
+	case 4: w0(r); w2(9); w2(0);
+		w0(val); w2(1); w2(3); w2(0);
+		break;
+
+	}
+}
+
+/* These macros access the bpck registers in native addressing */
+
+#define WR(r,v)		bpck_write_regr(pi,2,r,v)
+#define RR(r)		(bpck_read_regr(pi,2,r))
+
+static void bpck_write_block( PIA *pi, char * buf, int count )
+
+{	int i;
+
+	switch (pi->mode) {
+
+	case 0: WR(4,0x40);
+		w0(0x40); t2(2); t2(1);
+		for (i=0;i<count;i++) { w0(buf[i]); t2(4); }
+		WR(4,0);
+		break;
+
+	case 1: WR(4,0x50);
+                w0(0x40); t2(2); t2(1);
+                for (i=0;i<count;i++) { w0(buf[i]); t2(4); }
+                WR(4,0x10);
+		break;
+
+	case 2: WR(4,0x48);
+		w0(0x40); w2(9); w2(0); w2(1);
+		for (i=0;i<count;i++) w4(buf[i]);
+		w2(0);
+		WR(4,8);
+		break;
+
+        case 3: WR(4,0x48);
+                w0(0x40); w2(9); w2(0); w2(1);
+                for (i=0;i<count/2;i++) w4w(((u16 *)buf)[i]);
+                w2(0);
+                WR(4,8);
+                break;
+ 
+        case 4: WR(4,0x48);
+                w0(0x40); w2(9); w2(0); w2(1);
+                for (i=0;i<count/4;i++) w4l(((u32 *)buf)[i]);
+                w2(0);
+                WR(4,8);
+                break;
+ 	}
+}
+
+static void bpck_read_block( PIA *pi, char * buf, int count )
+
+{	int i, l, h;
+
+	switch (pi->mode) {
+
+      	case 0: WR(4,0x40);
+		w0(0x40); t2(2);
+		for (i=0;i<count;i++) {
+		    t2(4); l = r1();
+		    t2(4); h = r1();
+		    buf[i] = j44(l,h);
+		}
+		WR(4,0);
+		break;
+
+	case 1: WR(4,0x50);
+		w0(0x40); t2(2); t2(0x20);
+      	        for(i=0;i<count;i++) { t2(4); buf[i] = r0(); }
+	        t2(1); t2(0x20);
+	        WR(4,0x10);
+		break;
+
+	case 2: WR(4,0x48);
+		w0(0x40); w2(9); w2(0); w2(0x20);
+		for (i=0;i<count;i++) buf[i] = r4();
+		w2(0);
+		WR(4,8);
+		break;
+
+        case 3: WR(4,0x48);
+                w0(0x40); w2(9); w2(0); w2(0x20);
+                for (i=0;i<count/2;i++) ((u16 *)buf)[i] = r4w();
+                w2(0);
+                WR(4,8);
+                break;
+
+        case 4: WR(4,0x48);
+                w0(0x40); w2(9); w2(0); w2(0x20);
+                for (i=0;i<count/4;i++) ((u32 *)buf)[i] = r4l();
+                w2(0);
+                WR(4,8);
+                break;
+
+	}
+}
+
+static int bpck_probe_unit ( PIA *pi )
+
+{	int o1, o0, f7, id;
+	int t, s;
+
+	id = pi->unit;
+	s = 0;
+	w2(4); w2(0xe); r2(); t2(2); 
+	o1 = r1()&0xf8;
+	o0 = r0();
+	w0(255-id); w2(4); w0(id);
+	t2(8); t2(8); t2(8);
+	t2(2); t = r1()&0xf8;
+	f7 = ((id % 8) == 7);
+	if ((f7) || (t != o1)) { t2(2); s = r1()&0xf8; }
+	if ((t == o1) && ((!f7) || (s == o1)))  {
+		w2(0x4c); w0(o0);
+		return 0;	
+	}
+	t2(8); w0(0); t2(2); w2(0x4c); w0(o0);
+	return 1;
+}
+	
+static void bpck_connect ( PIA *pi  )
+
+{       pi->saved_r0 = r0();
+	w0(0xff-pi->unit); w2(4); w0(pi->unit);
+	t2(8); t2(8); t2(8); 
+	t2(2); t2(2);
+	
+	switch (pi->mode) {
+
+	case 0: t2(8); WR(4,0);
+		break;
+
+	case 1: t2(8); WR(4,0x10);
+		break;
+
+	case 2:
+        case 3:
+	case 4: w2(0); WR(4,8);
+		break;
+
+	}
+
+	WR(5,8);
+
+	if (pi->devtype == PI_PCD) {
+		WR(0x46,0x10);		/* fiddle with ESS logic ??? */
+		WR(0x4c,0x38);
+		WR(0x4d,0x88);
+		WR(0x46,0xa0);
+		WR(0x41,0);
+		WR(0x4e,8);
+		}
+}
+
+static void bpck_disconnect ( PIA *pi )
+
+{	w0(0); 
+	if (pi->mode >= 2) { w2(9); w2(0); } else t2(2);
+	w2(0x4c); w0(pi->saved_r0);
+} 
+
+static void bpck_force_spp ( PIA *pi )
+
+/* This fakes the EPP protocol to turn off EPP ... */
+
+{       pi->saved_r0 = r0();
+        w0(0xff-pi->unit); w2(4); w0(pi->unit);
+        t2(8); t2(8); t2(8); 
+        t2(2); t2(2);
+
+        w2(0); 
+        w0(4); w2(9); w2(0); 
+        w0(0); w2(1); w2(3); w2(0);     
+        w0(0); w2(9); w2(0);
+        w2(0x4c); w0(pi->saved_r0);
+}
+
+#define TEST_LEN  16
+
+static int bpck_test_proto( PIA *pi, char * scratch, int verbose )
+
+{	int i, e, l, h, om;
+	char buf[TEST_LEN];
+
+	bpck_force_spp(pi);
+
+	switch (pi->mode) {
+
+	case 0: bpck_connect(pi);
+		WR(0x13,0x7f);
+		w0(0x13); t2(2);
+		for(i=0;i<TEST_LEN;i++) {
+                    t2(4); l = r1();
+                    t2(4); h = r1();
+                    buf[i] = j44(l,h);
+		}
+		bpck_disconnect(pi);
+		break;
+
+        case 1: bpck_connect(pi);
+		WR(0x13,0x7f);
+                w0(0x13); t2(2); t2(0x20);
+                for(i=0;i<TEST_LEN;i++) { t2(4); buf[i] = r0(); }
+                t2(1); t2(0x20);
+		bpck_disconnect(pi);
+		break;
+
+	case 2:
+	case 3:
+	case 4: om = pi->mode;
+		pi->mode = 0;
+		bpck_connect(pi);
+		WR(7,3);
+		WR(4,8);
+		bpck_disconnect(pi);
+
+		pi->mode = om;
+		bpck_connect(pi);
+		w0(0x13); w2(9); w2(1); w0(0); w2(3); w2(0); w2(0xe0);
+
+		switch (pi->mode) {
+		  case 2: for (i=0;i<TEST_LEN;i++) buf[i] = r4();
+			  break;
+		  case 3: for (i=0;i<TEST_LEN/2;i++) ((u16 *)buf)[i] = r4w();
+                          break;
+		  case 4: for (i=0;i<TEST_LEN/4;i++) ((u32 *)buf)[i] = r4l();
+                          break;
+		}
+
+		w2(0);
+		WR(7,0);
+		bpck_disconnect(pi);
+
+		break;
+
+	}
+
+	if (verbose) {
+	    printk("%s: bpck: 0x%x unit %d mode %d: ",
+		   pi->device,pi->port,pi->unit,pi->mode);
+	    for (i=0;i<TEST_LEN;i++) printk("%3d",buf[i]);
+	    printk("\n");
+	}
+
+	e = 0;
+	for (i=0;i<TEST_LEN;i++) if (buf[i] != (i+1)) e++;
+	return e;
+}
+
+static void bpck_read_eeprom ( PIA *pi, char * buf )
+
+{       int i,j,k,n,p,v,f, om, od;
+
+	bpck_force_spp(pi);
+
+	om = pi->mode;  od = pi->delay;
+	pi->mode = 0; pi->delay = 6;
+
+	bpck_connect(pi);
+	
+	n = 0;
+	WR(4,0);
+	for (i=0;i<64;i++) {
+	    WR(6,8);  
+	    WR(6,0xc);
+	    p = 0x100;
+	    for (k=0;k<9;k++) {
+		f = (((i + 0x180) & p) != 0) * 2;
+		WR(6,f+0xc); 
+		WR(6,f+0xd); 
+		WR(6,f+0xc);
+		p = (p >> 1);
+	    }
+	    for (j=0;j<2;j++) {
+		v = 0;
+		for (k=0;k<8;k++) {
+		    WR(6,0xc); 
+		    WR(6,0xd); 
+		    WR(6,0xc); 
+		    f = RR(0);
+		    v = 2*v + (f == 0x84);
+		}
+		buf[2*i+1-j] = v;
+	    }
+	}
+	WR(6,8);
+	WR(6,0);
+	WR(5,8);
+
+	bpck_disconnect(pi);
+
+        if (om >= 2) {
+                bpck_connect(pi);
+                WR(7,3);
+                WR(4,8);
+                bpck_disconnect(pi);
+        }
+
+	pi->mode = om; pi->delay = od;
+}
+
+static int bpck_test_port ( PIA *pi ) 	/* check for 8-bit port */
+
+{	int	i, r, m;
+
+	w2(0x2c); i = r0(); w0(255-i); r = r0(); w0(i);
+	m = -1;
+	if (r == i) m = 2;
+	if (r == (255-i)) m = 0;
+
+	w2(0xc); i = r0(); w0(255-i); r = r0(); w0(i);
+	if (r != (255-i)) m = -1;
+	
+	if (m == 0) { w2(6); w2(0xc); r = r0(); w0(0xaa); w0(r); w0(0xaa); }
+	if (m == 2) { w2(0x26); w2(0xc); }
+
+	if (m == -1) return 0;
+	return 5;
+}
+
+static void bpck_log_adapter( PIA *pi, char * scratch, int verbose )
+
+{	char	*mode_string[5] = { "4-bit","8-bit","EPP-8",
+				    "EPP-16","EPP-32" };
+
+#ifdef DUMP_EEPROM
+	int i;
+#endif
+
+	bpck_read_eeprom(pi,scratch);
+
+#ifdef DUMP_EEPROM
+	if (verbose) {
+	   for(i=0;i<128;i++)
+		if ((scratch[i] < ' ') || (scratch[i] > '~'))
+		    scratch[i] = '.';
+	   printk("%s: bpck EEPROM: %64.64s\n",pi->device,scratch);
+	   printk("%s:              %64.64s\n",pi->device,&scratch[64]);
+	}
+#endif
+
+	printk("%s: bpck %s, backpack %8.8s unit %d",
+		pi->device,BPCK_VERSION,&scratch[110],pi->unit);
+	printk(" at 0x%x, mode %d (%s), delay %d\n",pi->port,
+		pi->mode,mode_string[pi->mode],pi->delay);
+}
+
+static struct pi_protocol bpck = {
+	.owner		= THIS_MODULE,
+	.name		= "bpck",
+	.max_mode	= 5,
+	.epp_first	= 2,
+	.default_delay	= 4,
+	.max_units	= 255,
+	.write_regr	= bpck_write_regr,
+	.read_regr	= bpck_read_regr,
+	.write_block	= bpck_write_block,
+	.read_block	= bpck_read_block,
+	.connect	= bpck_connect,
+	.disconnect	= bpck_disconnect,
+	.test_port	= bpck_test_port,
+	.probe_unit	= bpck_probe_unit,
+	.test_proto	= bpck_test_proto,
+	.log_adapter	= bpck_log_adapter,
+};
+
+static int __init bpck_init(void)
+{
+	return paride_register(&bpck);
+}
+
+static void __exit bpck_exit(void)
+{
+	paride_unregister(&bpck);
+}
+
+MODULE_LICENSE("GPL");
+module_init(bpck_init)
+module_exit(bpck_exit)

drivers/block/paride/bpck6.c

@@ -0,0 +1,267 @@
+/*
+	backpack.c (c) 2001 Micro Solutions Inc.
+		Released under the terms of the GNU General Public license
+
+	backpack.c is a low-level protocol driver for the Micro Solutions
+		"BACKPACK" parallel port IDE adapter
+		(Works on Series 6 drives)
+
+	Written by: Ken Hahn     (linux-dev@micro-solutions.com)
+	            Clive Turvey (linux-dev@micro-solutions.com)
+
+*/
+
+/*
+   This is Ken's linux wrapper for the PPC library
+   Version 1.0.0 is the backpack driver for which source is not available
+   Version 2.0.0 is the first to have source released 
+   Version 2.0.1 is the "Cox-ified" source code 
+   Version 2.0.2 - fixed version string usage, and made ppc functions static 
+*/
+
+
+#define BACKPACK_VERSION "2.0.2"
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/types.h>
+#include <asm/io.h>
+#include <linux/parport.h>
+
+#include "ppc6lnx.c"
+#include "paride.h"
+
+/* PARAMETERS */
+static bool verbose; /* set this to 1 to see debugging messages and whatnot */
+ 
+
+#define PPCSTRUCT(pi) ((Interface *)(pi->private))
+
+/****************************************************************/
+/*
+ ATAPI CDROM DRIVE REGISTERS
+*/
+#define ATAPI_DATA       0      /* data port                  */
+#define ATAPI_ERROR      1      /* error register (read)      */
+#define ATAPI_FEATURES   1      /* feature register (write)   */
+#define ATAPI_INT_REASON 2      /* interrupt reason register  */
+#define ATAPI_COUNT_LOW  4      /* byte count register (low)  */
+#define ATAPI_COUNT_HIGH 5      /* byte count register (high) */
+#define ATAPI_DRIVE_SEL  6      /* drive select register      */
+#define ATAPI_STATUS     7      /* status port (read)         */
+#define ATAPI_COMMAND    7      /* command port (write)       */
+#define ATAPI_ALT_STATUS 0x0e /* alternate status reg (read) */
+#define ATAPI_DEVICE_CONTROL 0x0e /* device control (write)   */
+/****************************************************************/
+
+static int bpck6_read_regr(PIA *pi, int cont, int reg)
+{
+	unsigned int out;
+
+	/* check for bad settings */
+	if (reg<0 || reg>7 || cont<0 || cont>2)
+	{
+		return(-1);
+	}
+	out=ppc6_rd_port(PPCSTRUCT(pi),cont?reg|8:reg);
+	return(out);
+}
+
+static void bpck6_write_regr(PIA *pi, int cont, int reg, int val)
+{
+	/* check for bad settings */
+	if (reg>=0 && reg<=7 && cont>=0 && cont<=1)
+	{
+		ppc6_wr_port(PPCSTRUCT(pi),cont?reg|8:reg,(u8)val);
+	}
+}
+
+static void bpck6_write_block( PIA *pi, char * buf, int len )
+{
+	ppc6_wr_port16_blk(PPCSTRUCT(pi),ATAPI_DATA,buf,(u32)len>>1); 
+}
+
+static void bpck6_read_block( PIA *pi, char * buf, int len )
+{
+	ppc6_rd_port16_blk(PPCSTRUCT(pi),ATAPI_DATA,buf,(u32)len>>1);
+}
+
+static void bpck6_connect ( PIA *pi  )
+{
+	if(verbose)
+	{
+		printk(KERN_DEBUG "connect\n");
+	}
+
+	if(pi->mode >=2)
+  	{
+		PPCSTRUCT(pi)->mode=4+pi->mode-2;	
+	}
+	else if(pi->mode==1)
+	{
+		PPCSTRUCT(pi)->mode=3;	
+	}
+	else
+	{
+		PPCSTRUCT(pi)->mode=1;		
+	}
+
+	ppc6_open(PPCSTRUCT(pi));  
+	ppc6_wr_extout(PPCSTRUCT(pi),0x3);
+}
+
+static void bpck6_disconnect ( PIA *pi )
+{
+	if(verbose)
+	{
+		printk("disconnect\n");
+	}
+	ppc6_wr_extout(PPCSTRUCT(pi),0x0);
+	ppc6_close(PPCSTRUCT(pi));
+}
+
+static int bpck6_test_port ( PIA *pi )   /* check for 8-bit port */
+{
+	if(verbose)
+	{
+		printk(KERN_DEBUG "PARPORT indicates modes=%x for lp=0x%lx\n",
+               		((struct pardevice*)(pi->pardev))->port->modes,
+			((struct pardevice *)(pi->pardev))->port->base); 
+	}
+
+	/*copy over duplicate stuff.. initialize state info*/
+	PPCSTRUCT(pi)->ppc_id=pi->unit;
+	PPCSTRUCT(pi)->lpt_addr=pi->port;
+
+	/* look at the parport device to see if what modes we can use */
+	if(((struct pardevice *)(pi->pardev))->port->modes & 
+		(PARPORT_MODE_EPP)
+          )
+	{
+		return 5; /* Can do EPP*/
+	}
+	else if(((struct pardevice *)(pi->pardev))->port->modes & 
+			(PARPORT_MODE_TRISTATE)
+               )
+	{
+		return 2;
+	}
+	else /*Just flat SPP*/
+	{
+		return 1;
+	}
+}
+
+static int bpck6_probe_unit ( PIA *pi )
+{
+	int out;
+
+	if(verbose)
+	{
+		printk(KERN_DEBUG "PROBE UNIT %x on port:%x\n",pi->unit,pi->port);
+	}
+
+	/*SET PPC UNIT NUMBER*/
+	PPCSTRUCT(pi)->ppc_id=pi->unit;
+
+	/*LOWER DOWN TO UNIDIRECTIONAL*/
+	PPCSTRUCT(pi)->mode=1;		
+
+	out=ppc6_open(PPCSTRUCT(pi));
+
+	if(verbose)
+	{
+		printk(KERN_DEBUG "ppc_open returned %2x\n",out);
+	}
+
+  	if(out)
+ 	{
+		ppc6_close(PPCSTRUCT(pi));
+		if(verbose)
+		{
+			printk(KERN_DEBUG "leaving probe\n");
+		}
+               return(1);
+	}
+  	else
+  	{
+		if(verbose)
+		{
+			printk(KERN_DEBUG "Failed open\n");
+		}
+    		return(0);
+  	}
+}
+
+static void bpck6_log_adapter( PIA *pi, char * scratch, int verbose )
+{
+	char *mode_string[5]=
+		{"4-bit","8-bit","EPP-8","EPP-16","EPP-32"};
+
+	printk("%s: BACKPACK Protocol Driver V"BACKPACK_VERSION"\n",pi->device);
+	printk("%s: Copyright 2001 by Micro Solutions, Inc., DeKalb IL.\n",pi->device);
+	printk("%s: BACKPACK %s, Micro Solutions BACKPACK Drive at 0x%x\n",
+		pi->device,BACKPACK_VERSION,pi->port);
+	printk("%s: Unit: %d Mode:%d (%s) Delay %d\n",pi->device,
+		pi->unit,pi->mode,mode_string[pi->mode],pi->delay);
+}
+
+static int bpck6_init_proto(PIA *pi)
+{
+	Interface *p = kzalloc(sizeof(Interface), GFP_KERNEL);
+
+	if (p) {
+		pi->private = (unsigned long)p;
+		return 0;
+	}
+
+	printk(KERN_ERR "%s: ERROR COULDN'T ALLOCATE MEMORY\n", pi->device); 
+	return -1;
+}
+
+static void bpck6_release_proto(PIA *pi)
+{
+	kfree((void *)(pi->private)); 
+}
+
+static struct pi_protocol bpck6 = {
+	.owner		= THIS_MODULE,
+	.name		= "bpck6",
+	.max_mode	= 5,
+	.epp_first	= 2, /* 2-5 use epp (need 8 ports) */
+	.max_units	= 255,
+	.write_regr	= bpck6_write_regr,
+	.read_regr	= bpck6_read_regr,
+	.write_block	= bpck6_write_block,
+	.read_block	= bpck6_read_block,
+	.connect	= bpck6_connect,
+	.disconnect	= bpck6_disconnect,
+	.test_port	= bpck6_test_port,
+	.probe_unit	= bpck6_probe_unit,
+	.log_adapter	= bpck6_log_adapter,
+	.init_proto	= bpck6_init_proto,
+	.release_proto	= bpck6_release_proto,
+};
+
+static int __init bpck6_init(void)
+{
+	printk(KERN_INFO "bpck6: BACKPACK Protocol Driver V"BACKPACK_VERSION"\n");
+	printk(KERN_INFO "bpck6: Copyright 2001 by Micro Solutions, Inc., DeKalb IL. USA\n");
+	if(verbose)
+		printk(KERN_DEBUG "bpck6: verbose debug enabled.\n");
+	return paride_register(&bpck6);
+}
+
+static void __exit bpck6_exit(void)
+{
+	paride_unregister(&bpck6);
+}
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Micro Solutions Inc.");
+MODULE_DESCRIPTION("BACKPACK Protocol module, compatible with PARIDE");
+module_param(verbose, bool, 0644);
+module_init(bpck6_init)
+module_exit(bpck6_exit)

drivers/block/paride/comm.c

@@ -0,0 +1,218 @@
+/* 
+        comm.c    (c) 1997-8  Grant R. Guenther <grant@torque.net>
+                              Under the terms of the GNU General Public License.
+
+	comm.c is a low-level protocol driver for some older models
+	of the DataStor "Commuter" parallel to IDE adapter.  Some of
+	the parallel port devices marketed by Arista currently
+	use this adapter.
+*/
+
+/* Changes:
+
+	1.01	GRG 1998.05.05  init_proto, release_proto
+
+*/
+
+#define COMM_VERSION      "1.01"
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/wait.h>
+#include <asm/io.h>
+
+#include "paride.h"
+
+/* mode codes:  0  nybble reads, 8-bit writes
+                1  8-bit reads and writes
+                2  8-bit EPP mode
+*/
+
+#define j44(a,b)	(((a>>3)&0x0f)|((b<<1)&0xf0))
+
+#define P1	w2(5);w2(0xd);w2(0xd);w2(5);w2(4);
+#define P2	w2(5);w2(7);w2(7);w2(5);w2(4);
+
+/* cont = 0 - access the IDE register file 
+   cont = 1 - access the IDE command set 
+*/
+
+static int  cont_map[2] = { 0x08, 0x10 };
+
+static int comm_read_regr( PIA *pi, int cont, int regr )
+
+{       int     l, h, r;
+
+        r = regr + cont_map[cont];
+
+        switch (pi->mode)  {
+
+        case 0: w0(r); P1; w0(0);
+        	w2(6); l = r1(); w0(0x80); h = r1(); w2(4);
+                return j44(l,h);
+
+        case 1: w0(r+0x20); P1; 
+        	w0(0); w2(0x26); h = r0(); w2(4);
+                return h;
+
+	case 2:
+	case 3:
+        case 4: w3(r+0x20); (void)r1();
+        	w2(0x24); h = r4(); w2(4);
+                return h;
+
+        }
+        return -1;
+}       
+
+static void comm_write_regr( PIA *pi, int cont, int regr, int val )
+
+{       int  r;
+
+        r = regr + cont_map[cont];
+
+        switch (pi->mode)  {
+
+        case 0:
+        case 1: w0(r); P1; w0(val); P2;
+		break;
+
+	case 2:
+	case 3:
+        case 4: w3(r); (void)r1(); w4(val);
+                break;
+        }
+}
+
+static void comm_connect ( PIA *pi  )
+
+{       pi->saved_r0 = r0();
+        pi->saved_r2 = r2();
+        w2(4); w0(0xff); w2(6);
+        w2(4); w0(0xaa); w2(6);
+        w2(4); w0(0x00); w2(6);
+        w2(4); w0(0x87); w2(6);
+        w2(4); w0(0xe0); w2(0xc); w2(0xc); w2(4);
+}
+
+static void comm_disconnect ( PIA *pi )
+
+{       w2(0); w2(0); w2(0); w2(4); 
+	w0(pi->saved_r0);
+        w2(pi->saved_r2);
+} 
+
+static void comm_read_block( PIA *pi, char * buf, int count )
+
+{       int     i, l, h;
+
+        switch (pi->mode) {
+        
+        case 0: w0(0x48); P1;
+                for(i=0;i<count;i++) {
+                        w0(0); w2(6); l = r1();
+                        w0(0x80); h = r1(); w2(4);
+                        buf[i] = j44(l,h);
+                }
+                break;
+
+        case 1: w0(0x68); P1; w0(0);
+                for(i=0;i<count;i++) {
+                        w2(0x26); buf[i] = r0(); w2(0x24);
+                }
+		w2(4);
+		break;
+		
+	case 2: w3(0x68); (void)r1(); w2(0x24);
+		for (i=0;i<count;i++) buf[i] = r4();
+		w2(4);
+		break;
+
+        case 3: w3(0x68); (void)r1(); w2(0x24);
+                for (i=0;i<count/2;i++) ((u16 *)buf)[i] = r4w();
+                w2(4);
+                break;
+
+        case 4: w3(0x68); (void)r1(); w2(0x24);
+                for (i=0;i<count/4;i++) ((u32 *)buf)[i] = r4l();
+                w2(4);
+                break;
+		
+	}
+}
+
+/* NB: Watch out for the byte swapped writes ! */
+
+static void comm_write_block( PIA *pi, char * buf, int count )
+
+{       int	k;
+
+        switch (pi->mode) {
+
+        case 0:
+        case 1: w0(0x68); P1;
+        	for (k=0;k<count;k++) {
+                        w2(5); w0(buf[k^1]); w2(7);
+                }
+                w2(5); w2(4);
+                break;
+
+        case 2: w3(0x48); (void)r1();
+                for (k=0;k<count;k++) w4(buf[k^1]);
+                break;
+
+        case 3: w3(0x48); (void)r1();
+                for (k=0;k<count/2;k++) w4w(pi_swab16(buf,k));
+                break;
+
+        case 4: w3(0x48); (void)r1();
+                for (k=0;k<count/4;k++) w4l(pi_swab32(buf,k));
+                break;
+
+
+        }
+}
+
+static void comm_log_adapter( PIA *pi, char * scratch, int verbose )
+
+{       char    *mode_string[5] = {"4-bit","8-bit","EPP-8","EPP-16","EPP-32"};
+
+        printk("%s: comm %s, DataStor Commuter at 0x%x, ",
+                pi->device,COMM_VERSION,pi->port);
+        printk("mode %d (%s), delay %d\n",pi->mode,
+		mode_string[pi->mode],pi->delay);
+
+}
+
+static struct pi_protocol comm = {
+	.owner		= THIS_MODULE,
+	.name		= "comm",
+	.max_mode	= 5,
+	.epp_first	= 2,
+	.default_delay	= 1,
+	.max_units	= 1,
+	.write_regr	= comm_write_regr,
+	.read_regr	= comm_read_regr,
+	.write_block	= comm_write_block,
+	.read_block	= comm_read_block,
+	.connect	= comm_connect,
+	.disconnect	= comm_disconnect,
+	.log_adapter	= comm_log_adapter,
+};
+
+static int __init comm_init(void)
+{
+	return paride_register(&comm);
+}
+
+static void __exit comm_exit(void)
+{
+	paride_unregister(&comm);
+}
+
+MODULE_LICENSE("GPL");
+module_init(comm_init)
+module_exit(comm_exit)

drivers/block/paride/dstr.c

@@ -0,0 +1,233 @@
+/* 
+        dstr.c    (c) 1997-8  Grant R. Guenther <grant@torque.net>
+                              Under the terms of the GNU General Public License.
+
+        dstr.c is a low-level protocol driver for the 
+        DataStor EP2000 parallel to IDE adapter chip.
+
+*/
+
+/* Changes:
+
+        1.01    GRG 1998.05.06 init_proto, release_proto
+
+*/
+
+#define DSTR_VERSION      "1.01"
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/wait.h>
+#include <asm/io.h>
+
+#include "paride.h"
+
+/* mode codes:  0  nybble reads, 8-bit writes
+                1  8-bit reads and writes
+                2  8-bit EPP mode
+		3  EPP-16
+		4  EPP-32
+*/
+
+#define j44(a,b)  (((a>>3)&0x07)|((~a>>4)&0x08)|((b<<1)&0x70)|((~b)&0x80))
+
+#define P1	w2(5);w2(0xd);w2(5);w2(4);
+#define P2	w2(5);w2(7);w2(5);w2(4);
+#define P3      w2(6);w2(4);w2(6);w2(4);
+
+/* cont = 0 - access the IDE register file 
+   cont = 1 - access the IDE command set 
+*/
+
+static int  cont_map[2] = { 0x20, 0x40 };
+
+static int dstr_read_regr( PIA *pi, int cont, int regr )
+
+{       int     a, b, r;
+
+        r = regr + cont_map[cont];
+
+	w0(0x81); P1;
+	if (pi->mode) { w0(0x11); } else { w0(1); }
+	P2; w0(r); P1;
+
+        switch (pi->mode)  {
+
+        case 0: w2(6); a = r1(); w2(4); w2(6); b = r1(); w2(4);
+                return j44(a,b);
+
+        case 1: w0(0); w2(0x26); a = r0(); w2(4);
+                return a;
+
+	case 2:
+	case 3:
+        case 4: w2(0x24); a = r4(); w2(4);
+                return a;
+
+        }
+        return -1;
+}       
+
+static void dstr_write_regr(  PIA *pi, int cont, int regr, int val )
+
+{       int  r;
+
+        r = regr + cont_map[cont];
+
+	w0(0x81); P1; 
+	if (pi->mode >= 2) { w0(0x11); } else { w0(1); }
+	P2; w0(r); P1;
+	
+        switch (pi->mode)  {
+
+        case 0:
+        case 1: w0(val); w2(5); w2(7); w2(5); w2(4);
+		break;
+
+	case 2:
+	case 3:
+        case 4: w4(val); 
+                break;
+        }
+}
+
+#define  CCP(x)  w0(0xff);w2(0xc);w2(4);\
+		 w0(0xaa);w0(0x55);w0(0);w0(0xff);w0(0x87);w0(0x78);\
+		 w0(x);w2(5);w2(4);
+
+static void dstr_connect ( PIA *pi  )
+
+{       pi->saved_r0 = r0();
+        pi->saved_r2 = r2();
+        w2(4); CCP(0xe0); w0(0xff);
+}
+
+static void dstr_disconnect ( PIA *pi )
+
+{       CCP(0x30);
+        w0(pi->saved_r0);
+        w2(pi->saved_r2);
+} 
+
+static void dstr_read_block( PIA *pi, char * buf, int count )
+
+{       int     k, a, b;
+
+        w0(0x81); P1;
+        if (pi->mode) { w0(0x19); } else { w0(9); }
+	P2; w0(0x82); P1; P3; w0(0x20); P1;
+
+        switch (pi->mode) {
+
+        case 0: for (k=0;k<count;k++) {
+                        w2(6); a = r1(); w2(4);
+                        w2(6); b = r1(); w2(4);
+                        buf[k] = j44(a,b);
+                } 
+                break;
+
+        case 1: w0(0);
+                for (k=0;k<count;k++) {
+                        w2(0x26); buf[k] = r0(); w2(0x24);
+                }
+                w2(4);
+                break;
+
+        case 2: w2(0x24); 
+                for (k=0;k<count;k++) buf[k] = r4();
+                w2(4);
+                break;
+
+        case 3: w2(0x24); 
+                for (k=0;k<count/2;k++) ((u16 *)buf)[k] = r4w();
+                w2(4);
+                break;
+
+        case 4: w2(0x24); 
+                for (k=0;k<count/4;k++) ((u32 *)buf)[k] = r4l();
+                w2(4);
+                break;
+
+        }
+}
+
+static void dstr_write_block( PIA *pi, char * buf, int count )
+
+{       int	k;
+
+        w0(0x81); P1;
+        if (pi->mode) { w0(0x19); } else { w0(9); }
+        P2; w0(0x82); P1; P3; w0(0x20); P1;
+
+        switch (pi->mode) {
+
+        case 0:
+        case 1: for (k=0;k<count;k++) {
+                        w2(5); w0(buf[k]); w2(7);
+                }
+                w2(5); w2(4);
+                break;
+
+        case 2: w2(0xc5);
+                for (k=0;k<count;k++) w4(buf[k]);
+		w2(0xc4);
+                break;
+
+        case 3: w2(0xc5);
+                for (k=0;k<count/2;k++) w4w(((u16 *)buf)[k]);
+                w2(0xc4);
+                break;
+
+        case 4: w2(0xc5);
+                for (k=0;k<count/4;k++) w4l(((u32 *)buf)[k]);
+                w2(0xc4);
+                break;
+
+        }
+}
+
+
+static void dstr_log_adapter( PIA *pi, char * scratch, int verbose )
+
+{       char    *mode_string[5] = {"4-bit","8-bit","EPP-8",
+				   "EPP-16","EPP-32"};
+
+        printk("%s: dstr %s, DataStor EP2000 at 0x%x, ",
+                pi->device,DSTR_VERSION,pi->port);
+        printk("mode %d (%s), delay %d\n",pi->mode,
+		mode_string[pi->mode],pi->delay);
+
+}
+
+static struct pi_protocol dstr = {
+	.owner		= THIS_MODULE,
+	.name		= "dstr",
+	.max_mode	= 5,
+	.epp_first	= 2,
+	.default_delay	= 1,
+	.max_units	= 1,
+	.write_regr	= dstr_write_regr,
+	.read_regr	= dstr_read_regr,
+	.write_block	= dstr_write_block,
+	.read_block	= dstr_read_block,
+	.connect	= dstr_connect,
+	.disconnect	= dstr_disconnect,
+	.log_adapter	= dstr_log_adapter,
+};
+
+static int __init dstr_init(void)
+{
+	return paride_register(&dstr);
+}
+
+static void __exit dstr_exit(void)
+{
+	paride_unregister(&dstr);
+}
+
+MODULE_LICENSE("GPL");
+module_init(dstr_init)
+module_exit(dstr_exit)

drivers/block/paride/epat.c

@@ -0,0 +1,340 @@
+/* 
+        epat.c  (c) 1997-8  Grant R. Guenther <grant@torque.net>
+                            Under the terms of the GNU General Public License.
+
+	This is the low level protocol driver for the EPAT parallel
+        to IDE adapter from Shuttle Technologies.  This adapter is
+        used in many popular parallel port disk products such as the
+        SyQuest EZ drives, the Avatar Shark and the Imation SuperDisk.
+	
+*/
+
+/* Changes:
+
+        1.01    GRG 1998.05.06 init_proto, release_proto
+        1.02    Joshua b. Jore CPP(renamed), epat_connect, epat_disconnect
+
+*/
+
+#define EPAT_VERSION      "1.02"
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/wait.h>
+#include <asm/io.h>
+
+#include "paride.h"
+
+#define j44(a,b)		(((a>>4)&0x0f)+(b&0xf0))
+#define j53(a,b)		(((a>>3)&0x1f)+((b<<4)&0xe0))
+
+static int epatc8;
+
+module_param(epatc8, int, 0);
+MODULE_PARM_DESC(epatc8, "support for the Shuttle EP1284 chip, "
+	"used in any recent Imation SuperDisk (LS-120) drive.");
+
+/* cont =  0   IDE register file
+   cont =  1   IDE control registers
+   cont =  2   internal EPAT registers
+*/
+
+static int cont_map[3] = { 0x18, 0x10, 0 };
+
+static void epat_write_regr( PIA *pi, int cont, int regr, int val)
+
+{	int r;
+
+	r = regr + cont_map[cont];
+
+	switch (pi->mode) {
+
+	case 0:
+	case 1:
+	case 2:	w0(0x60+r); w2(1); w0(val); w2(4);
+		break;
+
+	case 3:
+	case 4:
+	case 5: w3(0x40+r); w4(val);
+		break;
+
+	}
+}
+
+static int epat_read_regr( PIA *pi, int cont, int regr )
+
+{	int  a, b, r;
+
+	r = regr + cont_map[cont];
+
+	switch (pi->mode) {
+
+	case 0:	w0(r); w2(1); w2(3); 
+		a = r1(); w2(4); b = r1();
+		return j44(a,b);
+
+	case 1: w0(0x40+r); w2(1); w2(4);
+		a = r1(); b = r2(); w0(0xff);
+		return j53(a,b);
+
+	case 2: w0(0x20+r); w2(1); w2(0x25);
+		a = r0(); w2(4);
+		return a;
+
+	case 3:
+	case 4:
+	case 5: w3(r); w2(0x24); a = r4(); w2(4);
+		return a;
+
+	}
+	return -1;	/* never gets here */
+}
+
+static void epat_read_block( PIA *pi, char * buf, int count )
+
+{	int  k, ph, a, b;
+
+	switch (pi->mode) {
+
+	case 0:	w0(7); w2(1); w2(3); w0(0xff);
+		ph = 0;
+		for(k=0;k<count;k++) {
+			if (k == count-1) w0(0xfd);
+			w2(6+ph); a = r1();
+			if (a & 8) b = a; 
+			  else { w2(4+ph); b = r1(); }
+			buf[k] = j44(a,b);
+			ph =  1 - ph;
+		}
+		w0(0); w2(4);
+		break;
+
+	case 1: w0(0x47); w2(1); w2(5); w0(0xff);
+		ph = 0;
+		for(k=0;k<count;k++) {
+			if (k == count-1) w0(0xfd); 
+			w2(4+ph);
+			a = r1(); b = r2();
+			buf[k] = j53(a,b);
+			ph = 1 - ph;
+		}
+		w0(0); w2(4);
+		break;
+
+	case 2: w0(0x27); w2(1); w2(0x25); w0(0);
+		ph = 0;
+		for(k=0;k<count-1;k++) {
+			w2(0x24+ph);
+			buf[k] = r0();
+			ph = 1 - ph;
+		}
+		w2(0x26); w2(0x27); buf[count-1] = r0(); 
+		w2(0x25); w2(4);
+		break;
+
+	case 3: w3(0x80); w2(0x24);
+		for(k=0;k<count-1;k++) buf[k] = r4();
+		w2(4); w3(0xa0); w2(0x24); buf[count-1] = r4();
+		w2(4);
+		break;
+
+	case 4: w3(0x80); w2(0x24);
+		for(k=0;k<(count/2)-1;k++) ((u16 *)buf)[k] = r4w();
+		buf[count-2] = r4();
+		w2(4); w3(0xa0); w2(0x24); buf[count-1] = r4();
+		w2(4);
+		break;
+
+	case 5: w3(0x80); w2(0x24);
+		for(k=0;k<(count/4)-1;k++) ((u32 *)buf)[k] = r4l();
+		for(k=count-4;k<count-1;k++) buf[k] = r4();
+		w2(4); w3(0xa0); w2(0x24); buf[count-1] = r4();
+		w2(4);
+		break;
+
+	}
+}
+
+static void epat_write_block( PIA *pi, char * buf, int count )   
+
+{	int ph, k;
+
+	switch (pi->mode) {
+
+	case 0:
+	case 1:
+	case 2: w0(0x67); w2(1); w2(5);
+		ph = 0;
+		for(k=0;k<count;k++) {
+		  	w0(buf[k]);
+			w2(4+ph);
+			ph = 1 - ph;
+		}
+		w2(7); w2(4);
+		break;
+
+	case 3: w3(0xc0); 
+		for(k=0;k<count;k++) w4(buf[k]);
+		w2(4);
+		break;
+
+	case 4: w3(0xc0); 
+		for(k=0;k<(count/2);k++) w4w(((u16 *)buf)[k]);
+		w2(4);
+		break;
+
+	case 5: w3(0xc0); 
+		for(k=0;k<(count/4);k++) w4l(((u32 *)buf)[k]);
+		w2(4);
+		break;
+
+	}
+}
+
+/* these macros access the EPAT registers in native addressing */
+
+#define	WR(r,v)		epat_write_regr(pi,2,r,v)
+#define	RR(r)		(epat_read_regr(pi,2,r))
+
+/* and these access the IDE task file */
+
+#define WRi(r,v)         epat_write_regr(pi,0,r,v)
+#define RRi(r)           (epat_read_regr(pi,0,r))
+
+/* FIXME:  the CPP stuff should be fixed to handle multiple EPATs on a chain */
+
+#define CPP(x) 	w2(4);w0(0x22);w0(0xaa);w0(0x55);w0(0);w0(0xff);\
+                w0(0x87);w0(0x78);w0(x);w2(4);w2(5);w2(4);w0(0xff);
+
+static void epat_connect ( PIA *pi )
+
+{       pi->saved_r0 = r0();
+        pi->saved_r2 = r2();
+
+ 	/* Initialize the chip */
+	CPP(0);
+
+	if (epatc8) {
+		CPP(0x40);CPP(0xe0);
+		w0(0);w2(1);w2(4);
+		WR(0x8,0x12);WR(0xc,0x14);WR(0x12,0x10);
+		WR(0xe,0xf);WR(0xf,4);
+		/* WR(0xe,0xa);WR(0xf,4); */
+		WR(0xe,0xd);WR(0xf,0);
+		/* CPP(0x30); */
+	}
+
+        /* Connect to the chip */
+	CPP(0xe0);
+        w0(0);w2(1);w2(4); /* Idle into SPP */
+        if (pi->mode >= 3) {
+          w0(0);w2(1);w2(4);w2(0xc);
+          /* Request EPP */
+          w0(0x40);w2(6);w2(7);w2(4);w2(0xc);w2(4);
+        }
+
+	if (!epatc8) {
+		WR(8,0x10); WR(0xc,0x14); WR(0xa,0x38); WR(0x12,0x10);
+	}
+}
+
+static void epat_disconnect (PIA *pi)
+{	CPP(0x30);
+	w0(pi->saved_r0);
+	w2(pi->saved_r2);
+}
+
+static int epat_test_proto( PIA *pi, char * scratch, int verbose )
+
+{       int     k, j, f, cc;
+	int	e[2] = {0,0};
+
+        epat_connect(pi);
+	cc = RR(0xd);
+	epat_disconnect(pi);
+
+	epat_connect(pi);
+	for (j=0;j<2;j++) {
+  	    WRi(6,0xa0+j*0x10);
+            for (k=0;k<256;k++) {
+                WRi(2,k^0xaa);
+                WRi(3,k^0x55);
+                if (RRi(2) != (k^0xaa)) e[j]++;
+                }
+	    }
+        epat_disconnect(pi);
+
+        f = 0;
+        epat_connect(pi);
+        WR(0x13,1); WR(0x13,0); WR(0xa,0x11);
+        epat_read_block(pi,scratch,512);
+	
+        for (k=0;k<256;k++) {
+            if ((scratch[2*k] & 0xff) != k) f++;
+            if ((scratch[2*k+1] & 0xff) != (0xff-k)) f++;
+        }
+        epat_disconnect(pi);
+
+        if (verbose)  {
+            printk("%s: epat: port 0x%x, mode %d, ccr %x, test=(%d,%d,%d)\n",
+		   pi->device,pi->port,pi->mode,cc,e[0],e[1],f);
+	}
+	
+        return (e[0] && e[1]) || f;
+}
+
+static void epat_log_adapter( PIA *pi, char * scratch, int verbose )
+
+{	int	ver;
+        char    *mode_string[6] = 
+		   {"4-bit","5/3","8-bit","EPP-8","EPP-16","EPP-32"};
+
+	epat_connect(pi);
+	WR(0xa,0x38);		/* read the version code */
+        ver = RR(0xb);
+        epat_disconnect(pi);
+
+	printk("%s: epat %s, Shuttle EPAT chip %x at 0x%x, ",
+		pi->device,EPAT_VERSION,ver,pi->port);
+	printk("mode %d (%s), delay %d\n",pi->mode,
+		mode_string[pi->mode],pi->delay);
+
+}
+
+static struct pi_protocol epat = {
+	.owner		= THIS_MODULE,
+	.name		= "epat",
+	.max_mode	= 6,
+	.epp_first	= 3,
+	.default_delay	= 1,
+	.max_units	= 1,
+	.write_regr	= epat_write_regr,
+	.read_regr	= epat_read_regr,
+	.write_block	= epat_write_block,
+	.read_block	= epat_read_block,
+	.connect	= epat_connect,
+	.disconnect	= epat_disconnect,
+	.test_proto	= epat_test_proto,
+	.log_adapter	= epat_log_adapter,
+};
+
+static int __init epat_init(void)
+{
+#ifdef CONFIG_PARIDE_EPATC8
+	epatc8 = 1;
+#endif
+	return paride_register(&epat);
+}
+
+static void __exit epat_exit(void)
+{
+	paride_unregister(&epat);
+}
+
+MODULE_LICENSE("GPL");
+module_init(epat_init)
+module_exit(epat_exit)

