From 1da177e4c3f41524e886b7f1b8a0c1fc7321cac2 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sat, 16 Apr 2005 15:20:36 -0700 Subject: Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip! --- drivers/md/Kconfig | 240 +++ drivers/md/Makefile | 107 ++ drivers/md/dm-bio-list.h | 68 + drivers/md/dm-bio-record.h | 45 + drivers/md/dm-crypt.c | 977 ++++++++++ drivers/md/dm-emc.c | 359 ++++ drivers/md/dm-exception-store.c | 648 +++++++ drivers/md/dm-hw-handler.c | 216 +++ drivers/md/dm-hw-handler.h | 61 + drivers/md/dm-io.c | 426 +++++ drivers/md/dm-io.h | 77 + drivers/md/dm-ioctl.c | 1416 +++++++++++++++ drivers/md/dm-linear.c | 123 ++ drivers/md/dm-log.c | 711 ++++++++ drivers/md/dm-log.h | 130 ++ drivers/md/dm-mpath.c | 1302 ++++++++++++++ drivers/md/dm-mpath.h | 25 + drivers/md/dm-path-selector.c | 156 ++ drivers/md/dm-path-selector.h | 93 + drivers/md/dm-raid1.c | 1269 +++++++++++++ drivers/md/dm-round-robin.c | 214 +++ drivers/md/dm-snap.c | 1208 +++++++++++++ drivers/md/dm-snap.h | 161 ++ drivers/md/dm-stripe.c | 234 +++ drivers/md/dm-table.c | 950 ++++++++++ drivers/md/dm-target.c | 196 ++ drivers/md/dm-zero.c | 81 + drivers/md/dm.c | 1194 +++++++++++++ drivers/md/dm.h | 195 ++ drivers/md/faulty.c | 343 ++++ drivers/md/kcopyd.c | 687 +++++++ drivers/md/kcopyd.h | 42 + drivers/md/linear.c | 343 ++++ drivers/md/md.c | 3766 +++++++++++++++++++++++++++++++++++++++ drivers/md/mktables.c | 125 ++ drivers/md/multipath.c | 584 ++++++ drivers/md/raid0.c | 539 ++++++ drivers/md/raid1.c | 1449 +++++++++++++++ drivers/md/raid10.c | 1787 +++++++++++++++++++ drivers/md/raid5.c | 1965 ++++++++++++++++++++ drivers/md/raid6.h | 135 ++ drivers/md/raid6algos.c | 153 ++ drivers/md/raid6altivec.uc | 122 ++ drivers/md/raid6int.uc | 117 ++ drivers/md/raid6main.c | 2136 ++++++++++++++++++++++ drivers/md/raid6mmx.c | 150 ++ drivers/md/raid6recov.c | 133 ++ drivers/md/raid6sse1.c | 171 ++ drivers/md/raid6sse2.c | 270 +++ drivers/md/raid6test/Makefile | 58 + drivers/md/raid6test/test.c | 103 ++ drivers/md/raid6x86.h | 245 +++ drivers/md/unroll.pl | 24 + drivers/md/xor.c | 154 ++ 54 files changed, 28483 insertions(+) create mode 100644 drivers/md/Kconfig create mode 100644 drivers/md/Makefile create mode 100644 drivers/md/dm-bio-list.h create mode 100644 drivers/md/dm-bio-record.h create mode 100644 drivers/md/dm-crypt.c create mode 100644 drivers/md/dm-emc.c create mode 100644 drivers/md/dm-exception-store.c create mode 100644 drivers/md/dm-hw-handler.c create mode 100644 drivers/md/dm-hw-handler.h create mode 100644 drivers/md/dm-io.c create mode 100644 drivers/md/dm-io.h create mode 100644 drivers/md/dm-ioctl.c create mode 100644 drivers/md/dm-linear.c create mode 100644 drivers/md/dm-log.c create mode 100644 drivers/md/dm-log.h create mode 100644 drivers/md/dm-mpath.c create mode 100644 drivers/md/dm-mpath.h create mode 100644 drivers/md/dm-path-selector.c create mode 100644 drivers/md/dm-path-selector.h create mode 100644 drivers/md/dm-raid1.c create mode 100644 drivers/md/dm-round-robin.c create mode 100644 drivers/md/dm-snap.c create mode 100644 drivers/md/dm-snap.h create mode 100644 drivers/md/dm-stripe.c create mode 100644 drivers/md/dm-table.c create mode 100644 drivers/md/dm-target.c create mode 100644 drivers/md/dm-zero.c create mode 100644 drivers/md/dm.c create mode 100644 drivers/md/dm.h create mode 100644 drivers/md/faulty.c create mode 100644 drivers/md/kcopyd.c create mode 100644 drivers/md/kcopyd.h create mode 100644 drivers/md/linear.c create mode 100644 drivers/md/md.c create mode 100644 drivers/md/mktables.c create mode 100644 drivers/md/multipath.c create mode 100644 drivers/md/raid0.c create mode 100644 drivers/md/raid1.c create mode 100644 drivers/md/raid10.c create mode 100644 drivers/md/raid5.c create mode 100644 drivers/md/raid6.h create mode 100644 drivers/md/raid6algos.c create mode 100644 drivers/md/raid6altivec.uc create mode 100644 drivers/md/raid6int.uc create mode 100644 drivers/md/raid6main.c create mode 100644 drivers/md/raid6mmx.c create mode 100644 drivers/md/raid6recov.c create mode 100644 drivers/md/raid6sse1.c create mode 100644 drivers/md/raid6sse2.c create mode 100644 drivers/md/raid6test/Makefile create mode 100644 drivers/md/raid6test/test.c create mode 100644 drivers/md/raid6x86.h create mode 100644 drivers/md/unroll.pl create mode 100644 drivers/md/xor.c (limited to 'drivers/md') diff --git a/drivers/md/Kconfig b/drivers/md/Kconfig new file mode 100644 index 000000000000..ac43f98062fd --- /dev/null +++ b/drivers/md/Kconfig @@ -0,0 +1,240 @@ +# +# Block device driver configuration +# + +menu "Multi-device support (RAID and LVM)" + +config MD + bool "Multiple devices driver support (RAID and LVM)" + help + Support multiple physical spindles through a single logical device. + Required for RAID and logical volume management. + +config BLK_DEV_MD + tristate "RAID support" + depends on MD + ---help--- + This driver lets you combine several hard disk partitions into one + logical block device. This can be used to simply append one + partition to another one or to combine several redundant hard disks + into a RAID1/4/5 device so as to provide protection against hard + disk failures. This is called "Software RAID" since the combining of + the partitions is done by the kernel. "Hardware RAID" means that the + combining is done by a dedicated controller; if you have such a + controller, you do not need to say Y here. + + More information about Software RAID on Linux is contained in the + Software RAID mini-HOWTO, available from + . There you will also learn + where to get the supporting user space utilities raidtools. + + If unsure, say N. + +config MD_LINEAR + tristate "Linear (append) mode" + depends on BLK_DEV_MD + ---help--- + If you say Y here, then your multiple devices driver will be able to + use the so-called linear mode, i.e. it will combine the hard disk + partitions by simply appending one to the other. + + To compile this as a module, choose M here: the module + will be called linear. + + If unsure, say Y. + +config MD_RAID0 + tristate "RAID-0 (striping) mode" + depends on BLK_DEV_MD + ---help--- + If you say Y here, then your multiple devices driver will be able to + use the so-called raid0 mode, i.e. it will combine the hard disk + partitions into one logical device in such a fashion as to fill them + up evenly, one chunk here and one chunk there. This will increase + the throughput rate if the partitions reside on distinct disks. + + Information about Software RAID on Linux is contained in the + Software-RAID mini-HOWTO, available from + . There you will also + learn where to get the supporting user space utilities raidtools. + + To compile this as a module, choose M here: the module + will be called raid0. + + If unsure, say Y. + +config MD_RAID1 + tristate "RAID-1 (mirroring) mode" + depends on BLK_DEV_MD + ---help--- + A RAID-1 set consists of several disk drives which are exact copies + of each other. In the event of a mirror failure, the RAID driver + will continue to use the operational mirrors in the set, providing + an error free MD (multiple device) to the higher levels of the + kernel. In a set with N drives, the available space is the capacity + of a single drive, and the set protects against a failure of (N - 1) + drives. + + Information about Software RAID on Linux is contained in the + Software-RAID mini-HOWTO, available from + . There you will also + learn where to get the supporting user space utilities raidtools. + + If you want to use such a RAID-1 set, say Y. To compile this code + as a module, choose M here: the module will be called raid1. + + If unsure, say Y. + +config MD_RAID10 + tristate "RAID-10 (mirrored striping) mode (EXPERIMENTAL)" + depends on BLK_DEV_MD && EXPERIMENTAL + ---help--- + RAID-10 provides a combination of striping (RAID-0) and + mirroring (RAID-1) with easier configuration and more flexable + layout. + Unlike RAID-0, but like RAID-1, RAID-10 requires all devices to + be the same size (or at least, only as much as the smallest device + will be used). + RAID-10 provides a variety of layouts that provide different levels + of redundancy and performance. + + RAID-10 requires mdadm-1.7.0 or later, available at: + + ftp://ftp.kernel.org/pub/linux/utils/raid/mdadm/ + + If unsure, say Y. + +config MD_RAID5 + tristate "RAID-4/RAID-5 mode" + depends on BLK_DEV_MD + ---help--- + A RAID-5 set of N drives with a capacity of C MB per drive provides + the capacity of C * (N - 1) MB, and protects against a failure + of a single drive. For a given sector (row) number, (N - 1) drives + contain data sectors, and one drive contains the parity protection. + For a RAID-4 set, the parity blocks are present on a single drive, + while a RAID-5 set distributes the parity across the drives in one + of the available parity distribution methods. + + Information about Software RAID on Linux is contained in the + Software-RAID mini-HOWTO, available from + . There you will also + learn where to get the supporting user space utilities raidtools. + + If you want to use such a RAID-4/RAID-5 set, say Y. To + compile this code as a module, choose M here: the module + will be called raid5. + + If unsure, say Y. + +config MD_RAID6 + tristate "RAID-6 mode" + depends on BLK_DEV_MD + ---help--- + A RAID-6 set of N drives with a capacity of C MB per drive + provides the capacity of C * (N - 2) MB, and protects + against a failure of any two drives. For a given sector + (row) number, (N - 2) drives contain data sectors, and two + drives contains two independent redundancy syndromes. Like + RAID-5, RAID-6 distributes the syndromes across the drives + in one of the available parity distribution methods. + + RAID-6 requires mdadm-1.5.0 or later, available at: + + ftp://ftp.kernel.org/pub/linux/utils/raid/mdadm/ + + If you want to use such a RAID-6 set, say Y. To compile + this code as a module, choose M here: the module will be + called raid6. + + If unsure, say Y. + +config MD_MULTIPATH + tristate "Multipath I/O support" + depends on BLK_DEV_MD + help + Multipath-IO is the ability of certain devices to address the same + physical disk over multiple 'IO paths'. The code ensures that such + paths can be defined and handled at runtime, and ensures that a + transparent failover to the backup path(s) happens if a IO errors + arrives on the primary path. + + If unsure, say N. + +config MD_FAULTY + tristate "Faulty test module for MD" + depends on BLK_DEV_MD + help + The "faulty" module allows for a block device that occasionally returns + read or write errors. It is useful for testing. + + In unsure, say N. + +config BLK_DEV_DM + tristate "Device mapper support" + depends on MD + ---help--- + Device-mapper is a low level volume manager. It works by allowing + people to specify mappings for ranges of logical sectors. Various + mapping types are available, in addition people may write their own + modules containing custom mappings if they wish. + + Higher level volume managers such as LVM2 use this driver. + + To compile this as a module, choose M here: the module will be + called dm-mod. + + If unsure, say N. + +config DM_CRYPT + tristate "Crypt target support" + depends on BLK_DEV_DM && EXPERIMENTAL + select CRYPTO + ---help--- + This device-mapper target allows you to create a device that + transparently encrypts the data on it. You'll need to activate + the ciphers you're going to use in the cryptoapi configuration. + + Information on how to use dm-crypt can be found on + + + + To compile this code as a module, choose M here: the module will + be called dm-crypt. + + If unsure, say N. + +config DM_SNAPSHOT + tristate "Snapshot target (EXPERIMENTAL)" + depends on BLK_DEV_DM && EXPERIMENTAL + ---help--- + Allow volume managers to take writeable snapshots of a device. + +config DM_MIRROR + tristate "Mirror target (EXPERIMENTAL)" + depends on BLK_DEV_DM && EXPERIMENTAL + ---help--- + Allow volume managers to mirror logical volumes, also + needed for live data migration tools such as 'pvmove'. + +config DM_ZERO + tristate "Zero target (EXPERIMENTAL)" + depends on BLK_DEV_DM && EXPERIMENTAL + ---help--- + A target that discards writes, and returns all zeroes for + reads. Useful in some recovery situations. + +config DM_MULTIPATH + tristate "Multipath target (EXPERIMENTAL)" + depends on BLK_DEV_DM && EXPERIMENTAL + ---help--- + Allow volume managers to support multipath hardware. + +config DM_MULTIPATH_EMC + tristate "EMC CX/AX multipath support (EXPERIMENTAL)" + depends on DM_MULTIPATH && BLK_DEV_DM && EXPERIMENTAL + ---help--- + Multipath support for EMC CX/AX series hardware. + +endmenu + diff --git a/drivers/md/Makefile b/drivers/md/Makefile new file mode 100644 index 000000000000..90de9c146a5f --- /dev/null +++ b/drivers/md/Makefile @@ -0,0 +1,107 @@ +# +# Makefile for the kernel software RAID and LVM drivers. +# + +dm-mod-objs := dm.o dm-table.o dm-target.o dm-linear.o dm-stripe.o \ + dm-ioctl.o dm-io.o kcopyd.o +dm-multipath-objs := dm-hw-handler.o dm-path-selector.o dm-mpath.o +dm-snapshot-objs := dm-snap.o dm-exception-store.o +dm-mirror-objs := dm-log.o dm-raid1.o +raid6-objs := raid6main.o raid6algos.o raid6recov.o raid6tables.o \ + raid6int1.o raid6int2.o raid6int4.o \ + raid6int8.o raid6int16.o raid6int32.o \ + raid6altivec1.o raid6altivec2.o raid6altivec4.o \ + raid6altivec8.o \ + raid6mmx.o raid6sse1.o raid6sse2.o +hostprogs-y := mktables + +# Note: link order is important. All raid personalities +# and xor.o must come before md.o, as they each initialise +# themselves, and md.o may use the personalities when it +# auto-initialised. + +obj-$(CONFIG_MD_LINEAR) += linear.o +obj-$(CONFIG_MD_RAID0) += raid0.o +obj-$(CONFIG_MD_RAID1) += raid1.o +obj-$(CONFIG_MD_RAID10) += raid10.o +obj-$(CONFIG_MD_RAID5) += raid5.o xor.o +obj-$(CONFIG_MD_RAID6) += raid6.o xor.o +obj-$(CONFIG_MD_MULTIPATH) += multipath.o +obj-$(CONFIG_MD_FAULTY) += faulty.o +obj-$(CONFIG_BLK_DEV_MD) += md.o +obj-$(CONFIG_BLK_DEV_DM) += dm-mod.o +obj-$(CONFIG_DM_CRYPT) += dm-crypt.o +obj-$(CONFIG_DM_MULTIPATH) += dm-multipath.o dm-round-robin.o +obj-$(CONFIG_DM_MULTIPATH_EMC) += dm-emc.o +obj-$(CONFIG_DM_SNAPSHOT) += dm-snapshot.o +obj-$(CONFIG_DM_MIRROR) += dm-mirror.o +obj-$(CONFIG_DM_ZERO) += dm-zero.o + +quiet_cmd_unroll = UNROLL $@ + cmd_unroll = $(PERL) $(srctree)/$(src)/unroll.pl $(UNROLL) \ + < $< > $@ || ( rm -f $@ && exit 1 ) + +ifeq ($(CONFIG_ALTIVEC),y) +altivec_flags := -maltivec -mabi=altivec +endif + +targets += raid6int1.c +$(obj)/raid6int1.c: UNROLL := 1 +$(obj)/raid6int1.c: $(src)/raid6int.uc $(src)/unroll.pl FORCE + $(call if_changed,unroll) + +targets += raid6int2.c +$(obj)/raid6int2.c: UNROLL := 2 +$(obj)/raid6int2.c: $(src)/raid6int.uc $(src)/unroll.pl FORCE + $(call if_changed,unroll) + +targets += raid6int4.c +$(obj)/raid6int4.c: UNROLL := 4 +$(obj)/raid6int4.c: $(src)/raid6int.uc $(src)/unroll.pl FORCE + $(call if_changed,unroll) + +targets += raid6int8.c +$(obj)/raid6int8.c: UNROLL := 8 +$(obj)/raid6int8.c: $(src)/raid6int.uc $(src)/unroll.pl FORCE + $(call if_changed,unroll) + +targets += raid6int16.c +$(obj)/raid6int16.c: UNROLL := 16 +$(obj)/raid6int16.c: $(src)/raid6int.uc $(src)/unroll.pl FORCE + $(call if_changed,unroll) + +targets += raid6int32.c +$(obj)/raid6int32.c: UNROLL := 32 +$(obj)/raid6int32.c: $(src)/raid6int.uc $(src)/unroll.pl FORCE + $(call if_changed,unroll) + +CFLAGS_raid6altivec1.o += $(altivec_flags) +targets += raid6altivec1.c +$(obj)/raid6altivec1.c: UNROLL := 1 +$(obj)/raid6altivec1.c: $(src)/raid6altivec.uc $(src)/unroll.pl FORCE + $(call if_changed,unroll) + +CFLAGS_raid6altivec2.o += $(altivec_flags) +targets += raid6altivec2.c +$(obj)/raid6altivec2.c: UNROLL := 2 +$(obj)/raid6altivec2.c: $(src)/raid6altivec.uc $(src)/unroll.pl FORCE + $(call if_changed,unroll) + +CFLAGS_raid6altivec4.o += $(altivec_flags) +targets += raid6altivec4.c +$(obj)/raid6altivec4.c: UNROLL := 4 +$(obj)/raid6altivec4.c: $(src)/raid6altivec.uc $(src)/unroll.pl FORCE + $(call if_changed,unroll) + +CFLAGS_raid6altivec8.o += $(altivec_flags) +targets += raid6altivec8.c +$(obj)/raid6altivec8.c: UNROLL := 8 +$(obj)/raid6altivec8.c: $(src)/raid6altivec.uc $(src)/unroll.pl FORCE + $(call if_changed,unroll) + +quiet_cmd_mktable = TABLE $@ + cmd_mktable = $(obj)/mktables > $@ || ( rm -f $@ && exit 1 ) + +targets += raid6tables.c +$(obj)/raid6tables.c: $(obj)/mktables FORCE + $(call if_changed,mktable) diff --git a/drivers/md/dm-bio-list.h b/drivers/md/dm-bio-list.h new file mode 100644 index 000000000000..bc021e1fd4d1 --- /dev/null +++ b/drivers/md/dm-bio-list.h @@ -0,0 +1,68 @@ +/* + * Copyright (C) 2004 Red Hat UK Ltd. + * + * This file is released under the GPL. + */ + +#ifndef DM_BIO_LIST_H +#define DM_BIO_LIST_H + +#include + +struct bio_list { + struct bio *head; + struct bio *tail; +}; + +static inline void bio_list_init(struct bio_list *bl) +{ + bl->head = bl->tail = NULL; +} + +static inline void bio_list_add(struct bio_list *bl, struct bio *bio) +{ + bio->bi_next = NULL; + + if (bl->tail) + bl->tail->bi_next = bio; + else + bl->head = bio; + + bl->tail = bio; +} + +static inline void bio_list_merge(struct bio_list *bl, struct bio_list *bl2) +{ + if (bl->tail) + bl->tail->bi_next = bl2->head; + else + bl->head = bl2->head; + + bl->tail = bl2->tail; +} + +static inline struct bio *bio_list_pop(struct bio_list *bl) +{ + struct bio *bio = bl->head; + + if (bio) { + bl->head = bl->head->bi_next; + if (!bl->head) + bl->tail = NULL; + + bio->bi_next = NULL; + } + + return bio; +} + +static inline struct bio *bio_list_get(struct bio_list *bl) +{ + struct bio *bio = bl->head; + + bl->head = bl->tail = NULL; + + return bio; +} + +#endif diff --git a/drivers/md/dm-bio-record.h b/drivers/md/dm-bio-record.h new file mode 100644 index 000000000000..d3ec217847d6 --- /dev/null +++ b/drivers/md/dm-bio-record.h @@ -0,0 +1,45 @@ +/* + * Copyright (C) 2004-2005 Red Hat, Inc. All rights reserved. + * + * This file is released under the GPL. + */ + +#ifndef DM_BIO_RECORD_H +#define DM_BIO_RECORD_H + +#include + +/* + * There are lots of mutable fields in the bio struct that get + * changed by the lower levels of the block layer. Some targets, + * such as multipath, may wish to resubmit a bio on error. The + * functions in this file help the target record and restore the + * original bio state. + */ +struct dm_bio_details { + sector_t bi_sector; + struct block_device *bi_bdev; + unsigned int bi_size; + unsigned short bi_idx; + unsigned long bi_flags; +}; + +static inline void dm_bio_record(struct dm_bio_details *bd, struct bio *bio) +{ + bd->bi_sector = bio->bi_sector; + bd->bi_bdev = bio->bi_bdev; + bd->bi_size = bio->bi_size; + bd->bi_idx = bio->bi_idx; + bd->bi_flags = bio->bi_flags; +} + +static inline void dm_bio_restore(struct dm_bio_details *bd, struct bio *bio) +{ + bio->bi_sector = bd->bi_sector; + bio->bi_bdev = bd->bi_bdev; + bio->bi_size = bd->bi_size; + bio->bi_idx = bd->bi_idx; + bio->bi_flags = bd->bi_flags; +} + +#endif diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c new file mode 100644 index 000000000000..77619a56e2bf --- /dev/null +++ b/drivers/md/dm-crypt.c @@ -0,0 +1,977 @@ +/* + * Copyright (C) 2003 Christophe Saout + * Copyright (C) 2004 Clemens Fruhwirth + * + * This file is released under the GPL. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include "dm.h" + +#define PFX "crypt: " + +/* + * per bio private data + */ +struct crypt_io { + struct dm_target *target; + struct bio *bio; + struct bio *first_clone; + struct work_struct work; + atomic_t pending; + int error; +}; + +/* + * context holding the current state of a multi-part conversion + */ +struct convert_context { + struct bio *bio_in; + struct bio *bio_out; + unsigned int offset_in; + unsigned int offset_out; + unsigned int idx_in; + unsigned int idx_out; + sector_t sector; + int write; +}; + +struct crypt_config; + +struct crypt_iv_operations { + int (*ctr)(struct crypt_config *cc, struct dm_target *ti, + const char *opts); + void (*dtr)(struct crypt_config *cc); + const char *(*status)(struct crypt_config *cc); + int (*generator)(struct crypt_config *cc, u8 *iv, sector_t sector); +}; + +/* + * Crypt: maps a linear range of a block device + * and encrypts / decrypts at the same time. + */ +struct crypt_config { + struct dm_dev *dev; + sector_t start; + + /* + * pool for per bio private data and + * for encryption buffer pages + */ + mempool_t *io_pool; + mempool_t *page_pool; + + /* + * crypto related data + */ + struct crypt_iv_operations *iv_gen_ops; + char *iv_mode; + void *iv_gen_private; + sector_t iv_offset; + unsigned int iv_size; + + struct crypto_tfm *tfm; + unsigned int key_size; + u8 key[0]; +}; + +#define MIN_IOS 256 +#define MIN_POOL_PAGES 32 +#define MIN_BIO_PAGES 8 + +static kmem_cache_t *_crypt_io_pool; + +/* + * Mempool alloc and free functions for the page + */ +static void *mempool_alloc_page(unsigned int __nocast gfp_mask, void *data) +{ + return alloc_page(gfp_mask); +} + +static void mempool_free_page(void *page, void *data) +{ + __free_page(page); +} + + +/* + * Different IV generation algorithms: + * + * plain: the initial vector is the 32-bit low-endian version of the sector + * number, padded with zeros if neccessary. + * + * ess_iv: "encrypted sector|salt initial vector", the sector number is + * encrypted with the bulk cipher using a salt as key. The salt + * should be derived from the bulk cipher's key via hashing. + * + * plumb: unimplemented, see: + * http://article.gmane.org/gmane.linux.kernel.device-mapper.dm-crypt/454 + */ + +static int crypt_iv_plain_gen(struct crypt_config *cc, u8 *iv, sector_t sector) +{ + memset(iv, 0, cc->iv_size); + *(u32 *)iv = cpu_to_le32(sector & 0xffffffff); + + return 0; +} + +static int crypt_iv_essiv_ctr(struct crypt_config *cc, struct dm_target *ti, + const char *opts) +{ + struct crypto_tfm *essiv_tfm; + struct crypto_tfm *hash_tfm; + struct scatterlist sg; + unsigned int saltsize; + u8 *salt; + + if (opts == NULL) { + ti->error = PFX "Digest algorithm missing for ESSIV mode"; + return -EINVAL; + } + + /* Hash the cipher key with the given hash algorithm */ + hash_tfm = crypto_alloc_tfm(opts, 0); + if (hash_tfm == NULL) { + ti->error = PFX "Error initializing ESSIV hash"; + return -EINVAL; + } + + if (crypto_tfm_alg_type(hash_tfm) != CRYPTO_ALG_TYPE_DIGEST) { + ti->error = PFX "Expected digest algorithm for ESSIV hash"; + crypto_free_tfm(hash_tfm); + return -EINVAL; + } + + saltsize = crypto_tfm_alg_digestsize(hash_tfm); + salt = kmalloc(saltsize, GFP_KERNEL); + if (salt == NULL) { + ti->error = PFX "Error kmallocing salt storage in ESSIV"; + crypto_free_tfm(hash_tfm); + return -ENOMEM; + } + + sg.page = virt_to_page(cc->key); + sg.offset = offset_in_page(cc->key); + sg.length = cc->key_size; + crypto_digest_digest(hash_tfm, &sg, 1, salt); + crypto_free_tfm(hash_tfm); + + /* Setup the essiv_tfm with the given salt */ + essiv_tfm = crypto_alloc_tfm(crypto_tfm_alg_name(cc->tfm), + CRYPTO_TFM_MODE_ECB); + if (essiv_tfm == NULL) { + ti->error = PFX "Error allocating crypto tfm for ESSIV"; + kfree(salt); + return -EINVAL; + } + if (crypto_tfm_alg_blocksize(essiv_tfm) + != crypto_tfm_alg_ivsize(cc->tfm)) { + ti->error = PFX "Block size of ESSIV cipher does " + "not match IV size of block cipher"; + crypto_free_tfm(essiv_tfm); + kfree(salt); + return -EINVAL; + } + if (crypto_cipher_setkey(essiv_tfm, salt, saltsize) < 0) { + ti->error = PFX "Failed to set key for ESSIV cipher"; + crypto_free_tfm(essiv_tfm); + kfree(salt); + return -EINVAL; + } + kfree(salt); + + cc->iv_gen_private = (void *)essiv_tfm; + return 0; +} + +static void crypt_iv_essiv_dtr(struct crypt_config *cc) +{ + crypto_free_tfm((struct crypto_tfm *)cc->iv_gen_private); + cc->iv_gen_private = NULL; +} + +static int crypt_iv_essiv_gen(struct crypt_config *cc, u8 *iv, sector_t sector) +{ + struct scatterlist sg = { NULL, }; + + memset(iv, 0, cc->iv_size); + *(u64 *)iv = cpu_to_le64(sector); + + sg.page = virt_to_page(iv); + sg.offset = offset_in_page(iv); + sg.length = cc->iv_size; + crypto_cipher_encrypt((struct crypto_tfm *)cc->iv_gen_private, + &sg, &sg, cc->iv_size); + + return 0; +} + +static struct crypt_iv_operations crypt_iv_plain_ops = { + .generator = crypt_iv_plain_gen +}; + +static struct crypt_iv_operations crypt_iv_essiv_ops = { + .ctr = crypt_iv_essiv_ctr, + .dtr = crypt_iv_essiv_dtr, + .generator = crypt_iv_essiv_gen +}; + + +static inline int +crypt_convert_scatterlist(struct crypt_config *cc, struct scatterlist *out, + struct scatterlist *in, unsigned int length, + int write, sector_t sector) +{ + u8 iv[cc->iv_size]; + int r; + + if (cc->iv_gen_ops) { + r = cc->iv_gen_ops->generator(cc, iv, sector); + if (r < 0) + return r; + + if (write) + r = crypto_cipher_encrypt_iv(cc->tfm, out, in, length, iv); + else + r = crypto_cipher_decrypt_iv(cc->tfm, out, in, length, iv); + } else { + if (write) + r = crypto_cipher_encrypt(cc->tfm, out, in, length); + else + r = crypto_cipher_decrypt(cc->tfm, out, in, length); + } + + return r; +} + +static void +crypt_convert_init(struct crypt_config *cc, struct convert_context *ctx, + struct bio *bio_out, struct bio *bio_in, + sector_t sector, int write) +{ + ctx->bio_in = bio_in; + ctx->bio_out = bio_out; + ctx->offset_in = 0; + ctx->offset_out = 0; + ctx->idx_in = bio_in ? bio_in->bi_idx : 0; + ctx->idx_out = bio_out ? bio_out->bi_idx : 0; + ctx->sector = sector + cc->iv_offset; + ctx->write = write; +} + +/* + * Encrypt / decrypt data from one bio to another one (can be the same one) + */ +static int crypt_convert(struct crypt_config *cc, + struct convert_context *ctx) +{ + int r = 0; + + while(ctx->idx_in < ctx->bio_in->bi_vcnt && + ctx->idx_out < ctx->bio_out->bi_vcnt) { + struct bio_vec *bv_in = bio_iovec_idx(ctx->bio_in, ctx->idx_in); + struct bio_vec *bv_out = bio_iovec_idx(ctx->bio_out, ctx->idx_out); + struct scatterlist sg_in = { + .page = bv_in->bv_page, + .offset = bv_in->bv_offset + ctx->offset_in, + .length = 1 << SECTOR_SHIFT + }; + struct scatterlist sg_out = { + .page = bv_out->bv_page, + .offset = bv_out->bv_offset + ctx->offset_out, + .length = 1 << SECTOR_SHIFT + }; + + ctx->offset_in += sg_in.length; + if (ctx->offset_in >= bv_in->bv_len) { + ctx->offset_in = 0; + ctx->idx_in++; + } + + ctx->offset_out += sg_out.length; + if (ctx->offset_out >= bv_out->bv_len) { + ctx->offset_out = 0; + ctx->idx_out++; + } + + r = crypt_convert_scatterlist(cc, &sg_out, &sg_in, sg_in.length, + ctx->write, ctx->sector); + if (r < 0) + break; + + ctx->sector++; + } + + return r; +} + +/* + * Generate a new unfragmented bio with the given size + * This should never violate the device limitations + * May return a smaller bio when running out of pages + */ +static struct bio * +crypt_alloc_buffer(struct crypt_config *cc, unsigned int size, + struct bio *base_bio, unsigned int *bio_vec_idx) +{ + struct bio *bio; + unsigned int nr_iovecs = (size + PAGE_SIZE - 1) >> PAGE_SHIFT; + int gfp_mask = GFP_NOIO | __GFP_HIGHMEM; + unsigned long flags = current->flags; + unsigned int i; + + /* + * Tell VM to act less aggressively and fail earlier. + * This is not necessary but increases throughput. + * FIXME: Is this really intelligent? + */ + current->flags &= ~PF_MEMALLOC; + + if (base_bio) + bio = bio_clone(base_bio, GFP_NOIO); + else + bio = bio_alloc(GFP_NOIO, nr_iovecs); + if (!bio) { + if (flags & PF_MEMALLOC) + current->flags |= PF_MEMALLOC; + return NULL; + } + + /* if the last bio was not complete, continue where that one ended */ + bio->bi_idx = *bio_vec_idx; + bio->bi_vcnt = *bio_vec_idx; + bio->bi_size = 0; + bio->bi_flags &= ~(1 << BIO_SEG_VALID); + + /* bio->bi_idx pages have already been allocated */ + size -= bio->bi_idx * PAGE_SIZE; + + for(i = bio->bi_idx; i < nr_iovecs; i++) { + struct bio_vec *bv = bio_iovec_idx(bio, i); + + bv->bv_page = mempool_alloc(cc->page_pool, gfp_mask); + if (!bv->bv_page) + break; + + /* + * if additional pages cannot be allocated without waiting, + * return a partially allocated bio, the caller will then try + * to allocate additional bios while submitting this partial bio + */ + if ((i - bio->bi_idx) == (MIN_BIO_PAGES - 1)) + gfp_mask = (gfp_mask | __GFP_NOWARN) & ~__GFP_WAIT; + + bv->bv_offset = 0; + if (size > PAGE_SIZE) + bv->bv_len = PAGE_SIZE; + else + bv->bv_len = size; + + bio->bi_size += bv->bv_len; + bio->bi_vcnt++; + size -= bv->bv_len; + } + + if (flags & PF_MEMALLOC) + current->flags |= PF_MEMALLOC; + + if (!bio->bi_size) { + bio_put(bio); + return NULL; + } + + /* + * Remember the last bio_vec allocated to be able + * to correctly continue after the splitting. + */ + *bio_vec_idx = bio->bi_vcnt; + + return bio; +} + +static void crypt_free_buffer_pages(struct crypt_config *cc, + struct bio *bio, unsigned int bytes) +{ + unsigned int i, start, end; + struct bio_vec *bv; + + /* + * This is ugly, but Jens Axboe thinks that using bi_idx in the + * endio function is too dangerous at the moment, so I calculate the + * correct position using bi_vcnt and bi_size. + * The bv_offset and bv_len fields might already be modified but we + * know that we always allocated whole pages. + * A fix to the bi_idx issue in the kernel is in the works, so + * we will hopefully be able to revert to the cleaner solution soon. + */ + i = bio->bi_vcnt - 1; + bv = bio_iovec_idx(bio, i); + end = (i << PAGE_SHIFT) + (bv->bv_offset + bv->bv_len) - bio->bi_size; + start = end - bytes; + + start >>= PAGE_SHIFT; + if (!bio->bi_size) + end = bio->bi_vcnt; + else + end >>= PAGE_SHIFT; + + for(i = start; i < end; i++) { + bv = bio_iovec_idx(bio, i); + BUG_ON(!bv->bv_page); + mempool_free(bv->bv_page, cc->page_pool); + bv->bv_page = NULL; + } +} + +/* + * One of the bios was finished. Check for completion of + * the whole request and correctly clean up the buffer. + */ +static void dec_pending(struct crypt_io *io, int error) +{ + struct crypt_config *cc = (struct crypt_config *) io->target->private; + + if (error < 0) + io->error = error; + + if (!atomic_dec_and_test(&io->pending)) + return; + + if (io->first_clone) + bio_put(io->first_clone); + + bio_endio(io->bio, io->bio->bi_size, io->error); + + mempool_free(io, cc->io_pool); +} + +/* + * kcryptd: + * + * Needed because it would be very unwise to do decryption in an + * interrupt context, so bios returning from read requests get + * queued here. + */ +static struct workqueue_struct *_kcryptd_workqueue; + +static void kcryptd_do_work(void *data) +{ + struct crypt_io *io = (struct crypt_io *) data; + struct crypt_config *cc = (struct crypt_config *) io->target->private; + struct convert_context ctx; + int r; + + crypt_convert_init(cc, &ctx, io->bio, io->bio, + io->bio->bi_sector - io->target->begin, 0); + r = crypt_convert(cc, &ctx); + + dec_pending(io, r); +} + +static void kcryptd_queue_io(struct