aboutsummaryrefslogtreecommitdiff
path: root/drivers/md
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/md')
-rw-r--r--drivers/md/Kconfig6
-rw-r--r--drivers/md/Makefile2
-rw-r--r--drivers/md/dm-bio-list.h4
-rw-r--r--drivers/md/dm-crypt.c33
-rw-r--r--drivers/md/dm-delay.c23
-rw-r--r--drivers/md/dm-exception-store.c76
-rw-r--r--drivers/md/dm-io.c5
-rw-r--r--drivers/md/dm-mpath-rdac.c700
-rw-r--r--drivers/md/dm-mpath.c34
-rw-r--r--drivers/md/dm-raid1.c75
-rw-r--r--drivers/md/dm-round-robin.c2
-rw-r--r--drivers/md/dm-snap.c116
-rw-r--r--drivers/md/dm-snap.h6
-rw-r--r--drivers/md/dm.c33
-rw-r--r--drivers/md/dm.h40
-rw-r--r--drivers/md/kcopyd.c11
16 files changed, 984 insertions, 182 deletions
diff --git a/drivers/md/Kconfig b/drivers/md/Kconfig
index bfd9b9c6252..64bf3a81db9 100644
--- a/drivers/md/Kconfig
+++ b/drivers/md/Kconfig
@@ -264,6 +264,12 @@ config DM_MULTIPATH_EMC
---help---
Multipath support for EMC CX/AX series hardware.
+config DM_MULTIPATH_RDAC
+ tristate "LSI/Engenio RDAC multipath support (EXPERIMENTAL)"
+ depends on DM_MULTIPATH && BLK_DEV_DM && EXPERIMENTAL
+ ---help---
+ Multipath support for LSI/Engenio RDAC.
+
config DM_DELAY
tristate "I/O delaying target (EXPERIMENTAL)"
depends on BLK_DEV_DM && EXPERIMENTAL
diff --git a/drivers/md/Makefile b/drivers/md/Makefile
index 71eb45f7417..c49366cdc05 100644
--- a/drivers/md/Makefile
+++ b/drivers/md/Makefile
@@ -7,6 +7,7 @@ dm-mod-objs := dm.o dm-table.o dm-target.o dm-linear.o dm-stripe.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
+dm-rdac-objs := dm-mpath-rdac.o
md-mod-objs := md.o bitmap.o
raid456-objs := raid5.o raid6algos.o raid6recov.o raid6tables.o \
raid6int1.o raid6int2.o raid6int4.o \
@@ -34,6 +35,7 @@ obj-$(CONFIG_DM_CRYPT) += dm-crypt.o
obj-$(CONFIG_DM_DELAY) += dm-delay.o
obj-$(CONFIG_DM_MULTIPATH) += dm-multipath.o dm-round-robin.o
obj-$(CONFIG_DM_MULTIPATH_EMC) += dm-emc.o
+obj-$(CONFIG_DM_MULTIPATH_RDAC) += dm-rdac.o
obj-$(CONFIG_DM_SNAPSHOT) += dm-snapshot.o
obj-$(CONFIG_DM_MIRROR) += dm-mirror.o
obj-$(CONFIG_DM_ZERO) += dm-zero.o
diff --git a/drivers/md/dm-bio-list.h b/drivers/md/dm-bio-list.h
index c6be88826fa..16ee3b018b3 100644
--- a/drivers/md/dm-bio-list.h
+++ b/drivers/md/dm-bio-list.h
@@ -8,7 +8,6 @@
#define DM_BIO_LIST_H
#include <linux/bio.h>
-#include <linux/prefetch.h>
struct bio_list {
struct bio *head;
@@ -31,8 +30,7 @@ static inline void bio_list_init(struct bio_list *bl)
}
#define bio_list_for_each(bio, bl) \
- for (bio = (bl)->head; bio && ({ prefetch(bio->bi_next); 1; }); \
- bio = bio->bi_next)
+ for (bio = (bl)->head; bio; bio = bio->bi_next)
static inline unsigned bio_list_size(const struct bio_list *bl)
{
diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c
index 7b0fcfc9eaa..ba952a03259 100644
--- a/drivers/md/dm-crypt.c
+++ b/drivers/md/dm-crypt.c
@@ -30,7 +30,7 @@
/*
* per bio private data
*/
-struct crypt_io {
+struct dm_crypt_io {
struct dm_target *target;
struct bio *base_bio;
struct work_struct work;
@@ -106,7 +106,7 @@ struct crypt_config {
static struct kmem_cache *_crypt_io_pool;
-static void clone_init(struct crypt_io *, struct bio *);
+static void clone_init(struct dm_crypt_io *, struct bio *);
/*
* Different IV generation algorithms:
@@ -382,7 +382,7 @@ static int crypt_convert(struct crypt_config *cc,
static void dm_crypt_bio_destructor(struct bio *bio)
{
- struct crypt_io *io = bio->bi_private;
+ struct dm_crypt_io *io = bio->bi_private;
struct crypt_config *cc = io->target->private;
bio_free(bio, cc->bs);
@@ -393,7 +393,7 @@ static int crypt_convert(struct crypt_config *cc,
* 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_io *io, unsigned int size)