drivers/block/paride/epia.c

@@ -0,0 +1,316 @@
+/* 
+        epia.c    (c) 1997-8  Grant R. Guenther <grant@torque.net>
+                              Under the terms of the GNU General Public License.
+
+        epia.c is a low-level protocol driver for Shuttle Technologies 
+	EPIA parallel to IDE adapter chip.  This device is now obsolete
+	and has been replaced with the EPAT chip, which is supported
+	by epat.c, however, some devices based on EPIA are still
+	available.
+
+*/
+
+/* Changes:
+
+        1.01    GRG 1998.05.06 init_proto, release_proto
+	1.02    GRG 1998.06.17 support older versions of EPIA
+
+*/
+
+#define EPIA_VERSION      "1.02"
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/wait.h>
+#include <asm/io.h>
+
+#include "paride.h"
+
+/* mode codes:  0  nybble reads on port 1, 8-bit writes
+                1  5/3 reads on ports 1 & 2, 8-bit writes
+                2  8-bit reads and writes
+                3  8-bit EPP mode
+		4  16-bit EPP
+		5  32-bit EPP
+*/
+
+#define j44(a,b)                (((a>>4)&0x0f)+(b&0xf0))
+#define j53(a,b)                (((a>>3)&0x1f)+((b<<4)&0xe0))
+
+/* cont =  0   IDE register file
+   cont =  1   IDE control registers
+*/
+
+static int cont_map[2] = { 0, 0x80 };
+
+static int epia_read_regr( PIA *pi, int cont, int regr )
+
+{       int     a, b, r;
+
+	regr += cont_map[cont];
+
+        switch (pi->mode)  {
+
+        case 0: r = regr^0x39;
+                w0(r); w2(1); w2(3); w0(r);
+                a = r1(); w2(1); b = r1(); w2(4);
+                return j44(a,b);
+
+        case 1: r = regr^0x31;
+                w0(r); w2(1); w0(r&0x37); 
+                w2(3); w2(5); w0(r|0xf0);
+                a = r1(); b = r2(); w2(4);
+                return j53(a,b);
+
+        case 2: r = regr^0x29;
+                w0(r); w2(1); w2(0X21); w2(0x23); 
+                a = r0(); w2(4);
+                return a;
+
+	case 3:
+	case 4:
+        case 5: w3(regr); w2(0x24); a = r4(); w2(4);
+                return a;
+
+        }
+        return -1;
+}       
+
+static void epia_write_regr( PIA *pi, int cont, int regr, int val)
+
+{       int  r;
+
+	regr += cont_map[cont];
+
+        switch (pi->mode)  {
+
+        case 0:
+        case 1:
+        case 2: r = regr^0x19;
+                w0(r); w2(1); w0(val); w2(3); w2(4);
+                break;
+
+	case 3:
+	case 4:
+        case 5: r = regr^0x40;
+                w3(r); w4(val); w2(4);
+                break;
+        }
+}
+
+#define WR(r,v)         epia_write_regr(pi,0,r,v)
+#define RR(r)           (epia_read_regr(pi,0,r))
+
+/* The use of register 0x84 is entirely unclear - it seems to control
+   some EPP counters ...  currently we know about 3 different block
+   sizes:  the standard 512 byte reads and writes, 12 byte writes and 
+   2048 byte reads (the last two being used in the CDrom drivers.
+*/
+
+static void epia_connect ( PIA *pi  )
+
+{       pi->saved_r0 = r0();
+        pi->saved_r2 = r2();
+
+        w2(4); w0(0xa0); w0(0x50); w0(0xc0); w0(0x30); w0(0xa0); w0(0);
+        w2(1); w2(4);
+        if (pi->mode >= 3) { 
+                w0(0xa); w2(1); w2(4); w0(0x82); w2(4); w2(0xc); w2(4);
+                w2(0x24); w2(0x26); w2(4);
+        }
+        WR(0x86,8);  
+}
+
+static void epia_disconnect ( PIA *pi )
+
+{       /* WR(0x84,0x10); */
+        w0(pi->saved_r0);
+        w2(1); w2(4);
+        w0(pi->saved_r0);
+        w2(pi->saved_r2);
+} 
+
+static void epia_read_block( PIA *pi, char * buf, int count )
+
+{       int     k, ph, a, b;
+
+        switch (pi->mode) {
+
+        case 0: w0(0x81); w2(1); w2(3); w0(0xc1);
+                ph = 1;
+                for (k=0;k<count;k++) {
+                        w2(2+ph); a = r1();
+                        w2(4+ph); b = r1();
+                        buf[k] = j44(a,b);
+                        ph = 1 - ph;
+                } 
+                w0(0); w2(4);
+                break;
+
+        case 1: w0(0x91); w2(1); w0(0x10); w2(3); 
+                w0(0x51); w2(5); w0(0xd1); 
+                ph = 1;
+                for (k=0;k<count;k++) {
+                        w2(4+ph);
+                        a = r1(); b = r2();
+                        buf[k] = j53(a,b);
+                        ph = 1 - ph;
+                }
+                w0(0); w2(4);
+                break;
+
+        case 2: w0(0x89); w2(1); w2(0x23); w2(0x21); 
+                ph = 1;
+                for (k=0;k<count;k++) {
+                        w2(0x24+ph);
+                        buf[k] = r0();
+                        ph = 1 - ph;
+                }
+                w2(6); w2(4);
+                break;
+
+        case 3: if (count > 512) WR(0x84,3);
+		w3(0); w2(0x24);
+                for (k=0;k<count;k++) buf[k] = r4();
+                w2(4); WR(0x84,0);
+                break;
+
+        case 4: if (count > 512) WR(0x84,3);
+		w3(0); w2(0x24);
+		for (k=0;k<count/2;k++) ((u16 *)buf)[k] = r4w();
+                w2(4); WR(0x84,0);
+                break;
+
+        case 5: if (count > 512) WR(0x84,3);
+		w3(0); w2(0x24);
+                for (k=0;k<count/4;k++) ((u32 *)buf)[k] = r4l();
+                w2(4); WR(0x84,0);
+                break;
+
+        }
+}
+
+static void epia_write_block( PIA *pi, char * buf, int count )
+
+{       int     ph, k, last, d;
+
+        switch (pi->mode) {
+
+        case 0:
+        case 1:
+        case 2: w0(0xa1); w2(1); w2(3); w2(1); w2(5);
+                ph = 0;  last = 0x8000;
+                for (k=0;k<count;k++) {
+                        d = buf[k];
+                        if (d != last) { last = d; w0(d); }
+                        w2(4+ph);
+                        ph = 1 - ph;
+                }
+                w2(7); w2(4);
+                break;
+
+        case 3: if (count < 512) WR(0x84,1);
+		w3(0x40);
+                for (k=0;k<count;k++) w4(buf[k]);
+		if (count < 512) WR(0x84,0);
+                break;
+
+        case 4: if (count < 512) WR(0x84,1);
+		w3(0x40);
+                for (k=0;k<count/2;k++) w4w(((u16 *)buf)[k]);
+		if (count < 512) WR(0x84,0);
+                break;
+
+        case 5: if (count < 512) WR(0x84,1);
+		w3(0x40);
+                for (k=0;k<count/4;k++) w4l(((u32 *)buf)[k]);
+		if (count < 512) WR(0x84,0);
+                break;
+
+        }
+
+}
+
+static int epia_test_proto( PIA *pi, char * scratch, int verbose )
+
+{       int     j, k, f;
+	int	e[2] = {0,0};
+
+        epia_connect(pi);
+        for (j=0;j<2;j++) {
+            WR(6,0xa0+j*0x10);
+            for (k=0;k<256;k++) {
+                WR(2,k^0xaa);
+                WR(3,k^0x55);
+                if (RR(2) != (k^0xaa)) e[j]++;
+                }
+	    WR(2,1); WR(3,1);
+            }
+        epia_disconnect(pi);
+
+        f = 0;
+        epia_connect(pi);
+        WR(0x84,8);
+        epia_read_block(pi,scratch,512);
+        for (k=0;k<256;k++) {
+            if ((scratch[2*k] & 0xff) != ((k+1) & 0xff)) f++;
+            if ((scratch[2*k+1] & 0xff) != ((-2-k) & 0xff)) f++;
+        }
+        WR(0x84,0);
+        epia_disconnect(pi);
+
+        if (verbose)  {
+            printk("%s: epia: port 0x%x, mode %d, test=(%d,%d,%d)\n",
+                   pi->device,pi->port,pi->mode,e[0],e[1],f);
+        }
+        
+        return (e[0] && e[1]) || f;
+
+}
+
+
+static void epia_log_adapter( PIA *pi, char * scratch, int verbose )
+
+{       char    *mode_string[6] = {"4-bit","5/3","8-bit",
+				   "EPP-8","EPP-16","EPP-32"};
+
+        printk("%s: epia %s, Shuttle EPIA at 0x%x, ",
+                pi->device,EPIA_VERSION,pi->port);
+        printk("mode %d (%s), delay %d\n",pi->mode,
+		mode_string[pi->mode],pi->delay);
+
+}
+
+static struct pi_protocol epia = {
+	.owner		= THIS_MODULE,
+	.name		= "epia",
+	.max_mode	= 6,
+	.epp_first	= 3,
+	.default_delay	= 1,
+	.max_units	= 1,
+	.write_regr	= epia_write_regr,
+	.read_regr	= epia_read_regr,
+	.write_block	= epia_write_block,
+	.read_block	= epia_read_block,
+	.connect	= epia_connect,
+	.disconnect	= epia_disconnect,
+	.test_proto	= epia_test_proto,
+	.log_adapter	= epia_log_adapter,
+};
+
+static int __init epia_init(void)
+{
+	return paride_register(&epia);
+}
+
+static void __exit epia_exit(void)
+{
+	paride_unregister(&epia);
+}
+
+MODULE_LICENSE("GPL");
+module_init(epia_init)
+module_exit(epia_exit)

drivers/block/paride/fit2.c

@@ -0,0 +1,151 @@
+/* 
+        fit2.c        (c) 1998  Grant R. Guenther <grant@torque.net>
+                          Under the terms of the GNU General Public License.
+
+	fit2.c is a low-level protocol driver for the older version
+        of the Fidelity International Technology parallel port adapter.  
+	This adapter is used in their TransDisk 2000 and older TransDisk
+	3000 portable hard-drives.  As far as I can tell, this device
+	supports 4-bit mode _only_.  
+
+	Newer models of the FIT products use an enhanced protocol.
+	The "fit3" protocol module should support current drives.
+
+*/
+
+#define FIT2_VERSION      "1.0"
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/wait.h>
+#include <asm/io.h>
+
+#include "paride.h"
+
+#define j44(a,b)                (((a>>4)&0x0f)|(b&0xf0))
+
+/* cont = 0 - access the IDE register file 
+   cont = 1 - access the IDE command set 
+
+NB:  The FIT adapter does not appear to use the control registers.
+So, we map ALT_STATUS to STATUS and NO-OP writes to the device
+control register - this means that IDE reset will not work on these
+devices.
+
+*/
+
+static void  fit2_write_regr( PIA *pi, int cont, int regr, int val)
+
+{	if (cont == 1) return;
+	w2(0xc); w0(regr); w2(4); w0(val); w2(5); w0(0); w2(4);
+}
+
+static int fit2_read_regr( PIA *pi, int cont, int regr )
+
+{	int  a, b, r;
+
+	if (cont) {
+	  if (regr != 6) return 0xff;
+	  r = 7;
+	} else r = regr + 0x10;
+
+	w2(0xc); w0(r); w2(4); w2(5); 
+	         w0(0); a = r1();
+	         w0(1); b = r1();
+	w2(4);
+
+	return j44(a,b);
+
+}
+
+static void fit2_read_block( PIA *pi, char * buf, int count )
+
+{	int  k, a, b, c, d;
+
+	w2(0xc); w0(0x10);
+
+	for (k=0;k<count/4;k++) {
+
+		w2(4); w2(5);
+		w0(0); a = r1(); w0(1); b = r1();
+		w0(3); c = r1(); w0(2); d = r1(); 
+		buf[4*k+0] = j44(a,b);
+		buf[4*k+1] = j44(d,c);
+
+                w2(4); w2(5);
+                       a = r1(); w0(3); b = r1();
+                w0(1); c = r1(); w0(0); d = r1(); 
+                buf[4*k+2] = j44(d,c);
+                buf[4*k+3] = j44(a,b);
+
+	}
+
+	w2(4);
+
+}
+
+static void fit2_write_block( PIA *pi, char * buf, int count )
+
+{	int k;
+
+
+	w2(0xc); w0(0); 
+	for (k=0;k<count/2;k++) {
+		w2(4); w0(buf[2*k]); 
+		w2(5); w0(buf[2*k+1]);
+	}
+	w2(4);
+}
+
+static void fit2_connect ( PIA *pi  )
+
+{       pi->saved_r0 = r0();
+        pi->saved_r2 = r2();
+	w2(0xcc); 
+}
+
+static void fit2_disconnect ( PIA *pi )
+
+{       w0(pi->saved_r0);
+        w2(pi->saved_r2);
+} 
+
+static void fit2_log_adapter( PIA *pi, char * scratch, int verbose )
+
+{       printk("%s: fit2 %s, FIT 2000 adapter at 0x%x, delay %d\n",
+                pi->device,FIT2_VERSION,pi->port,pi->delay);
+
+}
+
+static struct pi_protocol fit2 = {
+	.owner		= THIS_MODULE,
+	.name		= "fit2",
+	.max_mode	= 1,
+	.epp_first	= 2,
+	.default_delay	= 1,
+	.max_units	= 1,
+	.write_regr	= fit2_write_regr,
+	.read_regr	= fit2_read_regr,
+	.write_block	= fit2_write_block,
+	.read_block	= fit2_read_block,
+	.connect	= fit2_connect,
+	.disconnect	= fit2_disconnect,
+	.log_adapter	= fit2_log_adapter,
+};
+
+static int __init fit2_init(void)
+{
+	return paride_register(&fit2);
+}
+
+static void __exit fit2_exit(void)
+{
+	paride_unregister(&fit2);
+}
+
+MODULE_LICENSE("GPL");
+module_init(fit2_init)
+module_exit(fit2_exit)

drivers/block/paride/fit3.c

@@ -0,0 +1,211 @@
+/* 
+        fit3.c        (c) 1998  Grant R. Guenther <grant@torque.net>
+                          Under the terms of the GNU General Public License.
+
+	fit3.c is a low-level protocol driver for newer models 
+        of the Fidelity International Technology parallel port adapter.  
+	This adapter is used in their TransDisk 3000 portable 
+	hard-drives, as well as CD-ROM, PD-CD and other devices.
+
+	The TD-2000 and certain older devices use a different protocol.
+	Try the fit2 protocol module with them.
+
+        NB:  The FIT adapters do not appear to support the control 
+	registers.  So, we map ALT_STATUS to STATUS and NO-OP writes 
+	to the device control register - this means that IDE reset 
+	will not work on these devices.
+
+*/
+
+#define FIT3_VERSION      "1.0"
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/wait.h>
+#include <asm/io.h>
+
+#include "paride.h"
+
+#define j44(a,b)                (((a>>3)&0x0f)|((b<<1)&0xf0))
+
+#define w7(byte)                {out_p(7,byte);}
+#define r7()                    (in_p(7) & 0xff)
+
+/* cont = 0 - access the IDE register file 
+   cont = 1 - access the IDE command set 
+
+*/
+
+static void  fit3_write_regr( PIA *pi, int cont, int regr, int val)
+
+{	if (cont == 1) return;
+
+	switch (pi->mode) {
+
+	case 0:
+	case 1: w2(0xc); w0(regr); w2(0x8); w2(0xc); 
+		w0(val); w2(0xd); 
+		w0(0);   w2(0xc);
+		break;
+
+	case 2: w2(0xc); w0(regr); w2(0x8); w2(0xc);
+		w4(val); w4(0);
+		w2(0xc);
+		break;
+
+	}
+}
+
+static int fit3_read_regr( PIA *pi, int cont, int regr )
+
+{	int  a, b;
+
+	if (cont) {
+	  if (regr != 6) return 0xff;
+	  regr = 7;
+	} 
+
+	switch (pi->mode) {
+
+	case 0: w2(0xc); w0(regr + 0x10); w2(0x8); w2(0xc);
+		w2(0xd); a = r1();
+		w2(0xf); b = r1(); 
+		w2(0xc);
+		return j44(a,b);
+
+	case 1: w2(0xc); w0(regr + 0x90); w2(0x8); w2(0xc);
+		w2(0xec); w2(0xee); w2(0xef); a = r0(); 
+		w2(0xc);
+		return a;
+
+	case 2: w2(0xc); w0(regr + 0x90); w2(0x8); w2(0xc); 
+		w2(0xec); 
+		a = r4(); b = r4(); 
+		w2(0xc);
+		return a;
+
+	}
+	return -1; 
+
+}
+
+static void fit3_read_block( PIA *pi, char * buf, int count )
+
+{	int  k, a, b, c, d;
+
+	switch (pi->mode) {
+
+	case 0: w2(0xc); w0(0x10); w2(0x8); w2(0xc);
+		for (k=0;k<count/2;k++) {
+		    w2(0xd); a = r1();
+		    w2(0xf); b = r1();
+		    w2(0xc); c = r1();
+		    w2(0xe); d = r1();
+		    buf[2*k  ] = j44(a,b);
+		    buf[2*k+1] = j44(c,d);
+		}
+		w2(0xc);
+		break;
+
+	case 1: w2(0xc); w0(0x90); w2(0x8); w2(0xc); 
+		w2(0xec); w2(0xee);
+		for (k=0;k<count/2;k++) {
+		    w2(0xef); a = r0();
+		    w2(0xee); b = r0();
+                    buf[2*k  ] = a;
+                    buf[2*k+1] = b;
+		}
+		w2(0xec); 
+		w2(0xc);
+		break;
+
+	case 2: w2(0xc); w0(0x90); w2(0x8); w2(0xc); 
+                w2(0xec);
+		for (k=0;k<count;k++) buf[k] = r4();
+                w2(0xc);
+		break;
+
+	}
+}
+
+static void fit3_write_block( PIA *pi, char * buf, int count )
+
+{	int k;
+
+        switch (pi->mode) {
+
+	case 0:
+        case 1: w2(0xc); w0(0); w2(0x8); w2(0xc);
+                for (k=0;k<count/2;k++) {
+ 		    w0(buf[2*k  ]); w2(0xd);
+ 		    w0(buf[2*k+1]); w2(0xc);
+		}
+		break;
+
+        case 2: w2(0xc); w0(0); w2(0x8); w2(0xc); 
+                for (k=0;k<count;k++) w4(buf[k]);
+                w2(0xc);
+		break;
+	}
+}
+
+static void fit3_connect ( PIA *pi  )
+
+{       pi->saved_r0 = r0();
+        pi->saved_r2 = r2();
+	w2(0xc); w0(0); w2(0xa);
+	if (pi->mode == 2) { 
+		w2(0xc); w0(0x9); w2(0x8); w2(0xc); 
+		}
+}
+
+static void fit3_disconnect ( PIA *pi )
+
+{       w2(0xc); w0(0xa); w2(0x8); w2(0xc);
+	w0(pi->saved_r0);
+        w2(pi->saved_r2);
+} 
+
+static void fit3_log_adapter( PIA *pi, char * scratch, int verbose )
+
+{       char    *mode_string[3] = {"4-bit","8-bit","EPP"};
+
+	printk("%s: fit3 %s, FIT 3000 adapter at 0x%x, "
+	       "mode %d (%s), delay %d\n",
+                pi->device,FIT3_VERSION,pi->port,
+		pi->mode,mode_string[pi->mode],pi->delay);
+
+}
+
+static struct pi_protocol fit3 = {
+	.owner		= THIS_MODULE,
+	.name		= "fit3",
+	.max_mode	= 3,
+	.epp_first	= 2,
+	.default_delay	= 1,
+	.max_units	= 1,
+	.write_regr	= fit3_write_regr,
+	.read_regr	= fit3_read_regr,
+	.write_block	= fit3_write_block,
+	.read_block	= fit3_read_block,
+	.connect	= fit3_connect,
+	.disconnect	= fit3_disconnect,
+	.log_adapter	= fit3_log_adapter,
+};
+
+static int __init fit3_init(void)
+{
+	return paride_register(&fit3);
+}
+
+static void __exit fit3_exit(void)
+{
+	paride_unregister(&fit3);
+}
+
+MODULE_LICENSE("GPL");
+module_init(fit3_init)
+module_exit(fit3_exit)

drivers/block/paride/friq.c

@@ -0,0 +1,276 @@
+/* 
+	friq.c	(c) 1998    Grant R. Guenther <grant@torque.net>
+		            Under the terms of the GNU General Public License
+
+	friq.c is a low-level protocol driver for the Freecom "IQ"
+	parallel port IDE adapter.   Early versions of this adapter
+	use the 'frpw' protocol.
+	
+	Freecom uses this adapter in a battery powered external 
+	CD-ROM drive.  It is also used in LS-120 drives by
+	Maxell and Panasonic, and other devices.
+
+	The battery powered drive requires software support to
+	control the power to the drive.  This module enables the
+	drive power when the high level driver (pcd) is loaded
+	and disables it when the module is unloaded.  Note, if
+	the friq module is built in to the kernel, the power
+	will never be switched off, so other means should be
+	used to conserve battery power.
+
+*/
+
+/* Changes:
+
+	1.01	GRG 1998.12.20	 Added support for soft power switch
+*/
+
+#define	FRIQ_VERSION	"1.01" 
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/wait.h>
+#include <asm/io.h>
+
+#include "paride.h"
+
+#define CMD(x)		w2(4);w0(0xff);w0(0xff);w0(0x73);w0(0x73);\
+			w0(0xc9);w0(0xc9);w0(0x26);w0(0x26);w0(x);w0(x);
+
+#define j44(l,h)	(((l>>4)&0x0f)|(h&0xf0))
+
+/* cont = 0 - access the IDE register file 
+   cont = 1 - access the IDE command set 
+*/
+
+static int  cont_map[2] = { 0x08, 0x10 };
+
+static int friq_read_regr( PIA *pi, int cont, int regr )
+
+{	int	h,l,r;
+
+	r = regr + cont_map[cont];
+
+	CMD(r);
+	w2(6); l = r1();
+	w2(4); h = r1();
+	w2(4); 
+
+	return j44(l,h);
+
+}
+
+static void friq_write_regr( PIA *pi, int cont, int regr, int val)
+
+{	int r;
+
+        r = regr + cont_map[cont];
+
+	CMD(r);
+	w0(val);
+	w2(5);w2(7);w2(5);w2(4);
+}
+
+static void friq_read_block_int( PIA *pi, char * buf, int count, int regr )
+
+{       int     h, l, k, ph;
+
+        switch(pi->mode) {
+
+        case 0: CMD(regr); 
+                for (k=0;k<count;k++) {
+                        w2(6); l = r1();
+                        w2(4); h = r1();
+                        buf[k] = j44(l,h);
+                }
+                w2(4);
+                break;
+
+        case 1: ph = 2;
+                CMD(regr+0xc0); 
+                w0(0xff);
+                for (k=0;k<count;k++) {
+                        w2(0xa4 + ph); 
+                        buf[k] = r0();
+                        ph = 2 - ph;
+                } 
+                w2(0xac); w2(0xa4); w2(4);
+                break;
+
+	case 2: CMD(regr+0x80);
+		for (k=0;k<count-2;k++) buf[k] = r4();
+		w2(0xac); w2(0xa4);
+		buf[count-2] = r4();
+		buf[count-1] = r4();
+		w2(4);
+		break;
+
+	case 3: CMD(regr+0x80);
+                for (k=0;k<(count/2)-1;k++) ((u16 *)buf)[k] = r4w();
+                w2(0xac); w2(0xa4);
+                buf[count-2] = r4();
+                buf[count-1] = r4();
+                w2(4);
+                break;
+
+	case 4: CMD(regr+0x80);
+                for (k=0;k<(count/4)-1;k++) ((u32 *)buf)[k] = r4l();
+                buf[count-4] = r4();
+                buf[count-3] = r4();
+                w2(0xac); w2(0xa4);
+                buf[count-2] = r4();
+                buf[count-1] = r4();
+                w2(4);
+                break;
+
+        }
+}
+
+static void friq_read_block( PIA *pi, char * buf, int count)
+
+{	friq_read_block_int(pi,buf,count,0x08);
+}
+
+static void friq_write_block( PIA *pi, char * buf, int count )
+ 
+{	int	k;
+
+	switch(pi->mode) {
+
+	case 0:
+	case 1: CMD(8); w2(5);
+        	for (k=0;k<count;k++) {
+			w0(buf[k]);
+			w2(7);w2(5);
+		}
+		w2(4);
+		break;
+
+	case 2: CMD(0xc8); w2(5);
+		for (k=0;k<count;k++) w4(buf[k]);
+		w2(4);
+		break;
+
+        case 3: CMD(0xc8); w2(5);
+                for (k=0;k<count/2;k++) w4w(((u16 *)buf)[k]);
+                w2(4);
+                break;
+
+        case 4: CMD(0xc8); w2(5);
+                for (k=0;k<count/4;k++) w4l(((u32 *)buf)[k]);
+                w2(4);
+                break;
+	}
+}
+
+static void friq_connect ( PIA *pi  )
+
+{       pi->saved_r0 = r0();
+        pi->saved_r2 = r2();
+	w2(4);
+}
+
+static void friq_disconnect ( PIA *pi )
+
+{       CMD(0x20);
+	w0(pi->saved_r0);
+        w2(pi->saved_r2);
+} 
+
+static int friq_test_proto( PIA *pi, char * scratch, int verbose )
+
+{       int     j, k, r;
+	int	e[2] = {0,0};
+
+	pi->saved_r0 = r0();	
+	w0(0xff); udelay(20); CMD(0x3d); /* turn the power on */
+	udelay(500);
+	w0(pi->saved_r0);
+
+	friq_connect(pi);
+	for (j=0;j<2;j++) {
+                friq_write_regr(pi,0,6,0xa0+j*0x10);
+                for (k=0;k<256;k++) {
+                        friq_write_regr(pi,0,2,k^0xaa);
+                        friq_write_regr(pi,0,3,k^0x55);
+                        if (friq_read_regr(pi,0,2) != (k^0xaa)) e[j]++;
+                        }
+                }
+	friq_disconnect(pi);
+
+	friq_connect(pi);
+        friq_read_block_int(pi,scratch,512,0x10);
+        r = 0;
+        for (k=0;k<128;k++) if (scratch[k] != k) r++;
+	friq_disconnect(pi);
+
+        if (verbose)  {
+            printk("%s: friq: port 0x%x, mode %d, test=(%d,%d,%d)\n",
+                   pi->device,pi->port,pi->mode,e[0],e[1],r);
+        }
+
+        return (r || (e[0] && e[1]));
+}
+
+
+static void friq_log_adapter( PIA *pi, char * scratch, int verbose )
+
+{       char    *mode_string[6] = {"4-bit","8-bit",
+				   "EPP-8","EPP-16","EPP-32"};
+
+        printk("%s: friq %s, Freecom IQ ASIC-2 adapter at 0x%x, ", pi->device,
+		FRIQ_VERSION,pi->port);
+        printk("mode %d (%s), delay %d\n",pi->mode,
+		mode_string[pi->mode],pi->delay);
+
+	pi->private = 1;
+	friq_connect(pi);
+	CMD(0x9e);  		/* disable sleep timer */
+	friq_disconnect(pi);
+
+}
+
+static void friq_release_proto( PIA *pi)
+{
+	if (pi->private) {		/* turn off the power */
+		friq_connect(pi);
+		CMD(0x1d); CMD(0x1e);
+		friq_disconnect(pi);
+		pi->private = 0;
+	}
+}
+
+static struct pi_protocol friq = {
+	.owner		= THIS_MODULE,
+	.name		= "friq",
+	.max_mode	= 5,
+	.epp_first	= 2,
+	.default_delay	= 1,
+	.max_units	= 1,
+	.write_regr	= friq_write_regr,
+	.read_regr	= friq_read_regr,
+	.write_block	= friq_write_block,
+	.read_block	= friq_read_block,
+	.connect	= friq_connect,
+	.disconnect	= friq_disconnect,
+	.test_proto	= friq_test_proto,
+	.log_adapter	= friq_log_adapter,
+	.release_proto	= friq_release_proto,
+};
+
+static int __init friq_init(void)
+{
+	return paride_register(&friq);
+}
+
+static void __exit friq_exit(void)
+{
+	paride_unregister(&friq);
+}
+
+MODULE_LICENSE("GPL");
+module_init(friq_init)
+module_exit(friq_exit)

drivers/block/paride/frpw.c

@@ -0,0 +1,313 @@
+/* 
+	frpw.c	(c) 1996-8  Grant R. Guenther <grant@torque.net>
+		            Under the terms of the GNU General Public License
+
+	frpw.c is a low-level protocol driver for the Freecom "Power"
+	parallel port IDE adapter.
+	
+	Some applications of this adapter may require a "printer" reset
+	prior to loading the driver.  This can be done by loading and
+	unloading the "lp" driver, or it can be done by this driver
+	if you define FRPW_HARD_RESET.  The latter is not recommended
+	as it may upset devices on other ports.
+
+*/
+
+/* Changes:
+
+        1.01    GRG 1998.05.06 init_proto, release_proto
+			       fix chip detect
+			       added EPP-16 and EPP-32
+	1.02    GRG 1998.09.23 added hard reset to initialisation process
+	1.03    GRG 1998.12.14 made hard reset conditional
+
+*/
+
+#define	FRPW_VERSION	"1.03" 
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/wait.h>
+#include <asm/io.h>
+
+#include "paride.h"
+
+#define cec4		w2(0xc);w2(0xe);w2(0xe);w2(0xc);w2(4);w2(4);w2(4);
+#define j44(l,h)	(((l>>4)&0x0f)|(h&0xf0))
+
+/* cont = 0 - access the IDE register file 
+   cont = 1 - access the IDE command set 
+*/
+
+static int  cont_map[2] = { 0x08, 0x10 };
+
+static int frpw_read_regr( PIA *pi, int cont, int regr )
+
+{	int	h,l,r;
+
+	r = regr + cont_map[cont];
+
+	w2(4);
+	w0(r); cec4;
+	w2(6); l = r1();
+	w2(4); h = r1();
+	w2(4); 
+
+	return j44(l,h);
+
+}
+
+static void frpw_write_regr( PIA *pi, int cont, int regr, int val)
+
+{	int r;
+
+        r = regr + cont_map[cont];
+
+	w2(4); w0(r); cec4; 
+	w0(val);
+	w2(5);w2(7);w2(5);w2(4);
+}
+
+static void frpw_read_block_int( PIA *pi, char * buf, int count, int regr )
+
+{       int     h, l, k, ph;
+
+        switch(pi->mode) {
+
+        case 0: w2(4); w0(regr); cec4;
+                for (k=0;k<count;k++) {
+                        w2(6); l = r1();
+                        w2(4); h = r1();
+                        buf[k] = j44(l,h);
+                }
+                w2(4);
+                break;
+
+        case 1: ph = 2;
+                w2(4); w0(regr + 0xc0); cec4;
+                w0(0xff);
+                for (k=0;k<count;k++) {
+                        w2(0xa4 + ph); 
+                        buf[k] = r0();
+                        ph = 2 - ph;
+                } 
+                w2(0xac); w2(0xa4); w2(4);
+                break;
+
+        case 2: w2(4); w0(regr + 0x80); cec4;
+                for (k=0;k<count;k++) buf[k] = r4();
+                w2(0xac); w2(0xa4);
+                w2(4);
+                break;
+
+	case 3: w2(4); w0(regr + 0x80); cec4;
+		for (k=0;k<count-2;k++) buf[k] = r4();
+		w2(0xac); w2(0xa4);
+		buf[count-2] = r4();
+		buf[count-1] = r4();
+		w2(4);
+		break;
+
+	case 4: w2(4); w0(regr + 0x80); cec4;
+                for (k=0;k<(count/2)-1;k++) ((u16 *)buf)[k] = r4w();
+                w2(0xac); w2(0xa4);
+                buf[count-2] = r4();
+                buf[count-1] = r4();
+                w2(4);
+                break;
+
+	case 5: w2(4); w0(regr + 0x80); cec4;
+                for (k=0;k<(count/4)-1;k++) ((u32 *)buf)[k] = r4l();
+                buf[count-4] = r4();
+                buf[count-3] = r4();
+                w2(0xac); w2(0xa4);
+                buf[count-2] = r4();
+                buf[count-1] = r4();
+                w2(4);
+                break;
+
+        }
+}
+
+static void frpw_read_block( PIA *pi, char * buf, int count)
+
+{	frpw_read_block_int(pi,buf,count,0x08);
+}
+
+static void frpw_write_block( PIA *pi, char * buf, int count )
+ 
+{	int	k;
+
+	switch(pi->mode) {
+
+	case 0:
+	case 1:
+	case 2: w2(4); w0(8); cec4; w2(5);
+        	for (k=0;k<count;k++) {
+			w0(buf[k]);
+			w2(7);w2(5);
+		}
+		w2(4);
+		break;
+
+	case 3: w2(4); w0(0xc8); cec4; w2(5);
+		for (k=0;k<count;k++) w4(buf[k]);
+		w2(4);
+		break;
+
+        case 4: w2(4); w0(0xc8); cec4; w2(5);
+                for (k=0;k<count/2;k++) w4w(((u16 *)buf)[k]);
+                w2(4);
+                break;
+
+        case 5: w2(4); w0(0xc8); cec4; w2(5);
+                for (k=0;k<count/4;k++) w4l(((u32 *)buf)[k]);
+                w2(4);
+                break;
+	}
+}
+
+static void frpw_connect ( PIA *pi  )
+
+{       pi->saved_r0 = r0();
+        pi->saved_r2 = r2();
+	w2(4);
+}
+
+static void frpw_disconnect ( PIA *pi )
+
+{       w2(4); w0(0x20); cec4;
+	w0(pi->saved_r0);
+        w2(pi->saved_r2);
+} 
+
+/* Stub logic to see if PNP string is available - used to distinguish
+   between the Xilinx and ASIC implementations of the Freecom adapter.
+*/
+
+static int frpw_test_pnp ( PIA *pi )
+
+/*  returns chip_type:   0 = Xilinx, 1 = ASIC   */
+
+{	int olddelay, a, b;
+
+#ifdef FRPW_HARD_RESET
+        w0(0); w2(8); udelay(50); w2(0xc);   /* parallel bus reset */
+        mdelay(1500);
+#endif
+
+	olddelay = pi->delay;
+	pi->delay = 10;
+
+	pi->saved_r0 = r0();
+        pi->saved_r2 = r2();
+	
+	w2(4); w0(4); w2(6); w2(7);
+	a = r1() & 0xff; w2(4); b = r1() & 0xff;
+	w2(0xc); w2(0xe); w2(4);
+
+	pi->delay = olddelay;
+        w0(pi->saved_r0);
+        w2(pi->saved_r2);
+
+	return ((~a&0x40) && (b&0x40));
+} 
+
+/* We use the pi->private to remember the result of the PNP test.
+   To make this work, private = port*2 + chip.  Yes, I know it's
+   a hack :-(
+*/
+
+static int frpw_test_proto( PIA *pi, char * scratch, int verbose )
+
+{       int     j, k, r;
+	int	e[2] = {0,0};
+
+	if ((pi->private>>1) != pi->port)
+	   pi->private = frpw_test_pnp(pi) + 2*pi->port;
+
+	if (((pi->private%2) == 0) && (pi->mode > 2)) {
+	   if (verbose) 
+		printk("%s: frpw: Xilinx does not support mode %d\n",
+			pi->device, pi->mode);
+	   return 1;
+	}
+
+	if (((pi->private%2) == 1) && (pi->mode == 2)) {
+	   if (verbose)
+		printk("%s: frpw: ASIC does not support mode 2\n",
+			pi->device);
+	   return 1;
+	}
+
+	frpw_connect(pi);
+	for (j=0;j<2;j++) {
+                frpw_write_regr(pi,0,6,0xa0+j*0x10);
+                for (k=0;k<256;k++) {
+                        frpw_write_regr(pi,0,2,k^0xaa);
+                        frpw_write_regr(pi,0,3,k^0x55);
+                        if (frpw_read_regr(pi,0,2) != (k^0xaa)) e[j]++;
+                        }
+                }
+	frpw_disconnect(pi);
+
+	frpw_connect(pi);
+        frpw_read_block_int(pi,scratch,512,0x10);
+        r = 0;
+        for (k=0;k<128;k++) if (scratch[k] != k) r++;
+	frpw_disconnect(pi);
+
+        if (verbose)  {
+            printk("%s: frpw: port 0x%x, chip %ld, mode %d, test=(%d,%d,%d)\n",
+                   pi->device,pi->port,(pi->private%2),pi->mode,e[0],e[1],r);
+        }
+
+        return (r || (e[0] && e[1]));
+}
+
+
+static void frpw_log_adapter( PIA *pi, char * scratch, int verbose )
+
+{       char    *mode_string[6] = {"4-bit","8-bit","EPP",
+				   "EPP-8","EPP-16","EPP-32"};
+
+        printk("%s: frpw %s, Freecom (%s) adapter at 0x%x, ", pi->device,
+		FRPW_VERSION,((pi->private%2) == 0)?"Xilinx":"ASIC",pi->port);
+        printk("mode %d (%s), delay %d\n",pi->mode,
+		mode_string[pi->mode],pi->delay);
+
+}
+
+static struct pi_protocol frpw = {
+	.owner		= THIS_MODULE,
+	.name		= "frpw",
+	.max_mode	= 6,
+	.epp_first	= 2,
+	.default_delay	= 2,
+	.max_units	= 1,
+	.write_regr	= frpw_write_regr,
+	.read_regr	= frpw_read_regr,
+	.write_block	= frpw_write_block,
+	.read_block	= frpw_read_block,
+	.connect	= frpw_connect,
+	.disconnect	= frpw_disconnect,
+	.test_proto	= frpw_test_proto,
+	.log_adapter	= frpw_log_adapter,
+};
+
+static int __init frpw_init(void)
+{
+	return paride_register(&frpw);
+}
+
+static void __exit frpw_exit(void)
+{
+	paride_unregister(&frpw);
+}
+
+MODULE_LICENSE("GPL");
+module_init(frpw_init)
+module_exit(frpw_exit)