crypt_io *io) +{ + INIT_WORK(&io->work, kcryptd_do_work, io); + queue_work(_kcryptd_workqueue, &io->work); +} + +/* + * Decode key from its hex representation + */ +static int crypt_decode_key(u8 *key, char *hex, unsigned int size) +{ + char buffer[3]; + char *endp; + unsigned int i; + + buffer[2] = '\0'; + + for(i = 0; i < size; i++) { + buffer[0] = *hex++; + buffer[1] = *hex++; + + key[i] = (u8)simple_strtoul(buffer, &endp, 16); + + if (endp != &buffer[2]) + return -EINVAL; + } + + if (*hex != '\0') + return -EINVAL; + + return 0; +} + +/* + * Encode key into its hex representation + */ +static void crypt_encode_key(char *hex, u8 *key, unsigned int size) +{ + unsigned int i; + + for(i = 0; i < size; i++) { + sprintf(hex, "%02x", *key); + hex += 2; + key++; + } +} + +/* + * Construct an encryption mapping: + * + */ +static int crypt_ctr(struct dm_target *ti, unsigned int argc, char **argv) +{ + struct crypt_config *cc; + struct crypto_tfm *tfm; + char *tmp; + char *cipher; + char *chainmode; + char *ivmode; + char *ivopts; + unsigned int crypto_flags; + unsigned int key_size; + + if (argc != 5) { + ti->error = PFX "Not enough arguments"; + return -EINVAL; + } + + tmp = argv[0]; + cipher = strsep(&tmp, "-"); + chainmode = strsep(&tmp, "-"); + ivopts = strsep(&tmp, "-"); + ivmode = strsep(&ivopts, ":"); + + if (tmp) + DMWARN(PFX "Unexpected additional cipher options"); + + key_size = strlen(argv[1]) >> 1; + + cc = kmalloc(sizeof(*cc) + key_size * sizeof(u8), GFP_KERNEL); + if (cc == NULL) { + ti->error = + PFX "Cannot allocate transparent encryption context"; + return -ENOMEM; + } + + cc->key_size = key_size; + if ((!key_size && strcmp(argv[1], "-") != 0) || + (key_size && crypt_decode_key(cc->key, argv[1], key_size) < 0)) { + ti->error = PFX "Error decoding key"; + goto bad1; + } + + /* Compatiblity mode for old dm-crypt cipher strings */ + if (!chainmode || (strcmp(chainmode, "plain") == 0 && !ivmode)) { + chainmode = "cbc"; + ivmode = "plain"; + } + + /* Choose crypto_flags according to chainmode */ + if (strcmp(chainmode, "cbc") == 0) + crypto_flags = CRYPTO_TFM_MODE_CBC; + else if (strcmp(chainmode, "ecb") == 0) + crypto_flags = CRYPTO_TFM_MODE_ECB; + else { + ti->error = PFX "Unknown chaining mode"; + goto bad1; + } + + if (crypto_flags != CRYPTO_TFM_MODE_ECB && !ivmode) { + ti->error = PFX "This chaining mode requires an IV mechanism"; + goto bad1; + } + + tfm = crypto_alloc_tfm(cipher, crypto_flags); + if (!tfm) { + ti->error = PFX "Error allocating crypto tfm"; + goto bad1; + } + if (crypto_tfm_alg_type(tfm) != CRYPTO_ALG_TYPE_CIPHER) { + ti->error = PFX "Expected cipher algorithm"; + goto bad2; + } + + cc->tfm = tfm; + + /* + * Choose ivmode. Valid modes: "plain", "essiv:". + * See comments at iv code + */ + + if (ivmode == NULL) + cc->iv_gen_ops = NULL; + else if (strcmp(ivmode, "plain") == 0) + cc->iv_gen_ops = &crypt_iv_plain_ops; + else if (strcmp(ivmode, "essiv") == 0) + cc->iv_gen_ops = &crypt_iv_essiv_ops; + else { + ti->error = PFX "Invalid IV mode"; + goto bad2; + } + + if (cc->iv_gen_ops && cc->iv_gen_ops->ctr && + cc->iv_gen_ops->ctr(cc, ti, ivopts) < 0) + goto bad2; + + if (tfm->crt_cipher.cit_decrypt_iv && tfm->crt_cipher.cit_encrypt_iv) + /* at least a 64 bit sector number should fit in our buffer */ + cc->iv_size = max(crypto_tfm_alg_ivsize(tfm), + (unsigned int)(sizeof(u64) / sizeof(u8))); + else { + cc->iv_size = 0; + if (cc->iv_gen_ops) { + DMWARN(PFX "Selected cipher does not support IVs"); + if (cc->iv_gen_ops->dtr) + cc->iv_gen_ops->dtr(cc); + cc->iv_gen_ops = NULL; + } + } + + cc->io_pool = mempool_create(MIN_IOS, mempool_alloc_slab, + mempool_free_slab, _crypt_io_pool); + if (!cc->io_pool) { + ti->error = PFX "Cannot allocate crypt io mempool"; + goto bad3; + } + + cc->page_pool = mempool_create(MIN_POOL_PAGES, mempool_alloc_page, + mempool_free_page, NULL); + if (!cc->page_pool) { + ti->error = PFX "Cannot allocate page mempool"; + goto bad4; + } + + if (tfm->crt_cipher.cit_setkey(tfm, cc->key, key_size) < 0) { + ti->error = PFX "Error setting key"; + goto bad5; + } + + if (sscanf(argv[2], SECTOR_FORMAT, &cc->iv_offset) != 1) { + ti->error = PFX "Invalid iv_offset sector"; + goto bad5; + } + + if (sscanf(argv[4], SECTOR_FORMAT, &cc->start) != 1) { + ti->error = PFX "Invalid device sector"; + goto bad5; + } + + if (dm_get_device(ti, argv[3], cc->start, ti->len, + dm_table_get_mode(ti->table), &cc->dev)) { + ti->error = PFX "Device lookup failed"; + goto bad5; + } + + if (ivmode && cc->iv_gen_ops) { + if (ivopts) + *(ivopts - 1) = ':'; + cc->iv_mode = kmalloc(strlen(ivmode) + 1, GFP_KERNEL); + if (!cc->iv_mode) { + ti->error = PFX "Error kmallocing iv_mode string"; + goto bad5; + } + strcpy(cc->iv_mode, ivmode); + } else + cc->iv_mode = NULL; + + ti->private = cc; + return 0; + +bad5: + mempool_destroy(cc->page_pool); +bad4: + mempool_destroy(cc->io_pool); +bad3: + if (cc->iv_gen_ops && cc->iv_gen_ops->dtr) + cc->iv_gen_ops->dtr(cc); +bad2: + crypto_free_tfm(tfm); +bad1: + kfree(cc); + return -EINVAL; +} + +static void crypt_dtr(struct dm_target *ti) +{ + struct crypt_config *cc = (struct crypt_config *) ti->private; + + mempool_destroy(cc->page_pool); + mempool_destroy(cc->io_pool); + + if (cc->iv_mode) + kfree(cc->iv_mode); + if (cc->iv_gen_ops && cc->iv_gen_ops->dtr) + cc->iv_gen_ops->dtr(cc); + crypto_free_tfm(cc->tfm); + dm_put_device(ti, cc->dev); + kfree(cc); +} + +static int crypt_endio(struct bio *bio, unsigned int done, int error) +{ + struct crypt_io *io = (struct crypt_io *) bio->bi_private; + struct crypt_config *cc = (struct crypt_config *) io->target->private; + + if (bio_data_dir(bio) == WRITE) { + /* + * free the processed pages, even if + * it's only a partially completed write + */ + crypt_free_buffer_pages(cc, bio, done); + } + + if (bio->bi_size) + return 1; + + bio_put(bio); + + /* + * successful reads are decrypted by the worker thread + */ + if ((bio_data_dir(bio) == READ) + && bio_flagged(bio, BIO_UPTODATE)) { + kcryptd_queue_io(io); + return 0; + } + + dec_pending(io, error); + return error; +} + +static inline struct bio * +crypt_clone(struct crypt_config *cc, struct crypt_io *io, struct bio *bio, + sector_t sector, unsigned int *bvec_idx, + struct convert_context *ctx) +{ + struct bio *clone; + + if (bio_data_dir(bio) == WRITE) { + clone = crypt_alloc_buffer(cc, bio->bi_size, + io->first_clone, bvec_idx); + if (clone) { + ctx->bio_out = clone; + if (crypt_convert(cc, ctx) < 0) { + crypt_free_buffer_pages(cc, clone, + clone->bi_size); + bio_put(clone); + return NULL; + } + } + } else { + /* + * The block layer might modify the bvec array, so always + * copy the required bvecs because we need the original + * one in order to decrypt the whole bio data *afterwards*. + */ + clone = bio_alloc(GFP_NOIO, bio_segments(bio)); + if (clone) { + clone->bi_idx = 0; + clone->bi_vcnt = bio_segments(bio); + clone->bi_size = bio->bi_size; + memcpy(clone->bi_io_vec, bio_iovec(bio), + sizeof(struct bio_vec) * clone->bi_vcnt); + } + } + + if (!clone) + return NULL; + + clone->bi_private = io; + clone->bi_end_io = crypt_endio; + clone->bi_bdev = cc->dev->bdev; + clone->bi_sector = cc->start + sector; + clone->bi_rw = bio->bi_rw; + + return clone; +} + +static int crypt_map(struct dm_target *ti, struct bio *bio, + union map_info *map_context) +{ + struct crypt_config *cc = (struct crypt_config *) ti->private; + struct crypt_io *io = mempool_alloc(cc->io_pool, GFP_NOIO); + struct convert_context ctx; + struct bio *clone; + unsigned int remaining = bio->bi_size; + sector_t sector = bio->bi_sector - ti->begin; + unsigned int bvec_idx = 0; + + io->target = ti; + io->bio = bio; + io->first_clone = NULL; + io->error = 