drivers/block/paride/kbic.c

@@ -0,0 +1,305 @@
+/*
+        kbic.c    (c) 1997-8  Grant R. Guenther <grant@torque.net>
+                              Under the terms of the GNU General Public License.
+
+        This is a low-level driver for the KBIC-951A and KBIC-971A
+        parallel to IDE adapter chips from KingByte Information Systems.
+
+	The chips are almost identical, however, the wakeup code 
+	required for the 971A interferes with the correct operation of
+        the 951A, so this driver registers itself twice, once for
+	each chip.
+
+*/
+
+/* Changes:
+
+        1.01    GRG 1998.05.06 init_proto, release_proto
+
+*/
+
+#define KBIC_VERSION      "1.01"
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/wait.h>
+#include <asm/io.h>
+
+#include "paride.h"
+
+#define r12w()			(delay_p,inw(pi->port+1)&0xffff) 
+
+#define j44(a,b)                ((((a>>4)&0x0f)|(b&0xf0))^0x88)
+#define j53(w)                  (((w>>3)&0x1f)|((w>>4)&0xe0))
+
+
+/* cont = 0 - access the IDE register file 
+   cont = 1 - access the IDE command set 
+*/
+
+static int  cont_map[2] = { 0x80, 0x40 };
+
+static int kbic_read_regr( PIA *pi, int cont, int regr )
+
+{       int     a, b, s;
+
+        s = cont_map[cont];
+
+	switch (pi->mode) {
+
+	case 0: w0(regr|0x18|s); w2(4); w2(6); w2(4); w2(1); w0(8);
+	        a = r1(); w0(0x28); b = r1(); w2(4);
+		return j44(a,b);
+
+	case 1: w0(regr|0x38|s); w2(4); w2(6); w2(4); w2(5); w0(8);
+		a = r12w(); w2(4);
+		return j53(a);
+
+	case 2: w0(regr|0x08|s); w2(4); w2(6); w2(4); w2(0xa5); w2(0xa1);
+		a = r0(); w2(4);
+       		return a;
+
+	case 3:
+	case 4:
+	case 5: w0(0x20|s); w2(4); w2(6); w2(4); w3(regr);
+		a = r4(); b = r4(); w2(4); w2(0); w2(4);
+		return a;
+
+	}
+	return -1;
+}       
+
+static void  kbic_write_regr( PIA *pi, int cont, int regr, int val)
+
+{       int  s;
+
+        s = cont_map[cont];
+
+        switch (pi->mode) {
+
+	case 0: 
+        case 1:
+	case 2:	w0(regr|0x10|s); w2(4); w2(6); w2(4); 
+		w0(val); w2(5); w2(4);
+		break;
+
+	case 3:
+	case 4:
+	case 5: w0(0x20|s); w2(4); w2(6); w2(4); w3(regr);
+		w4(val); w4(val);
+		w2(4); w2(0); w2(4);
+                break;
+
+	}
+}
+
+static void k951_connect ( PIA *pi  )
+
+{ 	pi->saved_r0 = r0();
+        pi->saved_r2 = r2();
+        w2(4); 
+}
+
+static void k951_disconnect ( PIA *pi )
+
+{      	w0(pi->saved_r0);
+        w2(pi->saved_r2);
+}
+
+#define	CCP(x)	w2(0xc4);w0(0xaa);w0(0x55);w0(0);w0(0xff);w0(0x87);\
+		w0(0x78);w0(x);w2(0xc5);w2(0xc4);w0(0xff);
+
+static void k971_connect ( PIA *pi  )
+
+{ 	pi->saved_r0 = r0();
+        pi->saved_r2 = r2();
+	CCP(0x20);
+        w2(4); 
+}
+
+static void k971_disconnect ( PIA *pi )
+
+{       CCP(0x30);
+	w0(pi->saved_r0);
+        w2(pi->saved_r2);
+}
+
+/* counts must be congruent to 0 MOD 4, but all known applications
+   have this property.
+*/
+
+static void kbic_read_block( PIA *pi, char * buf, int count )
+
+{       int     k, a, b;
+
+        switch (pi->mode) {
+
+        case 0: w0(0x98); w2(4); w2(6); w2(4);
+                for (k=0;k<count/2;k++) {
+			w2(1); w0(8);    a = r1();
+			       w0(0x28); b = r1();
+			buf[2*k]   = j44(a,b);
+			w2(5);           b = r1();
+			       w0(8);    a = r1();
+			buf[2*k+1] = j44(a,b);
+			w2(4);
+                } 
+                break;
+
+        case 1: w0(0xb8); w2(4); w2(6); w2(4); 
+                for (k=0;k<count/4;k++) {
+                        w0(0xb8); 
+			w2(4); w2(5); 
+                        w0(8);    buf[4*k]   = j53(r12w());
+			w0(0xb8); buf[4*k+1] = j53(r12w());
+			w2(4); w2(5);
+			          buf[4*k+3] = j53(r12w());
+			w0(8);    buf[4*k+2] = j53(r12w());
+                }
+                w2(4);
+                break;
+
+        case 2: w0(0x88); w2(4); w2(6); w2(4);
+                for (k=0;k<count/2;k++) {
+                        w2(0xa0); w2(0xa1); buf[2*k] = r0();
+                        w2(0xa5); buf[2*k+1] = r0();
+                }
+                w2(4);
+                break;
+
+        case 3: w0(0xa0); w2(4); w2(6); w2(4); w3(0);
+                for (k=0;k<count;k++) buf[k] = r4();
+                w2(4); w2(0); w2(4);
+                break;
+
+	case 4: w0(0xa0); w2(4); w2(6); w2(4); w3(0);
+                for (k=0;k<count/2;k++) ((u16 *)buf)[k] = r4w();
+                w2(4); w2(0); w2(4);
+                break;
+
+        case 5: w0(0xa0); w2(4); w2(6); w2(4); w3(0);
+                for (k=0;k<count/4;k++) ((u32 *)buf)[k] = r4l();
+                w2(4); w2(0); w2(4);
+                break;
+
+
+        }
+}
+
+static void kbic_write_block( PIA *pi, char * buf, int count )
+
+{       int     k;
+
+        switch (pi->mode) {
+
+        case 0:
+        case 1:
+        case 2: w0(0x90); w2(4); w2(6); w2(4); 
+		for(k=0;k<count/2;k++) {
+			w0(buf[2*k+1]); w2(0); w2(4); 
+			w0(buf[2*k]);   w2(5); w2(4); 
+		}
+		break;
+
+        case 3: w0(0xa0); w2(4); w2(6); w2(4); w3(0);
+		for(k=0;k<count/2;k++) {
+			w4(buf[2*k+1]); 
+                        w4(buf[2*k]);
+                }
+		w2(4); w2(0); w2(4);
+		break;
+
+	case 4: w0(0xa0); w2(4); w2(6); w2(4); w3(0);
+                for(k=0;k<count/2;k++) w4w(pi_swab16(buf,k));
+                w2(4); w2(0); w2(4);
+                break;
+
+        case 5: w0(0xa0); w2(4); w2(6); w2(4); w3(0);
+                for(k=0;k<count/4;k++) w4l(pi_swab32(buf,k));
+                w2(4); w2(0); w2(4);
+                break;
+
+        }
+
+}
+
+static void kbic_log_adapter( PIA *pi, char * scratch, 
+			      int verbose, char * chip )
+
+{       char    *mode_string[6] = {"4-bit","5/3","8-bit",
+				   "EPP-8","EPP_16","EPP-32"};
+
+        printk("%s: kbic %s, KingByte %s at 0x%x, ",
+                pi->device,KBIC_VERSION,chip,pi->port);
+        printk("mode %d (%s), delay %d\n",pi->mode,
+		mode_string[pi->mode],pi->delay);
+
+}
+
+static void k951_log_adapter( PIA *pi, char * scratch, int verbose )
+
+{	kbic_log_adapter(pi,scratch,verbose,"KBIC-951A");
+}
+
+static void k971_log_adapter( PIA *pi, char * scratch, int verbose )
+
+{       kbic_log_adapter(pi,scratch,verbose,"KBIC-971A");
+}
+
+static struct pi_protocol k951 = {
+	.owner		= THIS_MODULE,
+	.name		= "k951",
+	.max_mode	= 6,
+	.epp_first	= 3,
+	.default_delay	= 1,
+	.max_units	= 1,
+	.write_regr	= kbic_write_regr,
+	.read_regr	= kbic_read_regr,
+	.write_block	= kbic_write_block,
+	.read_block	= kbic_read_block,
+	.connect	= k951_connect,
+	.disconnect	= k951_disconnect,
+	.log_adapter	= k951_log_adapter,
+};
+
+static struct pi_protocol k971 = {
+	.owner		= THIS_MODULE,
+	.name		= "k971",
+	.max_mode	= 6,
+	.epp_first	= 3,
+	.default_delay	= 1,
+	.max_units	= 1,
+	.write_regr	= kbic_write_regr,
+	.read_regr	= kbic_read_regr,
+	.write_block	= kbic_write_block,
+	.read_block	= kbic_read_block,
+	.connect	= k971_connect,
+	.disconnect	= k971_disconnect,
+	.log_adapter	= k971_log_adapter,
+};
+
+static int __init kbic_init(void)
+{
+	int rv;
+
+	rv = paride_register(&k951);
+	if (rv < 0)
+		return rv;
+	rv = paride_register(&k971);
+	if (rv < 0)
+		paride_unregister(&k951);
+	return rv;
+}
+
+static void __exit kbic_exit(void)
+{
+	paride_unregister(&k951);
+	paride_unregister(&k971);
+}
+
+MODULE_LICENSE("GPL");
+module_init(kbic_init)
+module_exit(kbic_exit)

drivers/block/paride/ktti.c

@@ -0,0 +1,128 @@
+/* 
+        ktti.c        (c) 1998  Grant R. Guenther <grant@torque.net>
+                          Under the terms of the GNU General Public License.
+
+	ktti.c is a low-level protocol driver for the KT Technology
+	parallel port adapter.  This adapter is used in the "PHd" 
+        portable hard-drives.  As far as I can tell, this device
+	supports 4-bit mode _only_.  
+
+*/
+
+#define KTTI_VERSION      "1.0"
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/wait.h>
+#include <asm/io.h>
+
+#include "paride.h"
+
+#define j44(a,b)                (((a>>4)&0x0f)|(b&0xf0))
+
+/* cont = 0 - access the IDE register file 
+   cont = 1 - access the IDE command set 
+*/
+
+static int  cont_map[2] = { 0x10, 0x08 };
+
+static void  ktti_write_regr( PIA *pi, int cont, int regr, int val)
+
+{	int r;
+
+	r = regr + cont_map[cont];
+
+	w0(r); w2(0xb); w2(0xa); w2(3); w2(6); 
+	w0(val); w2(3); w0(0); w2(6); w2(0xb);
+}
+
+static int ktti_read_regr( PIA *pi, int cont, int regr )
+
+{	int  a, b, r;
+
+        r = regr + cont_map[cont];
+
+        w0(r); w2(0xb); w2(0xa); w2(9); w2(0xc); w2(9); 
+	a = r1(); w2(0xc);  b = r1(); w2(9); w2(0xc); w2(9);
+	return j44(a,b);
+
+}
+
+static void ktti_read_block( PIA *pi, char * buf, int count )
+
+{	int  k, a, b;
+
+	for (k=0;k<count/2;k++) {
+		w0(0x10); w2(0xb); w2(0xa); w2(9); w2(0xc); w2(9);
+		a = r1(); w2(0xc); b = r1(); w2(9);
+		buf[2*k] = j44(a,b);
+		a = r1(); w2(0xc); b = r1(); w2(9);
+		buf[2*k+1] = j44(a,b);
+	}
+}
+
+static void ktti_write_block( PIA *pi, char * buf, int count )
+
+{	int k;
+
+	for (k=0;k<count/2;k++) {
+		w0(0x10); w2(0xb); w2(0xa); w2(3); w2(6);
+		w0(buf[2*k]); w2(3);
+		w0(buf[2*k+1]); w2(6);
+		w2(0xb);
+	}
+}
+
+static void ktti_connect ( PIA *pi  )
+
+{       pi->saved_r0 = r0();
+        pi->saved_r2 = r2();
+	w2(0xb); w2(0xa); w0(0); w2(3); w2(6);	
+}
+
+static void ktti_disconnect ( PIA *pi )
+
+{       w2(0xb); w2(0xa); w0(0xa0); w2(3); w2(4);
+	w0(pi->saved_r0);
+        w2(pi->saved_r2);
+} 
+
+static void ktti_log_adapter( PIA *pi, char * scratch, int verbose )
+
+{       printk("%s: ktti %s, KT adapter at 0x%x, delay %d\n",
+                pi->device,KTTI_VERSION,pi->port,pi->delay);
+
+}
+
+static struct pi_protocol ktti = {
+	.owner		= THIS_MODULE,
+	.name		= "ktti",
+	.max_mode	= 1,
+	.epp_first	= 2,
+	.default_delay	= 1,
+	.max_units	= 1,
+	.write_regr	= ktti_write_regr,
+	.read_regr	= ktti_read_regr,
+	.write_block	= ktti_write_block,
+	.read_block	= ktti_read_block,
+	.connect	= ktti_connect,
+	.disconnect	= ktti_disconnect,
+	.log_adapter	= ktti_log_adapter,
+};
+
+static int __init ktti_init(void)
+{
+	return paride_register(&ktti);
+}
+
+static void __exit ktti_exit(void)
+{
+	paride_unregister(&ktti);
+}
+
+MODULE_LICENSE("GPL");
+module_init(ktti_init)
+module_exit(ktti_exit)

drivers/block/paride/mkd

@@ -0,0 +1,30 @@
+#!/bin/bash
+#
+# mkd -- a script to create the device special files for the PARIDE subsystem
+#
+#  block devices:  	pd (45), pcd (46), pf (47)
+#  character devices:	pt (96), pg (97)
+#
+function mkdev {
+  mknod $1 $2 $3 $4 ; chmod 0660 $1 ; chown root:disk $1
+}
+#
+function pd {
+  D=$( printf \\$( printf "x%03x" $[ $1 + 97 ] ) )
+  mkdev pd$D b 45 $[ $1 * 16 ]
+  for P in 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
+  do mkdev pd$D$P b 45 $[ $1 * 16 + $P ]
+  done
+}
+#
+cd /dev
+#
+for u in 0 1 2 3 ; do pd $u ; done
+for u in 0 1 2 3 ; do mkdev pcd$u b 46 $u ; done 
+for u in 0 1 2 3 ; do mkdev pf$u  b 47 $u ; done 
+for u in 0 1 2 3 ; do mkdev pt$u  c 96 $u ; done 
+for u in 0 1 2 3 ; do mkdev npt$u c 96 $[ $u + 128 ] ; done 
+for u in 0 1 2 3 ; do mkdev pg$u  c 97 $u ; done
+#
+# end of mkd
+

drivers/block/paride/on20.c

@@ -0,0 +1,153 @@
+/* 
+	on20.c	(c) 1996-8  Grant R. Guenther <grant@torque.net>
+		            Under the terms of the GNU General Public License.
+
+        on20.c is a low-level protocol driver for the
+        Onspec 90c20 parallel to IDE adapter. 
+*/
+
+/* Changes:
+
+        1.01    GRG 1998.05.06 init_proto, release_proto
+
+*/
+
+#define	ON20_VERSION	"1.01"
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/wait.h>
+#include <asm/io.h>
+
+#include "paride.h"
+
+#define op(f)	w2(4);w0(f);w2(5);w2(0xd);w2(5);w2(0xd);w2(5);w2(4);
+#define vl(v)	w2(4);w0(v);w2(5);w2(7);w2(5);w2(4);
+
+#define j44(a,b)  (((a>>4)&0x0f)|(b&0xf0))
+
+/* cont = 0 - access the IDE register file 
+   cont = 1 - access the IDE command set 
+*/
+
+static int on20_read_regr( PIA *pi, int cont, int regr )
+
+{	int h,l, r ;
+
+        r = (regr<<2) + 1 + cont;
+
+        op(1); vl(r); op(0);
+
+	switch (pi->mode)  {
+
+        case 0:  w2(4); w2(6); l = r1();
+                 w2(4); w2(6); h = r1();
+                 w2(4); w2(6); w2(4); w2(6); w2(4);
+		 return j44(l,h);
+
+	case 1:  w2(4); w2(0x26); r = r0(); 
+                 w2(4); w2(0x26); w2(4);
+		 return r;
+
+	}
+	return -1;
+}	
+
+static void on20_write_regr( PIA *pi, int cont, int regr, int val )
+
+{	int r;
+
+	r = (regr<<2) + 1 + cont;
+
+	op(1); vl(r); 
+	op(0); vl(val); 
+	op(0); vl(val);
+}
+
+static void on20_connect ( PIA *pi)
+
+{	pi->saved_r0 = r0();
+        pi->saved_r2 = r2();
+
+	w2(4);w0(0);w2(0xc);w2(4);w2(6);w2(4);w2(6);w2(4); 
+	if (pi->mode) { op(2); vl(8); op(2); vl(9); }
+	       else   { op(2); vl(0); op(2); vl(8); }
+}
+
+static void on20_disconnect ( PIA *pi )
+
+{	w2(4);w0(7);w2(4);w2(0xc);w2(4);
+        w0(pi->saved_r0);
+        w2(pi->saved_r2);
+} 
+
+static void on20_read_block( PIA *pi, char * buf, int count )
+
+{	int     k, l, h; 
+
+	op(1); vl(1); op(0);
+
+	for (k=0;k<count;k++) 
+	    if (pi->mode) {
+		w2(4); w2(0x26); buf[k] = r0();
+	    } else {
+		w2(6); l = r1(); w2(4);
+		w2(6); h = r1(); w2(4);
+		buf[k] = j44(l,h);
+	    }
+	w2(4);
+}
+
+static void on20_write_block(  PIA *pi, char * buf, int count )
+
+{	int	k;
+
+	op(1); vl(1); op(0);
+
+	for (k=0;k<count;k++) { w2(5); w0(buf[k]); w2(7); }
+	w2(4);
+}
+
+static void on20_log_adapter( PIA *pi, char * scratch, int verbose )
+
+{       char    *mode_string[2] = {"4-bit","8-bit"};
+
+        printk("%s: on20 %s, OnSpec 90c20 at 0x%x, ",
+                pi->device,ON20_VERSION,pi->port);
+        printk("mode %d (%s), delay %d\n",pi->mode,
+		mode_string[pi->mode],pi->delay);
+
+}
+
+static struct pi_protocol on20 = {
+	.owner		= THIS_MODULE,
+	.name		= "on20",
+	.max_mode	= 2,
+	.epp_first	= 2,
+	.default_delay	= 1,
+	.max_units	= 1,
+	.write_regr	= on20_write_regr,
+	.read_regr	= on20_read_regr,
+	.write_block	= on20_write_block,
+	.read_block	= on20_read_block,
+	.connect	= on20_connect,
+	.disconnect	= on20_disconnect,
+	.log_adapter	= on20_log_adapter,
+};
+
+static int __init on20_init(void)
+{
+	return paride_register(&on20);
+}
+
+static void __exit on20_exit(void)
+{
+	paride_unregister(&on20);
+}
+
+MODULE_LICENSE("GPL");
+module_init(on20_init)
+module_exit(on20_exit)

drivers/block/paride/on26.c

@@ -0,0 +1,319 @@
+/* 
+        on26.c    (c) 1997-8  Grant R. Guenther <grant@torque.net>
+                              Under the terms of the GNU General Public License.
+
+        on26.c is a low-level protocol driver for the 
+        OnSpec 90c26 parallel to IDE adapter chip.
+
+*/
+
+/* Changes:
+
+        1.01    GRG 1998.05.06 init_proto, release_proto
+	1.02    GRG 1998.09.23 updates for the -E rev chip
+	1.03    GRG 1998.12.14 fix for slave drives
+	1.04    GRG 1998.12.20 yet another bug fix
+
+*/
+
+#define ON26_VERSION      "1.04"
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/types.h>
+#include <linux/wait.h>
+#include <asm/io.h>
+
+#include "paride.h"
+
+/* mode codes:  0  nybble reads, 8-bit writes
+                1  8-bit reads and writes
+                2  8-bit EPP mode
+		3  EPP-16
+		4  EPP-32
+*/
+
+#define j44(a,b)  (((a>>4)&0x0f)|(b&0xf0))
+
+#define P1	w2(5);w2(0xd);w2(5);w2(0xd);w2(5);w2(4);
+#define P2	w2(5);w2(7);w2(5);w2(4);
+
+/* cont = 0 - access the IDE register file 
+   cont = 1 - access the IDE command set 
+*/
+
+static int on26_read_regr( PIA *pi, int cont, int regr )
+
+{       int     a, b, r;
+
+	r = (regr<<2) + 1 + cont;
+
+        switch (pi->mode)  {
+
+        case 0: w0(1); P1; w0(r); P2; w0(0); P1; 
+		w2(6); a = r1(); w2(4);
+		w2(6); b = r1(); w2(4);
+		w2(6); w2(4); w2(6); w2(4);
+                return j44(a,b);
+
+        case 1: w0(1); P1; w0(r); P2; w0(0); P1;
+		w2(0x26); a = r0(); w2(4); w2(0x26); w2(4);
+                return a;
+
+	case 2:
+	case 3:
+        case 4: w3(1); w3(1); w2(5); w4(r); w2(4);
+		w3(0); w3(0); w2(0x24); a = r4(); w2(4);
+		w2(0x24); (void)r4(); w2(4);
+                return a;
+
+        }
+        return -1;
+}       
+
+static void on26_write_regr( PIA *pi, int cont, int regr, int val )
+
+{       int  r;
+
+        r = (regr<<2) + 1 + cont;
+
+        switch (pi->mode)  {
+
+        case 0:
+        case 1: w0(1); P1; w0(r); P2; w0(0); P1;
+		w0(val); P2; w0(val); P2;
+		break;
+
+	case 2:
+	case 3:
+        case 4: w3(1); w3(1); w2(5); w4(r); w2(4);
+		w3(0); w3(0); 
+		w2(5); w4(val); w2(4);
+		w2(5); w4(val); w2(4);
+                break;
+        }
+}
+
+#define  CCP(x)  w0(0xfe);w0(0xaa);w0(0x55);w0(0);w0(0xff);\
+		 w0(0x87);w0(0x78);w0(x);w2(4);w2(5);w2(4);w0(0xff);
+
+static void on26_connect ( PIA *pi )
+
+{       int	x;
+
+	pi->saved_r0 = r0();
+        pi->saved_r2 = r2();
+
+        CCP(0x20);
+	x = 8; if (pi->mode) x = 9;
+
+	w0(2); P1; w0(8); P2;
+	w0(2); P1; w0(x); P2;
+}
+
+static void on26_disconnect ( PIA *pi )
+
+{       if (pi->mode >= 2) { w3(4); w3(4); w3(4); w3(4); }
+	              else { w0(4); P1; w0(4); P1; }
+	CCP(0x30);
+        w0(pi->saved_r0);
+        w2(pi->saved_r2);
+} 
+
+#define	RESET_WAIT  200
+
+static int on26_test_port( PIA *pi)  /* hard reset */
+
+{       int     i, m, d, x=0, y=0;
+
+        pi->saved_r0 = r0();
+        pi->saved_r2 = r2();
+
+        d = pi->delay;
+        m = pi->mode;
+        pi->delay = 5;
+        pi->mode = 0;
+
+        w2(0xc);
+
+        CCP(0x30); CCP(0); 
+
+        w0(0xfe);w0(0xaa);w0(0x55);w0(0);w0(0xff);
+        i = ((r1() & 0xf0) << 4); w0(0x87);
+        i |= (r1() & 0xf0); w0(0x78);
+        w0(0x20);w2(4);w2(5);
+        i |= ((r1() & 0xf0) >> 4);
+        w2(4);w0(0xff);
+
+        if (i == 0xb5f) {
+
+            w0(2); P1; w0(0);   P2;
+            w0(3); P1; w0(0);   P2;
+            w0(2); P1; w0(8);   P2; udelay(100);
+            w0(2); P1; w0(0xa); P2; udelay(100);
+            w0(2); P1; w0(8);   P2; udelay(1000);
+            
+            on26_write_regr(pi,0,6,0xa0);
+
+            for (i=0;i<RESET_WAIT;i++) {
+                on26_write_regr(pi,0,6,0xa0);
+                x = on26_read_regr(pi,0,7);
+                on26_write_regr(pi,0,6,0xb0);
+                y = on26_read_regr(pi,0,7);
+                if (!((x&0x80)||(y&0x80))) break;
+                mdelay(100);
+            }
+
+	    if (i == RESET_WAIT) 
+		printk("on26: Device reset failed (%x,%x)\n",x,y);
+
+            w0(4); P1; w0(4); P1;
+        }
+
+        CCP(0x30);
+
+        pi->delay = d;
+        pi->mode = m;
+        w0(pi->saved_r0);
+        w2(pi->saved_r2);
+
+        return 5;
+}
+
+
+static void on26_read_block( PIA *pi, char * buf, int count )
+
+{       int     k, a, b;
+
+        switch (pi->mode) {
+
+        case 0: w0(1); P1; w0(1); P2; w0(2); P1; w0(0x18); P2; w0(0); P1;
+		udelay(10);
+		for (k=0;k<count;k++) {
+                        w2(6); a = r1();
+                        w2(4); b = r1();
+                        buf[k] = j44(a,b);
+                }
+		w0(2); P1; w0(8); P2; 
+                break;
+
+        case 1: w0(1); P1; w0(1); P2; w0(2); P1; w0(0x19); P2; w0(0); P1;
+		udelay(10);
+                for (k=0;k<count/2;k++) {
+                        w2(0x26); buf[2*k] = r0();  
+			w2(0x24); buf[2*k+1] = r0();
+                }
+                w0(2); P1; w0(9); P2;
+                break;
+
+        case 2: w3(1); w3(1); w2(5); w4(1); w2(4);
+		w3(0); w3(0); w2(0x24);
+		udelay(10);
+                for (k=0;k<count;k++) buf[k] = r4();
+                w2(4);
+                break;
+
+        case 3: w3(1); w3(1); w2(5); w4(1); w2(4);
+                w3(0); w3(0); w2(0x24);
+                udelay(10);
+                for (k=0;k<count/2;k++) ((u16 *)buf)[k] = r4w();
+                w2(4);
+                break;
+
+        case 4: w3(1); w3(1); w2(5); w4(1); w2(4);
+                w3(0); w3(0); w2(0x24);
+                udelay(10);
+                for (k=0;k<count/4;k++) ((u32 *)buf)[k] = r4l();
+                w2(4);
+                break;
+
+        }
+}
+
+static void on26_write_block( PIA *pi, char * buf, int count )
+
+{       int	k;
+
+        switch (pi->mode) {
+
+        case 0: 
+        case 1: w0(1); P1; w0(1); P2; 
+		w0(2); P1; w0(0x18+pi->mode); P2; w0(0); P1;
+		udelay(10);
+		for (k=0;k<count/2;k++) {
+                        w2(5); w0(buf[2*k]); 
+			w2(7); w0(buf[2*k+1]);
+                }
+                w2(5); w2(4);
+		w0(2); P1; w0(8+pi->mode); P2;
+                break;
+
+        case 2: w3(1); w3(1); w2(5); w4(1); w2(4);
+		w3(0); w3(0); w2(0xc5);
+		udelay(10);
+                for (k=0;k<count;k++) w4(buf[k]);
+		w2(0xc4);
+                break;
+
+        case 3: w3(1); w3(1); w2(5); w4(1); w2(4);
+                w3(0); w3(0); w2(0xc5);
+                udelay(10);
+                for (k=0;k<count/2;k++) w4w(((u16 *)buf)[k]);
+                w2(0xc4);
+                break;
+
+        case 4: w3(1); w3(1); w2(5); w4(1); w2(4);
+                w3(0); w3(0); w2(0xc5);
+                udelay(10);
+                for (k=0;k<count/4;k++) w4l(((u32 *)buf)[k]);
+                w2(0xc4);
+                break;
+
+        }
+
+}
+
+static void on26_log_adapter( PIA *pi, char * scratch, int verbose )
+
+{       char    *mode_string[5] = {"4-bit","8-bit","EPP-8",
+				   "EPP-16","EPP-32"};
+
+        printk("%s: on26 %s, OnSpec 90c26 at 0x%x, ",
+                pi->device,ON26_VERSION,pi->port);
+        printk("mode %d (%s), delay %d\n",pi->mode,
+		mode_string[pi->mode],pi->delay);
+
+}
+
+static struct pi_protocol on26 = {
+	.owner		= THIS_MODULE,
+	.name		= "on26",
+	.max_mode	= 5,
+	.epp_first	= 2,
+	.default_delay	= 1,
+	.max_units	= 1,
+	.write_regr	= on26_write_regr,
+	.read_regr	= on26_read_regr,
+	.write_block	= on26_write_block,
+	.read_block	= on26_read_block,
+	.connect	= on26_connect,
+	.disconnect	= on26_disconnect,
+	.test_port	= on26_test_port,
+	.log_adapter	= on26_log_adapter,
+};
+
+static int __init on26_init(void)
+{
+	return paride_register(&on26);
+}
+
+static void __exit on26_exit(void)
+{
+	paride_unregister(&on26);
+}
+
+MODULE_LICENSE("GPL");
+module_init(on26_init)
+module_exit(on26_exit)

drivers/block/paride/paride.c

@@ -0,0 +1,434 @@
+/* 
+        paride.c  (c) 1997-8  Grant R. Guenther <grant@torque.net>
+                              Under the terms of the GNU General Public License.
+
+	This is the base module for the family of device drivers
+        that support parallel port IDE devices.  
+
+*/
+
+/* Changes:
+
+	1.01	GRG 1998.05.03	Use spinlocks
+	1.02	GRG 1998.05.05  init_proto, release_proto, ktti
+	1.03	GRG 1998.08.15  eliminate compiler warning
+	1.04    GRG 1998.11.28  added support for FRIQ 
+	1.05    TMW 2000.06.06  use parport_find_number instead of
+				parport_enumerate
+	1.06    TMW 2001.03.26  more sane parport-or-not resource management
+*/
+
+#define PI_VERSION      "1.06"
+
+#include <linux/module.h>
+#include <linux/kmod.h>
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/ioport.h>
+#include <linux/string.h>
+#include <linux/spinlock.h>
+#include <linux/wait.h>
+#include <linux/sched.h>	/* TASK_* */
+#include <linux/parport.h>
+
+#include "paride.h"
+
+MODULE_LICENSE("GPL");
+
+#define MAX_PROTOS	32
+
+static struct pi_protocol *protocols[MAX_PROTOS];
+
+static DEFINE_SPINLOCK(pi_spinlock);
+
+void pi_write_regr(PIA * pi, int cont, int regr, int val)
+{
+	pi->proto->write_regr(pi, cont, regr, val);
+}
+
+EXPORT_SYMBOL(pi_write_regr);
+
+int pi_read_regr(PIA * pi, int cont, int regr)
+{
+	return pi->proto->read_regr(pi, cont, regr);
+}
+
+EXPORT_SYMBOL(pi_read_regr);
+
+void pi_write_block(PIA * pi, char *buf, int count)
+{
+	pi->proto->write_block(pi, buf, count);
+}
+
+EXPORT_SYMBOL(pi_write_block);
+
+void pi_read_block(PIA * pi, char *buf, int count)
+{
+	pi->proto->read_block(pi, buf, count);
+}
+
+EXPORT_SYMBOL(pi_read_block);
+
+static void pi_wake_up(void *p)
+{
+	PIA *pi = (PIA *) p;
+	unsigned long flags;
+	void (*cont) (void) = NULL;
+
+	spin_lock_irqsave(&pi_spinlock, flags);
+
+	if (pi->claim_cont && !parport_claim(pi->pardev)) {
+		cont = pi->claim_cont;
+		pi->claim_cont = NULL;
+		pi->claimed = 1;
+	}
+
+	spin_unlock_irqrestore(&pi_spinlock, flags);
+
+	wake_up(&(pi->parq));
+
+	if (cont)
+		cont();
+}
+
+int pi_schedule_claimed(PIA * pi, void (*cont) (void))
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&pi_spinlock, flags);
+	if (pi->pardev && parport_claim(pi->pardev)) {
+		pi->claim_cont = cont;
+		spin_unlock_irqrestore(&pi_spinlock, flags);
+		return 0;
+	}
+	pi->claimed = 1;
+	spin_unlock_irqrestore(&pi_spinlock, flags);
+	return 1;
+}
+EXPORT_SYMBOL(pi_schedule_claimed);
+
+void pi_do_claimed(PIA * pi, void (*cont) (void))
+{
+	if (pi_schedule_claimed(pi, cont))
+		cont();
+}
+
+EXPORT_SYMBOL(pi_do_claimed);
+
+static void pi_claim(PIA * pi)
+{
+	if (pi->claimed)
+		return;
+	pi->claimed = 1;
+	if (pi->pardev)
+		wait_event(pi->parq,
+			   !parport_claim((struct pardevice *) pi->pardev));
+}
+
+static void pi_unclaim(PIA * pi)
+{
+	pi->claimed = 0;
+	if (pi->pardev)
+		parport_release((struct pardevice *) (pi->pardev));
+}
+
+void pi_connect(PIA * pi)
+{
+	pi_claim(pi);
+	pi->proto->connect(pi);
+}
+
+EXPORT_SYMBOL(pi_connect);
+
+void pi_disconnect(PIA * pi)
+{
+	pi->proto->disconnect(pi);
+	pi_unclaim(pi);
+}
+
+EXPORT_SYMBOL(pi_disconnect);
+
+static void pi_unregister_parport(PIA * pi)
+{
+	if (pi->pardev) {
+		parport_unregister_device((struct pardevice *) (pi->pardev));
+		pi->pardev = NULL;
+	}
+}
+
+void pi_release(PIA * pi)
+{
+	pi_unregister_parport(pi);
+	if (pi->proto->release_proto)
+		pi->proto->release_proto(pi);
+	module_put(pi->proto->owner);
+}
+
+EXPORT_SYMBOL(pi_release);
+
+static int default_test_proto(PIA * pi, char *scratch, int verbose)
+{
+	int j, k;
+	int e[2] = { 0, 0 };
+
+	pi->proto->connect(pi);
+
+	for (j = 0; j < 2; j++) {
+		pi_write_regr(pi, 0, 6, 0xa0 + j * 0x10);
+		for (k = 0; k < 256; k++) {
+			pi_write_regr(pi, 0, 2, k ^ 0xaa);
+			pi_write_regr(pi, 0, 3, k ^ 0x55);
+			if (pi_read_regr(pi, 0, 2) != (k ^ 0xaa))
+				e[j]++;
+		}
+	}
+	pi->proto->disconnect(pi);
+
+	if (verbose)
+		printk("%s: %s: port 0x%x, mode  %d, test=(%d,%d)\n",
+		       pi->device, pi->proto->name, pi->port,
+		       pi->mode, e[0], e[1]);
+
+	return (e[0] && e[1]);	/* not here if both > 0 */
+}
+
+static int pi_test_proto(PIA * pi, char *scratch, int verbose)
+{
+	int res;
+
+	pi_claim(pi);
+	if (pi->proto->test_proto)
+		res = pi->proto->test_proto(pi, scratch, verbose);
+	else
+		res = default_test_proto(pi, scratch, verbose);
+	pi_unclaim(pi);
+
+	return res;
+}
+
+int paride_register(PIP * pr)
+{
+	int k;
+
+	for (k = 0; k < MAX_PROTOS; k++)
+		if (protocols[k] && !strcmp(pr->name, protocols[k]->name)) {
+			printk("paride: %s protocol already registered\n",
+			       pr->name);
+			return -1;
+		}
+	k = 0;
+	while ((k < MAX_PROTOS) && (protocols[k]))
+		k++;
+	if (k == MAX_PROTOS) {
+		printk("paride: protocol table full\n");
+		return -1;
+	}
+	protocols[k] = pr;
+	pr->index = k;
+	printk("paride: %s registered as protocol %d\n", pr->name, k);
+	return 0;
+}
+
+EXPORT_SYMBOL(paride_register);
+
+void paride_unregister(PIP * pr)
+{
+	if (!pr)
+		return;
+	if (protocols[pr->index] != pr) {
+		printk("paride: %s not registered\n", pr->name);
+		return;
+	}
+	protocols[pr->index] = NULL;
+}
+
+EXPORT_SYMBOL(paride_unregister);
+
+static int pi_register_parport(PIA * pi, int verbose)
+{
+	struct parport *port;
+
+	port = parport_find_base(pi->port);
+	if (!port)
+		return 0;
+
+	pi->pardev = parport_register_device(port,
+					     pi->device, NULL,
+					     pi_wake_up, NULL, 0, (void *) pi);
+	parport_put_port(port);
+	if (!pi->pardev)
+		return 0;
+
+	init_waitqueue_head(&pi->parq);
+
+	if (verbose)
+		printk("%s: 0x%x is %s\n", pi->device, pi->port, port->name);
+
+	pi->parname = (char *) port->name;
+
+	return 1;
+}
+
+static int pi_probe_mode(PIA * pi, int max, char *scratch, int verbose)
+{
+	int best, range;
+
+	if (pi->mode != -1) {
+		if (pi->mode >= max)
+			return 0;
+		range = 3;
+		if (pi->mode >= pi->proto->epp_first)
+			range = 8;
+		if ((range == 8) && (pi->port % 8))
+			return 0;
+		pi->reserved = range;
+		return (!pi_test_proto(pi, scratch, verbose));
+	}
+	best = -1;
+	for (pi->mode = 0; pi->mode < max; pi->mode++) {
+		range = 3;
+		if (pi->mode >= pi->proto->epp_first)
+			range = 8;
+		if ((range == 8) && (pi->port % 8))
+			break;
+		pi->reserved = range;
+		if (!pi_test_proto(pi, scratch, verbose))
+			best = pi->mode;
+	}
+	pi->mode = best;
+	return (best > -1);
+}
+
+static int pi_probe_unit(PIA * pi, int unit, char *scratch, int verbose)
+{
+	int max, s, e;
+
+	s = unit;
+	e = s + 1;
+
+	if (s == -1) {
+		s = 0;
+		e = pi->proto->max_units;
+	}
+
+	if (!pi_register_parport(pi, verbose))
+		return 0;
+
+	if (pi->proto->test_port) {
+		pi_claim(pi);
+		max = pi->proto->test_port(pi);
+		pi_unclaim(pi);
+	} else
+		max = pi->proto->max_mode;
+
+	if (pi->proto->probe_unit) {
+		pi_claim(pi);
+		for (pi->unit = s; pi->unit < e; pi->unit++)
+			if (pi->proto->probe_unit(pi)) {
+				pi_unclaim(pi);
+				if (pi_probe_mode(pi, max, scratch, verbose))
+					return 1;
+				pi_unregister_parport(pi);
+				return 0;
+			}
+		pi_unclaim(pi);
+		pi_unregister_parport(pi);
+		return 0;
+	}
+
+	if (!pi_probe_mode(pi, max, scratch, verbose)) {
+		pi_unregister_parport(pi);
+		return 0;
+	}
+	return 1;
+
+}
+
+int pi_init(PIA * pi, int autoprobe, int port, int mode,
+	int unit, int protocol, int delay, char *scratch,
+	int devtype, int verbose, char *device)
+{
+	int p, k, s, e;
+	int lpts[7] = { 0x3bc, 0x378, 0x278, 0x268, 0x27c, 0x26c, 0 };
+
+	s = protocol;
+	e = s + 1;
+
+	if (!protocols[0])
+		request_module("paride_protocol");
+
+	if (autoprobe) {
+		s = 0;
+		e = MAX_PROTOS;
+	} else if ((s < 0) || (s >= MAX_PROTOS) || (port <= 0) ||
+		   (!protocols[s]) || (unit < 0) ||
+		   (unit >= protocols[s]->max_units)) {
+		printk("%s: Invalid parameters\n", device);
+		return 0;
+	}
+
+	for (p = s; p < e; p++) {
+		struct pi_protocol *proto = protocols[p];
+		if (!proto)
+			continue;
+		/* still racy */
+		if (!try_module_get(proto->owner))
+			continue;
+		pi->proto = proto;
+		pi->private = 0;
+		if (proto->init_proto && proto->init_proto(pi) < 0) {
+			pi->proto = NULL;
+			module_put(proto->owner);
+			continue;
+		}
+		if (delay == -1)
+			pi->delay = pi->proto->default_delay;
+		else
+			pi->delay = delay;
+		pi->devtype = devtype;
+		pi->device = device;
+
+		pi->parname = NULL;
+		pi->pardev = NULL;
+		init_waitqueue_head(&pi->parq);
+		pi->claimed = 0;
+		pi->claim_cont = NULL;
+
+		pi->mode = mode;
+		if (port != -1) {
+			pi->port = port;
+			if (pi_probe_unit(pi, unit, scratch, verbose))
+				break;
+			pi->port = 0;
+		} else {
+			k = 0;
+			while ((pi->port = lpts[k++]))
+				if (pi_probe_unit
+				    (pi, unit, scratch, verbose))
+					break;
+			if (pi->port)
+				break;
+		}
+		if (pi->proto->release_proto)
+			pi->proto->release_proto(pi);
+		module_put(proto->owner);
+	}
+
+	if (!pi->port) {
+		if (autoprobe)
+			printk("%s: Autoprobe failed\n", device);
+		else
+			printk("%s: Adapter not found\n", device);
+		return 0;
+	}
+
+	if (pi->parname)
+		printk("%s: Sharing %s at 0x%x\n", pi->device,
+		       pi->parname, pi->port);
+
+	pi->proto->log_adapter(pi, scratch, verbose);
+
+	return 1;
+}
+
+EXPORT_SYMBOL(pi_init);

drivers/block/paride/paride.h

@@ -0,0 +1,170 @@
+#ifndef __DRIVERS_PARIDE_H__
+#define __DRIVERS_PARIDE_H__
+
+/* 
+	paride.h	(c) 1997-8  Grant R. Guenther <grant@torque.net>
+   		                    Under the terms of the GPL.
+
+   This file defines the interface between the high-level parallel
+   IDE device drivers (pd, pf, pcd, pt) and the adapter chips.
+
+*/
+
+/* Changes:
+
+	1.01	GRG 1998.05.05	init_proto, release_proto
+*/
+
+#define PARIDE_H_VERSION 	"1.01"
+
+/* Some adapters need to know what kind of device they are in
+
+   Values for devtype:
+*/
+
+#define	PI_PD	0	/* IDE disk */
+#define PI_PCD	1	/* ATAPI CDrom */
+#define PI_PF   2	/* ATAPI disk */
+#define PI_PT	3	/* ATAPI tape */
+#define PI_PG   4       /* ATAPI generic */
+
+/* The paride module contains no state, instead the drivers allocate
+   a pi_adapter data structure and pass it to paride in every operation.
+
+*/
+
+struct pi_adapter  {
+
+	struct pi_protocol *proto;   /* adapter protocol */
+	int	port;		     /* base address of parallel port */
+	int	mode;		     /* transfer mode in use */
+	int     delay;		     /* adapter delay setting */
+	int	devtype;	     /* device type: PI_PD etc. */
+	char    *device;	     /* name of driver */
+	int     unit;		     /* unit number for chained adapters */
+	int	saved_r0;	     /* saved port state */
+	int	saved_r2;	     /* saved port state */
+	int	reserved;	     /* number of ports reserved */
+	unsigned long	private;     /* for protocol module */
+
+	wait_queue_head_t parq;     /* semaphore for parport sharing */
+	void	*pardev;	     /* pointer to pardevice */
+	char	*parname;	     /* parport name */
+	int	claimed;	     /* parport has already been claimed */
+	void (*claim_cont)(void);    /* continuation for parport wait */
+};
+
+typedef struct pi_adapter PIA;
+
+/* functions exported by paride to the high level drivers */
+
+extern int pi_init(PIA *pi, 
+	int autoprobe,		/* 1 to autoprobe */
+	int port, 		/* base port address */
+	int mode, 		/* -1 for autoprobe */
+	int unit,		/* unit number, if supported */
+	int protocol, 		/* protocol to use */
+	int delay, 		/* -1 to use adapter specific default */
+	char * scratch, 	/* address of 512 byte buffer */
+	int devtype,		/* device type: PI_PD, PI_PCD, etc ... */
+	int verbose,		/* log verbose data while probing */
+	char *device		/* name of the driver */
+	);			/* returns 0 on failure, 1 on success */
+
+extern void pi_release(PIA *pi);
+
+/* registers are addressed as (cont,regr)
+
+       	cont: 0 for command register file, 1 for control register(s)
+	regr: 0-7 for register number.
+
+*/
+
+extern void pi_write_regr(PIA *pi, int cont, int regr, int val);
+
+extern int pi_read_regr(PIA *pi, int cont, int regr);
+
+extern void pi_write_block(PIA *pi, char * buf, int count);
+
+extern void pi_read_block(PIA *pi, char * buf, int count);
+
+extern void pi_connect(PIA *pi);
+
+extern void pi_disconnect(PIA *pi);
+
+extern void pi_do_claimed(PIA *pi, void (*cont)(void));
+extern int pi_schedule_claimed(PIA *pi, void (*cont)(void));
+
+/* macros and functions exported to the protocol modules */
+
+#define delay_p			(pi->delay?udelay(pi->delay):(void)0)
+#define out_p(offs,byte)	outb(byte,pi->port+offs); delay_p;
+#define in_p(offs)		(delay_p,inb(pi->port+offs))
+
+#define w0(byte)                {out_p(0,byte);}
+#define r0()                    (in_p(0) & 0xff)
+#define w1(byte)                {out_p(1,byte);}
+#define r1()                    (in_p(1) & 0xff)
+#define w2(byte)                {out_p(2,byte);}
+#define r2()                    (in_p(2) & 0xff)
+#define w3(byte)                {out_p(3,byte);}
+#define w4(byte)                {out_p(4,byte);}
+#define r4()                    (in_p(4) & 0xff)
+#define w4w(data)     		{outw(data,pi->port+4); delay_p;}
+#define w4l(data)     		{outl(data,pi->port+4); delay_p;}
+#define r4w()         		(delay_p,inw(pi->port+4)&0xffff)
+#define r4l()         		(delay_p,inl(pi->port+4)&0xffffffff)
+
+static inline u16 pi_swab16( char *b, int k)
+
+{ 	union { u16 u; char t[2]; } r;
+
+	r.t[0]=b[2*k+1]; r.t[1]=b[2*k];
+        return r.u;
+}
+
+static inline u32 pi_swab32( char *b, int k)
+
+{ 	union { u32 u; char f[4]; } r;
+
+	r.f[0]=b[4*k+1]; r.f[1]=b[4*k];
+	r.f[2]=b[4*k+3]; r.f[3]=b[4*k+2];
+        return r.u;
+}
+
+struct pi_protocol {
+
+	char	name[8];	/* name for this protocol */
+	int	index;		/* index into protocol table */
+
+	int	max_mode;	/* max mode number */
+	int	epp_first;	/* modes >= this use 8 ports */
+	
+	int	default_delay;  /* delay parameter if not specified */
+	int	max_units;	/* max chained units probed for */
+
+	void (*write_regr)(PIA *,int,int,int);
+	int  (*read_regr)(PIA *,int,int);
+	void (*write_block)(PIA *,char *,int);
+	void (*read_block)(PIA *,char *,int);
+
+	void (*connect)(PIA *);
+	void (*disconnect)(PIA *);
+	
+	int  (*test_port)(PIA *);
+	int  (*probe_unit)(PIA *);
+	int  (*test_proto)(PIA *,char *,int);
+	void (*log_adapter)(PIA *,char *,int);
+	
+	int (*init_proto)(PIA *);
+	void (*release_proto)(PIA *);
+	struct module *owner;
+};
+
+typedef struct pi_protocol PIP;
+
+extern int paride_register( PIP * );
+extern void paride_unregister ( PIP * );
+
+#endif /* __DRIVERS_PARIDE_H__ */
+/* end of paride.h */

drivers/block/paride/pcd.c

@@ -0,0 +1,991 @@
+/* 
+	pcd.c	(c) 1997-8  Grant R. Guenther <grant@torque.net>
+		            Under the terms of the GNU General Public License.
+
+	This is a high-level driver for parallel port ATAPI CD-ROM
+        drives based on chips supported by the paride module.
+
+        By default, the driver will autoprobe for a single parallel
+        port ATAPI CD-ROM drive, but if their individual parameters are
+        specified, the driver can handle up to 4 drives.
+
+        The behaviour of the pcd driver can be altered by setting
+        some parameters from the insmod command line.  The following
+        parameters are adjustable:
+
+            drive0      These four arguments can be arrays of       
+            drive1      1-6 integers as follows:
+            drive2
+            drive3      <prt>,<pro>,<uni>,<mod>,<slv>,<dly>
+
+                        Where,
+
+                <prt>   is the base of the parallel port address for
+                        the corresponding drive.  (required)
+
+                <pro>   is the protocol number for the adapter that
+                        supports this drive.  These numbers are
+                        logged by 'paride' when the protocol modules
+                        are initialised.  (0 if not given)
+
+                <uni>   for those adapters that support chained
+                        devices, this is the unit selector for the
+                        chain of devices on the given port.  It should
+                        be zero for devices that don't support chaining.
+                        (0 if not given)
+
+                <mod>   this can be -1 to choose the best mode, or one
+                        of the mode numbers supported by the adapter.
+                        (-1 if not given)
+
+		<slv>   ATAPI CD-ROMs can be jumpered to master or slave.
+			Set this to 0 to choose the master drive, 1 to
+                        choose the slave, -1 (the default) to choose the
+			first drive found.
+
+                <dly>   some parallel ports require the driver to 
+                        go more slowly.  -1 sets a default value that
+                        should work with the chosen protocol.  Otherwise,
+                        set this to a small integer, the larger it is
+                        the slower the port i/o.  In some cases, setting
+                        this to zero will speed up the device. (default -1)
+                        
+            major       You may use this parameter to overide the
+                        default major number (46) that this driver
+                        will use.  Be sure to change the device
+                        name as well.
+
+            name        This parameter is a character string that
+                        contains the name the kernel will use for this
+                        device (in /proc output, for instance).
+                        (default "pcd")
+
+            verbose     This parameter controls the amount of logging
+                        that the driver will do.  Set it to 0 for
+                        normal operation, 1 to see autoprobe progress
+                        messages, or 2 to see additional debugging
+                        output.  (default 0)
+  
+            nice        This parameter controls the driver's use of
+                        idle CPU time, at the expense of some speed.
+ 
+	If this driver is built into the kernel, you can use the
+        following kernel command line parameters, with the same values
+        as the corresponding module parameters listed above:
+
+	    pcd.drive0
+	    pcd.drive1
+	    pcd.drive2
+	    pcd.drive3
+	    pcd.nice
+
+        In addition, you can use the parameter pcd.disable to disable
+        the driver entirely.
+
+*/
+
+/* Changes:
+
+	1.01	GRG 1998.01.24	Added test unit ready support
+	1.02    GRG 1998.05.06  Changes to pcd_completion, ready_wait,
+				and loosen interpretation of ATAPI
+			        standard for clearing error status.
+				Use spinlocks. Eliminate sti().
+	1.03    GRG 1998.06.16  Eliminated an Ugh
+	1.04	GRG 1998.08.15  Added extra debugging, improvements to
+				pcd_completion, use HZ in loop timing
+	1.05	GRG 1998.08.16	Conformed to "Uniform CD-ROM" standard
+	1.06    GRG 1998.08.19  Added audio ioctl support
+	1.07    GRG 1998.09.24  Increased reset timeout, added jumbo support
+
+*/
+
+#define	PCD_VERSION	"1.07"
+#define PCD_MAJOR	46
+#define PCD_NAME	"pcd"
+#define PCD_UNITS	4
+
+/* Here are things one can override from the insmod command.
+   Most are autoprobed by paride unless set here.  Verbose is off
+   by default.
+
+*/
+
+static int verbose = 0;
+static int major = PCD_MAJOR;
+static char *name = PCD_NAME;
+static int nice = 0;
+static int disable = 0;
+
+static int drive0[6] = { 0, 0, 0, -1, -1, -1 };
+static int drive1[6] = { 0, 0, 0, -1, -1, -1 };
+static int drive2[6] = { 0, 0, 0, -1, -1, -1 };
+static int drive3[6] = { 0, 0, 0, -1, -1, -1 };
+
+static int (*drives[4])[6] = {&drive0, &drive1, &drive2, &drive3};
+static int pcd_drive_count;
+
+enum {D_PRT, D_PRO, D_UNI, D_MOD, D_SLV, D_DLY};
+
+/* end of parameters */
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/errno.h>
+#include <linux/fs.h>
+#include <linux/kernel.h>
+#include <linux/delay.h>
+#include <linux/cdrom.h>
+#include <linux/spinlock.h>
+#include <linux/blkdev.h>
+#include <linux/mutex.h>
+#include <asm/uaccess.h>
+
+static DEFINE_MUTEX(pcd_mutex);
+static DEFINE_SPINLOCK(pcd_lock);
+
+module_param(verbose, int, 0644);
+module_param(major, int, 0);
+module_param(name, charp, 0);
+module_param(nice, int, 0);
+module_param_array(drive0, int, NULL, 0);
+module_param_array(drive1, int, NULL, 0);
+module_param_array(drive2, int, NULL, 0);
+module_param_array(drive3, int, NULL, 0);
+
+#include "paride.h"
+#include "pseudo.h"
+
+#define PCD_RETRIES	     5
+#define PCD_TMO		   800	/* timeout in jiffies */
+#define PCD_DELAY           50	/* spin delay in uS */
+#define PCD_READY_TMO	    20	/* in seconds */
+#define PCD_RESET_TMO	   100	/* in tenths of a second */
+
+#define PCD_SPIN	(1000000*PCD_TMO)/(HZ*PCD_DELAY)
+
+#define IDE_ERR		0x01
+#define IDE_DRQ         0x08
+#define IDE_READY       0x40
+#define IDE_BUSY        0x80
+
+static int pcd_open(struct cdrom_device_info *cdi, int purpose);
+static void pcd_release(struct cdrom_device_info *cdi);
+static int pcd_drive_status(struct cdrom_device_info *cdi, int slot_nr);
+static unsigned int pcd_check_events(struct cdrom_device_info *cdi,
+				     unsigned int clearing, int slot_nr);
+static int pcd_tray_move(struct cdrom_device_info *cdi, int position);
+static int pcd_lock_door(struct cdrom_device_info *cdi, int lock);
+static int pcd_drive_reset(struct cdrom_device_info *cdi);
+static int pcd_get_mcn(struct cdrom_device_info *cdi, struct cdrom_mcn *mcn);
+static int pcd_audio_ioctl(struct cdrom_device_info *cdi,
+			   unsigned int cmd, void *arg);
+static int pcd_packet(struct cdrom_device_info *cdi,
+		      struct packet_command *cgc);
+
+static int pcd_detect(void);
+static void pcd_probe_capabilities(void);
+static void do_pcd_read_drq(void);
+static void do_pcd_request(struct request_queue * q);
+static void do_pcd_read(void);
+
+struct pcd_unit {
+	struct pi_adapter pia;	/* interface to paride layer */
+	struct pi_adapter *pi;
+	int drive;		/* master/slave */
+	int last_sense;		/* result of last request sense */
+	int changed;		/* media change seen */
+	int present;		/* does this unit exist ? */
+	char *name;		/* pcd0, pcd1, etc */
+	struct cdrom_device_info info;	/* uniform cdrom interface */
+	struct gendisk *disk;
+};
+
+static struct pcd_unit pcd[PCD_UNITS];
+
+static char pcd_scratch[64];
+static char pcd_buffer[2048];	/* raw block buffer */
+static int pcd_bufblk = -1;	/* block in buffer, in CD units,
+				   -1 for nothing there. See also
+				   pd_unit.
+				 */
+
+/* the variables below are used mainly in the I/O request engine, which
+   processes only one request at a time.
+*/
+
+static struct pcd_unit *pcd_current; /* current request's drive */
+static struct request *pcd_req;
+static int pcd_retries;		/* retries on current request */
+static int pcd_busy;		/* request being processed ? */
+static int pcd_sector;		/* address of next requested sector */
+static int pcd_count;		/* number of blocks still to do */
+static char *pcd_buf;		/* buffer for request in progress */
+
+/* kernel glue structures */
+
+static int pcd_block_open(struct block_device *bdev, fmode_t mode)
+{
+	struct pcd_unit *cd = bdev->bd_disk->private_data;
+	int ret;
+
+	mutex_lock(&pcd_mutex);
+	ret = cdrom_open(&cd->info, bdev, mode);
+	mutex_unlock(&pcd_mutex);
+
+	return ret;
+}
+
+static void pcd_block_release(struct gendisk *disk, fmode_t mode)
+{
+	struct pcd_unit *cd = disk->private_data;
+	mutex_lock(&pcd_mutex);
+	cdrom_release(&cd->info, mode);
+	mutex_unlock(&pcd_mutex);
+}
+
+static int pcd_block_ioctl(struct block_device *bdev, fmode_t mode,
+				unsigned cmd, unsigned long arg)
+{
+	struct pcd_unit *cd = bdev->bd_disk->private_data;
+	int ret;
+
+	mutex_lock(&pcd_mutex);
+	ret = cdrom_ioctl(&cd->info, bdev, mode, cmd, arg);
+	mutex_unlock(&pcd_mutex);
+
+	return ret;
+}
+
+static unsigned int pcd_block_check_events(struct gendisk *disk,
+					   unsigned int clearing)
+{
+	struct pcd_unit *cd = disk->private_data;
+	return cdrom_check_events(&cd->info, clearing);
+}
+
+static const struct block_device_operations pcd_bdops = {
+	.owner		= THIS_MODULE,
+	.open		= pcd_block_open,
+	.release	= pcd_block_release,
+	.ioctl		= pcd_block_ioctl,
+	.check_events	= pcd_block_check_events,
+};
+
+static struct cdrom_device_ops pcd_dops = {
+	.open		= pcd_open,
+	.release	= pcd_release,
+	.drive_status	= pcd_drive_status,
+	.check_events	= pcd_check_events,
+	.tray_move	= pcd_tray_move,
+	.lock_door	= pcd_lock_door,
+	.get_mcn	= pcd_get_mcn,
+	.reset		= pcd_drive_reset,
+	.audio_ioctl	= pcd_audio_ioctl,
+	.generic_packet	= pcd_packet,
+	.capability	= CDC_CLOSE_TRAY | CDC_OPEN_TRAY | CDC_LOCK |
+			  CDC_MCN | CDC_MEDIA_CHANGED | CDC_RESET |
+			  CDC_PLAY_AUDIO | CDC_GENERIC_PACKET | CDC_CD_R |
+			  CDC_CD_RW,
+};
+
+static void pcd_init_units(void)
+{
+	struct pcd_unit *cd;
+	int unit;
+
+	pcd_drive_count = 0;
+	for (unit = 0, cd = pcd; unit < PCD_UNITS; unit++, cd++) {
+		struct gendisk *disk = alloc_disk(1);
+		if (!disk)
+			continue;
+		cd->disk = disk;
+		cd->pi = &cd->pia;
+		cd->present = 0;
+		cd->last_sense = 0;
+		cd->changed = 1;
+		cd->drive = (*drives[unit])[D_SLV];
+		if ((*drives[unit])[D_PRT])
+			pcd_drive_count++;
+
+		cd->name = &cd->info.name[0];
+		snprintf(cd->name, sizeof(cd->info.name), "%s%d", name, unit);
+		cd->info.ops = &pcd_dops;
+		cd->info.handle = cd;
+		cd->info.speed = 0;
+		cd->info.capacity = 1;
+		cd->info.mask = 0;
+		disk->major = major;
+		disk->first_minor = unit;
+		strcpy(disk->disk_name, cd->name);	/* umm... */
+		disk->fops = &pcd_bdops;
+		disk->flags = GENHD_FL_BLOCK_EVENTS_ON_EXCL_WRITE;
+	}
+}
+
+static int pcd_open(struct cdrom_device_info *cdi, int purpose)
+{
+	struct pcd_unit *cd = cdi->handle;
+	if (!cd->present)
+		return -ENODEV;
+	return 0;
+}
+
+static void pcd_release(struct cdrom_device_info *cdi)
+{
+}
+
+static inline int status_reg(struct pcd_unit *cd)
+{
+	return pi_read_regr(cd->pi, 1, 6);
+}
+
+static inline int read_reg(struct pcd_unit *cd, int reg)
+{
+	return pi_read_regr(cd->pi, 0, reg);
+}
+
+static inline void write_reg(struct pcd_unit *cd, int reg, int val)
+{
+	pi_write_regr(cd->pi, 0, reg, val);
+}
+
+static int pcd_wait(struct pcd_unit *cd, int go, int stop, char *fun, char *msg)
+{
+	int j, r, e, s, p;
+
+	j = 0;
+	while ((((r = status_reg(cd)) & go) || (stop && (!(r & stop))))
+	       && (j++ < PCD_SPIN))
+		udelay(PCD_DELAY);
+
+	if ((r & (IDE_ERR & stop)) || (j > PCD_SPIN)) {
+		s = read_reg(cd, 7);
+		e = read_reg(cd, 1);
+		p = read_reg(cd, 2);
+		if (j > PCD_SPIN)
+			e |= 0x100;
+		if (fun)
+			printk("%s: %s %s: alt=0x%x stat=0x%x err=0x%x"
+			       " loop=%d phase=%d\n",
+			       cd->name, fun, msg, r, s, e, j, p);
+		return (s << 8) + r;
+	}
+	return 0;
+}
+
+static int pcd_command(struct pcd_unit *cd, char *cmd, int dlen, char *fun)
+{
+	pi_connect(cd->pi);
+
+	write_reg(cd, 6, 0xa0 + 0x10 * cd->drive);
+
+	if (pcd_wait(cd, IDE_BUSY | IDE_DRQ, 0, fun, "before command")) {
+		pi_disconnect(cd->pi);
+		return -1;
+	}
+
+	write_reg(cd, 4, dlen % 256);
+	write_reg(cd, 5, dlen / 256);
+	write_reg(cd, 7, 0xa0);	/* ATAPI packet command */
+
+	if (pcd_wait(cd, IDE_BUSY, IDE_DRQ, fun, "command DRQ")) {
+		pi_disconnect(cd->pi);
+		return -1;
+	}
+
+	if (read_reg(cd, 2) != 1) {
+		printk("%s: %s: command phase error\n", cd->name, fun);
+		pi_disconnect(cd->pi);
+		return -1;
+	}
+
+	pi_write_block(cd->pi, cmd, 12);
+
+	return 0;
+}
+
+static int pcd_completion(struct pcd_unit *cd, char *buf, char *fun)
+{
+	int r, d, p, n, k, j;
+
+	r = -1;
+	k = 0;
+	j = 0;
+
+	if (!pcd_wait(cd, IDE_BUSY, IDE_DRQ | IDE_READY | IDE_ERR,
+		      fun, "completion")) {
+		r = 0;
+		while (read_reg(cd, 7) & IDE_DRQ) {
+			d = read_reg(cd, 4) + 256 * read_reg(cd, 5);
+			n = (d + 3) & 0xfffc;
+			p = read_reg(cd, 2) & 3;
+
+			if ((p == 2) && (n > 0) && (j == 0)) {
+				pi_read_block(cd->pi, buf, n);
+				if (verbose > 1)
+					printk("%s: %s: Read %d bytes\n",
+					       cd->name, fun, n);
+				r = 0;
+				j++;
+			} else {
+				if (verbose > 1)
+					printk
+					    ("%s: %s: Unexpected phase %d, d=%d, k=%d\n",
+					     cd->name, fun, p, d, k);
+				if (verbose < 2)
+					printk_once(
+					    "%s: WARNING: ATAPI phase errors\n",
+					    cd->name);
+				mdelay(1);
+			}
+			if (k++ > PCD_TMO) {
+				printk("%s: Stuck DRQ\n", cd->name);
+				break;
+			}
+			if (pcd_wait
+			    (cd, IDE_BUSY, IDE_DRQ | IDE_READY | IDE_ERR, fun,
+			     "completion")) {
+				r = -1;
+				break;
+			}
+		}
+	}
+
+	pi_disconnect(cd->pi);
+
+	return r;
+}
+
+static void pcd_req_sense(struct pcd_unit *cd, char *fun)
+{
+	char rs_cmd[12] = { 0x03, 0, 0, 0, 16, 0, 0, 0, 0, 0, 0, 0 };
+	char buf[16];
+	int r, c;
+
+	r = pcd_command(cd, rs_cmd, 16, "Request sense");
+	mdelay(1);
+	if (!r)
+		pcd_completion(cd, buf, "Request sense");
+
+	cd->last_sense = -1;
+	c = 2;
+	if (!r) {
+		if (fun)
+			printk("%s: %s: Sense key: %x, ASC: %x, ASQ: %x\n",
+			       cd->name, fun, buf[2] & 0xf, buf[12], buf[13]);
+		c = buf[2] & 0xf;
+		cd->last_sense =
+		    c | ((buf[12] & 0xff) << 8) | ((buf[13] & 0xff) << 16);
+	}
+	if ((c == 2) || (c == 6))
+		cd->changed = 1;
+}
+
+static int pcd_atapi(struct pcd_unit *cd, char *cmd, int dlen, char *buf, char *fun)
+{
+	int r;
+
+	r = pcd_command(cd, cmd, dlen, fun);
+	mdelay(1);
+	if (!r)
+		r = pcd_completion(cd, buf, fun);
+	if (r)
+		pcd_req_sense(cd, fun);
+
+	return r;
+}
+
+static int pcd_packet(struct cdrom_device_info *cdi, struct packet_command *cgc)
+{
+	return pcd_atapi(cdi->handle, cgc->cmd, cgc->buflen, cgc->buffer,
+			 "generic packet");
+}
+
+#define DBMSG(msg)	((verbose>1)?(msg):NULL)
+
+static unsigned int pcd_check_events(struct cdrom_device_info *cdi,
+				     unsigned int clearing, int slot_nr)
+{
+	struct pcd_unit *cd = cdi->handle;
+	int res = cd->changed;
+	if (res)
+		cd->changed = 0;
+	return res ? DISK_EVENT_MEDIA_CHANGE : 0;
+}
+
+static int pcd_lock_door(struct cdrom_device_info *cdi, int lock)
+{
+	char un_cmd[12] = { 0x1e, 0, 0, 0, lock, 0, 0, 0, 0, 0, 0, 0 };
+
+	return pcd_atapi(cdi->handle, un_cmd, 0, pcd_scratch,
+			 lock ? "lock door" : "unlock door");
+}
+
+static int pcd_tray_move(struct cdrom_device_info *cdi, int position)
+{
+	char ej_cmd[12] = { 0x1b, 0, 0, 0, 3 - position, 0, 0, 0, 0, 0, 0, 0 };
+
+	return pcd_atapi(cdi->handle, ej_cmd, 0, pcd_scratch,
+			 position ? "eject" : "close tray");
+}
+
+static void pcd_sleep(int cs)
+{
+	schedule_timeout_interruptible(cs);
+}
+
+static int pcd_reset(struct pcd_unit *cd)
+{
+	int i, k, flg;
+	int expect[5] = { 1, 1, 1, 0x14, 0xeb };
+
+	pi_connect(cd->pi);
+	write_reg(cd, 6, 0xa0 + 0x10 * cd->drive);
+	write_reg(cd, 7, 8);
+
+	pcd_sleep(20 * HZ / 1000);	/* delay a bit */
+
+	k = 0;
+	while ((k++ < PCD_RESET_TMO) && (status_reg(cd) & IDE_BUSY))
+		pcd_sleep(HZ / 10);
+
+	flg = 1;
+	for (i = 0; i < 5; i++)
+		flg &= (read_reg(cd, i + 1) == expect[i]);
+
+	if (verbose) {
+		printk("%s: Reset (%d) signature = ", cd->name, k);
+		for (i = 0; i < 5; i++)
+			printk("%3x", read_reg(cd, i + 1));
+		if (!flg)
+			printk(" (incorrect)");
+		printk("\n");
+	}
+
+	pi_disconnect(cd->pi);
+	return flg - 1;
+}
+
+static int pcd_drive_reset(struct cdrom_device_info *cdi)
+{
+	return pcd_reset(cdi->handle);
+}
+
+static int pcd_ready_wait(struct pcd_unit *cd, int tmo)
+{
+	char tr_cmd[12] = { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 };
+	int k, p;
+
+	k = 0;
+	while (k < tmo) {
+		cd->last_sense = 0;
+		pcd_atapi(cd, tr_cmd, 0, NULL, DBMSG("test unit ready"));
+		p = cd->last_sense;
+		if (!p)
+			return 0;
+		if (!(((p & 0xffff) == 0x0402) || ((p & 0xff) == 6)))
+			return p;
+		k++;
+		pcd_sleep(HZ);
+	}
+	return 0x000020;	/* timeout */
+}
+
+static int pcd_drive_status(struct cdrom_device_info *cdi, int slot_nr)
+{
+	char rc_cmd[12] = { 0x25, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 };
+	struct pcd_unit *cd = cdi->handle;
+
+	if (pcd_ready_wait(cd, PCD_READY_TMO))
+		return CDS_DRIVE_NOT_READY;
+	if (pcd_atapi(cd, rc_cmd, 8, pcd_scratch, DBMSG("check media")))
+		return CDS_NO_DISC;
+	return CDS_DISC_OK;
+}
+
+static int pcd_identify(struct pcd_unit *cd, char *id)
+{
+	int k, s;
+	char id_cmd[12] = { 0x12, 0, 0, 0, 36, 0, 0, 0, 0, 0, 0, 0 };
+
+	pcd_bufblk = -1;
+
+	s = pcd_atapi(cd, id_cmd, 36, pcd_buffer, "identify");
+
+	if (s)
+		return -1;
+	if ((pcd_buffer[0] & 0x1f) != 5) {
+		if (verbose)
+			printk("%s: %s is not a CD-ROM\n",
+			       cd->name, cd->drive ? "Slave" : "Master");
+		return -1;
+	}
+	memcpy(id, pcd_buffer + 16, 16);
+	id[16] = 0;
+	k = 16;
+	while ((k >= 0) && (id[k] <= 0x20)) {
+		id[k] = 0;
+		k--;
+	}
+
+	printk("%s: %s: %s\n", cd->name, cd->drive ? "Slave" : "Master", id);
+
+	return 0;
+}
+
+/*
+ * returns  0, with id set if drive is detected
+ *	    -1, if drive detection failed
+ */
+static int pcd_probe(struct pcd_unit *cd, int ms, char *id)
+{
+	if (ms == -1) {
+		for (cd->drive = 0; cd->drive <= 1; cd->drive++)
+			if (!pcd_reset(cd) && !pcd_identify(cd, id))
+				return 0;
+	} else {
+		cd->drive = ms;
+		if (!pcd_reset(cd) && !pcd_identify(cd, id))
+			return 0;
+	}
+	return -1;
+}
+
+static void pcd_probe_capabilities(void)
+{
+	int unit, r;
+	char buffer[32];
+	char cmd[12] = { 0x5a, 1 << 3, 0x2a, 0, 0, 0, 0, 18, 0, 0, 0, 0 };
+	struct pcd_unit *cd;
+
+	for (unit = 0, cd = pcd; unit < PCD_UNITS; unit++, cd++) {
+		if (!cd->present)
+			continue;
+		r = pcd_atapi(cd, cmd, 18, buffer, "mode sense capabilities");
+		if (r)
+			continue;
+		/* we should now have the cap page */
+		if ((buffer[11] & 1) == 0)
+			cd->info.mask |= CDC_CD_R;
+		if ((buffer[11] & 2) == 0)
+			cd->info.mask |= CDC_CD_RW;
+		if ((buffer[12] & 1) == 0)
+			cd->info.mask |= CDC_PLAY_AUDIO;
+		if ((buffer[14] & 1) == 0)
+			cd->info.mask |= CDC_LOCK;
+		if ((buffer[14] & 8) == 0)
+			cd->info.mask |= CDC_OPEN_TRAY;
+		if ((buffer[14] >> 6) == 0)
+			cd->info.mask |= CDC_CLOSE_TRAY;
+	}
+}
+
+static int pcd_detect(void)
+{
+	char id[18];
+	int k, unit;
+	struct pcd_unit *cd;
+
+	printk("%s: %s version %s, major %d, nice %d\n",
+	       name, name, PCD_VERSION, major, nice);
+
+	k = 0;
+	if (pcd_drive_count == 0) { /* nothing spec'd - so autoprobe for 1 */
+		cd = pcd;
+		if (pi_init(cd->pi, 1, -1, -1, -1, -1, -1, pcd_buffer,
+			    PI_PCD, verbose, cd->name)) {
+			if (!pcd_probe(cd, -1, id) && cd->disk) {
+				cd->present = 1;
+				k++;
+			} else
+				pi_release(cd->pi);
+		}
+	} else {
+		for (unit = 0, cd = pcd; unit < PCD_UNITS; unit++, cd++) {
+			int *conf = *drives[unit];
+			if (!conf[D_PRT])
+				continue;
+			if (!pi_init(cd->pi, 0, conf[D_PRT], conf[D_MOD],
+				     conf[D_UNI], conf[D_PRO], conf[D_DLY],
+				     pcd_buffer, PI_PCD, verbose, cd->name)) 
+				continue;
+			if (!pcd_probe(cd, conf[D_SLV], id) && cd->disk) {
+				cd->present = 1;
+				k++;
+			} else
+				pi_release(cd->pi);
+		}
+	}
+	if (k)
+		return 0;
+
+	printk("%s: No CD-ROM drive found\n", name);
+	for (unit = 0, cd = pcd; unit < PCD_UNITS; unit++, cd++)
+		put_disk(cd->disk);
+	return -1;
+}
+
+/* I/O request processing */
+static struct request_queue *pcd_queue;
+
+static void do_pcd_request(struct request_queue * q)
+{
+	if (pcd_busy)
+		return;
+	while (1) {
+		if (!pcd_req) {
+			pcd_req = blk_fetch_request(q);
+			if (!pcd_req)
+				return;
+		}
+
+		if (rq_data_dir(pcd_req) == READ) {
+			struct pcd_unit *cd = pcd_req->rq_disk->private_data;
+			if (cd != pcd_current)
+				pcd_bufblk = -1;
+			pcd_current = cd;
+			pcd_sector = blk_rq_pos(pcd_req);
+			pcd_count = blk_rq_cur_sectors(pcd_req);
+			pcd_buf = bio_data(pcd_req->bio);
+			pcd_busy = 1;
+			ps_set_intr(do_pcd_read, NULL, 0, nice);
+			return;
+		} else {
+			__blk_end_request_all(pcd_req, -EIO);
+			pcd_req = NULL;
+		}
+	}
+}
+
+static inline void next_request(int err)
+{
+	unsigned long saved_flags;
+
+	spin_lock_irqsave(&pcd_lock, saved_flags);
+	if (!__blk_end_request_cur(pcd_req, err))
+		pcd_req = NULL;
+	pcd_busy = 0;
+	do_pcd_request(pcd_queue);
+	spin_unlock_irqrestore(&pcd_lock, saved_flags);
+}
+
+static int pcd_ready(void)
+{
+	return (((status_reg(pcd_current) & (IDE_BUSY | IDE_DRQ)) == IDE_DRQ));
+}
+
+static void pcd_transfer(void)
+{
+
+	while (pcd_count && (pcd_sector / 4 == pcd_bufblk)) {
+		int o = (pcd_sector % 4) * 512;
+		memcpy(pcd_buf, pcd_buffer + o, 512);
+		pcd_count--;
+		pcd_buf += 512;
+		pcd_sector++;
+	}
+}
+
+static void pcd_start(void)
+{
+	int b, i;
+	char rd_cmd[12] = { 0xa8, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0 };
+
+	pcd_bufblk = pcd_sector / 4;
+	b = pcd_bufblk;
+	for (i = 0; i < 4; i++) {
+		rd_cmd[5 - i] = b & 0xff;
+		b = b >> 8;
+	}
+
+	if (pcd_command(pcd_current, rd_cmd, 2048, "read block")) {
+		pcd_bufblk = -1;
+		next_request(-EIO);
+		return;
+	}
+
+	mdelay(1);
+
+	ps_set_intr(do_pcd_read_drq, pcd_ready, PCD_TMO, nice);
+}
+
+static void do_pcd_read(void)
+{
+	pcd_busy = 1;
+	pcd_retries = 0;
+	pcd_transfer();
+	if (!pcd_count) {
+		next_request(0);
+		return;
+	}
+
+	pi_do_claimed(pcd_current->pi, pcd_start);
+}
+
+static void do_pcd_read_drq(void)
+{
+	unsigned long saved_flags;
+
+	if (pcd_completion(pcd_current, pcd_buffer, "read block")) {
+		if (pcd_retries < PCD_RETRIES) {
+			mdelay(1);
+			pcd_retries++;
+			pi_do_claimed(pcd_current->pi, pcd_start);
+			return;
+		}
+		pcd_bufblk = -1;
+		next_request(-EIO);
+		return;
+	}
+
+	do_pcd_read();
+	spin_lock_irqsave(&pcd_lock, saved_flags);
+	do_pcd_request(pcd_queue);
+	spin_unlock_irqrestore(&pcd_lock, saved_flags);
+}
+
+/* the audio_ioctl stuff is adapted from sr_ioctl.c */
+
+static int pcd_audio_ioctl(struct cdrom_device_info *cdi, unsigned int cmd, void *arg)
+{
+	struct pcd_unit *cd = cdi->handle;
+
+	switch (cmd) {
+
+	case CDROMREADTOCHDR:
+
+		{
+			char cmd[12] =
+			    { GPCMD_READ_TOC_PMA_ATIP, 0, 0, 0, 0, 0, 0, 0, 12,
+			 0, 0, 0 };
+			struct cdrom_tochdr *tochdr =
+			    (struct cdrom_tochdr *) arg;
+			char buffer[32];
+			int r;
+
+			r = pcd_atapi(cd, cmd, 12, buffer, "read toc header");
+
+			tochdr->cdth_trk0 = buffer[2];
+			tochdr->cdth_trk1 = buffer[3];
+
+			return r ? -EIO : 0;
+		}
+
+	case CDROMREADTOCENTRY:
+
+		{
+			char cmd[12] =
+			    { GPCMD_READ_TOC_PMA_ATIP, 0, 0, 0, 0, 0, 0, 0, 12,
+			 0, 0, 0 };
+
+			struct cdrom_tocentry *tocentry =
+			    (struct cdrom_tocentry *) arg;
+			unsigned char buffer[32];
+			int r;
+
+			cmd[1] =
+			    (tocentry->cdte_format == CDROM_MSF ? 0x02 : 0);
+			cmd[6] = tocentry->cdte_track;
+
+			r = pcd_atapi(cd, cmd, 12, buffer, "read toc entry");
+
+			tocentry->cdte_ctrl = buffer[5] & 0xf;
+			tocentry->cdte_adr = buffer[5] >> 4;
+			tocentry->cdte_datamode =
+			    (tocentry->cdte_ctrl & 0x04) ? 1 : 0;
+			if (tocentry->cdte_format == CDROM_MSF) {
+				tocentry->cdte_addr.msf.minute = buffer[9];
+				tocentry->cdte_addr.msf.second = buffer[10];
+				tocentry->cdte_addr.msf.frame = buffer[11];
+			} else
+				tocentry->cdte_addr.lba =
+				    (((((buffer[8] << 8) + buffer[9]) << 8)
+				      + buffer[10]) << 8) + buffer[11];
+
+			return r ? -EIO : 0;
+		}
+
+	default:
+
+		return -ENOSYS;
+	}
+}
+
+static int pcd_get_mcn(struct cdrom_device_info *cdi, struct cdrom_mcn *mcn)
+{
+	char cmd[12] =
+	    { GPCMD_READ_SUBCHANNEL, 0, 0x40, 2, 0, 0, 0, 0, 24, 0, 0, 0 };
+	char buffer[32];
+
+	if (pcd_atapi(cdi->handle, cmd, 24, buffer, "get mcn"))
+		return -EIO;
+
+	memcpy(mcn->medium_catalog_number, buffer + 9, 13);
+	mcn->medium_catalog_number[13] = 0;
+
+	return 0;
+}
+
+static int __init pcd_init(void)
+{
+	struct pcd_unit *cd;
+	int unit;
+
+	if (disable)
+		return -EINVAL;
+
+	pcd_init_units();
+
+	if (pcd_detect())
+		return -ENODEV;
+
+	/* get the atapi capabilities page */
+	pcd_probe_capabilities();
+
+	if (register_blkdev(major, name)) {
+		for (unit = 0, cd = pcd; unit < PCD_UNITS; unit++, cd++)
+			put_disk(cd->disk);
+		return -EBUSY;
+	}
+
+	pcd_queue = blk_init_queue(do_pcd_request, &pcd_lock);
+	if (!pcd_queue) {
+		unregister_blkdev(major, name);
+		for (unit = 0, cd = pcd; unit < PCD_UNITS; unit++, cd++)
+			put_disk(cd->disk);
+		return -ENOMEM;
+	}
+
+	for (unit = 0, cd = pcd; unit < PCD_UNITS; unit++, cd++) {
+		if (cd->present) {
+			register_cdrom(&cd->info);
+			cd->disk->private_data = cd;
+			cd->disk->queue = pcd_queue;
+			add_disk(cd->disk);
+		}
+	}
+
+	return 0;
+}
+
+static void __exit pcd_exit(void)
+{
+	struct pcd_unit *cd;
+	int unit;
+
+	for (unit = 0, cd = pcd; unit < PCD_UNITS; unit++, cd++) {
+		if (cd->present) {
+			del_gendisk(cd->disk);
+			pi_release(cd->pi);
+			unregister_cdrom(&cd->info);
+		}
+		put_disk(cd->disk);
+	}
+	blk_cleanup_queue(pcd_queue);
+	unregister_blkdev(major, name);
+}
+
+MODULE_LICENSE("GPL");
+module_init(pcd_init)
+module_exit(pcd_exit)