0; + atomic_set(&io->pending, 1); /* hold a reference */ + + if (bio_data_dir(bio) == WRITE) + crypt_convert_init(cc, &ctx, NULL, bio, sector, 1); + + /* + * The allocated buffers can be smaller than the whole bio, + * so repeat the whole process until all the data can be handled. + */ + while (remaining) { + clone = crypt_clone(cc, io, bio, sector, &bvec_idx, &ctx); + if (!clone) + goto cleanup; + + if (!io->first_clone) { + /* + * hold a reference to the first clone, because it + * holds the bio_vec array and that can't be freed + * before all other clones are released + */ + bio_get(clone); + io->first_clone = clone; + } + atomic_inc(&io->pending); + + remaining -= clone->bi_size; + sector += bio_sectors(clone); + + generic_make_request(clone); + + /* out of memory -> run queues */ + if (remaining) + blk_congestion_wait(bio_data_dir(clone), HZ/100); + } + + /* drop reference, clones could have returned before we reach this */ + dec_pending(io, 0); + return 0; + +cleanup: + if (io->first_clone) { + dec_pending(io, -ENOMEM); + return 0; + } + + /* if no bio has been dispatched yet, we can directly return the error */ + mempool_free(io, cc->io_pool); + return -ENOMEM; +} + +static int crypt_status(struct dm_target *ti, status_type_t type, + char *result, unsigned int maxlen) +{ + struct crypt_config *cc = (struct crypt_config *) ti->private; + const char *cipher; + const char *chainmode = NULL; + unsigned int sz = 0; + + switch (type) { + case STATUSTYPE_INFO: + result[0] = '\0'; + break; + + case STATUSTYPE_TABLE: + cipher = crypto_tfm_alg_name(cc->tfm); + + switch(cc->tfm->crt_cipher.cit_mode) { + case CRYPTO_TFM_MODE_CBC: + chainmode = "cbc"; + break; + case CRYPTO_TFM_MODE_ECB: + chainmode = "ecb"; + break; + default: + BUG(); + } + + if (cc->iv_mode) + DMEMIT("%s-%s-%s ", cipher, chainmode, cc->iv_mode); + else + DMEMIT("%s-%s ", cipher, chainmode); + + if (cc->key_size > 0) { + if ((maxlen - sz) < ((cc->key_size << 1) + 1)) + return -ENOMEM; + + crypt_encode_key(result + sz, cc->key, cc->key_size); + sz += cc->key_size << 1; + } else { + if (sz >= maxlen) + return -ENOMEM; + result[sz++] = '-'; + } + + DMEMIT(" " SECTOR_FORMAT " %s " SECTOR_FORMAT, + cc->iv_offset, cc->dev->name, cc->start); + break; + } + return 0; +} + +static struct target_type crypt_target = { + .name = "crypt", + .version= {1, 1, 0}, + .module = THIS_MODULE, + .ctr = crypt_ctr, + .dtr = crypt_dtr, + .map = crypt_map, + .status = crypt_status, +}; + +static int __init dm_crypt_init(void) +{ + int r; + + _crypt_io_pool = kmem_cache_create("dm-crypt_io", + sizeof(struct crypt_io), + 0, 0, NULL, NULL); + if (!_crypt_io_pool) + return -ENOMEM; + + _kcryptd_workqueue = create_workqueue("kcryptd"); + if (!_kcryptd_workqueue) { + r = -ENOMEM; + DMERR(PFX "couldn't create kcryptd"); + goto bad1; + } + + r = dm_register_target(&crypt_target); + if (r < 0) { + DMERR(PFX "register failed %d", r); + goto bad2; + } + + return 0; + +bad2: + destroy_workqueue(_kcryptd_workqueue); +bad1: + kmem_cache_destroy(_crypt_io_pool); + return r; +} + +static void __exit dm_crypt_exit(void) +{ + int r = dm_unregister_target(&crypt_target); + + if (r < 0) + DMERR(PFX "unregister failed %d", r); + + destroy_workqueue(_kcryptd_workqueue); + kmem_cache_destroy(_crypt_io_pool); +} + +module_init(dm_crypt_init); +module_exit(dm_crypt_exit); + +MODULE_AUTHOR("Christophe Saout "); +MODULE_DESCRIPTION(DM_NAME " target for transparent encryption / decryption"); +MODULE_LICENSE("GPL"); diff --git a/drivers/md/dm-emc.c b/drivers/md/dm-emc.c new file mode 100644 index 000000000000..700658664594 --- /dev/null +++ b/drivers/md/dm-emc.c @@ -0,0 +1,359 @@ +/* + * Copyright (C) 2004 SUSE LINUX Products GmbH. All rights reserved. + * Copyright (C) 2004 Red Hat, Inc. All rights reserved. + * + * This file is released under the GPL. + * + * Multipath support for EMC CLARiiON AX/CX-series hardware. + */ + +#include "dm.h" +#include "dm-hw-handler.h" +#include +#include + +struct emc_handler { + spinlock_t lock; + + /* Whether we should send the short trespass command (FC-series) + * or the long version (default for AX/CX CLARiiON arrays). */ + unsigned short_trespass; + /* Whether or not to honor SCSI reservations when initiating a + * switch-over. Default: Don't. */ + unsigned hr; + + unsigned char sense[SCSI_SENSE_BUFFERSIZE]; +}; + +#define TRESPASS_PAGE 0x22 +#define EMC_FAILOVER_TIMEOUT (60 * HZ) + +/* Code borrowed from dm-lsi-rdac by Mike Christie */ + +static inline void free_bio(struct bio *bio) +{ + __free_page(bio->bi_io_vec[0].bv_page); + bio_put(bio); +} + +static int emc_endio(struct bio *bio, unsigned int bytes_done, int error) +{ + struct path *path = bio->bi_private; + + if (bio->bi_size) + return 1; + + /* We also need to look at the sense keys here whether or not to + * switch to the next PG etc. + * + * For now simple logic: either it works or it doesn't. + */ + if (error) + dm_pg_init_complete(path, MP_FAIL_PATH); + else + dm_pg_init_complete(path, 0); + + /* request is freed in block layer */ + free_bio(bio); + + return 0; +} + +static struct bio *get_failover_bio(struct path *path, unsigned data_size) +{ + struct bio *bio; + struct page *page; + + bio = bio_alloc(GFP_ATOMIC, 1); + if (!bio) { + DMERR("dm-emc: get_failover_bio: bio_alloc() failed."); + return NULL; + } + + bio->bi_rw |= (1 << BIO_RW); + bio->bi_bdev = path->dev->bdev; + bio->bi_sector = 0; + bio->bi_private = path; + bio->bi_end_io = emc_endio; + + page = alloc_page(GFP_ATOMIC); + if (!page) { + DMERR("dm-emc: get_failover_bio: alloc_page() failed."); + bio_put(bio); + return NULL; + } + + if (bio_add_page(bio, page, data_size, 0) != data_size) { + DMERR("dm-emc: get_failover_bio: alloc_page() failed."); + __free_page(page); + bio_put(bio); + return NULL; + } + + return bio; +} + +static struct request *get_failover_req(struct emc_handler *h, + struct bio *bio, struct path *path) +{ + struct request *rq; + struct block_device *bdev = bio->bi_bdev; + struct request_queue *q = bdev_get_queue(bdev); + + /* FIXME: Figure out why it fails with GFP_ATOMIC. */ + rq = blk_get_request(q, WRITE, __GFP_WAIT); + if (!rq) { + DMERR("dm-emc: get_failover_req: blk_get_request failed"); + return NULL; + } + + rq->bio = rq->biotail = bio; + blk_rq_bio_prep(q, rq, bio); + + rq->rq_disk = bdev->bd_contains->bd_disk; + + /* bio backed don't set data */ + rq->buffer = rq->data = NULL; + /* rq data_len used for pc cmd's request_bufflen */ + rq->data_len = bio->bi_size; + + rq->sense = h->sense; + memset(rq->sense, 0, SCSI_SENSE_BUFFERSIZE); + rq->sense_len = 0; + + memset(&rq->cmd, 0, BLK_MAX_CDB); + + rq->timeout = EMC_FAILOVER_TIMEOUT; + rq->flags |= (REQ_BLOCK_PC | REQ_FAILFAST | REQ_NOMERGE); + + return rq; +} + +static struct request *emc_trespass_get(struct emc_handler *h, + struct path *path) +{ + struct bio *bio; + struct request *rq; + unsigned char *page22; + unsigned char long_trespass_pg[] = { + 0, 0, 0, 0, + TRESPASS_PAGE, /* Page code */ + 0x09, /* Page length - 2 */ + h->hr ? 0x01 : 0x81, /* Trespass code + Honor reservation bit */ + 0xff, 0xff, /* Trespass target */ + 0, 0, 0, 0, 0, 0 /* Reserved bytes / unknown */ + }; + unsigned char short_trespass_pg[] = { + 0, 0, 0, 0, + TRESPASS_PAGE, /* Page code */ + 0x02, /* Page length - 2 */ + h->hr ? 