drivers/block/paride/pd.c

@@ -0,0 +1,958 @@
+/* 
+        pd.c    (c) 1997-8  Grant R. Guenther <grant@torque.net>
+                            Under the terms of the GNU General Public License.
+
+        This is the high-level driver for parallel port IDE hard
+        drives based on chips supported by the paride module.
+
+	By default, the driver will autoprobe for a single parallel
+	port IDE drive, but if their individual parameters are
+        specified, the driver can handle up to 4 drives.
+
+        The behaviour of the pd driver can be altered by setting
+        some parameters from the insmod command line.  The following
+        parameters are adjustable:
+ 
+	    drive0  	These four arguments can be arrays of	    
+	    drive1	1-8 integers as follows:
+	    drive2
+	    drive3	<prt>,<pro>,<uni>,<mod>,<geo>,<sby>,<dly>,<slv>
+
+			Where,
+
+		<prt>	is the base of the parallel port address for
+			the corresponding drive.  (required)
+
+		<pro>   is the protocol number for the adapter that
+			supports this drive.  These numbers are
+                        logged by 'paride' when the protocol modules
+			are initialised.  (0 if not given)
+
+		<uni>   for those adapters that support chained
+			devices, this is the unit selector for the
+		        chain of devices on the given port.  It should
+			be zero for devices that don't support chaining.
+			(0 if not given)
+
+		<mod>   this can be -1 to choose the best mode, or one
+		        of the mode numbers supported by the adapter.
+			(-1 if not given)
+
+		<geo>   this defaults to 0 to indicate that the driver
+			should use the CHS geometry provided by the drive
+			itself.  If set to 1, the driver will provide
+			a logical geometry with 64 heads and 32 sectors
+			per track, to be consistent with most SCSI
+		        drivers.  (0 if not given)
+
+		<sby>   set this to zero to disable the power saving
+			standby mode, if needed.  (1 if not given)
+
+		<dly>   some parallel ports require the driver to 
+			go more slowly.  -1 sets a default value that
+			should work with the chosen protocol.  Otherwise,
+			set this to a small integer, the larger it is
+			the slower the port i/o.  In some cases, setting
+			this to zero will speed up the device. (default -1)
+
+		<slv>   IDE disks can be jumpered to master or slave.
+                        Set this to 0 to choose the master drive, 1 to
+                        choose the slave, -1 (the default) to choose the
+                        first drive found.
+			
+
+            major       You may use this parameter to overide the
+                        default major number (45) that this driver
+                        will use.  Be sure to change the device
+                        name as well.
+
+            name        This parameter is a character string that
+                        contains the name the kernel will use for this
+                        device (in /proc output, for instance).
+			(default "pd")
+
+	    cluster	The driver will attempt to aggregate requests
+			for adjacent blocks into larger multi-block
+			clusters.  The maximum cluster size (in 512
+			byte sectors) is set with this parameter.
+			(default 64)
+
+	    verbose	This parameter controls the amount of logging
+			that the driver will do.  Set it to 0 for 
+			normal operation, 1 to see autoprobe progress
+			messages, or 2 to see additional debugging
+			output.  (default 0)
+
+            nice        This parameter controls the driver's use of
+                        idle CPU time, at the expense of some speed.
+
+        If this driver is built into the kernel, you can use kernel
+        the following command line parameters, with the same values
+        as the corresponding module parameters listed above:
+
+            pd.drive0
+            pd.drive1
+            pd.drive2
+            pd.drive3
+            pd.cluster
+            pd.nice
+
+        In addition, you can use the parameter pd.disable to disable
+        the driver entirely.
+ 
+*/
+
+/* Changes:
+
+	1.01	GRG 1997.01.24	Restored pd_reset()
+				Added eject ioctl
+	1.02    GRG 1998.05.06  SMP spinlock changes, 
+				Added slave support
+	1.03    GRG 1998.06.16  Eliminate an Ugh.
+	1.04	GRG 1998.08.15  Extra debugging, use HZ in loop timing
+	1.05    GRG 1998.09.24  Added jumbo support
+
+*/
+
+#define PD_VERSION      "1.05"
+#define PD_MAJOR	45
+#define PD_NAME		"pd"
+#define PD_UNITS	4
+
+/* Here are things one can override from the insmod command.
+   Most are autoprobed by paride unless set here.  Verbose is off
+   by default.
+
+*/
+#include <linux/types.h>
+
+static bool verbose = 0;
+static int major = PD_MAJOR;
+static char *name = PD_NAME;
+static int cluster = 64;
+static int nice = 0;
+static int disable = 0;
+
+static int drive0[8] = { 0, 0, 0, -1, 0, 1, -1, -1 };
+static int drive1[8] = { 0, 0, 0, -1, 0, 1, -1, -1 };
+static int drive2[8] = { 0, 0, 0, -1, 0, 1, -1, -1 };
+static int drive3[8] = { 0, 0, 0, -1, 0, 1, -1, -1 };
+
+static int (*drives[4])[8] = {&drive0, &drive1, &drive2, &drive3};
+
+enum {D_PRT, D_PRO, D_UNI, D_MOD, D_GEO, D_SBY, D_DLY, D_SLV};
+
+/* end of parameters */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/gfp.h>
+#include <linux/fs.h>
+#include <linux/delay.h>
+#include <linux/hdreg.h>
+#include <linux/cdrom.h>	/* for the eject ioctl */
+#include <linux/blkdev.h>
+#include <linux/blkpg.h>
+#include <linux/kernel.h>
+#include <linux/mutex.h>
+#include <asm/uaccess.h>
+#include <linux/workqueue.h>
+
+static DEFINE_MUTEX(pd_mutex);
+static DEFINE_SPINLOCK(pd_lock);
+
+module_param(verbose, bool, 0);
+module_param(major, int, 0);
+module_param(name, charp, 0);
+module_param(cluster, int, 0);
+module_param(nice, int, 0);
+module_param_array(drive0, int, NULL, 0);
+module_param_array(drive1, int, NULL, 0);
+module_param_array(drive2, int, NULL, 0);
+module_param_array(drive3, int, NULL, 0);
+
+#include "paride.h"
+
+#define PD_BITS    4
+
+/* numbers for "SCSI" geometry */
+
+#define PD_LOG_HEADS    64
+#define PD_LOG_SECTS    32
+
+#define PD_ID_OFF       54
+#define PD_ID_LEN       14
+
+#define PD_MAX_RETRIES  5
+#define PD_TMO          800	/* interrupt timeout in jiffies */
+#define PD_SPIN_DEL     50	/* spin delay in micro-seconds  */
+
+#define PD_SPIN         (1000000*PD_TMO)/(HZ*PD_SPIN_DEL)
+
+#define STAT_ERR        0x00001
+#define STAT_INDEX      0x00002
+#define STAT_ECC        0x00004
+#define STAT_DRQ        0x00008
+#define STAT_SEEK       0x00010
+#define STAT_WRERR      0x00020
+#define STAT_READY      0x00040
+#define STAT_BUSY       0x00080
+
+#define ERR_AMNF        0x00100
+#define ERR_TK0NF       0x00200
+#define ERR_ABRT        0x00400
+#define ERR_MCR         0x00800
+#define ERR_IDNF        0x01000
+#define ERR_MC          0x02000
+#define ERR_UNC         0x04000
+#define ERR_TMO         0x10000
+
+#define IDE_READ        	0x20
+#define IDE_WRITE       	0x30
+#define IDE_READ_VRFY		0x40
+#define IDE_INIT_DEV_PARMS	0x91
+#define IDE_STANDBY     	0x96
+#define IDE_ACKCHANGE   	0xdb
+#define IDE_DOORLOCK    	0xde
+#define IDE_DOORUNLOCK  	0xdf
+#define IDE_IDENTIFY    	0xec
+#define IDE_EJECT		0xed
+
+#define PD_NAMELEN	8
+
+struct pd_unit {
+	struct pi_adapter pia;	/* interface to paride layer */
+	struct pi_adapter *pi;
+	int access;		/* count of active opens ... */
+	int capacity;		/* Size of this volume in sectors */
+	int heads;		/* physical geometry */
+	int sectors;
+	int cylinders;
+	int can_lba;
+	int drive;		/* master=0 slave=1 */
+	int changed;		/* Have we seen a disk change ? */
+	int removable;		/* removable media device  ?  */
+	int standby;
+	int alt_geom;
+	char name[PD_NAMELEN];	/* pda, pdb, etc ... */
+	struct gendisk *gd;
+};
+
+static struct pd_unit pd[PD_UNITS];
+
+static char pd_scratch[512];	/* scratch block buffer */
+
+static char *pd_errs[17] = { "ERR", "INDEX", "ECC", "DRQ", "SEEK", "WRERR",
+	"READY", "BUSY", "AMNF", "TK0NF", "ABRT", "MCR",
+	"IDNF", "MC", "UNC", "???", "TMO"
+};
+
+static inline int status_reg(struct pd_unit *disk)
+{
+	return pi_read_regr(disk->pi, 1, 6);
+}
+
+static inline int read_reg(struct pd_unit *disk, int reg)
+{
+	return pi_read_regr(disk->pi, 0, reg);
+}
+
+static inline void write_status(struct pd_unit *disk, int val)
+{
+	pi_write_regr(disk->pi, 1, 6, val);
+}
+
+static inline void write_reg(struct pd_unit *disk, int reg, int val)
+{
+	pi_write_regr(disk->pi, 0, reg, val);
+}
+
+static inline u8 DRIVE(struct pd_unit *disk)
+{
+	return 0xa0+0x10*disk->drive;
+}
+
+/*  ide command interface */
+
+static void pd_print_error(struct pd_unit *disk, char *msg, int status)
+{
+	int i;
+
+	printk("%s: %s: status = 0x%x =", disk->name, msg, status);
+	for (i = 0; i < ARRAY_SIZE(pd_errs); i++)
+		if (status & (1 << i))
+			printk(" %s", pd_errs[i]);
+	printk("\n");
+}
+
+static void pd_reset(struct pd_unit *disk)
+{				/* called only for MASTER drive */
+	write_status(disk, 4);
+	udelay(50);
+	write_status(disk, 0);
+	udelay(250);
+}
+
+#define DBMSG(msg)	((verbose>1)?(msg):NULL)
+
+static int pd_wait_for(struct pd_unit *disk, int w, char *msg)
+{				/* polled wait */
+	int k, r, e;
+
+	k = 0;
+	while (k < PD_SPIN) {
+		r = status_reg(disk);
+		k++;
+		if (((r & w) == w) && !(r & STAT_BUSY))
+			break;
+		udelay(PD_SPIN_DEL);
+	}
+	e = (read_reg(disk, 1) << 8) + read_reg(disk, 7);
+	if (k >= PD_SPIN)
+		e |= ERR_TMO;
+	if ((e & (STAT_ERR | ERR_TMO)) && (msg != NULL))
+		pd_print_error(disk, msg, e);
+	return e;
+}
+
+static void pd_send_command(struct pd_unit *disk, int n, int s, int h, int c0, int c1, int func)
+{
+	write_reg(disk, 6, DRIVE(disk) + h);
+	write_reg(disk, 1, 0);		/* the IDE task file */
+	write_reg(disk, 2, n);
+	write_reg(disk, 3, s);
+	write_reg(disk, 4, c0);
+	write_reg(disk, 5, c1);
+	write_reg(disk, 7, func);
+
+	udelay(1);
+}
+
+static void pd_ide_command(struct pd_unit *disk, int func, int block, int count)
+{
+	int c1, c0, h, s;
+
+	if (disk->can_lba) {
+		s = block & 255;
+		c0 = (block >>= 8) & 255;
+		c1 = (block >>= 8) & 255;
+		h = ((block >>= 8) & 15) + 0x40;
+	} else {
+		s = (block % disk->sectors) + 1;
+		h = (block /= disk->sectors) % disk->heads;
+		c0 = (block /= disk->heads) % 256;
+		c1 = (block >>= 8);
+	}
+	pd_send_command(disk, count, s, h, c0, c1, func);
+}
+
+/* The i/o request engine */
+
+enum action {Fail = 0, Ok = 1, Hold, Wait};
+
+static struct request *pd_req;	/* current request */
+static enum action (*phase)(void);
+
+static void run_fsm(void);
+
+static void ps_tq_int(struct work_struct *work);
+
+static DECLARE_DELAYED_WORK(fsm_tq, ps_tq_int);
+
+static void schedule_fsm(void)
+{
+	if (!nice)
+		schedule_delayed_work(&fsm_tq, 0);
+	else
+		schedule_delayed_work(&fsm_tq, nice-1);
+}
+
+static void ps_tq_int(struct work_struct *work)
+{
+	run_fsm();
+}
+
+static enum action do_pd_io_start(void);
+static enum action pd_special(void);
+static enum action do_pd_read_start(void);
+static enum action do_pd_write_start(void);
+static enum action do_pd_read_drq(void);
+static enum action do_pd_write_done(void);
+
+static struct request_queue *pd_queue;
+static int pd_claimed;
+
+static struct pd_unit *pd_current; /* current request's drive */
+static PIA *pi_current; /* current request's PIA */
+
+static void run_fsm(void)
+{
+	while (1) {
+		enum action res;
+		unsigned long saved_flags;
+		int stop = 0;
+
+		if (!phase) {
+			pd_current = pd_req->rq_disk->private_data;
+			pi_current = pd_current->pi;
+			phase = do_pd_io_start;
+		}
+
+		switch (pd_claimed) {
+			case 0:
+				pd_claimed = 1;
+				if (!pi_schedule_claimed(pi_current, run_fsm))
+					return;
+			case 1:
+				pd_claimed = 2;
+				pi_current->proto->connect(pi_current);
+		}
+
+		switch(res = phase()) {
+			case Ok: case Fail:
+				pi_disconnect(pi_current);
+				pd_claimed = 0;
+				phase = NULL;
+				spin_lock_irqsave(&pd_lock, saved_flags);
+				if (!__blk_end_request_cur(pd_req,
+						res == Ok ? 0 : -EIO)) {
+					pd_req = blk_fetch_request(pd_queue);
+					if (!pd_req)
+						stop = 1;
+				}
+				spin_unlock_irqrestore(&pd_lock, saved_flags);
+				if (stop)
+					return;
+			case Hold:
+				schedule_fsm();
+				return;
+			case Wait:
+				pi_disconnect(pi_current);
+				pd_claimed = 0;
+		}
+	}
+}
+
+static int pd_retries = 0;	/* i/o error retry count */
+static int pd_block;		/* address of next requested block */
+static int pd_count;		/* number of blocks still to do */
+static int pd_run;		/* sectors in current cluster */
+static int pd_cmd;		/* current command READ/WRITE */
+static char *pd_buf;		/* buffer for request in progress */
+
+static enum action do_pd_io_start(void)
+{
+	if (pd_req->cmd_type == REQ_TYPE_SPECIAL) {
+		phase = pd_special;
+		return pd_special();
+	}
+
+	pd_cmd = rq_data_dir(pd_req);
+	if (pd_cmd == READ || pd_cmd == WRITE) {
+		pd_block = blk_rq_pos(pd_req);
+		pd_count = blk_rq_cur_sectors(pd_req);
+		if (pd_block + pd_count > get_capacity(pd_req->rq_disk))
+			return Fail;
+		pd_run = blk_rq_sectors(pd_req);
+		pd_buf = bio_data(pd_req->bio);
+		pd_retries = 0;
+		if (pd_cmd == READ)
+			return do_pd_read_start();
+		else
+			return do_pd_write_start();
+	}
+	return Fail;
+}
+
+static enum action pd_special(void)
+{
+	enum action (*func)(struct pd_unit *) = pd_req->special;
+	return func(pd_current);
+}
+
+static int pd_next_buf(void)
+{
+	unsigned long saved_flags;
+
+	pd_count--;
+	pd_run--;
+	pd_buf += 512;
+	pd_block++;
+	if (!pd_run)
+		return 1;
+	if (pd_count)
+		return 0;
+	spin_lock_irqsave(&pd_lock, saved_flags);
+	__blk_end_request_cur(pd_req, 0);
+	pd_count = blk_rq_cur_sectors(pd_req);
+	pd_buf = bio_data(pd_req->bio);
+	spin_unlock_irqrestore(&pd_lock, saved_flags);
+	return 0;
+}
+
+static unsigned long pd_timeout;
+
+static enum action do_pd_read_start(void)
+{
+	if (pd_wait_for(pd_current, STAT_READY, "do_pd_read") & STAT_ERR) {
+		if (pd_retries < PD_MAX_RETRIES) {
+			pd_retries++;
+			return Wait;
+		}
+		return Fail;
+	}
+	pd_ide_command(pd_current, IDE_READ, pd_block, pd_run);
+	phase = do_pd_read_drq;
+	pd_timeout = jiffies + PD_TMO;
+	return Hold;
+}
+
+static enum action do_pd_write_start(void)
+{
+	if (pd_wait_for(pd_current, STAT_READY, "do_pd_write") & STAT_ERR) {
+		if (pd_retries < PD_MAX_RETRIES) {
+			pd_retries++;
+			return Wait;
+		}
+		return Fail;
+	}
+	pd_ide_command(pd_current, IDE_WRITE, pd_block, pd_run);
+	while (1) {
+		if (pd_wait_for(pd_current, STAT_DRQ, "do_pd_write_drq") & STAT_ERR) {
+			if (pd_retries < PD_MAX_RETRIES) {
+				pd_retries++;
+				return Wait;
+			}
+			return Fail;
+		}
+		pi_write_block(pd_current->pi, pd_buf, 512);
+		if (pd_next_buf())
+			break;
+	}
+	phase = do_pd_write_done;
+	pd_timeout = jiffies + PD_TMO;
+	return Hold;
+}
+
+static inline int pd_ready(void)
+{
+	return !(status_reg(pd_current) & STAT_BUSY);
+}
+
+static enum action do_pd_read_drq(void)
+{
+	if (!pd_ready() && !time_after_eq(jiffies, pd_timeout))
+		return Hold;
+
+	while (1) {
+		if (pd_wait_for(pd_current, STAT_DRQ, "do_pd_read_drq") & STAT_ERR) {
+			if (pd_retries < PD_MAX_RETRIES) {
+				pd_retries++;
+				phase = do_pd_read_start;
+				return Wait;
+			}
+			return Fail;
+		}
+		pi_read_block(pd_current->pi, pd_buf, 512);
+		if (pd_next_buf())
+			break;
+	}
+	return Ok;
+}
+
+static enum action do_pd_write_done(void)
+{
+	if (!pd_ready() && !time_after_eq(jiffies, pd_timeout))
+		return Hold;
+
+	if (pd_wait_for(pd_current, STAT_READY, "do_pd_write_done") & STAT_ERR) {
+		if (pd_retries < PD_MAX_RETRIES) {
+			pd_retries++;
+			phase = do_pd_write_start;
+			return Wait;
+		}
+		return Fail;
+	}
+	return Ok;
+}
+
+/* special io requests */
+
+/* According to the ATA standard, the default CHS geometry should be
+   available following a reset.  Some Western Digital drives come up
+   in a mode where only LBA addresses are accepted until the device
+   parameters are initialised.
+*/
+
+static void pd_init_dev_parms(struct pd_unit *disk)
+{
+	pd_wait_for(disk, 0, DBMSG("before init_dev_parms"));
+	pd_send_command(disk, disk->sectors, 0, disk->heads - 1, 0, 0,
+			IDE_INIT_DEV_PARMS);
+	udelay(300);
+	pd_wait_for(disk, 0, "Initialise device parameters");
+}
+
+static enum action pd_door_lock(struct pd_unit *disk)
+{
+	if (!(pd_wait_for(disk, STAT_READY, "Lock") & STAT_ERR)) {
+		pd_send_command(disk, 1, 0, 0, 0, 0, IDE_DOORLOCK);
+		pd_wait_for(disk, STAT_READY, "Lock done");
+	}
+	return Ok;
+}
+
+static enum action pd_door_unlock(struct pd_unit *disk)
+{
+	if (!(pd_wait_for(disk, STAT_READY, "Lock") & STAT_ERR)) {
+		pd_send_command(disk, 1, 0, 0, 0, 0, IDE_DOORUNLOCK);
+		pd_wait_for(disk, STAT_READY, "Lock done");
+	}
+	return Ok;
+}
+
+static enum action pd_eject(struct pd_unit *disk)
+{
+	pd_wait_for(disk, 0, DBMSG("before unlock on eject"));
+	pd_send_command(disk, 1, 0, 0, 0, 0, IDE_DOORUNLOCK);
+	pd_wait_for(disk, 0, DBMSG("after unlock on eject"));
+	pd_wait_for(disk, 0, DBMSG("before eject"));
+	pd_send_command(disk, 0, 0, 0, 0, 0, IDE_EJECT);
+	pd_wait_for(disk, 0, DBMSG("after eject"));
+	return Ok;
+}
+
+static enum action pd_media_check(struct pd_unit *disk)
+{
+	int r = pd_wait_for(disk, STAT_READY, DBMSG("before media_check"));
+	if (!(r & STAT_ERR)) {
+		pd_send_command(disk, 1, 1, 0, 0, 0, IDE_READ_VRFY);
+		r = pd_wait_for(disk, STAT_READY, DBMSG("RDY after READ_VRFY"));
+	} else
+		disk->changed = 1;	/* say changed if other error */
+	if (r & ERR_MC) {
+		disk->changed = 1;
+		pd_send_command(disk, 1, 0, 0, 0, 0, IDE_ACKCHANGE);
+		pd_wait_for(disk, STAT_READY, DBMSG("RDY after ACKCHANGE"));
+		pd_send_command(disk, 1, 1, 0, 0, 0, IDE_READ_VRFY);
+		r = pd_wait_for(disk, STAT_READY, DBMSG("RDY after VRFY"));
+	}
+	return Ok;
+}
+
+static void pd_standby_off(struct pd_unit *disk)
+{
+	pd_wait_for(disk, 0, DBMSG("before STANDBY"));
+	pd_send_command(disk, 0, 0, 0, 0, 0, IDE_STANDBY);
+	pd_wait_for(disk, 0, DBMSG("after STANDBY"));
+}
+
+static enum action pd_identify(struct pd_unit *disk)
+{
+	int j;
+	char id[PD_ID_LEN + 1];
+
+/* WARNING:  here there may be dragons.  reset() applies to both drives,
+   but we call it only on probing the MASTER. This should allow most
+   common configurations to work, but be warned that a reset can clear
+   settings on the SLAVE drive.
+*/
+
+	if (disk->drive == 0)
+		pd_reset(disk);
+
+	write_reg(disk, 6, DRIVE(disk));
+	pd_wait_for(disk, 0, DBMSG("before IDENT"));
+	pd_send_command(disk, 1, 0, 0, 0, 0, IDE_IDENTIFY);
+
+	if (pd_wait_for(disk, STAT_DRQ, DBMSG("IDENT DRQ")) & STAT_ERR)
+		return Fail;
+	pi_read_block(disk->pi, pd_scratch, 512);
+	disk->can_lba = pd_scratch[99] & 2;
+	disk->sectors = le16_to_cpu(*(__le16 *) (pd_scratch + 12));
+	disk->heads = le16_to_cpu(*(__le16 *) (pd_scratch + 6));
+	disk->cylinders = le16_to_cpu(*(__le16 *) (pd_scratch + 2));
+	if (disk->can_lba)
+		disk->capacity = le32_to_cpu(*(__le32 *) (pd_scratch + 120));
+	else
+		disk->capacity = disk->sectors * disk->heads * disk->cylinders;
+
+	for (j = 0; j < PD_ID_LEN; j++)
+		id[j ^ 1] = pd_scratch[j + PD_ID_OFF];
+	j = PD_ID_LEN - 1;
+	while ((j >= 0) && (id[j] <= 0x20))
+		j--;
+	j++;
+	id[j] = 0;
+
+	disk->removable = pd_scratch[0] & 0x80;
+
+	printk("%s: %s, %s, %d blocks [%dM], (%d/%d/%d), %s media\n",
+	       disk->name, id,
+	       disk->drive ? "slave" : "master",
+	       disk->capacity, disk->capacity / 2048,
+	       disk->cylinders, disk->heads, disk->sectors,
+	       disk->removable ? "removable" : "fixed");
+
+	if (disk->capacity)
+		pd_init_dev_parms(disk);
+	if (!disk->standby)
+		pd_standby_off(disk);
+
+	return Ok;
+}
+
+/* end of io request engine */
+
+static void do_pd_request(struct request_queue * q)
+{
+	if (pd_req)
+		return;
+	pd_req = blk_fetch_request(q);
+	if (!pd_req)
+		return;
+
+	schedule_fsm();
+}
+
+static int pd_special_command(struct pd_unit *disk,
+		      enum action (*func)(struct pd_unit *disk))
+{
+	struct request *rq;
+	int err = 0;
+
+	rq = blk_get_request(disk->gd->queue, READ, __GFP_WAIT);
+	if (IS_ERR(rq))
+		return PTR_ERR(rq);
+
+	rq->cmd_type = REQ_TYPE_SPECIAL;
+	rq->special = func;
+
+	err = blk_execute_rq(disk->gd->queue, disk->gd, rq, 0);
+
+	blk_put_request(rq);
+	return err;
+}
+
+/* kernel glue structures */
+
+static int pd_open(struct block_device *bdev, fmode_t mode)
+{
+	struct pd_unit *disk = bdev->bd_disk->private_data;
+
+	mutex_lock(&pd_mutex);
+	disk->access++;
+
+	if (disk->removable) {
+		pd_special_command(disk, pd_media_check);
+		pd_special_command(disk, pd_door_lock);
+	}
+	mutex_unlock(&pd_mutex);
+	return 0;
+}
+
+static int pd_getgeo(struct block_device *bdev, struct hd_geometry *geo)
+{
+	struct pd_unit *disk = bdev->bd_disk->private_data;
+
+	if (disk->alt_geom) {
+		geo->heads = PD_LOG_HEADS;
+		geo->sectors = PD_LOG_SECTS;
+		geo->cylinders = disk->capacity / (geo->heads * geo->sectors);
+	} else {
+		geo->heads = disk->heads;
+		geo->sectors = disk->sectors;
+		geo->cylinders = disk->cylinders;
+	}
+
+	return 0;
+}
+
+static int pd_ioctl(struct block_device *bdev, fmode_t mode,
+	 unsigned int cmd, unsigned long arg)
+{
+	struct pd_unit *disk = bdev->bd_disk->private_data;
+
+	switch (cmd) {
+	case CDROMEJECT:
+		mutex_lock(&pd_mutex);
+		if (disk->access == 1)
+			pd_special_command(disk, pd_eject);
+		mutex_unlock(&pd_mutex);
+		return 0;
+	default:
+		return -EINVAL;
+	}
+}
+
+static void pd_release(struct gendisk *p, fmode_t mode)
+{
+	struct pd_unit *disk = p->private_data;
+
+	mutex_lock(&pd_mutex);
+	if (!--disk->access && disk->removable)
+		pd_special_command(disk, pd_door_unlock);
+	mutex_unlock(&pd_mutex);
+}
+
+static unsigned int pd_check_events(struct gendisk *p, unsigned int clearing)
+{
+	struct pd_unit *disk = p->private_data;
+	int r;
+	if (!disk->removable)
+		return 0;
+	pd_special_command(disk, pd_media_check);
+	r = disk->changed;
+	disk->changed = 0;
+	return r ? DISK_EVENT_MEDIA_CHANGE : 0;
+}
+
+static int pd_revalidate(struct gendisk *p)
+{
+	struct pd_unit *disk = p->private_data;
+	if (pd_special_command(disk, pd_identify) == 0)
+		set_capacity(p, disk->capacity);
+	else
+		set_capacity(p, 0);
+	return 0;
+}
+
+static const struct block_device_operations pd_fops = {
+	.owner		= THIS_MODULE,
+	.open		= pd_open,
+	.release	= pd_release,
+	.ioctl		= pd_ioctl,
+	.getgeo		= pd_getgeo,
+	.check_events	= pd_check_events,
+	.revalidate_disk= pd_revalidate
+};
+
+/* probing */
+
+static void pd_probe_drive(struct pd_unit *disk)
+{
+	struct gendisk *p = alloc_disk(1 << PD_BITS);
+	if (!p)
+		return;
+	strcpy(p->disk_name, disk->name);
+	p->fops = &pd_fops;
+	p->major = major;
+	p->first_minor = (disk - pd) << PD_BITS;
+	disk->gd = p;
+	p->private_data = disk;
+	p->queue = pd_queue;
+
+	if (disk->drive == -1) {
+		for (disk->drive = 0; disk->drive <= 1; disk->drive++)
+			if (pd_special_command(disk, pd_identify) == 0)
+				return;
+	} else if (pd_special_command(disk, pd_identify) == 0)
+		return;
+	disk->gd = NULL;
+	put_disk(p);
+}
+
+static int pd_detect(void)
+{
+	int found = 0, unit, pd_drive_count = 0;
+	struct pd_unit *disk;
+
+	for (unit = 0; unit < PD_UNITS; unit++) {
+		int *parm = *drives[unit];
+		struct pd_unit *disk = pd + unit;
+		disk->pi = &disk->pia;
+		disk->access = 0;
+		disk->changed = 1;
+		disk->capacity = 0;
+		disk->drive = parm[D_SLV];
+		snprintf(disk->name, PD_NAMELEN, "%s%c", name, 'a'+unit);
+		disk->alt_geom = parm[D_GEO];
+		disk->standby = parm[D_SBY];
+		if (parm[D_PRT])
+			pd_drive_count++;
+	}
+
+	if (pd_drive_count == 0) { /* nothing spec'd - so autoprobe for 1 */
+		disk = pd;
+		if (pi_init(disk->pi, 1, -1, -1, -1, -1, -1, pd_scratch,
+			    PI_PD, verbose, disk->name)) {
+			pd_probe_drive(disk);
+			if (!disk->gd)
+				pi_release(disk->pi);
+		}
+
+	} else {
+		for (unit = 0, disk = pd; unit < PD_UNITS; unit++, disk++) {
+			int *parm = *drives[unit];
+			if (!parm[D_PRT])
+				continue;
+			if (pi_init(disk->pi, 0, parm[D_PRT], parm[D_MOD],
+				     parm[D_UNI], parm[D_PRO], parm[D_DLY],
+				     pd_scratch, PI_PD, verbose, disk->name)) {
+				pd_probe_drive(disk);
+				if (!disk->gd)
+					pi_release(disk->pi);
+			}
+		}
+	}
+	for (unit = 0, disk = pd; unit < PD_UNITS; unit++, disk++) {
+		if (disk->gd) {
+			set_capacity(disk->gd, disk->capacity);
+			add_disk(disk->gd);
+			found = 1;
+		}
+	}
+	if (!found)
+		printk("%s: no valid drive found\n", name);
+	return found;
+}
+
+static int __init pd_init(void)
+{
+	if (disable)
+		goto out1;
+
+	pd_queue = blk_init_queue(do_pd_request, &pd_lock);
+	if (!pd_queue)
+		goto out1;
+
+	blk_queue_max_hw_sectors(pd_queue, cluster);
+
+	if (register_blkdev(major, name))
+		goto out2;
+
+	printk("%s: %s version %s, major %d, cluster %d, nice %d\n",
+	       name, name, PD_VERSION, major, cluster, nice);
+	if (!pd_detect())
+		goto out3;
+
+	return 0;
+
+out3:
+	unregister_blkdev(major, name);
+out2:
+	blk_cleanup_queue(pd_queue);
+out1:
+	return -ENODEV;
+}
+
+static void __exit pd_exit(void)
+{
+	struct pd_unit *disk;
+	int unit;
+	unregister_blkdev(major, name);
+	for (unit = 0, disk = pd; unit < PD_UNITS; unit++, disk++) {
+		struct gendisk *p = disk->gd;
+		if (p) {
+			disk->gd = NULL;
+			del_gendisk(p);
+			put_disk(p);
+			pi_release(disk->pi);
+		}
+	}
+	blk_cleanup_queue(pd_queue);
+}
+
+MODULE_LICENSE("GPL");
+module_init(pd_init)
+module_exit(pd_exit)

drivers/block/paride/pg.c

@@ -0,0 +1,726 @@
+/* 
+	pg.c    (c) 1998  Grant R. Guenther <grant@torque.net>
+			  Under the terms of the GNU General Public License.
+
+	The pg driver provides a simple character device interface for
+	sending ATAPI commands to a device.  With the exception of the
+	ATAPI reset operation, all operations are performed by a pair
+	of read and write operations to the appropriate /dev/pgN device.
+	A write operation delivers a command and any outbound data in
+	a single buffer.  Normally, the write will succeed unless the
+	device is offline or malfunctioning, or there is already another
+	command pending.  If the write succeeds, it should be followed
+	immediately by a read operation, to obtain any returned data and
+	status information.  A read will fail if there is no operation
+	in progress.
+
+	As a special case, the device can be reset with a write operation,
+	and in this case, no following read is expected, or permitted.
+
+	There are no ioctl() operations.  Any single operation
+	may transfer at most PG_MAX_DATA bytes.  Note that the driver must
+	copy the data through an internal buffer.  In keeping with all
+	current ATAPI devices, command packets are assumed to be exactly
+	12 bytes in length.
+
+	To permit future changes to this interface, the headers in the
+	read and write buffers contain a single character "magic" flag.
+	Currently this flag must be the character "P".
+
+	By default, the driver will autoprobe for a single parallel
+	port ATAPI device, but if their individual parameters are
+	specified, the driver can handle up to 4 devices.
+
+	To use this device, you must have the following device 
+	special files defined:
+
+		/dev/pg0 c 97 0
+		/dev/pg1 c 97 1
+		/dev/pg2 c 97 2
+		/dev/pg3 c 97 3
+
+	(You'll need to change the 97 to something else if you use
+	the 'major' parameter to install the driver on a different
+	major number.)
+
+	The behaviour of the pg driver can be altered by setting
+	some parameters from the insmod command line.  The following
+	parameters are adjustable:
+
+	    drive0      These four arguments can be arrays of       
+	    drive1      1-6 integers as follows:
+	    drive2
+	    drive3      <prt>,<pro>,<uni>,<mod>,<slv>,<dly>
+
+			Where,
+
+		<prt>   is the base of the parallel port address for
+			the corresponding drive.  (required)
+
+		<pro>   is the protocol number for the adapter that
+			supports this drive.  These numbers are
+			logged by 'paride' when the protocol modules
+			are initialised.  (0 if not given)
+
+		<uni>   for those adapters that support chained
+			devices, this is the unit selector for the
+			chain of devices on the given port.  It should
+			be zero for devices that don't support chaining.
+			(0 if not given)
+
+		<mod>   this can be -1 to choose the best mode, or one
+			of the mode numbers supported by the adapter.
+			(-1 if not given)
+
+		<slv>   ATAPI devices can be jumpered to master or slave.
+			Set this to 0 to choose the master drive, 1 to
+			choose the slave, -1 (the default) to choose the
+			first drive found.
+
+		<dly>   some parallel ports require the driver to 
+			go more slowly.  -1 sets a default value that
+			should work with the chosen protocol.  Otherwise,
+			set this to a small integer, the larger it is
+			the slower the port i/o.  In some cases, setting
+			this to zero will speed up the device. (default -1)
+
+	    major	You may use this parameter to overide the
+			default major number (97) that this driver
+			will use.  Be sure to change the device
+			name as well.
+
+	    name	This parameter is a character string that
+			contains the name the kernel will use for this
+			device (in /proc output, for instance).
+			(default "pg").
+
+	    verbose     This parameter controls the amount of logging
+			that is done by the driver.  Set it to 0 for 
+			quiet operation, to 1 to enable progress
+			messages while the driver probes for devices,
+			or to 2 for full debug logging.  (default 0)
+
+	If this driver is built into the kernel, you can use 
+	the following command line parameters, with the same values
+	as the corresponding module parameters listed above:
+
+	    pg.drive0
+	    pg.drive1
+	    pg.drive2
+	    pg.drive3
+
+	In addition, you can use the parameter pg.disable to disable
+	the driver entirely.
+
+*/
+
+/* Changes:
+
+	1.01	GRG 1998.06.16	Bug fixes
+	1.02    GRG 1998.09.24  Added jumbo support
+
+*/
+
+#define PG_VERSION      "1.02"
+#define PG_MAJOR	97
+#define PG_NAME		"pg"
+#define PG_UNITS	4
+
+#ifndef PI_PG
+#define PI_PG	4
+#endif
+
+#include <linux/types.h>
+/* Here are things one can override from the insmod command.
+   Most are autoprobed by paride unless set here.  Verbose is 0
+   by default.
+
+*/
+
+static bool verbose = 0;
+static int major = PG_MAJOR;
+static char *name = PG_NAME;
+static int disable = 0;
+
+static int drive0[6] = { 0, 0, 0, -1, -1, -1 };
+static int drive1[6] = { 0, 0, 0, -1, -1, -1 };
+static int drive2[6] = { 0, 0, 0, -1, -1, -1 };
+static int drive3[6] = { 0, 0, 0, -1, -1, -1 };
+
+static int (*drives[4])[6] = {&drive0, &drive1, &drive2, &drive3};
+static int pg_drive_count;
+
+enum {D_PRT, D_PRO, D_UNI, D_MOD, D_SLV, D_DLY};
+
+/* end of parameters */
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/fs.h>
+#include <linux/delay.h>
+#include <linux/slab.h>
+#include <linux/mtio.h>
+#include <linux/pg.h>
+#include <linux/device.h>
+#include <linux/sched.h>	/* current, TASK_* */
+#include <linux/mutex.h>
+#include <linux/jiffies.h>
+
+#include <asm/uaccess.h>
+
+module_param(verbose, bool, 0644);
+module_param(major, int, 0);
+module_param(name, charp, 0);
+module_param_array(drive0, int, NULL, 0);
+module_param_array(drive1, int, NULL, 0);
+module_param_array(drive2, int, NULL, 0);
+module_param_array(drive3, int, NULL, 0);
+
+#include "paride.h"
+
+#define PG_SPIN_DEL     50	/* spin delay in micro-seconds  */
+#define PG_SPIN         200
+#define PG_TMO		HZ
+#define PG_RESET_TMO	10*HZ
+
+#define STAT_ERR        0x01
+#define STAT_INDEX      0x02
+#define STAT_ECC        0x04
+#define STAT_DRQ        0x08
+#define STAT_SEEK       0x10
+#define STAT_WRERR      0x20
+#define STAT_READY      0x40
+#define STAT_BUSY       0x80
+
+#define ATAPI_IDENTIFY		0x12
+
+static DEFINE_MUTEX(pg_mutex);
+static int pg_open(struct inode *inode, struct file *file);
+static int pg_release(struct inode *inode, struct file *file);
+static ssize_t pg_read(struct file *filp, char __user *buf,
+		       size_t count, loff_t * ppos);
+static ssize_t pg_write(struct file *filp, const char __user *buf,
+			size_t count, loff_t * ppos);
+static int pg_detect(void);
+
+#define PG_NAMELEN      8
+
+struct pg {
+	struct pi_adapter pia;	/* interface to paride layer */
+	struct pi_adapter *pi;
+	int busy;		/* write done, read expected */
+	int start;		/* jiffies at command start */
+	int dlen;		/* transfer size requested */
+	unsigned long timeout;	/* timeout requested */
+	int status;		/* last sense key */
+	int drive;		/* drive */
+	unsigned long access;	/* count of active opens ... */
+	int present;		/* device present ? */
+	char *bufptr;
+	char name[PG_NAMELEN];	/* pg0, pg1, ... */
+};
+
+static struct pg devices[PG_UNITS];
+
+static int pg_identify(struct pg *dev, int log);
+
+static char pg_scratch[512];	/* scratch block buffer */
+
+static struct class *pg_class;
+
+/* kernel glue structures */
+
+static const struct file_operations pg_fops = {
+	.owner = THIS_MODULE,
+	.read = pg_read,
+	.write = pg_write,
+	.open = pg_open,
+	.release = pg_release,
+	.llseek = noop_llseek,
+};
+
+static void pg_init_units(void)
+{
+	int unit;
+
+	pg_drive_count = 0;
+	for (unit = 0; unit < PG_UNITS; unit++) {
+		int *parm = *drives[unit];
+		struct pg *dev = &devices[unit];
+		dev->pi = &dev->pia;
+		clear_bit(0, &dev->access);
+		dev->busy = 0;
+		dev->present = 0;
+		dev->bufptr = NULL;
+		dev->drive = parm[D_SLV];
+		snprintf(dev->name, PG_NAMELEN, "%s%c", name, 'a'+unit);
+		if (parm[D_PRT])
+			pg_drive_count++;
+	}
+}
+
+static inline int status_reg(struct pg *dev)
+{
+	return pi_read_regr(dev->pi, 1, 6);
+}
+
+static inline int read_reg(struct pg *dev, int reg)
+{
+	return pi_read_regr(dev->pi, 0, reg);
+}
+
+static inline void write_reg(struct pg *dev, int reg, int val)
+{
+	pi_write_regr(dev->pi, 0, reg, val);
+}
+
+static inline u8 DRIVE(struct pg *dev)
+{
+	return 0xa0+0x10*dev->drive;
+}
+
+static void pg_sleep(int cs)
+{
+	schedule_timeout_interruptible(cs);
+}
+
+static int pg_wait(struct pg *dev, int go, int stop, unsigned long tmo, char *msg)
+{
+	int j, r, e, s, p, to;
+
+	dev->status = 0;
+
+	j = 0;
+	while ((((r = status_reg(dev)) & go) || (stop && (!(r & stop))))
+	       && time_before(jiffies, tmo)) {
+		if (j++ < PG_SPIN)
+			udelay(PG_SPIN_DEL);
+		else
+			pg_sleep(1);
+	}
+
+	to = time_after_eq(jiffies, tmo);
+
+	if ((r & (STAT_ERR & stop)) || to) {
+		s = read_reg(dev, 7);
+		e = read_reg(dev, 1);
+		p = read_reg(dev, 2);
+		if (verbose > 1)
+			printk("%s: %s: stat=0x%x err=0x%x phase=%d%s\n",
+			       dev->name, msg, s, e, p, to ? " timeout" : "");
+		if (to)
+			e |= 0x100;
+		dev->status = (e >> 4) & 0xff;
+		return -1;
+	}
+	return 0;
+}
+
+static int pg_command(struct pg *dev, char *cmd, int dlen, unsigned long tmo)
+{
+	int k;
+
+	pi_connect(dev->pi);
+
+	write_reg(dev, 6, DRIVE(dev));
+
+	if (pg_wait(dev, STAT_BUSY | STAT_DRQ, 0, tmo, "before command"))
+		goto fail;
+
+	write_reg(dev, 4, dlen % 256);
+	write_reg(dev, 5, dlen / 256);
+	write_reg(dev, 7, 0xa0);	/* ATAPI packet command */
+
+	if (pg_wait(dev, STAT_BUSY, STAT_DRQ, tmo, "command DRQ"))
+		goto fail;
+
+	if (read_reg(dev, 2) != 1) {
+		printk("%s: command phase error\n", dev->name);
+		goto fail;
+	}
+
+	pi_write_block(dev->pi, cmd, 12);
+
+	if (verbose > 1) {
+		printk("%s: Command sent, dlen=%d packet= ", dev->name, dlen);
+		for (k = 0; k < 12; k++)
+			printk("%02x ", cmd[k] & 0xff);
+		printk("\n");
+	}
+	return 0;
+fail:
+	pi_disconnect(dev->pi);
+	return -1;
+}
+
+static int pg_completion(struct pg *dev, char *buf, unsigned long tmo)
+{
+	int r, d, n, p;
+
+	r = pg_wait(dev, STAT_BUSY, STAT_DRQ | STAT_READY | STAT_ERR,
+		    tmo, "completion");
+
+	dev->dlen = 0;
+
+	while (read_reg(dev, 7) & STAT_DRQ) {
+		d = (read_reg(dev, 4) + 256 * read_reg(dev, 5));
+		n = ((d + 3) & 0xfffc);
+		p = read_reg(dev, 2) & 3;
+		if (p == 0)
+			pi_write_block(dev->pi, buf, n);
+		if (p == 2)
+			pi_read_block(dev->pi, buf, n);
+		if (verbose > 1)
+			printk("%s: %s %d bytes\n", dev->name,
+			       p ? "Read" : "Write", n);
+		dev->dlen += (1 - p) * d;
+		buf += d;
+		r = pg_wait(dev, STAT_BUSY, STAT_DRQ | STAT_READY | STAT_ERR,
+			    tmo, "completion");
+	}
+
+	pi_disconnect(dev->pi);
+
+	return r;
+}
+
+static int pg_reset(struct pg *dev)
+{
+	int i, k, err;
+	int expect[5] = { 1, 1, 1, 0x14, 0xeb };
+	int got[5];
+
+	pi_connect(dev->pi);
+	write_reg(dev, 6, DRIVE(dev));
+	write_reg(dev, 7, 8);
+
+	pg_sleep(20 * HZ / 1000);
+
+	k = 0;
+	while ((k++ < PG_RESET_TMO) && (status_reg(dev) & STAT_BUSY))
+		pg_sleep(1);
+
+	for (i = 0; i < 5; i++)
+		got[i] = read_reg(dev, i + 1);
+
+	err = memcmp(expect, got, sizeof(got)) ? -1 : 0;
+
+	if (verbose) {
+		printk("%s: Reset (%d) signature = ", dev->name, k);
+		for (i = 0; i < 5; i++)
+			printk("%3x", got[i]);
+		if (err)
+			printk(" (incorrect)");
+		printk("\n");
+	}
+
+	pi_disconnect(dev->pi);
+	return err;
+}
+
+static void xs(char *buf, char *targ, int len)
+{
+	char l = '\0';
+	int k;
+
+	for (k = 0; k < len; k++) {
+		char c = *buf++;
+		if (c != ' ' && c != l)
+			l = *targ++ = c;
+	}
+	if (l == ' ')
+		targ--;
+	*targ = '\0';
+}
+
+static int pg_identify(struct pg *dev, int log)
+{
+	int s;
+	char *ms[2] = { "master", "slave" };
+	char mf[10], id[18];
+	char id_cmd[12] = { ATAPI_IDENTIFY, 0, 0, 0, 36, 0, 0, 0, 0, 0, 0, 0 };
+	char buf[36];
+
+	s = pg_command(dev, id_cmd, 36, jiffies + PG_TMO);
+	if (s)
+		return -1;
+	s = pg_completion(dev, buf, jiffies + PG_TMO);
+	if (s)
+		return -1;
+
+	if (log) {
+		xs(buf + 8, mf, 8);
+		xs(buf + 16, id, 16);
+		printk("%s: %s %s, %s\n", dev->name, mf, id, ms[dev->drive]);
+	}
+
+	return 0;
+}
+
+/*
+ * returns  0, with id set if drive is detected
+ *	   -1, if drive detection failed
+ */
+static int pg_probe(struct pg *dev)
+{
+	if (dev->drive == -1) {
+		for (dev->drive = 0; dev->drive <= 1; dev->drive++)
+			if (!pg_reset(dev))
+				return pg_identify(dev, 1);
+	} else {
+		if (!pg_reset(dev))
+			return pg_identify(dev, 1);
+	}
+	return -1;
+}
+
+static int pg_detect(void)
+{
+	struct pg *dev = &devices[0];
+	int k, unit;
+
+	printk("%s: %s version %s, major %d\n", name, name, PG_VERSION, major);
+
+	k = 0;
+	if (pg_drive_count == 0) {
+		if (pi_init(dev->pi, 1, -1, -1, -1, -1, -1, pg_scratch,
+			    PI_PG, verbose, dev->name)) {
+			if (!pg_probe(dev)) {
+				dev->present = 1;
+				k++;
+			} else
+				pi_release(dev->pi);
+		}
+
+	} else
+		for (unit = 0; unit < PG_UNITS; unit++, dev++) {
+			int *parm = *drives[unit];
+			if (!parm[D_PRT])
+				continue;
+			if (pi_init(dev->pi, 0, parm[D_PRT], parm[D_MOD],
+				    parm[D_UNI], parm[D_PRO], parm[D_DLY],
+				    pg_scratch, PI_PG, verbose, dev->name)) {
+				if (!pg_probe(dev)) {
+					dev->present = 1;
+					k++;
+				} else
+					pi_release(dev->pi);
+			}
+		}
+
+	if (k)
+		return 0;
+
+	printk("%s: No ATAPI device detected\n", name);
+	return -1;
+}
+
+static int pg_open(struct inode *inode, struct file *file)
+{
+	int unit = iminor(inode) & 0x7f;
+	struct pg *dev = &devices[unit];
+	int ret = 0;
+
+	mutex_lock(&pg_mutex);
+	if ((unit >= PG_UNITS) || (!dev->present)) {
+		ret = -ENODEV;
+		goto out;
+	}
+
+	if (test_and_set_bit(0, &dev->access)) {
+		ret = -EBUSY;
+		goto out;
+	}
+
+	if (dev->busy) {
+		pg_reset(dev);
+		dev->busy = 0;
+	}
+
+	pg_identify(dev, (verbose > 1));
+
+	dev->bufptr = kmalloc(PG_MAX_DATA, GFP_KERNEL);
+	if (dev->bufptr == NULL) {
+		clear_bit(0, &dev->access);
+		printk("%s: buffer allocation failed\n", dev->name);
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	file->private_data = dev;
+
+out:
+	mutex_unlock(&pg_mutex);
+	return ret;
+}
+
+static int pg_release(struct inode *inode, struct file *file)
+{
+	struct pg *dev = file->private_data;
+
+	kfree(dev->bufptr);
+	dev->bufptr = NULL;
+	clear_bit(0, &dev->access);
+
+	return 0;
+}
+
+static ssize_t pg_write(struct file *filp, const char __user *buf, size_t count, loff_t *ppos)
+{
+	struct pg *dev = filp->private_data;
+	struct pg_write_hdr hdr;
+	int hs = sizeof (hdr);
+
+	if (dev->busy)
+		return -EBUSY;
+	if (count < hs)
+		return -EINVAL;
+
+	if (copy_from_user(&hdr, buf, hs))
+		return -EFAULT;
+
+	if (hdr.magic != PG_MAGIC)
+		return -EINVAL;
+	if (hdr.dlen < 0 || hdr.dlen > PG_MAX_DATA)
+		return -EINVAL;
+	if ((count - hs) > PG_MAX_DATA)
+		return -EINVAL;
+
+	if (hdr.func == PG_RESET) {
+		if (count != hs)
+			return -EINVAL;
+		if (pg_reset(dev))
+			return -EIO;
+		return count;
+	}
+
+	if (hdr.func != PG_COMMAND)
+		return -EINVAL;
+
+	dev->start = jiffies;
+	dev->timeout = hdr.timeout * HZ + HZ / 2 + jiffies;
+
+	if (pg_command(dev, hdr.packet, hdr.dlen, jiffies + PG_TMO)) {
+		if (dev->status & 0x10)
+			return -ETIME;
+		return -EIO;
+	}
+
+	dev->busy = 1;
+
+	if (copy_from_user(dev->bufptr, buf + hs, count - hs))
+		return -EFAULT;
+	return count;
+}
+
+static ssize_t pg_read(struct file *filp, char __user *buf, size_t count, loff_t *ppos)
+{
+	struct pg *dev = filp->private_data;
+	struct pg_read_hdr hdr;
+	int hs = sizeof (hdr);
+	int copy;
+
+	if (!dev->busy)
+		return -EINVAL;
+	if (count < hs)
+		return -EINVAL;
+
+	dev->busy = 0;
+
+	if (pg_completion(dev, dev->bufptr, dev->timeout))
+		if (dev->status & 0x10)
+			return -ETIME;
+
+	memset(&hdr, 0, sizeof(hdr));
+	hdr.magic = PG_MAGIC;
+	hdr.dlen = dev->dlen;
+	copy = 0;
+
+	if (hdr.dlen < 0) {
+		hdr.dlen = -1 * hdr.dlen;
+		copy = hdr.dlen;
+		if (copy > (count - hs))
+			copy = count - hs;
+	}
+
+	hdr.duration = (jiffies - dev->start + HZ / 2) / HZ;
+	hdr.scsi = dev->status & 0x0f;
+
+	if (copy_to_user(buf, &hdr, hs))
+		return -EFAULT;
+	if (copy > 0)
+		if (copy_to_user(buf + hs, dev->bufptr, copy))
+			return -EFAULT;
+	return copy + hs;
+}
+
+static int __init pg_init(void)
+{
+	int unit;
+	int err;
+
+	if (disable){
+		err = -EINVAL;
+		goto out;
+	}
+
+	pg_init_units();
+
+	if (pg_detect()) {
+		err = -ENODEV;
+		goto out;
+	}
+
+	err = register_chrdev(major, name, &pg_fops);
+	if (err < 0) {
+		printk("pg_init: unable to get major number %d\n", major);
+		for (unit = 0; unit < PG_UNITS; unit++) {
+			struct pg *dev = &devices[unit];
+			if (dev->present)
+				pi_release(dev->pi);
+		}
+		goto out;
+	}
+	major = err;	/* In case the user specified `major=0' (dynamic) */
+	pg_class = class_create(THIS_MODULE, "pg");
+	if (IS_ERR(pg_class)) {
+		err = PTR_ERR(pg_class);
+		goto out_chrdev;
+	}
+	for (unit = 0; unit < PG_UNITS; unit++) {
+		struct pg *dev = &devices[unit];
+		if (dev->present)
+			device_create(pg_class, NULL, MKDEV(major, unit), NULL,
+				      "pg%u", unit);
+	}
+	err = 0;
+	goto out;
+
+out_chrdev:
+	unregister_chrdev(major, "pg");
+out:
+	return err;
+}
+
+static void __exit pg_exit(void)
+{
+	int unit;
+
+	for (unit = 0; unit < PG_UNITS; unit++) {
+		struct pg *dev = &devices[unit];
+		if (dev->present)
+			device_destroy(pg_class, MKDEV(major, unit));
+	}
+	class_destroy(pg_class);
+	unregister_chrdev(major, name);
+
+	for (unit = 0; unit < PG_UNITS; unit++) {
+		struct pg *dev = &devices[unit];
+		if (dev->present)
+			pi_release(dev->pi);
+	}
+}
+
+MODULE_LICENSE("GPL");
+module_init(pg_init)
+module_exit(pg_exit)

drivers/block/paride/ppc6lnx.c

@@ -0,0 +1,726 @@
+/*
+	ppc6lnx.c (c) 2001 Micro Solutions Inc.
+		Released under the terms of the GNU General Public license
+
+	ppc6lnx.c  is a par of the protocol driver for the Micro Solutions
+		"BACKPACK" parallel port IDE adapter
+		(Works on Series 6 drives)
+
+*/
+
+//***************************************************************************
+
+// PPC 6 Code in C sanitized for LINUX
+// Original x86 ASM by Ron, Converted to C by Clive
+
+//***************************************************************************
+
+
+#define port_stb					1
+#define port_afd					2
+#define cmd_stb						port_afd
+#define port_init					4
+#define data_stb					port_init
+#define port_sel					8
+#define port_int					16
+#define port_dir					0x20
+
+#define ECR_EPP	0x80
+#define ECR_BI	0x20
+
+//***************************************************************************
+
+//  60772 Commands
+
+#define ACCESS_REG				0x00
+#define ACCESS_PORT				0x40
+
+#define ACCESS_READ				0x00
+#define ACCESS_WRITE			0x20
+
+//  60772 Command Prefix
+
+#define CMD_PREFIX_SET		0xe0		// Special command that modifies the next command's operation
+#define CMD_PREFIX_RESET	0xc0		// Resets current cmd modifier reg bits
+ #define PREFIX_IO16			0x01		// perform 16-bit wide I/O
+ #define PREFIX_FASTWR		0x04		// enable PPC mode fast-write
+ #define PREFIX_BLK				0x08		// enable block transfer mode
+
+// 60772 Registers
+
+#define REG_STATUS				0x00		// status register
+ #define STATUS_IRQA			0x01		// Peripheral IRQA line
+ #define STATUS_EEPROM_DO	0x40		// Serial EEPROM data bit
+#define REG_VERSION				0x01		// PPC version register (read)
+#define REG_HWCFG					0x02		// Hardware Config register
+#define REG_RAMSIZE				0x03		// Size of RAM Buffer
+ #define RAMSIZE_128K			0x02
+#define REG_EEPROM				0x06		// EEPROM control register
+ #define EEPROM_SK				0x01		// eeprom SK bit
+ #define EEPROM_DI				0x02		// eeprom DI bit
+ #define EEPROM_CS				0x04		// eeprom CS bit
+ #define EEPROM_EN				0x08		// eeprom output enable
+#define REG_BLKSIZE				0x08		// Block transfer len (24 bit)
+
+//***************************************************************************
+
+typedef struct ppc_storage {
+	u16	lpt_addr;				// LPT base address
+	u8	ppc_id;
+	u8	mode;						// operating mode
+					// 0 = PPC Uni SW
+					// 1 = PPC Uni FW
+					// 2 = PPC Bi SW
+					// 3 = PPC Bi FW
+					// 4 = EPP Byte
+					// 5 = EPP Word
+					// 6 = EPP Dword
+	u8	ppc_flags;
+	u8	org_data;				// original LPT data port contents
+	u8	org_ctrl;				// original LPT control port contents
+	u8	cur_ctrl;				// current control port contents
+} Interface;
+
+//***************************************************************************
+
+// ppc_flags
+
+#define fifo_wait					0x10
+
+//***************************************************************************
+
+// DONT CHANGE THESE LEST YOU BREAK EVERYTHING - BIT FIELD DEPENDENCIES
+
+#define PPCMODE_UNI_SW		0
+#define PPCMODE_UNI_FW		1
+#define PPCMODE_BI_SW			2
+#define PPCMODE_BI_FW			3
+#define PPCMODE_EPP_BYTE	4
+#define PPCMODE_EPP_WORD	5
+#define PPCMODE_EPP_DWORD	6
+
+//***************************************************************************
+
+static int ppc6_select(Interface *ppc);
+static void ppc6_deselect(Interface *ppc);
+static void ppc6_send_cmd(Interface *ppc, u8 cmd);
+static void ppc6_wr_data_byte(Interface *ppc, u8 data);
+static u8 ppc6_rd_data_byte(Interface *ppc);
+static u8 ppc6_rd_port(Interface *ppc, u8 port);
+static void ppc6_wr_port(Interface *ppc, u8 port, u8 data);
+static void ppc6_rd_data_blk(Interface *ppc, u8 *data, long count);
+static void ppc6_wait_for_fifo(Interface *ppc);
+static void ppc6_wr_data_blk(Interface *ppc, u8 *data, long count);
+static void ppc6_rd_port16_blk(Interface *ppc, u8 port, u8 *data, long length);
+static void ppc6_wr_port16_blk(Interface *ppc, u8 port, u8 *data, long length);
+static void ppc6_wr_extout(Interface *ppc, u8 regdata);
+static int ppc6_open(Interface *ppc);
+static void ppc6_close(Interface *ppc);
+
+//***************************************************************************
+
+static int ppc6_select(Interface *ppc)
+{
+	u8 i, j, k;
+
+	i = inb(ppc->lpt_addr + 1);
+
+	if (i & 1)
+		outb(i, ppc->lpt_addr + 1);
+
+	ppc->org_data = inb(ppc->lpt_addr);
+
+	ppc->org_ctrl = inb(ppc->lpt_addr + 2) & 0x5F; // readback ctrl
+
+	ppc->cur_ctrl = ppc->org_ctrl;
+
+	ppc->cur_ctrl |= port_sel;
+
+	outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+	if (ppc->org_data == 'b')
+		outb('x', ppc->lpt_addr);
+
+	outb('b', ppc->lpt_addr);
+	outb('p', ppc->lpt_addr);
+	outb(ppc->ppc_id, ppc->lpt_addr);
+	outb(~ppc->ppc_id,ppc->lpt_addr);
+
+	ppc->cur_ctrl &= ~port_sel;
+
+	outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+	ppc->cur_ctrl = (ppc->cur_ctrl & port_int) | port_init;
+
+	outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+	i = ppc->mode & 0x0C;
+
+	if (i == 0)
+		i = (ppc->mode & 2) | 1;
+
+	outb(i, ppc->lpt_addr);
+
+	ppc->cur_ctrl |= port_sel;
+
+	outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+	// DELAY
+
+	ppc->cur_ctrl |= port_afd;
+
+	outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+	j = ((i & 0x08) << 4) | ((i & 0x07) << 3);
+
+	k = inb(ppc->lpt_addr + 1) & 0xB8;
+
+	if (j == k)
+	{
+		ppc->cur_ctrl &= ~port_afd;
+
+		outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+		k = (inb(ppc->lpt_addr + 1) & 0xB8) ^ 0xB8;
+
+		if (j == k)
+		{
+			if (i & 4)	// EPP
+				ppc->cur_ctrl &= ~(port_sel | port_init);
+			else				// PPC/ECP
+				ppc->cur_ctrl &= ~port_sel;
+
+			outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+			return(1);
+		}
+	}
+
+	outb(ppc->org_ctrl, ppc->lpt_addr + 2);
+
+	outb(ppc->org_data, ppc->lpt_addr);
+
+	return(0); // FAIL
+}
+
+//***************************************************************************
+
+static void ppc6_deselect(Interface *ppc)
+{
+	if (ppc->mode & 4)	// EPP
+		ppc->cur_ctrl |= port_init;
+	else								// PPC/ECP
+		ppc->cur_ctrl |= port_sel;
+
+	outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+	outb(ppc->org_data, ppc->lpt_addr);
+
+	outb((ppc->org_ctrl | port_sel), ppc->lpt_addr + 2);
+
+	outb(ppc->org_ctrl, ppc->lpt_addr + 2);
+}
+
+//***************************************************************************
+
+static void ppc6_send_cmd(Interface *ppc, u8 cmd)
+{
+	switch(ppc->mode)
+	{
+		case PPCMODE_UNI_SW :
+		case PPCMODE_UNI_FW :
+		case PPCMODE_BI_SW :
+		case PPCMODE_BI_FW :
+		{
+			outb(cmd, ppc->lpt_addr);
+
+			ppc->cur_ctrl ^= cmd_stb;
+
+			outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+			break;
+		}
+
+		case PPCMODE_EPP_BYTE :
+		case PPCMODE_EPP_WORD :
+		case PPCMODE_EPP_DWORD :
+		{
+			outb(cmd, ppc->lpt_addr + 3);
+
+			break;
+		}
+	}
+}
+
+//***************************************************************************
+
+static void ppc6_wr_data_byte(Interface *ppc, u8 data)
+{
+	switch(ppc->mode)
+	{
+		case PPCMODE_UNI_SW :
+		case PPCMODE_UNI_FW :
+		case PPCMODE_BI_SW :
+		case PPCMODE_BI_FW :
+		{
+			outb(data, ppc->lpt_addr);
+
+			ppc->cur_ctrl ^= data_stb;
+
+			outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+			break;
+		}
+
+		case PPCMODE_EPP_BYTE :
+		case PPCMODE_EPP_WORD :
+		case PPCMODE_EPP_DWORD :
+		{
+			outb(data, ppc->lpt_addr + 4);
+
+			break;
+		}
+	}
+}
+
+//***************************************************************************
+
+static u8 ppc6_rd_data_byte(Interface *ppc)
+{
+	u8 data = 0;
+
+	switch(ppc->mode)
+	{
+		case PPCMODE_UNI_SW :
+		case PPCMODE_UNI_FW :
+		{
+			ppc->cur_ctrl = (ppc->cur_ctrl & ~port_stb) ^ data_stb;
+
+			outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+			// DELAY
+
+			data = inb(ppc->lpt_addr + 1);
+
+			data = ((data & 0x80) >> 1) | ((data & 0x38) >> 3);
+
+			ppc->cur_ctrl |= port_stb;
+
+			outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+			// DELAY
+
+			data |= inb(ppc->lpt_addr + 1) & 0xB8;
+
+			break;
+		}
+
+		case PPCMODE_BI_SW :
+		case PPCMODE_BI_FW :
+		{
+			ppc->cur_ctrl |= port_dir;
+
+			outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+			ppc->cur_ctrl = (ppc->cur_ctrl | port_stb) ^ data_stb;
+
+			outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+			data = inb(ppc->lpt_addr);
+
+			ppc->cur_ctrl &= ~port_stb;
+
+			outb(ppc->cur_ctrl,ppc->lpt_addr + 2);
+
+			ppc->cur_ctrl &= ~port_dir;
+
+			outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+			break;
+		}
+
+		case PPCMODE_EPP_BYTE :
+		case PPCMODE_EPP_WORD :
+		case PPCMODE_EPP_DWORD :
+		{
+			outb((ppc->cur_ctrl | port_dir),ppc->lpt_addr + 2);
+
+			data = inb(ppc->lpt_addr + 4);
+
+			outb(ppc->cur_ctrl,ppc->lpt_addr + 2);
+
+			break;
+		}
+	}
+
+	return(data);
+}
+
+//***************************************************************************
+
+static u8 ppc6_rd_port(Interface *ppc, u8 port)
+{
+	ppc6_send_cmd(ppc,(u8)(port | ACCESS_PORT | ACCESS_READ));
+
+	return(ppc6_rd_data_byte(ppc));
+}
+
+//***************************************************************************
+
+static void ppc6_wr_port(Interface *ppc, u8 port, u8 data)
+{
+	ppc6_send_cmd(ppc,(u8)(port | ACCESS_PORT | ACCESS_WRITE));
+
+	ppc6_wr_data_byte(ppc, data);
+}
+
+//***************************************************************************
+
+static void ppc6_rd_data_blk(Interface *ppc, u8 *data, long count)
+{
+	switch(ppc->mode)
+	{
+		case PPCMODE_UNI_SW :
+		case PPCMODE_UNI_FW :
+		{
+			while(count)
+			{
+				u8 d;
+
+				ppc->cur_ctrl = (ppc->cur_ctrl & ~port_stb) ^ data_stb;
+
+				outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+				// DELAY
+
+				d = inb(ppc->lpt_addr + 1);
+
+				d = ((d & 0x80) >> 1) | ((d & 0x38) >> 3);
+
+				ppc->cur_ctrl |= port_stb;
+
+				outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+				// DELAY
+
+				d |= inb(ppc->lpt_addr + 1) & 0xB8;
+
+				*data++ = d;
+				count--;
+			}
+
+			break;
+		}
+
+		case PPCMODE_BI_SW :
+		case PPCMODE_BI_FW :
+		{
+			ppc->cur_ctrl |= port_dir;
+
+			outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+			ppc->cur_ctrl |= port_stb;
+
+			while(count)
+			{
+				ppc->cur_ctrl ^= data_stb;
+
+				outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+				*data++ = inb(ppc->lpt_addr);
+				count--;
+			}
+
+			ppc->cur_ctrl &= ~port_stb;
+
+			outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+			ppc->cur_ctrl &= ~port_dir;
+
+			outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+			break;
+		}
+
+		case PPCMODE_EPP_BYTE :
+		{
+			outb((ppc->cur_ctrl | port_dir), ppc->lpt_addr + 2);
+
+			// DELAY
+
+			while(count)
+			{
+				*data++ = inb(ppc->lpt_addr + 4);
+				count--;
+			}
+
+			outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+			break;
+		}
+
+		case PPCMODE_EPP_WORD :
+		{
+			outb((ppc->cur_ctrl | port_dir), ppc->lpt_addr + 2);
+
+			// DELAY
+
+			while(count > 1)
+			{
+				*((u16 *)data) = inw(ppc->lpt_addr + 4);
+				data  += 2;
+				count -= 2;
+			}
+
+			while(count)
+			{
+				*data++ = inb(ppc->lpt_addr + 4);
+				count--;
+			}
+
+			outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+			break;
+		}
+
+		case PPCMODE_EPP_DWORD :
+		{
+			outb((ppc->cur_ctrl | port_dir),ppc->lpt_addr + 2);
+
+			// DELAY
+
+			while(count > 3)
+			{
+				*((u32 *)data) = inl(ppc->lpt_addr + 4);
+				data  += 4;
+				count -= 4;
+			}
+
+			while(count)
+			{
+				*data++ = inb(ppc->lpt_addr + 4);
+				count--;
+			}
+
+			outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+			break;
+		}
+	}
+
+}
+
+//***************************************************************************
+
+static void ppc6_wait_for_fifo(Interface *ppc)
+{
+	int i;
+
+	if (ppc->ppc_flags & fifo_wait)
+	{
+		for(i=0; i<20; i++)
+			inb(ppc->lpt_addr + 1);
+	}
+}
+
+//***************************************************************************
+
+static void ppc6_wr_data_blk(Interface *ppc, u8 *data, long count)
+{
+	switch(ppc->mode)
+	{
+		case PPCMODE_UNI_SW :
+		case PPCMODE_BI_SW :
+		{
+			while(count--)
+			{
+				outb(*data++, ppc->lpt_addr);
+
+				ppc->cur_ctrl ^= data_stb;
+
+				outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+			}
+
+			break;
+		}
+
+		case PPCMODE_UNI_FW :
+		case PPCMODE_BI_FW :
+		{
+			u8 this, last;
+
+			ppc6_send_cmd(ppc,(CMD_PREFIX_SET | PREFIX_FASTWR));
+
+			ppc->cur_ctrl |= port_stb;
+
+			outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+			last = *data;
+
+			outb(last, ppc->lpt_addr);
+
+			while(count)
+			{
+				this = *data++;
+				count--;
+
+				if (this == last)
+				{
+					ppc->cur_ctrl ^= data_stb;
+
+					outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+				}
+				else
+				{
+					outb(this, ppc->lpt_addr);
+
+					last = this;
+				}
+			}
+
+			ppc->cur_ctrl &= ~port_stb;
+
+			outb(ppc->cur_ctrl, ppc->lpt_addr + 2);
+
+			ppc6_send_cmd(ppc,(CMD_PREFIX_RESET | PREFIX_FASTWR));
+
+			break;
+		}
+
+		case PPCMODE_EPP_BYTE :
+		{
+			while(count)
+			{
+				outb(*data++,ppc->lpt_addr + 4);
+				count--;
+			}
+
+			ppc6_wait_for_fifo(ppc);
+
+			break;
+		}
+
+		case PPCMODE_EPP_WORD :
+		{
+			while(count > 1)
+			{
+				outw(*((u16 *)data),ppc->lpt_addr + 4);
+				data  += 2;
+				count -= 2;
+			}
+
+			while(count)
+			{
+				outb(*data++,ppc->lpt_addr + 4);
+				count--;
+			}
+
+			ppc6_wait_for_fifo(ppc);
+
+			break;
+		}
+
+		case PPCMODE_EPP_DWORD :
+		{
+			while(count > 3)
+			{
+				outl(*((u32 *)data),ppc->lpt_addr + 4);
+				data  += 4;
+				count -= 4;
+			}
+
+			while(count)
+			{
+				outb(*data++,ppc->lpt_addr + 4);
+				count--;
+			}
+
+			ppc6_wait_for_fifo(ppc);
+
+			break;
+		}
+	}
+}
+
+//***************************************************************************
+
+static void ppc6_rd_port16_blk(Interface *ppc, u8 port, u8 *data, long length)
+{
+	length = length << 1;
+
+	ppc6_send_cmd(ppc, (REG_BLKSIZE | ACCESS_REG | ACCESS_WRITE));
+	ppc6_wr_data_byte(ppc,(u8)length);
+	ppc6_wr_data_byte(ppc,(u8)(length >> 8));
+	ppc6_wr_data_byte(ppc,0);
+
+	ppc6_send_cmd(ppc, (CMD_PREFIX_SET | PREFIX_IO16 | PREFIX_BLK));
+
+	ppc6_send_cmd(ppc, (u8)(port | ACCESS_PORT | ACCESS_READ));
+
+	ppc6_rd_data_blk(ppc, data, length);
+
+	ppc6_send_cmd(ppc, (CMD_PREFIX_RESET | PREFIX_IO16 | PREFIX_BLK));
+}
+
+//***************************************************************************
+
+static void ppc6_wr_port16_blk(Interface *ppc, u8 port, u8 *data, long length)
+{
+	length = length << 1;
+
+	ppc6_send_cmd(ppc, (REG_BLKSIZE | ACCESS_REG | ACCESS_WRITE));
+	ppc6_wr_data_byte(ppc,(u8)length);
+	ppc6_wr_data_byte(ppc,(u8)(length >> 8));
+	ppc6_wr_data_byte(ppc,0);
+
+	ppc6_send_cmd(ppc, (CMD_PREFIX_SET | PREFIX_IO16 | PREFIX_BLK));
+
+	ppc6_send_cmd(ppc, (u8)(port | ACCESS_PORT | ACCESS_WRITE));
+
+	ppc6_wr_data_blk(ppc, data, length);
+
+	ppc6_send_cmd(ppc, (CMD_PREFIX_RESET | PREFIX_IO16 | PREFIX_BLK));
+}
+
+//***************************************************************************
+
+static void ppc6_wr_extout(Interface *ppc, u8 regdata)
+{
+	ppc6_send_cmd(ppc,(REG_VERSION | ACCESS_REG | ACCESS_WRITE));
+
+	ppc6_wr_data_byte(ppc, (u8)((regdata & 0x03) << 6));
+}
+
+//***************************************************************************
+
+static int ppc6_open(Interface *ppc)
+{
+	int ret;
+
+	ret = ppc6_select(ppc);
+
+	if (ret == 0)
+		return(ret);
+
+	ppc->ppc_flags &= ~fifo_wait;
+
+	ppc6_send_cmd(ppc, (ACCESS_REG | ACCESS_WRITE | REG_RAMSIZE));
+	ppc6_wr_data_byte(ppc, RAMSIZE_128K);
+
+	ppc6_send_cmd(ppc, (ACCESS_REG | ACCESS_READ | REG_VERSION));
+
+	if ((ppc6_rd_data_byte(ppc) & 0x3F) == 0x0C)
+		ppc->ppc_flags |= fifo_wait;
+
+	return(ret);
+}
+
+//***************************************************************************
+
+static void ppc6_close(Interface *ppc)
+{
+	ppc6_deselect(ppc);
+}
+
+//***************************************************************************
+

drivers/block/paride/pseudo.h

@@ -0,0 +1,102 @@
+/* 
+        pseudo.h    (c) 1997-8  Grant R. Guenther <grant@torque.net>
+                                Under the terms of the GNU General Public License.
+
+	This is the "pseudo-interrupt" logic for parallel port drivers.
+
+        This module is #included into each driver.  It makes one
+        function available:
+
+		ps_set_intr( void (*continuation)(void),
+			     int  (*ready)(void),
+			     int timeout,
+			     int nice )
+
+	Which will arrange for ready() to be evaluated frequently and
+	when either it returns true, or timeout jiffies have passed,
+	continuation() will be invoked.
+
+	If nice is 1, the test will done approximately once a
+	jiffy.  If nice is 0, the test will also be done whenever
+	the scheduler runs (by adding it to a task queue).  If
+	nice is greater than 1, the test will be done once every
+	(nice-1) jiffies. 
+
+*/
+
+/* Changes:
+
+	1.01	1998.05.03	Switched from cli()/sti() to spinlocks
+	1.02    1998.12.14      Added support for nice > 1
+*/
+	
+#define PS_VERSION	"1.02"
+
+#include <linux/sched.h>
+#include <linux/workqueue.h>
+
+static void ps_tq_int(struct work_struct *work);
+
+static void (* ps_continuation)(void);
+static int (* ps_ready)(void);
+static unsigned long ps_timeout;
+static int ps_tq_active = 0;
+static int ps_nice = 0;
+
+static DEFINE_SPINLOCK(ps_spinlock __attribute__((unused)));
+
+static DECLARE_DELAYED_WORK(ps_tq, ps_tq_int);
+
+static void ps_set_intr(void (*continuation)(void), 
+			int (*ready)(void),
+			int timeout, int nice)
+{
+	unsigned long	flags;
+
+	spin_lock_irqsave(&ps_spinlock,flags);
+
+	ps_continuation = continuation;
+	ps_ready = ready;
+	ps_timeout = jiffies + timeout;
+	ps_nice = nice;
+
+	if (!ps_tq_active) {
+		ps_tq_active = 1;
+		if (!ps_nice)
+			schedule_delayed_work(&ps_tq, 0);
+		else
+			schedule_delayed_work(&ps_tq, ps_nice-1);
+	}
+	spin_unlock_irqrestore(&ps_spinlock,flags);
+}
+
+static void ps_tq_int(struct work_struct *work)
+{
+	void (*con)(void);
+	unsigned long flags;
+
+	spin_lock_irqsave(&ps_spinlock,flags);
+
+	con = ps_continuation;
+	ps_tq_active = 0;
+
+	if (!con) {
+		spin_unlock_irqrestore(&ps_spinlock,flags);
+		return;
+	}
+	if (!ps_ready || ps_ready() || time_after_eq(jiffies, ps_timeout)) {
+		ps_continuation = NULL;
+		spin_unlock_irqrestore(&ps_spinlock,flags);
+		con();
+		return;
+	}
+	ps_tq_active = 1;
+	if (!ps_nice)
+		schedule_delayed_work(&ps_tq, 0);
+	else
+		schedule_delayed_work(&ps_tq, ps_nice-1);
+	spin_unlock_irqrestore(&ps_spinlock,flags);
+}
+
+/* end of pseudo.h */
+