0x01 : 0x81, /* Trespass code + Honor reservation bit */ + 0xff, /* Trespass target */ + }; + unsigned data_size = h->short_trespass ? sizeof(short_trespass_pg) : + sizeof(long_trespass_pg); + + /* get bio backing */ + if (data_size > PAGE_SIZE) + /* this should never happen */ + return NULL; + + bio = get_failover_bio(path, data_size); + if (!bio) { + DMERR("dm-emc: emc_trespass_get: no bio"); + return NULL; + } + + page22 = (unsigned char *)bio_data(bio); + memset(page22, 0, data_size); + + memcpy(page22, h->short_trespass ? + short_trespass_pg : long_trespass_pg, data_size); + + /* get request for block layer packet command */ + rq = get_failover_req(h, bio, path); + if (!rq) { + DMERR("dm-emc: emc_trespass_get: no rq"); + free_bio(bio); + return NULL; + } + + /* Prepare the command. */ + rq->cmd[0] = MODE_SELECT; + rq->cmd[1] = 0x10; + rq->cmd[4] = data_size; + rq->cmd_len = COMMAND_SIZE(rq->cmd[0]); + + return rq; +} + +static void emc_pg_init(struct hw_handler *hwh, unsigned bypassed, + struct path *path) +{ + struct request *rq; + struct request_queue *q = bdev_get_queue(path->dev->bdev); + + /* + * We can either blindly init the pg (then look at the sense), + * or we can send some commands to get the state here (then + * possibly send the fo cmnd), or we can also have the + * initial state passed into us and then get an update here. + */ + if (!q) { + DMINFO("dm-emc: emc_pg_init: no queue"); + goto fail_path; + } + + /* FIXME: The request should be pre-allocated. */ + rq = emc_trespass_get(hwh->context, path); + if (!rq) { + DMERR("dm-emc: emc_pg_init: no rq"); + goto fail_path; + } + + DMINFO("dm-emc: emc_pg_init: sending switch-over command"); + elv_add_request(q, rq, ELEVATOR_INSERT_FRONT, 1); + return; + +fail_path: + dm_pg_init_complete(path, MP_FAIL_PATH); +} + +static struct emc_handler *alloc_emc_handler(void) +{ + struct emc_handler *h = kmalloc(sizeof(*h), GFP_KERNEL); + + if (h) + spin_lock_init(&h->lock); + + return h; +} + +static int emc_create(struct hw_handler *hwh, unsigned argc, char **argv) +{ + struct emc_handler *h; + unsigned hr, short_trespass; + + if (argc == 0) { + /* No arguments: use defaults */ + hr = 0; + short_trespass = 0; + } else if (argc != 2) { + DMWARN("dm-emc hwhandler: incorrect number of arguments"); + return -EINVAL; + } else { + if ((sscanf(argv[0], "%u", &short_trespass) != 1) + || (short_trespass > 1)) { + DMWARN("dm-emc: invalid trespass mode selected"); + return -EINVAL; + } + + if ((sscanf(argv[1], "%u", &hr) != 1) + || (hr > 1)) { + DMWARN("dm-emc: invalid honor reservation flag selected"); + return -EINVAL; + } + } + + h = alloc_emc_handler(); + if (!h) + return -ENOMEM; + + memset(h, 0, sizeof(*h)); + + hwh->context = h; + + if ((h->short_trespass = short_trespass)) + DMWARN("dm-emc: short trespass command will be send"); + else + DMWARN("dm-emc: long trespass command will be send"); + + if ((h->hr = hr)) + DMWARN("dm-emc: honor reservation bit will be set"); + else + DMWARN("dm-emc: honor reservation bit will not be set (default)"); + + return 0; +} + +static void emc_destroy(struct hw_handler *hwh) +{ + struct emc_handler *h = (struct emc_handler *) hwh->context; + + kfree(h); + hwh->context = NULL; +} + +static unsigned emc_error(struct hw_handler *hwh, struct bio *bio) +{ + /* FIXME: Patch from axboe still missing */ +#if 0 + int sense; + + if (bio->bi_error & BIO_SENSE) { + sense = bio->bi_error & 0xffffff; /* sense key / asc / ascq */ + + if (sense == 0x020403) { + /* LUN Not Ready - Manual Intervention Required + * indicates this is a passive path. + * + * FIXME: However, if this is seen and EVPD C0 + * indicates that this is due to a NDU in + * progress, we should set FAIL_PATH too. + * This indicates we might have to do a SCSI + * inquiry in the end_io path. Ugh. */ + return MP_BYPASS_PG | MP_RETRY_IO; + } else if (sense == 0x052501) { + /* An array based copy is in progress. Do not + * fail the path, do not bypass to another PG, + * do not retry. Fail the IO immediately. + * (Actually this is the same conclusion as in + * the default handler, but lets make sure.) */ + return 0; + } else if (sense == 0x062900) { + /* Unit Attention Code. This is the first IO + * to the new path, so just retry. */ + return MP_RETRY_IO; + } + } +#endif + + /* Try default handler */ + return dm_scsi_err_handler(hwh, bio); +} + +static struct hw_handler_type emc_hwh = { + .name = "emc", + .module = THIS_MODULE, + .create = emc_create, + .destroy = emc_destroy, + .pg_init = emc_pg_init, + .error = emc_error, +}; + +static int __init dm_emc_init(void) +{ + int r = dm_register_hw_handler(&emc_hwh); + + if (r < 0) + DMERR("emc: register failed %d", r); + + DMINFO("dm-emc version 0.0.3 loaded"); + + return r; +} + +static void __exit dm_emc_exit(void) +{ + int r = dm_unregister_hw_handler(&emc_hwh); + + if (r < 0) + DMERR("emc: unregister failed %d", r); +} + +module_init(dm_emc_init); +module_exit(dm_emc_exit); + +MODULE_DESCRIPTION(DM_NAME " EMC CX/AX/FC-family multipath"); +MODULE_AUTHOR("Lars Marowsky-Bree "); +MODULE_LICENSE("GPL"); diff --git a/drivers/md/dm-exception-store.c b/drivers/md/dm-exception-store.c new file mode 100644 index 000000000000..17212b4201a1 --- /dev/null +++ b/drivers/md/dm-exception-store.c @@ -0,0 +1,648 @@ +/* + * dm-snapshot.c + * + * Copyright (C) 2001-2002 Sistina Software (UK) Limited. + * + * This file is released under the GPL. + */ + +#include "dm.h" +#include "dm-snap.h" +#include "dm-io.h" +#include "kcopyd.h" + +#include +#include +#include +#include + +/*----------------------------------------------------------------- + * Persistent snapshots, by persistent we mean that the snapshot + * will survive a reboot. + *---------------------------------------------------------------*/ + +/* + * We need to store a record of which parts of the origin have + * been copied to the snapshot device. The snapshot code + * requires that we copy exception chunks to chunk aligned areas + * of the COW store. It makes sense therefore, to store the + * metadata in chunk size blocks. + * + * There is no backward or forward compatibility implemented, + * snapshots with different disk versions than the kernel will + * not be usable. It is expected that "lvcreate" will blank out + * the start of a fresh COW device before calling the snapshot + * constructor. + * + * The first chunk of the COW device just contains the header. + * After this there is a chunk filled with exception metadata, + * followed by as many exception chunks as can fit in the + * metadata areas. + * + * All on disk structures are in little-endian format. The end + * of the exceptions info is indicated by an exception with a + * new_chunk of 0, which is invalid since it would point to the + * header chunk. + */ + +/* + * Magic for persistent snapshots: "SnAp" - Feeble isn't it. + */ +#define SNAP_MAGIC 0x70416e53 + +/* + * The on-disk version of the metadata. + */ +#define SNAPSHOT_DISK_VERSION 1 + +struct disk_header { + uint32_t magic; + + /* + * Is this snapshot valid. There is no way of recovering + * an invalid snapshot. + */ + uint32_t valid; + + /* + * Simple, incrementing version. no backward + * compatibility. + */ + uint32_t version; + + /* In sectors */ + uint32_t chunk_size; +}; + +struct disk_exception { + uint64_t old_chunk; + uint64_t new_chunk; +}; + +struct commit_callback { + void (*callback)(void *, int success); + void *context; +}; + +/* + * The top level structure for a persistent exception store. + */ +struct pstore { + struct dm_snapshot *snap; /* up pointer to my snapshot */ + int version; + int valid; + uint32_t chunk_size; + uint32_t exceptions_per_area; + + /* + * Now that we have an asynchronous kcopyd there is no + * need for large chunk sizes, so it wont hurt to have a + * whole chunks worth of metadata in memory at once. + */ + void *area; + + /* + * Used to keep track of which metadata area the data in + * 'chunk' refers to. + */ + uint32_t current_area; + + /* + * The next free chunk for an exception. + */ + uint32_t next_free; + + /* + * The index of next free exception in the current + * metadata area. + */ + uint32_t current_committed; + + atomic_t pending_count; + uint32_t callback_count; + struct commit_callback *callbacks; +}; + +static inline unsigned int sectors_to_pages(unsigned int sectors) +{ + return sectors / (PAGE_SIZE >> 9); +} + +static int alloc_area(struct pstore *ps) +{ + int r = -ENOMEM; + size_t len; + + len = ps->chunk_size << SECTOR_SHIFT; + + /* + * Allocate the chunk_size block of memory that will hold + * a single metadata area. + */ + ps->area = vmalloc(len); + if (!ps->area) + return r; + + return 0; +} + +static void free_area(struct pstore *ps) +{ + vfree(ps->area); +} + +/* + * Read or write a chunk aligned and sized block of data from a device. + */ +static int chunk_io(struct pstore *ps, uint32_t chunk, int rw) +{ + struct io_region where; + unsigned long bits; + + where.bdev = ps->snap->cow->bdev; + where.sector = ps->chunk_size * chunk; + where.count = ps->chunk_size; + + return dm_io_sync_vm(1, &where, rw, ps->area, &bits); +} + +/* + * Read or write a metadata area. Remembering to skip the first + * chunk which holds the header. + */ +static int area_io(struct pstore *ps, uint32_t area, int rw) +{ + int r; + uint32_t chunk; + + /* convert a metadata area index to a chunk index */ + chunk = 1 + ((ps->exceptions_per_area + 1) * area); + + r = chunk_io(ps, chunk, rw); + if (r) + return r; + + ps->current_area = area; + return 0; +} + +static int zero_area(struct pstore *ps, uint32_t area) +{ + memset(ps->area, 0, ps->chunk_size << SECTOR_SHIFT); + return area_io(ps, area, WRITE); +} + +static int read_header(struct pstore *ps, int *new_snapshot) +{ + int r; + struct disk_header *dh; + + r = chunk_io(ps, 0, READ); + if (r) + return r; + + dh = (struct disk_header *) ps->area; + + if (le32_to_cpu(dh->magic) == 0) { + *new_snapshot = 1; + + } else if (le32_to_cpu(dh->magic) == SNAP_MAGIC) { + *new_snapshot = 0; + ps->valid = le32_to_cpu(dh->valid); + ps->version = le32_to_cpu(dh->version); + ps->chunk_size = le32_to_cpu(dh->chunk_size); + + } else { + DMWARN("Invalid/corrupt snapshot"); + r = -ENXIO; + } + + return r; +} + +static int write_header(struct pstore *ps) +{ + struct disk_header *dh; + + memset(ps->area, 0, ps->chunk_size << SECTOR_SHIFT); + + dh = (struct disk_header *) ps->area; + dh->magic = cpu_to_le32(SNAP_MAGIC); + dh->valid = cpu_to_le32(ps->valid); + dh->version = cpu_to_le32(ps->version); + dh->chunk_size = cpu_to_le32(ps->chunk_size); + + return chunk_io(ps, 0, WRITE); +} + +/* + * Access functions for the disk exceptions, these do the endian conversions. + */ +static struct disk_exception *get_exception(struct pstore *ps, uint32_t index) +{ + if (index >= ps->exceptions_per_area) + return NULL; + + return ((struct disk_exception *) ps->area) + index; +} + +static int read_exception(struct pstore *ps, + uint32_t index, struct disk_exception *result) +{ + struct disk_exception *e; + + e = get_exception(ps, index); + if (!e) + return -EINVAL; + + /* copy it */ + result->old_chunk = le64_to_cpu(e->old_chunk); + result->new_chunk = le64_to_cpu(e->new_chunk); + + return 0; +} + +static int write_exception(struct pstore *ps, + uint32_t index, struct disk_exception *de) +{ + struct disk_exception *e; + + e = get_exception(ps, index); + if (!e) + return -EINVAL; + + /* copy it */ + e->old_chunk = cpu_to_le64(de->old_chunk); + e->new_chunk = cpu_to_le64(de->new_chunk); + + return 0; +} + +/* + * Registers the exceptions that are present in the current area. + * 'full' is filled in to indicate if the area has been + * filled. + */ +static int insert_exceptions(struct pstore *ps, int *full) +{ + int r; + unsigned int i; + struct disk_exception de; + + /* presume the area is full */ + *full = 1; + + for (i = 0; i < ps->exceptions_per_area; i++) { + r = read_exception(ps, i, &de); + + if (r) + return r; + + /* + * If the new_chunk is pointing at the start of + * the COW device, where the first metadata area + * is we know that we've hit the end of the + * exceptions. Therefore the area is not full. + */ + if (de.new_chunk == 0LL) { + ps->current_committed = i; + *full = 0; + break; + } + + /* + * Keep track of the start of the free chunks. + */ + if (ps->next_free <= de.new_chunk) + ps->next_free = de.new_chunk + 1; + + /* + * Otherwise we add the exception to the snapshot. + */ + r = dm_add_exception(ps->snap, de.old_chunk, de.new_chunk); + if (r) + return r; + } + + return 0; +} + +static int read_exceptions(struct pstore *ps) +{ + uint32_t area; + int r, full = 1; + + /* + * Keeping reading chunks and inserting exceptions until + * we find a partially full area. + */ + for (area = 0; full; area++) { + r = area_io(ps, area, READ); + if (r) + return r; + + r = insert_exceptions(ps, &full); + if (r) + return r; + } + + return 0; +} + +static inline struct pstore *get_info(struct exception_store *store) +{ + return (struct pstore *) store->context; +} + +static void persistent_fraction_full(struct exception_store *store, + sector_t *numerator, sector_t *denominator) +{ + *numerator = get_info(store)->next_free * store->snap->chunk_size; + *denominator = get_dev_size(store->snap->cow->bdev); +} + +static void persistent_destroy(struct exception_store *store) +{ + struct pstore *ps = get_info(store); + + dm_io_put(sectors_to_pages(ps->chunk_size)); + vfree(ps->callbacks); + free_area(ps); + kfree(ps); +} + +static int persistent_read_metadata(struct exception_store *store) +{ + int r, new_snapshot; + struct pstore *ps = get_info(store); + + /* + * Read the snapshot header. + */ + r = read_header(ps, &new_snapshot); + if (r) + return r; + + /* + * Do we need to setup a new snapshot ? + */ + if (new_snapshot) { + r = write_header(ps); + if (r) { + DMWARN("write_header failed"); + return r; + } + + r = zero_area(ps, 0); + if (r) { + DMWARN("zero_area(0) failed"); + return r; + } + + } else { + /* + * Sanity checks. + */ + if (!ps->valid) { + DMWARN("snapshot is marked invalid"); + return -EINVAL; + } + + if (ps->version != SNAPSHOT_DISK_VERSION) { + DMWARN("unable to handle snapshot disk version %d", + ps->version); + return -EINVAL; + } + + /* + * Read the metadata. + */ + r = read_exceptions(ps); + if (r) + return r; + } + + return 0; +} + +static int persistent_prepare(struct exception_store *store, + struct exception *e) +{ + struct pstore *ps = get_info(store); + uint32_t stride; + sector_t size = get_dev_size(store->snap->cow->bdev); + + /* Is there enough room ? */ + if (size < ((ps->next_free + 1) * store->snap->chunk_size)) + return -ENOSPC; + + e->new_chunk = ps->next_free; + + /* + * Move onto the next free pending, making sure to take + * into account the location of the metadata chunks. + */ + stride = (ps->exceptions_per_area + 1); + if ((++ps->next_free % stride) == 1) + ps->next_free++; + + atomic_inc(&ps->pending_count); + return 0; +} + +static void persistent_commit(struct exception_store *store, + struct exception *e, + void (*callback) (void *, int success), + void *callback_context) +{ + int r; + unsigned int i; + struct pstore *ps = get_info(store); + struct disk_exception de; + struct commit_callback *cb; + + de.old_chunk = e->old_chunk; + de.new_chunk = e->new_chunk; + write_exception(ps, ps->current_committed++, &de); + + /* + * Add the callback to the back of the array. This code + * is the only place where the callback array is + * manipulated, and we know that it will never be called + * multiple times concurrently. + */ + cb = ps->callbacks + ps->callback_count++; + cb->callback = callback; + cb->context = callback_context; + + /* + * If there are no more exceptions in flight, or we have + * filled this metadata area we commit the exceptions to + * disk. + */ + if (atomic_dec_and_test(&ps->pending_count)