drivers/block/ps3disk.c

@@ -0,0 +1,589 @@
+/*
+ * PS3 Disk Storage Driver
+ *
+ * Copyright (C) 2007 Sony Computer Entertainment Inc.
+ * Copyright 2007 Sony Corp.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published
+ * by the Free Software Foundation; version 2 of the License.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+ */
+
+#include <linux/ata.h>
+#include <linux/blkdev.h>
+#include <linux/slab.h>
+#include <linux/module.h>
+
+#include <asm/lv1call.h>
+#include <asm/ps3stor.h>
+#include <asm/firmware.h>
+
+
+#define DEVICE_NAME		"ps3disk"
+
+#define BOUNCE_SIZE		(64*1024)
+
+#define PS3DISK_MAX_DISKS	16
+#define PS3DISK_MINORS		16
+
+
+#define PS3DISK_NAME		"ps3d%c"
+
+
+struct ps3disk_private {
+	spinlock_t lock;		/* Request queue spinlock */
+	struct request_queue *queue;
+	struct gendisk *gendisk;
+	unsigned int blocking_factor;
+	struct request *req;
+	u64 raw_capacity;
+	unsigned char model[ATA_ID_PROD_LEN+1];
+};
+
+
+#define LV1_STORAGE_SEND_ATA_COMMAND	(2)
+#define LV1_STORAGE_ATA_HDDOUT		(0x23)
+
+struct lv1_ata_cmnd_block {
+	u16	features;
+	u16	sector_count;
+	u16	LBA_low;
+	u16	LBA_mid;
+	u16	LBA_high;
+	u8	device;
+	u8	command;
+	u32	is_ext;
+	u32	proto;
+	u32	in_out;
+	u32	size;
+	u64	buffer;
+	u32	arglen;
+};
+
+enum lv1_ata_proto {
+	NON_DATA_PROTO     = 0,
+	PIO_DATA_IN_PROTO  = 1,
+	PIO_DATA_OUT_PROTO = 2,
+	DMA_PROTO = 3
+};
+
+enum lv1_ata_in_out {
+	DIR_WRITE = 0,			/* memory -> device */
+	DIR_READ = 1			/* device -> memory */
+};
+
+static int ps3disk_major;
+
+
+static const struct block_device_operations ps3disk_fops = {
+	.owner		= THIS_MODULE,
+};
+
+
+static void ps3disk_scatter_gather(struct ps3_storage_device *dev,
+				   struct request *req, int gather)
+{
+	unsigned int offset = 0;
+	struct req_iterator iter;
+	struct bio_vec bvec;
+	unsigned int i = 0;
+	size_t size;
+	void *buf;
+
+	rq_for_each_segment(bvec, req, iter) {
+		unsigned long flags;
+		dev_dbg(&dev->sbd.core, "%s:%u: bio %u: %u sectors from %lu\n",
+			__func__, __LINE__, i, bio_sectors(iter.bio),
+			iter.bio->bi_iter.bi_sector);
+
+		size = bvec.bv_len;
+		buf = bvec_kmap_irq(&bvec, &flags);
+		if (gather)
+			memcpy(dev->bounce_buf+offset, buf, size);
+		else
+			memcpy(buf, dev->bounce_buf+offset, size);
+		offset += size;
+		flush_kernel_dcache_page(bvec.bv_page);
+		bvec_kunmap_irq(buf, &flags);
+		i++;
+	}
+}
+
+static int ps3disk_submit_request_sg(struct ps3_storage_device *dev,
+				     struct request *req)
+{
+	struct ps3disk_private *priv = ps3_system_bus_get_drvdata(&dev->sbd);
+	int write = rq_data_dir(req), res;
+	const char *op = write ? "write" : "read";
+	u64 start_sector, sectors;
+	unsigned int region_id = dev->regions[dev->region_idx].id;
+
+#ifdef DEBUG
+	unsigned int n = 0;
+	struct bio_vec bv;
+	struct req_iterator iter;
+
+	rq_for_each_segment(bv, req, iter)
+		n++;
+	dev_dbg(&dev->sbd.core,
+		"%s:%u: %s req has %u bvecs for %u sectors\n",
+		__func__, __LINE__, op, n, blk_rq_sectors(req));
+#endif
+
+	start_sector = blk_rq_pos(req) * priv->blocking_factor;
+	sectors = blk_rq_sectors(req) * priv->blocking_factor;
+	dev_dbg(&dev->sbd.core, "%s:%u: %s %llu sectors starting at %llu\n",
+		__func__, __LINE__, op, sectors, start_sector);
+
+	if (write) {
+		ps3disk_scatter_gather(dev, req, 1);
+
+		res = lv1_storage_write(dev->sbd.dev_id, region_id,
+					start_sector, sectors, 0,
+					dev->bounce_lpar, &dev->tag);
+	} else {
+		res = lv1_storage_read(dev->sbd.dev_id, region_id,
+				       start_sector, sectors, 0,
+				       dev->bounce_lpar, &dev->tag);
+	}
+	if (res) {
+		dev_err(&dev->sbd.core, "%s:%u: %s failed %d\n", __func__,
+			__LINE__, op, res);
+		__blk_end_request_all(req, -EIO);
+		return 0;
+	}
+
+	priv->req = req;
+	return 1;
+}
+
+static int ps3disk_submit_flush_request(struct ps3_storage_device *dev,
+					struct request *req)
+{
+	struct ps3disk_private *priv = ps3_system_bus_get_drvdata(&dev->sbd);
+	u64 res;
+
+	dev_dbg(&dev->sbd.core, "%s:%u: flush request\n", __func__, __LINE__);
+
+	res = lv1_storage_send_device_command(dev->sbd.dev_id,
+					      LV1_STORAGE_ATA_HDDOUT, 0, 0, 0,
+					      0, &dev->tag);
+	if (res) {
+		dev_err(&dev->sbd.core, "%s:%u: sync cache failed 0x%llx\n",
+			__func__, __LINE__, res);
+		__blk_end_request_all(req, -EIO);
+		return 0;
+	}
+
+	priv->req = req;
+	return 1;
+}
+
+static void ps3disk_do_request(struct ps3_storage_device *dev,
+			       struct request_queue *q)
+{
+	struct request *req;
+
+	dev_dbg(&dev->sbd.core, "%s:%u\n", __func__, __LINE__);
+
+	while ((req = blk_fetch_request(q))) {
+		if (req->cmd_flags & REQ_FLUSH) {
+			if (ps3disk_submit_flush_request(dev, req))
+				break;
+		} else if (req->cmd_type == REQ_TYPE_FS) {
+			if (ps3disk_submit_request_sg(dev, req))
+				break;
+		} else {
+			blk_dump_rq_flags(req, DEVICE_NAME " bad request");
+			__blk_end_request_all(req, -EIO);
+			continue;
+		}
+	}
+}
+
+static void ps3disk_request(struct request_queue *q)
+{
+	struct ps3_storage_device *dev = q->queuedata;
+	struct ps3disk_private *priv = ps3_system_bus_get_drvdata(&dev->sbd);
+
+	if (priv->req) {
+		dev_dbg(&dev->sbd.core, "%s:%u busy\n", __func__, __LINE__);
+		return;
+	}
+
+	ps3disk_do_request(dev, q);
+}
+
+static irqreturn_t ps3disk_interrupt(int irq, void *data)
+{
+	struct ps3_storage_device *dev = data;
+	struct ps3disk_private *priv;
+	struct request *req;
+	int res, read, error;
+	u64 tag, status;
+	const char *op;
+
+	res = lv1_storage_get_async_status(dev->sbd.dev_id, &tag, &status);
+
+	if (tag != dev->tag)
+		dev_err(&dev->sbd.core,
+			"%s:%u: tag mismatch, got %llx, expected %llx\n",
+			__func__, __LINE__, tag, dev->tag);
+
+	if (res) {
+		dev_err(&dev->sbd.core, "%s:%u: res=%d status=0x%llx\n",
+			__func__, __LINE__, res, status);
+		return IRQ_HANDLED;
+	}
+
+	priv = ps3_system_bus_get_drvdata(&dev->sbd);
+	req = priv->req;
+	if (!req) {
+		dev_dbg(&dev->sbd.core,
+			"%s:%u non-block layer request completed\n", __func__,
+			__LINE__);
+		dev->lv1_status = status;
+		complete(&dev->done);
+		return IRQ_HANDLED;
+	}
+
+	if (req->cmd_flags & REQ_FLUSH) {
+		read = 0;
+		op = "flush";
+	} else {
+		read = !rq_data_dir(req);
+		op = read ? "read" : "write";
+	}
+	if (status) {
+		dev_dbg(&dev->sbd.core, "%s:%u: %s failed 0x%llx\n", __func__,
+			__LINE__, op, status);
+		error = -EIO;
+	} else {
+		dev_dbg(&dev->sbd.core, "%s:%u: %s completed\n", __func__,
+			__LINE__, op);
+		error = 0;
+		if (read)
+			ps3disk_scatter_gather(dev, req, 0);
+	}
+
+	spin_lock(&priv->lock);
+	__blk_end_request_all(req, error);
+	priv->req = NULL;
+	ps3disk_do_request(dev, priv->queue);
+	spin_unlock(&priv->lock);
+
+	return IRQ_HANDLED;
+}
+
+static int ps3disk_sync_cache(struct ps3_storage_device *dev)
+{
+	u64 res;
+
+	dev_dbg(&dev->sbd.core, "%s:%u: sync cache\n", __func__, __LINE__);
+
+	res = ps3stor_send_command(dev, LV1_STORAGE_ATA_HDDOUT, 0, 0, 0, 0);
+	if (res) {
+		dev_err(&dev->sbd.core, "%s:%u: sync cache failed 0x%llx\n",
+			__func__, __LINE__, res);
+		return -EIO;
+	}
+	return 0;
+}
+
+
+/* ATA helpers copied from drivers/ata/libata-core.c */
+
+static void swap_buf_le16(u16 *buf, unsigned int buf_words)
+{
+#ifdef __BIG_ENDIAN
+	unsigned int i;
+
+	for (i = 0; i < buf_words; i++)
+		buf[i] = le16_to_cpu(buf[i]);
+#endif /* __BIG_ENDIAN */
+}
+
+static u64 ata_id_n_sectors(const u16 *id)
+{
+	if (ata_id_has_lba(id)) {
+		if (ata_id_has_lba48(id))
+			return ata_id_u64(id, 100);
+		else
+			return ata_id_u32(id, 60);
+	} else {
+		if (ata_id_current_chs_valid(id))
+			return ata_id_u32(id, 57);
+		else
+			return id[1] * id[3] * id[6];
+	}
+}
+
+static void ata_id_string(const u16 *id, unsigned char *s, unsigned int ofs,
+			  unsigned int len)
+{
+	unsigned int c;
+
+	while (len > 0) {
+		c = id[ofs] >> 8;
+		*s = c;
+		s++;
+
+		c = id[ofs] & 0xff;
+		*s = c;
+		s++;
+
+		ofs++;
+		len -= 2;
+	}
+}
+
+static void ata_id_c_string(const u16 *id, unsigned char *s, unsigned int ofs,
+			    unsigned int len)
+{
+	unsigned char *p;
+
+	WARN_ON(!(len & 1));
+
+	ata_id_string(id, s, ofs, len - 1);
+
+	p = s + strnlen(s, len - 1);
+	while (p > s && p[-1] == ' ')
+		p--;
+	*p = '\0';
+}
+
+static int ps3disk_identify(struct ps3_storage_device *dev)
+{
+	struct ps3disk_private *priv = ps3_system_bus_get_drvdata(&dev->sbd);
+	struct lv1_ata_cmnd_block ata_cmnd;
+	u16 *id = dev->bounce_buf;
+	u64 res;
+
+	dev_dbg(&dev->sbd.core, "%s:%u: identify disk\n", __func__, __LINE__);
+
+	memset(&ata_cmnd, 0, sizeof(struct lv1_ata_cmnd_block));
+	ata_cmnd.command = ATA_CMD_ID_ATA;
+	ata_cmnd.sector_count = 1;
+	ata_cmnd.size = ata_cmnd.arglen = ATA_ID_WORDS * 2;
+	ata_cmnd.buffer = dev->bounce_lpar;
+	ata_cmnd.proto = PIO_DATA_IN_PROTO;
+	ata_cmnd.in_out = DIR_READ;
+
+	res = ps3stor_send_command(dev, LV1_STORAGE_SEND_ATA_COMMAND,
+				   ps3_mm_phys_to_lpar(__pa(&ata_cmnd)),
+				   sizeof(ata_cmnd), ata_cmnd.buffer,
+				   ata_cmnd.arglen);
+	if (res) {
+		dev_err(&dev->sbd.core, "%s:%u: identify disk failed 0x%llx\n",
+			__func__, __LINE__, res);
+		return -EIO;
+	}
+
+	swap_buf_le16(id, ATA_ID_WORDS);
+
+	/* All we're interested in are raw capacity and model name */
+	priv->raw_capacity = ata_id_n_sectors(id);
+	ata_id_c_string(id, priv->model, ATA_ID_PROD, sizeof(priv->model));
+	return 0;
+}
+
+static unsigned long ps3disk_mask;
+
+static DEFINE_MUTEX(ps3disk_mask_mutex);
+
+static int ps3disk_probe(struct ps3_system_bus_device *_dev)
+{
+	struct ps3_storage_device *dev = to_ps3_storage_device(&_dev->core);
+	struct ps3disk_private *priv;
+	int error;
+	unsigned int devidx;
+	struct request_queue *queue;
+	struct gendisk *gendisk;
+
+	if (dev->blk_size < 512) {
+		dev_err(&dev->sbd.core,
+			"%s:%u: cannot handle block size %llu\n", __func__,
+			__LINE__, dev->blk_size);
+		return -EINVAL;
+	}
+
+	BUILD_BUG_ON(PS3DISK_MAX_DISKS > BITS_PER_LONG);
+	mutex_lock(&ps3disk_mask_mutex);
+	devidx = find_first_zero_bit(&ps3disk_mask, PS3DISK_MAX_DISKS);
+	if (devidx >= PS3DISK_MAX_DISKS) {
+		dev_err(&dev->sbd.core, "%s:%u: Too many disks\n", __func__,
+			__LINE__);
+		mutex_unlock(&ps3disk_mask_mutex);
+		return -ENOSPC;
+	}
+	__set_bit(devidx, &ps3disk_mask);
+	mutex_unlock(&ps3disk_mask_mutex);
+
+	priv = kzalloc(sizeof(*priv), GFP_KERNEL);
+	if (!priv) {
+		error = -ENOMEM;
+		goto fail;
+	}
+
+	ps3_system_bus_set_drvdata(_dev, priv);
+	spin_lock_init(&priv->lock);
+
+	dev->bounce_size = BOUNCE_SIZE;
+	dev->bounce_buf = kmalloc(BOUNCE_SIZE, GFP_DMA);
+	if (!dev->bounce_buf) {
+		error = -ENOMEM;
+		goto fail_free_priv;
+	}
+
+	error = ps3stor_setup(dev, ps3disk_interrupt);
+	if (error)
+		goto fail_free_bounce;
+
+	ps3disk_identify(dev);
+
+	queue = blk_init_queue(ps3disk_request, &priv->lock);
+	if (!queue) {
+		dev_err(&dev->sbd.core, "%s:%u: blk_init_queue failed\n",
+			__func__, __LINE__);
+		error = -ENOMEM;
+		goto fail_teardown;
+	}
+
+	priv->queue = queue;
+	queue->queuedata = dev;
+
+	blk_queue_bounce_limit(queue, BLK_BOUNCE_HIGH);
+
+	blk_queue_max_hw_sectors(queue, dev->bounce_size >> 9);
+	blk_queue_segment_boundary(queue, -1UL);
+	blk_queue_dma_alignment(queue, dev->blk_size-1);
+	blk_queue_logical_block_size(queue, dev->blk_size);
+
+	blk_queue_flush(queue, REQ_FLUSH);
+
+	blk_queue_max_segments(queue, -1);
+	blk_queue_max_segment_size(queue, dev->bounce_size);
+
+	gendisk = alloc_disk(PS3DISK_MINORS);
+	if (!gendisk) {
+		dev_err(&dev->sbd.core, "%s:%u: alloc_disk failed\n", __func__,
+			__LINE__);
+		error = -ENOMEM;
+		goto fail_cleanup_queue;
+	}
+
+	priv->gendisk = gendisk;
+	gendisk->major = ps3disk_major;
+	gendisk->first_minor = devidx * PS3DISK_MINORS;
+	gendisk->fops = &ps3disk_fops;
+	gendisk->queue = queue;
+	gendisk->private_data = dev;
+	gendisk->driverfs_dev = &dev->sbd.core;
+	snprintf(gendisk->disk_name, sizeof(gendisk->disk_name), PS3DISK_NAME,
+		 devidx+'a');
+	priv->blocking_factor = dev->blk_size >> 9;
+	set_capacity(gendisk,
+		     dev->regions[dev->region_idx].size*priv->blocking_factor);
+
+	dev_info(&dev->sbd.core,
+		 "%s is a %s (%llu MiB total, %lu MiB for OtherOS)\n",
+		 gendisk->disk_name, priv->model, priv->raw_capacity >> 11,
+		 get_capacity(gendisk) >> 11);
+
+	add_disk(gendisk);
+	return 0;
+
+fail_cleanup_queue:
+	blk_cleanup_queue(queue);
+fail_teardown:
+	ps3stor_teardown(dev);
+fail_free_bounce:
+	kfree(dev->bounce_buf);
+fail_free_priv:
+	kfree(priv);
+	ps3_system_bus_set_drvdata(_dev, NULL);
+fail:
+	mutex_lock(&ps3disk_mask_mutex);
+	__clear_bit(devidx, &ps3disk_mask);
+	mutex_unlock(&ps3disk_mask_mutex);
+	return error;
+}
+
+static int ps3disk_remove(struct ps3_system_bus_device *_dev)
+{
+	struct ps3_storage_device *dev = to_ps3_storage_device(&_dev->core);
+	struct ps3disk_private *priv = ps3_system_bus_get_drvdata(&dev->sbd);
+
+	mutex_lock(&ps3disk_mask_mutex);
+	__clear_bit(MINOR(disk_devt(priv->gendisk)) / PS3DISK_MINORS,
+		    &ps3disk_mask);
+	mutex_unlock(&ps3disk_mask_mutex);
+	del_gendisk(priv->gendisk);
+	blk_cleanup_queue(priv->queue);
+	put_disk(priv->gendisk);
+	dev_notice(&dev->sbd.core, "Synchronizing disk cache\n");
+	ps3disk_sync_cache(dev);
+	ps3stor_teardown(dev);
+	kfree(dev->bounce_buf);
+	kfree(priv);
+	ps3_system_bus_set_drvdata(_dev, NULL);
+	return 0;
+}
+
+static struct ps3_system_bus_driver ps3disk = {
+	.match_id	= PS3_MATCH_ID_STOR_DISK,
+	.core.name	= DEVICE_NAME,
+	.core.owner	= THIS_MODULE,
+	.probe		= ps3disk_probe,
+	.remove		= ps3disk_remove,
+	.shutdown	= ps3disk_remove,
+};
+
+
+static int __init ps3disk_init(void)
+{
+	int error;
+
+	if (!firmware_has_feature(FW_FEATURE_PS3_LV1))
+		return -ENODEV;
+
+	error = register_blkdev(0, DEVICE_NAME);
+	if (error <= 0) {
+		printk(KERN_ERR "%s:%u: register_blkdev failed %d\n", __func__,
+		       __LINE__, error);
+		return error;
+	}
+	ps3disk_major = error;
+
+	pr_info("%s:%u: registered block device major %d\n", __func__,
+		__LINE__, ps3disk_major);
+
+	error = ps3_system_bus_driver_register(&ps3disk);
+	if (error)
+		unregister_blkdev(ps3disk_major, DEVICE_NAME);
+
+	return error;
+}
+
+static void __exit ps3disk_exit(void)
+{
+	ps3_system_bus_driver_unregister(&ps3disk);
+	unregister_blkdev(ps3disk_major, DEVICE_NAME);
+}
+
+module_init(ps3disk_init);
+module_exit(ps3disk_exit);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("PS3 Disk Storage Driver");
+MODULE_AUTHOR("Sony Corporation");
+MODULE_ALIAS(PS3_MODULE_ALIAS_STOR_DISK);

drivers/block/ps3vram.c

@@ -0,0 +1,878 @@
+/*
+ * ps3vram - Use extra PS3 video ram as MTD block device.
+ *
+ * Copyright 2009 Sony Corporation
+ *
+ * Based on the MTD ps3vram driver, which is
+ * Copyright (c) 2007-2008 Jim Paris <jim@jtan.com>
+ * Added support RSX DMA Vivien Chappelier <vivien.chappelier@free.fr>
+ */
+
+#include <linux/blkdev.h>
+#include <linux/delay.h>
+#include <linux/module.h>
+#include <linux/proc_fs.h>
+#include <linux/seq_file.h>
+#include <linux/slab.h>
+
+#include <asm/cell-regs.h>
+#include <asm/firmware.h>
+#include <asm/lv1call.h>
+#include <asm/ps3.h>
+#include <asm/ps3gpu.h>
+
+
+#define DEVICE_NAME		"ps3vram"
+
+
+#define XDR_BUF_SIZE (2 * 1024 * 1024) /* XDR buffer (must be 1MiB aligned) */
+#define XDR_IOIF 0x0c000000
+
+#define FIFO_BASE XDR_IOIF
+#define FIFO_SIZE (64 * 1024)
+
+#define DMA_PAGE_SIZE (4 * 1024)
+
+#define CACHE_PAGE_SIZE (256 * 1024)
+#define CACHE_PAGE_COUNT ((XDR_BUF_SIZE - FIFO_SIZE) / CACHE_PAGE_SIZE)
+
+#define CACHE_OFFSET CACHE_PAGE_SIZE
+#define FIFO_OFFSET 0
+
+#define CTRL_PUT 0x10
+#define CTRL_GET 0x11
+#define CTRL_TOP 0x15
+
+#define UPLOAD_SUBCH	1
+#define DOWNLOAD_SUBCH	2
+
+#define NV_MEMORY_TO_MEMORY_FORMAT_OFFSET_IN	0x0000030c
+#define NV_MEMORY_TO_MEMORY_FORMAT_NOTIFY	0x00000104
+
+#define CACHE_PAGE_PRESENT 1
+#define CACHE_PAGE_DIRTY   2
+
+struct ps3vram_tag {
+	unsigned int address;
+	unsigned int flags;
+};
+
+struct ps3vram_cache {
+	unsigned int page_count;
+	unsigned int page_size;
+	struct ps3vram_tag *tags;
+	unsigned int hit;
+	unsigned int miss;
+};
+
+struct ps3vram_priv {
+	struct request_queue *queue;
+	struct gendisk *gendisk;
+
+	u64 size;
+
+	u64 memory_handle;
+	u64 context_handle;
+	u32 *ctrl;
+	void *reports;
+	u8 *xdr_buf;
+
+	u32 *fifo_base;
+	u32 *fifo_ptr;
+
+	struct ps3vram_cache cache;
+
+	spinlock_t lock;	/* protecting list of bios */
+	struct bio_list list;
+};
+
+
+static int ps3vram_major;
+
+
+static const struct block_device_operations ps3vram_fops = {
+	.owner		= THIS_MODULE,
+};
+
+
+#define DMA_NOTIFIER_HANDLE_BASE 0x66604200 /* first DMA notifier handle */
+#define DMA_NOTIFIER_OFFSET_BASE 0x1000     /* first DMA notifier offset */
+#define DMA_NOTIFIER_SIZE        0x40
+#define NOTIFIER 7	/* notifier used for completion report */
+
+static char *size = "256M";
+module_param(size, charp, 0);
+MODULE_PARM_DESC(size, "memory size");
+
+static u32 *ps3vram_get_notifier(void *reports, int notifier)
+{
+	return reports + DMA_NOTIFIER_OFFSET_BASE +
+	       DMA_NOTIFIER_SIZE * notifier;
+}
+
+static void ps3vram_notifier_reset(struct ps3_system_bus_device *dev)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+	u32 *notify = ps3vram_get_notifier(priv->reports, NOTIFIER);
+	int i;
+
+	for (i = 0; i < 4; i++)
+		notify[i] = 0xffffffff;
+}
+
+static int ps3vram_notifier_wait(struct ps3_system_bus_device *dev,
+				 unsigned int timeout_ms)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+	u32 *notify = ps3vram_get_notifier(priv->reports, NOTIFIER);
+	unsigned long timeout;
+
+	for (timeout = 20; timeout; timeout--) {
+		if (!notify[3])
+			return 0;
+		udelay(10);
+	}
+
+	timeout = jiffies + msecs_to_jiffies(timeout_ms);
+
+	do {
+		if (!notify[3])
+			return 0;
+		msleep(1);
+	} while (time_before(jiffies, timeout));
+
+	return -ETIMEDOUT;
+}
+
+static void ps3vram_init_ring(struct ps3_system_bus_device *dev)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+
+	priv->ctrl[CTRL_PUT] = FIFO_BASE + FIFO_OFFSET;
+	priv->ctrl[CTRL_GET] = FIFO_BASE + FIFO_OFFSET;
+}
+
+static int ps3vram_wait_ring(struct ps3_system_bus_device *dev,
+			     unsigned int timeout_ms)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+	unsigned long timeout = jiffies + msecs_to_jiffies(timeout_ms);
+
+	do {
+		if (priv->ctrl[CTRL_PUT] == priv->ctrl[CTRL_GET])
+			return 0;
+		msleep(1);
+	} while (time_before(jiffies, timeout));
+
+	dev_warn(&dev->core, "FIFO timeout (%08x/%08x/%08x)\n",
+		 priv->ctrl[CTRL_PUT], priv->ctrl[CTRL_GET],
+		 priv->ctrl[CTRL_TOP]);
+
+	return -ETIMEDOUT;
+}
+
+static void ps3vram_out_ring(struct ps3vram_priv *priv, u32 data)
+{
+	*(priv->fifo_ptr)++ = data;
+}
+
+static void ps3vram_begin_ring(struct ps3vram_priv *priv, u32 chan, u32 tag,
+			       u32 size)
+{
+	ps3vram_out_ring(priv, (size << 18) | (chan << 13) | tag);
+}
+
+static void ps3vram_rewind_ring(struct ps3_system_bus_device *dev)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+	int status;
+
+	ps3vram_out_ring(priv, 0x20000000 | (FIFO_BASE + FIFO_OFFSET));
+
+	priv->ctrl[CTRL_PUT] = FIFO_BASE + FIFO_OFFSET;
+
+	/* asking the HV for a blit will kick the FIFO */
+	status = lv1_gpu_fb_blit(priv->context_handle, 0, 0, 0, 0);
+	if (status)
+		dev_err(&dev->core, "%s: lv1_gpu_fb_blit failed %d\n",
+			__func__, status);
+
+	priv->fifo_ptr = priv->fifo_base;
+}
+
+static void ps3vram_fire_ring(struct ps3_system_bus_device *dev)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+	int status;
+
+	mutex_lock(&ps3_gpu_mutex);
+
+	priv->ctrl[CTRL_PUT] = FIFO_BASE + FIFO_OFFSET +
+			       (priv->fifo_ptr - priv->fifo_base) * sizeof(u32);
+
+	/* asking the HV for a blit will kick the FIFO */
+	status = lv1_gpu_fb_blit(priv->context_handle, 0, 0, 0, 0);
+	if (status)
+		dev_err(&dev->core, "%s: lv1_gpu_fb_blit failed %d\n",
+			__func__, status);
+
+	if ((priv->fifo_ptr - priv->fifo_base) * sizeof(u32) >
+	    FIFO_SIZE - 1024) {
+		dev_dbg(&dev->core, "FIFO full, rewinding\n");
+		ps3vram_wait_ring(dev, 200);
+		ps3vram_rewind_ring(dev);
+	}
+
+	mutex_unlock(&ps3_gpu_mutex);
+}
+
+static void ps3vram_bind(struct ps3_system_bus_device *dev)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+
+	ps3vram_begin_ring(priv, UPLOAD_SUBCH, 0, 1);
+	ps3vram_out_ring(priv, 0x31337303);
+	ps3vram_begin_ring(priv, UPLOAD_SUBCH, 0x180, 3);
+	ps3vram_out_ring(priv, DMA_NOTIFIER_HANDLE_BASE + NOTIFIER);
+	ps3vram_out_ring(priv, 0xfeed0001);	/* DMA system RAM instance */
+	ps3vram_out_ring(priv, 0xfeed0000);     /* DMA video RAM instance */
+
+	ps3vram_begin_ring(priv, DOWNLOAD_SUBCH, 0, 1);
+	ps3vram_out_ring(priv, 0x3137c0de);
+	ps3vram_begin_ring(priv, DOWNLOAD_SUBCH, 0x180, 3);
+	ps3vram_out_ring(priv, DMA_NOTIFIER_HANDLE_BASE + NOTIFIER);
+	ps3vram_out_ring(priv, 0xfeed0000);	/* DMA video RAM instance */
+	ps3vram_out_ring(priv, 0xfeed0001);	/* DMA system RAM instance */
+
+	ps3vram_fire_ring(dev);
+}
+
+static int ps3vram_upload(struct ps3_system_bus_device *dev,
+			  unsigned int src_offset, unsigned int dst_offset,
+			  int len, int count)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+
+	ps3vram_begin_ring(priv, UPLOAD_SUBCH,
+			   NV_MEMORY_TO_MEMORY_FORMAT_OFFSET_IN, 8);
+	ps3vram_out_ring(priv, XDR_IOIF + src_offset);
+	ps3vram_out_ring(priv, dst_offset);
+	ps3vram_out_ring(priv, len);
+	ps3vram_out_ring(priv, len);
+	ps3vram_out_ring(priv, len);
+	ps3vram_out_ring(priv, count);
+	ps3vram_out_ring(priv, (1 << 8) | 1);
+	ps3vram_out_ring(priv, 0);
+
+	ps3vram_notifier_reset(dev);
+	ps3vram_begin_ring(priv, UPLOAD_SUBCH,
+			   NV_MEMORY_TO_MEMORY_FORMAT_NOTIFY, 1);
+	ps3vram_out_ring(priv, 0);
+	ps3vram_begin_ring(priv, UPLOAD_SUBCH, 0x100, 1);
+	ps3vram_out_ring(priv, 0);
+	ps3vram_fire_ring(dev);
+	if (ps3vram_notifier_wait(dev, 200) < 0) {
+		dev_warn(&dev->core, "%s: Notifier timeout\n", __func__);
+		return -1;
+	}
+
+	return 0;
+}
+
+static int ps3vram_download(struct ps3_system_bus_device *dev,
+			    unsigned int src_offset, unsigned int dst_offset,
+			    int len, int count)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+
+	ps3vram_begin_ring(priv, DOWNLOAD_SUBCH,
+			   NV_MEMORY_TO_MEMORY_FORMAT_OFFSET_IN, 8);
+	ps3vram_out_ring(priv, src_offset);
+	ps3vram_out_ring(priv, XDR_IOIF + dst_offset);
+	ps3vram_out_ring(priv, len);
+	ps3vram_out_ring(priv, len);
+	ps3vram_out_ring(priv, len);
+	ps3vram_out_ring(priv, count);
+	ps3vram_out_ring(priv, (1 << 8) | 1);
+	ps3vram_out_ring(priv, 0);
+
+	ps3vram_notifier_reset(dev);
+	ps3vram_begin_ring(priv, DOWNLOAD_SUBCH,
+			   NV_MEMORY_TO_MEMORY_FORMAT_NOTIFY, 1);
+	ps3vram_out_ring(priv, 0);
+	ps3vram_begin_ring(priv, DOWNLOAD_SUBCH, 0x100, 1);
+	ps3vram_out_ring(priv, 0);
+	ps3vram_fire_ring(dev);
+	if (ps3vram_notifier_wait(dev, 200) < 0) {
+		dev_warn(&dev->core, "%s: Notifier timeout\n", __func__);
+		return -1;
+	}
+
+	return 0;
+}
+
+static void ps3vram_cache_evict(struct ps3_system_bus_device *dev, int entry)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+	struct ps3vram_cache *cache = &priv->cache;
+
+	if (!(cache->tags[entry].flags & CACHE_PAGE_DIRTY))
+		return;
+
+	dev_dbg(&dev->core, "Flushing %d: 0x%08x\n", entry,
+		cache->tags[entry].address);
+	if (ps3vram_upload(dev, CACHE_OFFSET + entry * cache->page_size,
+			   cache->tags[entry].address, DMA_PAGE_SIZE,
+			   cache->page_size / DMA_PAGE_SIZE) < 0) {
+		dev_err(&dev->core,
+			"Failed to upload from 0x%x to " "0x%x size 0x%x\n",
+			entry * cache->page_size, cache->tags[entry].address,
+			cache->page_size);
+	}
+	cache->tags[entry].flags &= ~CACHE_PAGE_DIRTY;
+}
+
+static void ps3vram_cache_load(struct ps3_system_bus_device *dev, int entry,
+			       unsigned int address)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+	struct ps3vram_cache *cache = &priv->cache;
+
+	dev_dbg(&dev->core, "Fetching %d: 0x%08x\n", entry, address);
+	if (ps3vram_download(dev, address,
+			     CACHE_OFFSET + entry * cache->page_size,
+			     DMA_PAGE_SIZE,
+			     cache->page_size / DMA_PAGE_SIZE) < 0) {
+		dev_err(&dev->core,
+			"Failed to download from 0x%x to 0x%x size 0x%x\n",
+			address, entry * cache->page_size, cache->page_size);
+	}
+
+	cache->tags[entry].address = address;
+	cache->tags[entry].flags |= CACHE_PAGE_PRESENT;
+}
+
+
+static void ps3vram_cache_flush(struct ps3_system_bus_device *dev)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+	struct ps3vram_cache *cache = &priv->cache;
+	int i;
+
+	dev_dbg(&dev->core, "FLUSH\n");
+	for (i = 0; i < cache->page_count; i++) {
+		ps3vram_cache_evict(dev, i);
+		cache->tags[i].flags = 0;
+	}
+}
+
+static unsigned int ps3vram_cache_match(struct ps3_system_bus_device *dev,
+					loff_t address)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+	struct ps3vram_cache *cache = &priv->cache;
+	unsigned int base;
+	unsigned int offset;
+	int i;
+	static int counter;
+
+	offset = (unsigned int) (address & (cache->page_size - 1));
+	base = (unsigned int) (address - offset);
+
+	/* fully associative check */
+	for (i = 0; i < cache->page_count; i++) {
+		if ((cache->tags[i].flags & CACHE_PAGE_PRESENT) &&
+		    cache->tags[i].address == base) {
+			cache->hit++;
+			dev_dbg(&dev->core, "Found entry %d: 0x%08x\n", i,
+				cache->tags[i].address);
+			return i;
+		}
+	}
+
+	/* choose a random entry */
+	i = (jiffies + (counter++)) % cache->page_count;
+	dev_dbg(&dev->core, "Using entry %d\n", i);
+
+	ps3vram_cache_evict(dev, i);
+	ps3vram_cache_load(dev, i, base);
+
+	cache->miss++;
+	return i;
+}
+
+static int ps3vram_cache_init(struct ps3_system_bus_device *dev)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+
+	priv->cache.page_count = CACHE_PAGE_COUNT;
+	priv->cache.page_size = CACHE_PAGE_SIZE;
+	priv->cache.tags = kzalloc(sizeof(struct ps3vram_tag) *
+				   CACHE_PAGE_COUNT, GFP_KERNEL);
+	if (priv->cache.tags == NULL) {
+		dev_err(&dev->core, "Could not allocate cache tags\n");
+		return -ENOMEM;
+	}
+
+	dev_info(&dev->core, "Created ram cache: %d entries, %d KiB each\n",
+		CACHE_PAGE_COUNT, CACHE_PAGE_SIZE / 1024);
+
+	return 0;
+}
+
+static void ps3vram_cache_cleanup(struct ps3_system_bus_device *dev)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+
+	ps3vram_cache_flush(dev);
+	kfree(priv->cache.tags);
+}
+
+static int ps3vram_read(struct ps3_system_bus_device *dev, loff_t from,
+			size_t len, size_t *retlen, u_char *buf)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+	unsigned int cached, count;
+
+	dev_dbg(&dev->core, "%s: from=0x%08x len=0x%zx\n", __func__,
+		(unsigned int)from, len);
+
+	if (from >= priv->size)
+		return -EIO;
+
+	if (len > priv->size - from)
+		len = priv->size - from;
+
+	/* Copy from vram to buf */
+	count = len;
+	while (count) {
+		unsigned int offset, avail;
+		unsigned int entry;
+
+		offset = (unsigned int) (from & (priv->cache.page_size - 1));
+		avail  = priv->cache.page_size - offset;
+
+		entry = ps3vram_cache_match(dev, from);
+		cached = CACHE_OFFSET + entry * priv->cache.page_size + offset;
+
+		dev_dbg(&dev->core, "%s: from=%08x cached=%08x offset=%08x "
+			"avail=%08x count=%08x\n", __func__,
+			(unsigned int)from, cached, offset, avail, count);
+
+		if (avail > count)
+			avail = count;
+		memcpy(buf, priv->xdr_buf + cached, avail);
+
+		buf += avail;
+		count -= avail;
+		from += avail;
+	}
+
+	*retlen = len;
+	return 0;
+}
+
+static int ps3vram_write(struct ps3_system_bus_device *dev, loff_t to,
+			 size_t len, size_t *retlen, const u_char *buf)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+	unsigned int cached, count;
+
+	if (to >= priv->size)
+		return -EIO;
+
+	if (len > priv->size - to)
+		len = priv->size - to;
+
+	/* Copy from buf to vram */
+	count = len;
+	while (count) {
+		unsigned int offset, avail;
+		unsigned int entry;
+
+		offset = (unsigned int) (to & (priv->cache.page_size - 1));
+		avail  = priv->cache.page_size - offset;
+
+		entry = ps3vram_cache_match(dev, to);
+		cached = CACHE_OFFSET + entry * priv->cache.page_size + offset;
+
+		dev_dbg(&dev->core, "%s: to=%08x cached=%08x offset=%08x "
+			"avail=%08x count=%08x\n", __func__, (unsigned int)to,
+			cached, offset, avail, count);
+
+		if (avail > count)
+			avail = count;
+		memcpy(priv->xdr_buf + cached, buf, avail);
+
+		priv->cache.tags[entry].flags |= CACHE_PAGE_DIRTY;
+
+		buf += avail;
+		count -= avail;
+		to += avail;
+	}
+
+	*retlen = len;
+	return 0;
+}
+
+static int ps3vram_proc_show(struct seq_file *m, void *v)
+{
+	struct ps3vram_priv *priv = m->private;
+
+	seq_printf(m, "hit:%u\nmiss:%u\n", priv->cache.hit, priv->cache.miss);
+	return 0;
+}
+
+static int ps3vram_proc_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, ps3vram_proc_show, PDE_DATA(inode));
+}
+
+static const struct file_operations ps3vram_proc_fops = {
+	.owner		= THIS_MODULE,
+	.open		= ps3vram_proc_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= single_release,
+};
+
+static void ps3vram_proc_init(struct ps3_system_bus_device *dev)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+	struct proc_dir_entry *pde;
+
+	pde = proc_create_data(DEVICE_NAME, 0444, NULL, &ps3vram_proc_fops,
+			       priv);
+	if (!pde)
+		dev_warn(&dev->core, "failed to create /proc entry\n");
+}
+
+static struct bio *ps3vram_do_bio(struct ps3_system_bus_device *dev,
+				  struct bio *bio)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+	int write = bio_data_dir(bio) == WRITE;
+	const char *op = write ? "write" : "read";
+	loff_t offset = bio->bi_iter.bi_sector << 9;
+	int error = 0;
+	struct bio_vec bvec;
+	struct bvec_iter iter;
+	struct bio *next;
+
+	bio_for_each_segment(bvec, bio, iter) {
+		/* PS3 is ppc64, so we don't handle highmem */
+		char *ptr = page_address(bvec.bv_page) + bvec.bv_offset;
+		size_t len = bvec.bv_len, retlen;
+
+		dev_dbg(&dev->core, "    %s %zu bytes at offset %llu\n", op,
+			len, offset);
+		if (write)
+			error = ps3vram_write(dev, offset, len, &retlen, ptr);
+		else
+			error = ps3vram_read(dev, offset, len, &retlen, ptr);
+
+		if (error) {
+			dev_err(&dev->core, "%s failed\n", op);
+			goto out;
+		}
+
+		if (retlen != len) {
+			dev_err(&dev->core, "Short %s\n", op);
+			error = -EIO;
+			goto out;
+		}
+
+		offset += len;
+	}
+
+	dev_dbg(&dev->core, "%s completed\n", op);
+
+out:
+	spin_lock_irq(&priv->lock);
+	bio_list_pop(&priv->list);
+	next = bio_list_peek(&priv->list);
+	spin_unlock_irq(&priv->lock);
+
+	bio_endio(bio, error);
+	return next;
+}
+
+static void ps3vram_make_request(struct request_queue *q, struct bio *bio)
+{
+	struct ps3_system_bus_device *dev = q->queuedata;
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+	int busy;
+
+	dev_dbg(&dev->core, "%s\n", __func__);
+
+	spin_lock_irq(&priv->lock);
+	busy = !bio_list_empty(&priv->list);
+	bio_list_add(&priv->list, bio);
+	spin_unlock_irq(&priv->lock);
+
+	if (busy)
+		return;
+
+	do {
+		bio = ps3vram_do_bio(dev, bio);
+	} while (bio);
+}
+
+static int ps3vram_probe(struct ps3_system_bus_device *dev)
+{
+	struct ps3vram_priv *priv;
+	int error, status;
+	struct request_queue *queue;
+	struct gendisk *gendisk;
+	u64 ddr_size, ddr_lpar, ctrl_lpar, info_lpar, reports_lpar,
+	    reports_size, xdr_lpar;
+	char *rest;
+
+	priv = kzalloc(sizeof(*priv), GFP_KERNEL);
+	if (!priv) {
+		error = -ENOMEM;
+		goto fail;
+	}
+
+	spin_lock_init(&priv->lock);
+	bio_list_init(&priv->list);
+	ps3_system_bus_set_drvdata(dev, priv);
+
+	/* Allocate XDR buffer (1MiB aligned) */
+	priv->xdr_buf = (void *)__get_free_pages(GFP_KERNEL,
+		get_order(XDR_BUF_SIZE));
+	if (priv->xdr_buf == NULL) {
+		dev_err(&dev->core, "Could not allocate XDR buffer\n");
+		error = -ENOMEM;
+		goto fail_free_priv;
+	}
+
+	/* Put FIFO at begginning of XDR buffer */
+	priv->fifo_base = (u32 *) (priv->xdr_buf + FIFO_OFFSET);
+	priv->fifo_ptr = priv->fifo_base;
+
+	/* XXX: Need to open GPU, in case ps3fb or snd_ps3 aren't loaded */
+	if (ps3_open_hv_device(dev)) {
+		dev_err(&dev->core, "ps3_open_hv_device failed\n");
+		error = -EAGAIN;
+		goto out_free_xdr_buf;
+	}
+
+	/* Request memory */
+	status = -1;
+	ddr_size = ALIGN(memparse(size, &rest), 1024*1024);
+	if (!ddr_size) {
+		dev_err(&dev->core, "Specified size is too small\n");
+		error = -EINVAL;
+		goto out_close_gpu;
+	}
+
+	while (ddr_size > 0) {
+		status = lv1_gpu_memory_allocate(ddr_size, 0, 0, 0, 0,
+						 &priv->memory_handle,
+						 &ddr_lpar);
+		if (!status)
+			break;
+		ddr_size -= 1024*1024;
+	}
+	if (status) {
+		dev_err(&dev->core, "lv1_gpu_memory_allocate failed %d\n",
+			status);
+		error = -ENOMEM;
+		goto out_close_gpu;
+	}
+
+	/* Request context */
+	status = lv1_gpu_context_allocate(priv->memory_handle, 0,
+					  &priv->context_handle, &ctrl_lpar,
+					  &info_lpar, &reports_lpar,
+					  &reports_size);
+	if (status) {
+		dev_err(&dev->core, "lv1_gpu_context_allocate failed %d\n",
+			status);
+		error = -ENOMEM;
+		goto out_free_memory;
+	}
+
+	/* Map XDR buffer to RSX */
+	xdr_lpar = ps3_mm_phys_to_lpar(__pa(priv->xdr_buf));
+	status = lv1_gpu_context_iomap(priv->context_handle, XDR_IOIF,
+				       xdr_lpar, XDR_BUF_SIZE,
+				       CBE_IOPTE_PP_W | CBE_IOPTE_PP_R |
+				       CBE_IOPTE_M);
+	if (status) {
+		dev_err(&dev->core, "lv1_gpu_context_iomap failed %d\n",
+			status);
+		error = -ENOMEM;
+		goto out_free_context;
+	}
+
+	priv->ctrl = ioremap(ctrl_lpar, 64 * 1024);
+	if (!priv->ctrl) {
+		dev_err(&dev->core, "ioremap CTRL failed\n");
+		error = -ENOMEM;
+		goto out_unmap_context;
+	}
+
+	priv->reports = ioremap(reports_lpar, reports_size);
+	if (!priv->reports) {
+		dev_err(&dev->core, "ioremap REPORTS failed\n");
+		error = -ENOMEM;
+		goto out_unmap_ctrl;
+	}
+
+	mutex_lock(&ps3_gpu_mutex);
+	ps3vram_init_ring(dev);
+	mutex_unlock(&ps3_gpu_mutex);
+
+	priv->size = ddr_size;
+
+	ps3vram_bind(dev);
+
+	mutex_lock(&ps3_gpu_mutex);
+	error = ps3vram_wait_ring(dev, 100);
+	mutex_unlock(&ps3_gpu_mutex);
+	if (error < 0) {
+		dev_err(&dev->core, "Failed to initialize channels\n");
+		error = -ETIMEDOUT;
+		goto out_unmap_reports;
+	}
+
+	ps3vram_cache_init(dev);
+	ps3vram_proc_init(dev);
+
+	queue = blk_alloc_queue(GFP_KERNEL);
+	if (!queue) {
+		dev_err(&dev->core, "blk_alloc_queue failed\n");
+		error = -ENOMEM;
+		goto out_cache_cleanup;
+	}
+
+	priv->queue = queue;
+	queue->queuedata = dev;
+	blk_queue_make_request(queue, ps3vram_make_request);
+	blk_queue_max_segments(queue, BLK_MAX_SEGMENTS);
+	blk_queue_max_segment_size(queue, BLK_MAX_SEGMENT_SIZE);
+	blk_queue_max_hw_sectors(queue, BLK_SAFE_MAX_SECTORS);
+
+	gendisk = alloc_disk(1);
+	if (!gendisk) {
+		dev_err(&dev->core, "alloc_disk failed\n");
+		error = -ENOMEM;
+		goto fail_cleanup_queue;
+	}
+
+	priv->gendisk = gendisk;
+	gendisk->major = ps3vram_major;
+	gendisk->first_minor = 0;
+	gendisk->fops = &ps3vram_fops;
+	gendisk->queue = queue;
+	gendisk->private_data = dev;
+	gendisk->driverfs_dev = &dev->core;
+	strlcpy(gendisk->disk_name, DEVICE_NAME, sizeof(gendisk->disk_name));
+	set_capacity(gendisk, priv->size >> 9);
+
+	dev_info(&dev->core, "%s: Using %lu MiB of GPU memory\n",
+		 gendisk->disk_name, get_capacity(gendisk) >> 11);
+
+	add_disk(gendisk);
+	return 0;
+
+fail_cleanup_queue:
+	blk_cleanup_queue(queue);
+out_cache_cleanup:
+	remove_proc_entry(DEVICE_NAME, NULL);
+	ps3vram_cache_cleanup(dev);
+out_unmap_reports:
+	iounmap(priv->reports);
+out_unmap_ctrl:
+	iounmap(priv->ctrl);
+out_unmap_context:
+	lv1_gpu_context_iomap(priv->context_handle, XDR_IOIF, xdr_lpar,
+			      XDR_BUF_SIZE, CBE_IOPTE_M);
+out_free_context:
+	lv1_gpu_context_free(priv->context_handle);
+out_free_memory:
+	lv1_gpu_memory_free(priv->memory_handle);
+out_close_gpu:
+	ps3_close_hv_device(dev);
+out_free_xdr_buf:
+	free_pages((unsigned long) priv->xdr_buf, get_order(XDR_BUF_SIZE));
+fail_free_priv:
+	kfree(priv);
+	ps3_system_bus_set_drvdata(dev, NULL);
+fail:
+	return error;
+}
+
+static int ps3vram_remove(struct ps3_system_bus_device *dev)
+{
+	struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
+
+	del_gendisk(priv->gendisk);
+	put_disk(priv->gendisk);
+	blk_cleanup_queue(priv->queue);
+	remove_proc_entry(DEVICE_NAME, NULL);
+	ps3vram_cache_cleanup(dev);
+	iounmap(priv->reports);
+	iounmap(priv->ctrl);
+	lv1_gpu_context_iomap(priv->context_handle, XDR_IOIF,
+			      ps3_mm_phys_to_lpar(__pa(priv->xdr_buf)),
+			      XDR_BUF_SIZE, CBE_IOPTE_M);
+	lv1_gpu_context_free(priv->context_handle);
+	lv1_gpu_memory_free(priv->memory_handle);
+	ps3_close_hv_device(dev);
+	free_pages((unsigned long) priv->xdr_buf, get_order(XDR_BUF_SIZE));
+	kfree(priv);
+	ps3_system_bus_set_drvdata(dev, NULL);
+	return 0;
+}
+
+static struct ps3_system_bus_driver ps3vram = {
+	.match_id	= PS3_MATCH_ID_GPU,
+	.match_sub_id	= PS3_MATCH_SUB_ID_GPU_RAMDISK,
+	.core.name	= DEVICE_NAME,
+	.core.owner	= THIS_MODULE,
+	.probe		= ps3vram_probe,
+	.remove		= ps3vram_remove,
+	.shutdown	= ps3vram_remove,
+};
+
+
+static int __init ps3vram_init(void)
+{
+	int error;
+
+	if (!firmware_has_feature(FW_FEATURE_PS3_LV1))
+		return -ENODEV;
+
+	error = register_blkdev(0, DEVICE_NAME);
+	if (error <= 0) {
+		pr_err("%s: register_blkdev failed %d\n", DEVICE_NAME, error);
+		return error;
+	}
+	ps3vram_major = error;
+
+	pr_info("%s: registered block device major %d\n", DEVICE_NAME,
+		ps3vram_major);
+
+	error = ps3_system_bus_driver_register(&ps3vram);
+	if (error)
+		unregister_blkdev(ps3vram_major, DEVICE_NAME);
+
+	return error;
+}
+
+static void __exit ps3vram_exit(void)
+{
+	ps3_system_bus_driver_unregister(&ps3vram);
+	unregister_blkdev(ps3vram_major, DEVICE_NAME);
+}
+
+module_init(ps3vram_init);
+module_exit(ps3vram_exit);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("PS3 Video RAM Storage Driver");
+MODULE_AUTHOR("Sony Corporation");
+MODULE_ALIAS(PS3_MODULE_ALIAS_GPU_RAMDISK);

drivers/block/rbd_types.h

@@ -0,0 +1,81 @@
+/*
+ * Ceph - scalable distributed file system
+ *
+ * Copyright (C) 2004-2010 Sage Weil <sage@newdream.net>
+ *
+ * This is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License version 2.1, as published by the Free Software
+ * Foundation.  See file COPYING.
+ *
+ */
+
+#ifndef CEPH_RBD_TYPES_H
+#define CEPH_RBD_TYPES_H
+
+#include <linux/types.h>
+
+/* For format version 2, rbd image 'foo' consists of objects
+ *   rbd_id.foo		- id of image
+ *   rbd_header.<id>	- image metadata
+ *   rbd_data.<id>.0000000000000000
+ *   rbd_data.<id>.0000000000000001
+ *   ...		- data
+ * Clients do not access header data directly in rbd format 2.
+ */
+
+#define RBD_HEADER_PREFIX      "rbd_header."
+#define RBD_DATA_PREFIX        "rbd_data."
+#define RBD_ID_PREFIX          "rbd_id."
+
+/*
+ * For format version 1, rbd image 'foo' consists of objects
+ *   foo.rbd		- image metadata
+ *   rb.<idhi>.<idlo>.00000000
+ *   rb.<idhi>.<idlo>.00000001
+ *   ...		- data
+ * There is no notion of a persistent image id in rbd format 1.
+ */
+
+#define RBD_SUFFIX		".rbd"
+
+#define RBD_DIRECTORY           "rbd_directory"
+#define RBD_INFO                "rbd_info"
+
+#define RBD_DEFAULT_OBJ_ORDER	22   /* 4MB */
+#define RBD_MIN_OBJ_ORDER       16
+#define RBD_MAX_OBJ_ORDER       30
+
+#define RBD_COMP_NONE		0
+#define RBD_CRYPT_NONE		0
+
+#define RBD_HEADER_TEXT		"<<< Rados Block Device Image >>>\n"
+#define RBD_HEADER_SIGNATURE	"RBD"
+#define RBD_HEADER_VERSION	"001.005"
+
+struct rbd_image_snap_ondisk {
+	__le64 id;
+	__le64 image_size;
+} __attribute__((packed));
+
+struct rbd_image_header_ondisk {
+	char text[40];
+	char object_prefix[24];
+	char signature[4];
+	char version[8];
+	struct {
+		__u8 order;
+		__u8 crypt_type;
+		__u8 comp_type;
+		__u8 unused;
+	} __attribute__((packed)) options;
+	__le64 image_size;
+	__le64 snap_seq;
+	__le32 snap_count;
+	__le32 reserved;
+	__le64 snap_names_len;
+	struct rbd_image_snap_ondisk snaps[0];
+} __attribute__((packed));
+
+
+#endif

drivers/block/rsxx/Makefile

@@ -0,0 +1,2 @@
+obj-$(CONFIG_BLK_DEV_RSXX) += rsxx.o
+rsxx-objs := config.o core.o cregs.o dev.o dma.o

drivers/block/rsxx/config.c

@@ -0,0 +1,211 @@
+/*
+* Filename: config.c
+*
+*
+* Authors: Joshua Morris <josh.h.morris@us.ibm.com>
+*	Philip Kelleher <pjk1939@linux.vnet.ibm.com>
+*
+* (C) Copyright 2013 IBM Corporation
+*
+* This program is free software; you can redistribute it and/or
+* modify it under the terms of the GNU General Public License as
+* published by the Free Software Foundation; either version 2 of the
+* License, or (at your option) any later version.
+*
+* This program is distributed in the hope that it will be useful, but
+* WITHOUT ANY WARRANTY; without even the implied warranty of
+* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+* General Public License for more details.
+*
+* You should have received a copy of the GNU General Public License
+* along with this program; if not, write to the Free Software Foundation,
+* Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+*/
+
+#include <linux/types.h>
+#include <linux/crc32.h>
+#include <linux/swab.h>
+
+#include "rsxx_priv.h"
+#include "rsxx_cfg.h"
+
+static void initialize_config(struct rsxx_card_cfg *cfg)
+{
+	cfg->hdr.version = RSXX_CFG_VERSION;
+
+	cfg->data.block_size        = RSXX_HW_BLK_SIZE;
+	cfg->data.stripe_size       = RSXX_HW_BLK_SIZE;
+	cfg->data.vendor_id         = RSXX_VENDOR_ID_IBM;
+	cfg->data.cache_order       = (-1);
+	cfg->data.intr_coal.mode    = RSXX_INTR_COAL_DISABLED;
+	cfg->data.intr_coal.count   = 0;
+	cfg->data.intr_coal.latency = 0;
+}
+
+static u32 config_data_crc32(struct rsxx_card_cfg *cfg)
+{
+	/*
+	 * Return the compliment of the CRC to ensure compatibility
+	 * (i.e. this is how early rsxx drivers did it.)
+	 */
+
+	return ~crc32(~0, &cfg->data, sizeof(cfg->data));
+}
+
+
+/*----------------- Config Byte Swap Functions -------------------*/
+static void config_hdr_be_to_cpu(struct card_cfg_hdr *hdr)
+{
+	hdr->version = be32_to_cpu((__force __be32) hdr->version);
+	hdr->crc     = be32_to_cpu((__force __be32) hdr->crc);
+}
+
+static void config_hdr_cpu_to_be(struct card_cfg_hdr *hdr)
+{
+	hdr->version = (__force u32) cpu_to_be32(hdr->version);
+	hdr->crc     = (__force u32) cpu_to_be32(hdr->crc);
+}
+
+static void config_data_swab(struct rsxx_card_cfg *cfg)
+{
+	u32 *data = (u32 *) &cfg->data;
+	int i;
+
+	for (i = 0; i < (sizeof(cfg->data) / 4); i++)
+		data[i] = swab32(data[i]);
+}
+
+static void config_data_le_to_cpu(struct rsxx_card_cfg *cfg)
+{
+	u32 *data = (u32 *) &cfg->data;
+	int i;
+
+	for (i = 0; i < (sizeof(cfg->data) / 4); i++)
+		data[i] = le32_to_cpu((__force __le32) data[i]);
+}
+
+static void config_data_cpu_to_le(struct rsxx_card_cfg *cfg)
+{
+	u32 *data = (u32 *) &cfg->data;
+	int i;
+
+	for (i = 0; i < (sizeof(cfg->data) / 4); i++)
+		data[i] = (__force u32) cpu_to_le32(data[i]);
+}
+
+
+/*----------------- Config Operations ------------------*/
+static int rsxx_save_config(struct rsxx_cardinfo *card)
+{
+	struct rsxx_card_cfg cfg;
+	int st;
+
+	memcpy(&cfg, &card->config, sizeof(cfg));
+
+	if (unlikely(cfg.hdr.version != RSXX_CFG_VERSION)) {
+		dev_err(CARD_TO_DEV(card),
+			"Cannot save config with invalid version %d\n",
+			cfg.hdr.version);
+		return -EINVAL;
+	}
+
+	/* Convert data to little endian for the CRC calculation. */
+	config_data_cpu_to_le(&cfg);
+
+	cfg.hdr.crc = config_data_crc32(&cfg);
+
+	/*
+	 * Swap the data from little endian to big endian so it can be
+	 * stored.
+	 */
+	config_data_swab(&cfg);
+	config_hdr_cpu_to_be(&cfg.hdr);
+
+	st = rsxx_creg_write(card, CREG_ADD_CONFIG, sizeof(cfg), &cfg, 1);
+	if (st)
+		return st;
+
+	return 0;
+}
+
+int rsxx_load_config(struct rsxx_cardinfo *card)
+{
+	int st;
+	u32 crc;
+
+	st = rsxx_creg_read(card, CREG_ADD_CONFIG, sizeof(card->config),
+				&card->config, 1);
+	if (st) {
+		dev_err(CARD_TO_DEV(card),
+			"Failed reading card config.\n");
+		return st;
+	}
+
+	config_hdr_be_to_cpu(&card->config.hdr);
+
+	if (card->config.hdr.version == RSXX_CFG_VERSION) {
+		/*
+		 * We calculate the CRC with the data in little endian, because
+		 * early drivers did not take big endian CPUs into account.
+		 * The data is always stored in big endian, so we need to byte
+		 * swap it before calculating the CRC.
+		 */
+
+		config_data_swab(&card->config);
+
+		/* Check the CRC */
+		crc = config_data_crc32(&card->config);
+		if (crc != card->config.hdr.crc) {
+			dev_err(CARD_TO_DEV(card),
+				"Config corruption detected!\n");
+			dev_info(CARD_TO_DEV(card),
+				"CRC (sb x%08x is x%08x)\n",
+				card->config.hdr.crc, crc);
+			return -EIO;
+		}
+
+		/* Convert the data to CPU byteorder */
+		config_data_le_to_cpu(&card->config);
+
+	} else if (card->config.hdr.version != 0) {
+		dev_err(CARD_TO_DEV(card),
+			"Invalid config version %d.\n",
+			card->config.hdr.version);
+		/*
+		 * Config version changes require special handling from the
+		 * user
+		 */
+		return -EINVAL;
+	} else {
+		dev_info(CARD_TO_DEV(card),
+			"Initializing card configuration.\n");
+		initialize_config(&card->config);
+		st = rsxx_save_config(card);
+		if (st)
+			return st;
+	}
+
+	card->config_valid = 1;
+
+	dev_dbg(CARD_TO_DEV(card), "version:     x%08x\n",
+		card->config.hdr.version);
+	dev_dbg(CARD_TO_DEV(card), "crc:         x%08x\n",
+		card->config.hdr.crc);
+	dev_dbg(CARD_TO_DEV(card), "block_size:  x%08x\n",
+		card->config.data.block_size);
+	dev_dbg(CARD_TO_DEV(card), "stripe_size: x%08x\n",
+		card->config.data.stripe_size);
+	dev_dbg(CARD_TO_DEV(card), "vendor_id:   x%08x\n",
+		card->config.data.vendor_id);
+	dev_dbg(CARD_TO_DEV(card), "cache_order: x%08x\n",
+		card->config.data.cache_order);
+	dev_dbg(CARD_TO_DEV(card), "mode:        x%08x\n",
+		card->config.data.intr_coal.mode);
+	dev_dbg(CARD_TO_DEV(card), "count:       x%08x\n",
+		card->config.data.intr_coal.count);
+	dev_dbg(CARD_TO_DEV(card), "latency:     x%08x\n",
+		 card->config.data.intr_coal.latency);
+
+	return 0;
+}
+

drivers/block/rsxx/cregs.c

@@ -0,0 +1,804 @@
+/*
+* Filename: cregs.c
+*
+*
+* Authors: Joshua Morris <josh.h.morris@us.ibm.com>
+*	Philip Kelleher <pjk1939@linux.vnet.ibm.com>
+*
+* (C) Copyright 2013 IBM Corporation
+*
+* This program is free software; you can redistribute it and/or
+* modify it under the terms of the GNU General Public License as
+* published by the Free Software Foundation; either version 2 of the
+* License, or (at your option) any later version.
+*
+* This program is distributed in the hope that it will be useful, but
+* WITHOUT ANY WARRANTY; without even the implied warranty of
+* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+* General Public License for more details.
+*
+* You should have received a copy of the GNU General Public License
+* along with this program; if not, write to the Free Software Foundation,
+* Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+*/
+
+#include <linux/completion.h>
+#include <linux/slab.h>
+
+#include "rsxx_priv.h"
+
+#define CREG_TIMEOUT_MSEC	10000
+
+typedef void (*creg_cmd_cb)(struct rsxx_cardinfo *card,
+			    struct creg_cmd *cmd,
+			    int st);
+
+struct creg_cmd {
+	struct list_head list;
+	creg_cmd_cb cb;
+	void *cb_private;
+	unsigned int op;
+	unsigned int addr;
+	int cnt8;
+	void *buf;
+	unsigned int stream;
+	unsigned int status;
+};
+
+static struct kmem_cache *creg_cmd_pool;
+
+
+/*------------ Private Functions --------------*/
+
+#if defined(__LITTLE_ENDIAN)
+#define LITTLE_ENDIAN 1
+#elif defined(__BIG_ENDIAN)
+#define LITTLE_ENDIAN 0
+#else
+#error Unknown endianess!!! Aborting...
+#endif
+
+static int copy_to_creg_data(struct rsxx_cardinfo *card,
+			      int cnt8,
+			      void *buf,
+			      unsigned int stream)
+{
+	int i = 0;
+	u32 *data = buf;
+
+	if (unlikely(card->eeh_state))
+		return -EIO;
+
+	for (i = 0; cnt8 > 0; i++, cnt8 -= 4) {
+		/*
+		 * Firmware implementation makes it necessary to byte swap on
+		 * little endian processors.
+		 */
+		if (LITTLE_ENDIAN && stream)
+			iowrite32be(data[i], card->regmap + CREG_DATA(i));
+		else
+			iowrite32(data[i], card->regmap + CREG_DATA(i));
+	}
+
+	return 0;
+}
+
+
+static int copy_from_creg_data(struct rsxx_cardinfo *card,
+				int cnt8,
+				void *buf,
+				unsigned int stream)
+{
+	int i = 0;
+	u32 *data = buf;
+
+	if (unlikely(card->eeh_state))
+		return -EIO;
+
+	for (i = 0; cnt8 > 0; i++, cnt8 -= 4) {
+		/*
+		 * Firmware implementation makes it necessary to byte swap on
+		 * little endian processors.
+		 */
+		if (LITTLE_ENDIAN && stream)
+			data[i] = ioread32be(card->regmap + CREG_DATA(i));
+		else
+			data[i] = ioread32(card->regmap + CREG_DATA(i));
+	}
+
+	return 0;
+}
+
+static void creg_issue_cmd(struct rsxx_cardinfo *card, struct creg_cmd *cmd)
+{
+	int st;
+
+	if (unlikely(card->eeh_state))
+		return;
+
+	iowrite32(cmd->addr, card->regmap + CREG_ADD);
+	iowrite32(cmd->cnt8, card->regmap + CREG_CNT);
+
+	if (cmd->op == CREG_OP_WRITE) {
+		if (cmd->buf) {
+			st = copy_to_creg_data(card, cmd->cnt8,
+					       cmd->buf, cmd->stream);
+			if (st)
+				return;
+		}
+	}
+
+	if (unlikely(card->eeh_state))
+		return;
+
+	/* Setting the valid bit will kick off the command. */
+	iowrite32(cmd->op, card->regmap + CREG_CMD);
+}
+
+static void creg_kick_queue(struct rsxx_cardinfo *card)
+{
+	if (card->creg_ctrl.active || list_empty(&card->creg_ctrl.queue))
+		return;
+
+	card->creg_ctrl.active = 1;
+	card->creg_ctrl.active_cmd = list_first_entry(&card->creg_ctrl.queue,
+						      struct creg_cmd, list);
+	list_del(&card->creg_ctrl.active_cmd->list);
+	card->creg_ctrl.q_depth--;
+
+	/*
+	 * We have to set the timer before we push the new command. Otherwise,
+	 * we could create a race condition that would occur if the timer
+	 * was not canceled, and expired after the new command was pushed,
+	 * but before the command was issued to hardware.
+	 */
+	mod_timer(&card->creg_ctrl.cmd_timer,
+				jiffies + msecs_to_jiffies(CREG_TIMEOUT_MSEC));
+
+	creg_issue_cmd(card, card->creg_ctrl.active_cmd);
+}
+
+static int creg_queue_cmd(struct rsxx_cardinfo *card,
+			  unsigned int op,
+			  unsigned int addr,
+			  unsigned int cnt8,
+			  void *buf,
+			  int stream,
+			  creg_cmd_cb callback,
+			  void *cb_private)
+{
+	struct creg_cmd *cmd;
+
+	/* Don't queue stuff up if we're halted. */
+	if (unlikely(card->halt))
+		return -EINVAL;
+
+	if (card->creg_ctrl.reset)
+		return -EAGAIN;
+
+	if (cnt8 > MAX_CREG_DATA8)
+		return -EINVAL;
+
+	cmd = kmem_cache_alloc(creg_cmd_pool, GFP_KERNEL);
+	if (!cmd)
+		return -ENOMEM;
+
+	INIT_LIST_HEAD(&cmd->list);
+
+	cmd->op		= op;
+	cmd->addr	= addr;
+	cmd->cnt8	= cnt8;
+	cmd->buf	= buf;
+	cmd->stream	= stream;
+	cmd->cb		= callback;
+	cmd->cb_private = cb_private;
+	cmd->status	= 0;
+
+	spin_lock_bh(&card->creg_ctrl.lock);
+	list_add_tail(&cmd->list, &card->creg_ctrl.queue);
+	card->creg_ctrl.q_depth++;
+	creg_kick_queue(card);
+	spin_unlock_bh(&card->creg_ctrl.lock);
+
+	return 0;
+}
+
+static void creg_cmd_timed_out(unsigned long data)
+{
+	struct rsxx_cardinfo *card = (struct rsxx_cardinfo *) data;
+	struct creg_cmd *cmd;
+
+	spin_lock(&card->creg_ctrl.lock);
+	cmd = card->creg_ctrl.active_cmd;
+	card->creg_ctrl.active_cmd = NULL;
+	spin_unlock(&card->creg_ctrl.lock);
+
+	if (cmd == NULL) {
+		card->creg_ctrl.creg_stats.creg_timeout++;
+		dev_warn(CARD_TO_DEV(card),
+			"No active command associated with timeout!\n");
+		return;
+	}
+
+	if (cmd->cb)
+		cmd->cb(card, cmd, -ETIMEDOUT);
+
+	kmem_cache_free(creg_cmd_pool, cmd);
+
+
+	spin_lock(&card->creg_ctrl.lock);
+	card->creg_ctrl.active = 0;
+	creg_kick_queue(card);
+	spin_unlock(&card->creg_ctrl.lock);
+}
+
+
+static void creg_cmd_done(struct work_struct *work)
+{
+	struct rsxx_cardinfo *card;
+	struct creg_cmd *cmd;
+	int st = 0;
+
+	card = container_of(work, struct rsxx_cardinfo,
+			    creg_ctrl.done_work);
+
+	/*
+	 * The timer could not be cancelled for some reason,
+	 * race to pop the active command.
+	 */
+	if (del_timer_sync(&card->creg_ctrl.cmd_timer) == 0)
+		card->creg_ctrl.creg_stats.failed_cancel_timer++;
+
+	spin_lock_bh(&card->creg_ctrl.lock);
+	cmd = card->creg_ctrl.active_cmd;
+	card->creg_ctrl.active_cmd = NULL;
+	spin_unlock_bh(&card->creg_ctrl.lock);
+
+	if (cmd == NULL) {
+		dev_err(CARD_TO_DEV(card),
+			"Spurious creg interrupt!\n");
+		return;
+	}
+
+	card->creg_ctrl.creg_stats.stat = ioread32(card->regmap + CREG_STAT);
+	cmd->status = card->creg_ctrl.creg_stats.stat;
+	if ((cmd->status & CREG_STAT_STATUS_MASK) == 0) {
+		dev_err(CARD_TO_DEV(card),
+			"Invalid status on creg command\n");
+		/*
+		 * At this point we're probably reading garbage from HW. Don't
+		 * do anything else that could mess up the system and let
+		 * the sync function return an error.
+		 */
+		st = -EIO;
+		goto creg_done;
+	} else if (cmd->status & CREG_STAT_ERROR) {
+		st = -EIO;
+	}
+
+	if ((cmd->op == CREG_OP_READ)) {
+		unsigned int cnt8 = ioread32(card->regmap + CREG_CNT);
+
+		/* Paranoid Sanity Checks */
+		if (!cmd->buf) {
+			dev_err(CARD_TO_DEV(card),
+				"Buffer not given for read.\n");
+			st = -EIO;
+			goto creg_done;
+		}
+		if (cnt8 != cmd->cnt8) {
+			dev_err(CARD_TO_DEV(card),
+				"count mismatch\n");
+			st = -EIO;
+			goto creg_done;
+		}
+
+		st = copy_from_creg_data(card, cnt8, cmd->buf, cmd->stream);
+	}
+
+creg_done:
+	if (cmd->cb)
+		cmd->cb(card, cmd, st);
+
+	kmem_cache_free(creg_cmd_pool, cmd);
+
+	spin_lock_bh(&card->creg_ctrl.lock);
+	card->creg_ctrl.active = 0;
+	creg_kick_queue(card);
+	spin_unlock_bh(&card->creg_ctrl.lock);
+}
+
+static void creg_reset(struct rsxx_cardinfo *card)
+{
+	struct creg_cmd *cmd = NULL;
+	struct creg_cmd *tmp;
+	unsigned long flags;
+
+	/*
+	 * mutex_trylock is used here because if reset_lock is taken then a
+	 * reset is already happening. So, we can just go ahead and return.
+	 */
+	if (!mutex_trylock(&card->creg_ctrl.reset_lock))
+		return;
+
+	card->creg_ctrl.reset = 1;
+	spin_lock_irqsave(&card->irq_lock, flags);
+	rsxx_disable_ier_and_isr(card, CR_INTR_CREG | CR_INTR_EVENT);
+	spin_unlock_irqrestore(&card->irq_lock, flags);
+
+	dev_warn(CARD_TO_DEV(card),
+		"Resetting creg interface for recovery\n");
+
+	/* Cancel outstanding commands */
+	spin_lock_bh(&card->creg_ctrl.lock);
+	list_for_each_entry_safe(cmd, tmp, &card->creg_ctrl.queue, list) {
+		list_del(&cmd->list);
+		card->creg_ctrl.q_depth--;
+		if (cmd->cb)
+			cmd->cb(card, cmd, -ECANCELED);
+		kmem_cache_free(creg_cmd_pool, cmd);
+	}
+
+	cmd = card->creg_ctrl.active_cmd;
+	card->creg_ctrl.active_cmd = NULL;
+	if (cmd) {
+		if (timer_pending(&card->creg_ctrl.cmd_timer))
+			del_timer_sync(&card->creg_ctrl.cmd_timer);
+
+		if (cmd->cb)
+			cmd->cb(card, cmd, -ECANCELED);
+		kmem_cache_free(creg_cmd_pool, cmd);
+
+		card->creg_ctrl.active = 0;
+	}
+	spin_unlock_bh(&card->creg_ctrl.lock);
+
+	card->creg_ctrl.reset = 0;
+	spin_lock_irqsave(&card->irq_lock, flags);
+	rsxx_enable_ier_and_isr(card, CR_INTR_CREG | CR_INTR_EVENT);
+	spin_unlock_irqrestore(&card->irq_lock, flags);
+
+	mutex_unlock(&card->creg_ctrl.reset_lock);
+}
+
+/* Used for synchronous accesses */
+struct creg_completion {
+	struct completion	*cmd_done;
+	int			st;
+	u32			creg_status;
+};
+
+static void creg_cmd_done_cb(struct rsxx_cardinfo *card,
+			     struct creg_cmd *cmd,
+			     int st)
+{
+	struct creg_completion *cmd_completion;
+
+	cmd_completion = cmd->cb_private;
+	BUG_ON(!cmd_completion);
+
+	cmd_completion->st = st;
+	cmd_completion->creg_status = cmd->status;
+	complete(cmd_completion->cmd_done);
+}
+
+static int __issue_creg_rw(struct rsxx_cardinfo *card,
+			   unsigned int op,
+			   unsigned int addr,
+			   unsigned int cnt8,
+			   void *buf,
+			   int stream,
+			   unsigned int *hw_stat)
+{
+	DECLARE_COMPLETION_ONSTACK(cmd_done);
+	struct creg_completion completion;
+	unsigned long timeout;
+	int st;
+
+	completion.cmd_done = &cmd_done;
+	completion.st = 0;
+	completion.creg_status = 0;
+
+	st = creg_queue_cmd(card, op, addr, cnt8, buf, stream, creg_cmd_done_cb,
+			    &completion);
+	if (st)
+		return st;
+
+	/*
+	 * This timeout is necessary for unresponsive hardware. The additional
+	 * 20 seconds to used to guarantee that each cregs requests has time to
+	 * complete.
+	 */
+	timeout = msecs_to_jiffies(CREG_TIMEOUT_MSEC *
+				   card->creg_ctrl.q_depth + 20000);
+
+	/*
+	 * The creg interface is guaranteed to complete. It has a timeout
+	 * mechanism that will kick in if hardware does not respond.
+	 */
+	st = wait_for_completion_timeout(completion.cmd_done, timeout);
+	if (st == 0) {
+		/*
+		 * This is really bad, because the kernel timer did not
+		 * expire and notify us of a timeout!
+		 */
+		dev_crit(CARD_TO_DEV(card),
+			"cregs timer failed\n");
+		creg_reset(card);
+		return -EIO;
+	}
+
+	*hw_stat = completion.creg_status;
+
+	if (completion.st) {
+		/*
+		* This read is needed to verify that there has not been any
+		* extreme errors that might have occurred, i.e. EEH. The
+		* function iowrite32 will not detect EEH errors, so it is
+		* necessary that we recover if such an error is the reason
+		* for the timeout. This is a dummy read.
+		*/
+		ioread32(card->regmap + SCRATCH);
+
+		dev_warn(CARD_TO_DEV(card),
+			"creg command failed(%d x%08x)\n",
+			completion.st, addr);
+		return completion.st;
+	}
+
+	return 0;
+}
+
+static int issue_creg_rw(struct rsxx_cardinfo *card,
+			 u32 addr,
+			 unsigned int size8,
+			 void *data,
+			 int stream,
+			 int read)
+{
+	unsigned int hw_stat;
+	unsigned int xfer;
+	unsigned int op;
+	int st;
+
+	op = read ? CREG_OP_READ : CREG_OP_WRITE;
+
+	do {
+		xfer = min_t(unsigned int, size8, MAX_CREG_DATA8);
+
+		st = __issue_creg_rw(card, op, addr, xfer,
+				     data, stream, &hw_stat);
+		if (st)
+			return st;
+
+		data   = (char *)data + xfer;
+		addr  += xfer;
+		size8 -= xfer;
+	} while (size8);
+
+	return 0;
+}
+
+/* ---------------------------- Public API ---------------------------------- */
+int rsxx_creg_write(struct rsxx_cardinfo *card,
+			u32 addr,
+			unsigned int size8,
+			void *data,
+			int byte_stream)
+{
+	return issue_creg_rw(card, addr, size8, data, byte_stream, 0);
+}
+
+int rsxx_creg_read(struct rsxx_cardinfo *card,
+		       u32 addr,
+		       unsigned int size8,
+		       void *data,
+		       int byte_stream)
+{
+	return issue_creg_rw(card, addr, size8, data, byte_stream, 1);
+}
+
+int rsxx_get_card_state(struct rsxx_cardinfo *card, unsigned int *state)
+{
+	return rsxx_creg_read(card, CREG_ADD_CARD_STATE,
+				  sizeof(*state), state, 0);
+}
+
+int rsxx_get_card_size8(struct rsxx_cardinfo *card, u64 *size8)
+{
+	unsigned int size;
+	int st;
+
+	st = rsxx_creg_read(card, CREG_ADD_CARD_SIZE,
+				sizeof(size), &size, 0);
+	if (st)
+		return st;
+
+	*size8 = (u64)size * RSXX_HW_BLK_SIZE;
+	return 0;
+}
+
+int rsxx_get_num_targets(struct rsxx_cardinfo *card,
+			     unsigned int *n_targets)
+{
+	return rsxx_creg_read(card, CREG_ADD_NUM_TARGETS,
+				  sizeof(*n_targets), n_targets, 0);
+}
+
+int rsxx_get_card_capabilities(struct rsxx_cardinfo *card,
+				   u32 *capabilities)
+{
+	return rsxx_creg_read(card, CREG_ADD_CAPABILITIES,
+				  sizeof(*capabilities), capabilities, 0);
+}
+
+int rsxx_issue_card_cmd(struct rsxx_cardinfo *card, u32 cmd)
+{
+	return rsxx_creg_write(card, CREG_ADD_CARD_CMD,
+				   sizeof(cmd), &cmd, 0);
+}
+
+
+/*----------------- HW Log Functions -------------------*/
+static void hw_log_msg(struct rsxx_cardinfo *card, const char *str, int len)
+{
+	static char level;
+
+	/*
+	 * New messages start with "<#>", where # is the log level. Messages
+	 * that extend past the log buffer will use the previous level
+	 */
+	if ((len > 3) && (str[0] == '<') && (str[2] == '>')) {
+		level = str[1];
+		str += 3; /* Skip past the log level. */
+		len -= 3;
+	}
+
+	switch (level) {
+	case '0':
+		dev_emerg(CARD_TO_DEV(card), "HW: %.*s", len, str);
+		break;
+	case '1':
+		dev_alert(CARD_TO_DEV(card), "HW: %.*s", len, str);
+		break;
+	case '2':
+		dev_crit(CARD_TO_DEV(card), "HW: %.*s", len, str);
+		break;
+	case '3':
+		dev_err(CARD_TO_DEV(card), "HW: %.*s", len, str);
+		break;
+	case '4':
+		dev_warn(CARD_TO_DEV(card), "HW: %.*s", len, str);
+		break;
+	case '5':
+		dev_notice(CARD_TO_DEV(card), "HW: %.*s", len, str);
+		break;
+	case '6':
+		dev_info(CARD_TO_DEV(card), "HW: %.*s", len, str);
+		break;
+	case '7':
+		dev_dbg(CARD_TO_DEV(card), "HW: %.*s", len, str);
+		break;
+	default:
+		dev_info(CARD_TO_DEV(card), "HW: %.*s", len, str);
+		break;
+	}
+}
+
+/*
+ * The substrncpy function copies the src string (which includes the
+ * terminating '\0' character), up to the count into the dest pointer.
+ * Returns the number of bytes copied to dest.
+ */
+static int substrncpy(char *dest, const char *src, int count)
+{
+	int max_cnt = count;
+
+	while (count) {
+		count--;
+		*dest = *src;
+		if (*dest == '\0')
+			break;
+		src++;
+		dest++;
+	}
+	return max_cnt - count;
+}
+
+
+static void read_hw_log_done(struct rsxx_cardinfo *card,
+			     struct creg_cmd *cmd,
+			     int st)
+{
+	char *buf;
+	char *log_str;
+	int cnt;
+	int len;
+	int off;
+
+	buf = cmd->buf;
+	off = 0;
+
+	/* Failed getting the log message */
+	if (st)
+		return;
+
+	while (off < cmd->cnt8) {
+		log_str = &card->log.buf[card->log.buf_len];
+		cnt = min(cmd->cnt8 - off, LOG_BUF_SIZE8 - card->log.buf_len);
+		len = substrncpy(log_str, &buf[off], cnt);
+
+		off += len;
+		card->log.buf_len += len;
+
+		/*
+		 * Flush the log if we've hit the end of a message or if we've
+		 * run out of buffer space.
+		 */
+		if ((log_str[len - 1] == '\0')  ||
+		    (card->log.buf_len == LOG_BUF_SIZE8)) {
+			if (card->log.buf_len != 1) /* Don't log blank lines. */
+				hw_log_msg(card, card->log.buf,
+					   card->log.buf_len);
+			card->log.buf_len = 0;
+		}
+
+	}
+
+	if (cmd->status & CREG_STAT_LOG_PENDING)
+		rsxx_read_hw_log(card);
+}
+
+int rsxx_read_hw_log(struct rsxx_cardinfo *card)
+{
+	int st;
+
+	st = creg_queue_cmd(card, CREG_OP_READ, CREG_ADD_LOG,
+			    sizeof(card->log.tmp), card->log.tmp,
+			    1, read_hw_log_done, NULL);
+	if (st)
+		dev_err(CARD_TO_DEV(card),
+			"Failed getting log text\n");
+
+	return st;
+}
+
+/*-------------- IOCTL REG Access ------------------*/
+static int issue_reg_cmd(struct rsxx_cardinfo *card,
+			 struct rsxx_reg_access *cmd,
+			 int read)
+{
+	unsigned int op = read ? CREG_OP_READ : CREG_OP_WRITE;
+
+	return __issue_creg_rw(card, op, cmd->addr, cmd->cnt, cmd->data,
+			       cmd->stream, &cmd->stat);
+}
+
+int rsxx_reg_access(struct rsxx_cardinfo *card,
+			struct rsxx_reg_access __user *ucmd,
+			int read)
+{
+	struct rsxx_reg_access cmd;
+	int st;
+
+	st = copy_from_user(&cmd, ucmd, sizeof(cmd));
+	if (st)
+		return -EFAULT;
+
+	if (cmd.cnt > RSXX_MAX_REG_CNT)
+		return -EFAULT;
+
+	st = issue_reg_cmd(card, &cmd, read);
+	if (st)
+		return st;
+
+	st = put_user(cmd.stat, &ucmd->stat);
+	if (st)
+		return -EFAULT;
+
+	if (read) {
+		st = copy_to_user(ucmd->data, cmd.data, cmd.cnt);
+		if (st)
+			return -EFAULT;
+	}
+
+	return 0;
+}
+
+void rsxx_eeh_save_issued_creg(struct rsxx_cardinfo *card)
+{
+	struct creg_cmd *cmd = NULL;
+
+	cmd = card->creg_ctrl.active_cmd;
+	card->creg_ctrl.active_cmd = NULL;
+
+	if (cmd) {
+		del_timer_sync(&card->creg_ctrl.cmd_timer);
+
+		spin_lock_bh(&card->creg_ctrl.lock);
+		list_add(&cmd->list, &card->creg_ctrl.queue);
+		card->creg_ctrl.q_depth++;
+		card->creg_ctrl.active = 0;
+		spin_unlock_bh(&card->creg_ctrl.lock);
+	}
+}
+
+void rsxx_kick_creg_queue(struct rsxx_cardinfo *card)
+{
+	spin_lock_bh(&card->creg_ctrl.lock);
+	if (!list_empty(&card->creg_ctrl.queue))
+		creg_kick_queue(card);
+	spin_unlock_bh(&card->creg_ctrl.lock);
+}
+
+/*------------ Initialization & Setup --------------*/
+int rsxx_creg_setup(struct rsxx_cardinfo *card)
+{
+	card->creg_ctrl.active_cmd = NULL;
+
+	card->creg_ctrl.creg_wq =
+			create_singlethread_workqueue(DRIVER_NAME"_creg");
+	if (!card->creg_ctrl.creg_wq)
+		return -ENOMEM;
+
+	INIT_WORK(&card->creg_ctrl.done_work, creg_cmd_done);
+	mutex_init(&card->creg_ctrl.reset_lock);
+	INIT_LIST_HEAD(&card->creg_ctrl.queue);
+	spin_lock_init(&card->creg_ctrl.lock);
+	setup_timer(&card->creg_ctrl.cmd_timer, creg_cmd_timed_out,
+		    (unsigned long) card);
+
+	return 0;
+}
+
+void rsxx_creg_destroy(struct rsxx_cardinfo *card)
+{
+	struct creg_cmd *cmd;
+	struct creg_cmd *tmp;
+	int cnt = 0;
+
+	/* Cancel outstanding commands */
+	spin_lock_bh(&card->creg_ctrl.lock);
+	list_for_each_entry_safe(cmd, tmp, &card->creg_ctrl.queue, list) {
+		list_del(&cmd->list);
+		if (cmd->cb)
+			cmd->cb(card, cmd, -ECANCELED);
+		kmem_cache_free(creg_cmd_pool, cmd);
+		cnt++;
+	}
+
+	if (cnt)
+		dev_info(CARD_TO_DEV(card),
+			"Canceled %d queue creg commands\n", cnt);
+
+	cmd = card->creg_ctrl.active_cmd;
+	card->creg_ctrl.active_cmd = NULL;
+	if (cmd) {
+		if (timer_pending(&card->creg_ctrl.cmd_timer))
+			del_timer_sync(&card->creg_ctrl.cmd_timer);
+
+		if (cmd->cb)
+			cmd->cb(card, cmd, -ECANCELED);
+		dev_info(CARD_TO_DEV(card),
+			"Canceled active creg command\n");
+		kmem_cache_free(creg_cmd_pool, cmd);
+	}
+	spin_unlock_bh(&card->creg_ctrl.lock);
+
+	cancel_work_sync(&card->creg_ctrl.done_work);
+}
+
+
+int rsxx_creg_init(void)
+{
+	creg_cmd_pool = KMEM_CACHE(creg_cmd, SLAB_HWCACHE_ALIGN);
+	if (!creg_cmd_pool)
+		return -ENOMEM;
+
+	return 0;
+}
+
+void rsxx_creg_cleanup(void)
+{
+	kmem_cache_destroy(creg_cmd_pool);
+}