+static struct bio *crypt_alloc_buffer(struct dm_crypt_io *io, unsigned size)
{
struct crypt_config *cc = io->target->private;
struct bio *clone;
@@ -479,7 +479,7 @@ static void crypt_free_buffer_pages(struct crypt_config *cc,
* 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)
+static void dec_pending(struct dm_crypt_io *io, int error)
{
struct crypt_config *cc = (struct crypt_config *) io->target->private;
@@ -503,7 +503,7 @@ static void dec_pending(struct crypt_io *io, int error)
static struct workqueue_struct *_kcryptd_workqueue;
static void kcryptd_do_work(struct work_struct *work);
-static void kcryptd_queue_io(struct crypt_io *io)
+static void kcryptd_queue_io(struct dm_crypt_io *io)
{
INIT_WORK(&io->work, kcryptd_do_work);
queue_work(_kcryptd_workqueue, &io->work);
@@ -511,7 +511,7 @@ static void kcryptd_queue_io(struct crypt_io *io)
static int crypt_endio(struct bio *clone, unsigned int done, int error)
{
- struct crypt_io *io = clone->bi_private;
+ struct dm_crypt_io *io = clone->bi_private;
struct crypt_config *cc = io->target->private;
unsigned read_io = bio_data_dir(clone) == READ;
@@ -545,7 +545,7 @@ out:
return error;
}
-static void clone_init(struct crypt_io *io, struct bio *clone)
+static void clone_init(struct dm_crypt_io *io, struct bio *clone)
{
struct crypt_config *cc = io->target->private;
@@ -556,7 +556,7 @@ static void clone_init(struct crypt_io *io, struct bio *clone)
clone->bi_destructor = dm_crypt_bio_destructor;
}
-static void process_read(struct crypt_io *io)
+static void process_read(struct dm_crypt_io *io)
{
struct crypt_config *cc = io->target->private;
struct bio *base_bio = io->base_bio;
@@ -587,7 +587,7 @@ static void process_read(struct crypt_io *io)
generic_make_request(clone);
}
-static void process_write(struct crypt_io *io)
+static void process_write(struct dm_crypt_io *io)
{
struct crypt_config *cc = io->target->private;
struct bio *base_bio = io->base_bio;
@@ -644,7 +644,7 @@ static void process_write(struct crypt_io *io)
}
}
-static void process_read_endio(struct crypt_io *io)
+static void process_read_endio(struct dm_crypt_io *io)
{
struct crypt_config *cc = io->target->private;
struct convert_context ctx;
@@ -657,7 +657,7 @@ static void process_read_endio(struct crypt_io *io)
static void kcryptd_do_work(struct work_struct *work)
{
- struct crypt_io *io = container_of(work, struct crypt_io, work);
+ struct dm_crypt_io *io = container_of(work, struct dm_crypt_io, work);
if (io->post_process)
process_read_endio(io);
@@ -939,10 +939,7 @@ static int crypt_map(struct dm_target *ti, struct bio *bio,
union map_info *map_context)
{
struct crypt_config *cc = ti->private;
- struct crypt_io *io;
-
- if (bio_barrier(bio))
- return -EOPNOTSUPP;
+ struct dm_crypt_io *io;
io = mempool_alloc(cc->io_pool, GFP_NOIO);
io->target = ti;
@@ -1062,9 +1059,7 @@ 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);
+ _crypt_io_pool = KMEM_CACHE(dm_crypt_io, 0);
if (!_crypt_io_pool)
return -ENOMEM;
diff --git a/drivers/md/dm-delay.c b/drivers/md/dm-delay.c
index 52c7cf9e580..6928c136d3c 100644
--- a/drivers/md/dm-delay.c
+++ b/drivers/md/dm-delay.c
@@ -20,7 +20,7 @@
struct delay_c {
struct timer_list delay_timer;
- struct semaphore timer_lock;
+ struct mutex timer_lock;
struct work_struct flush_expired_bios;
struct list_head delayed_bios;
atomic_t may_delay;
@@ -37,7 +37,7 @@ struct delay_c {
unsigned writes;
};
-struct delay_info {
+struct dm_delay_info {
struct delay_c *context;
struct list_head list;
struct bio *bio;
@@ -58,12 +58,12 @@ static void handle_delayed_timer(unsigned long data)
static void queue_timeout(struct delay_c *dc, unsigned long expires)
{
- down(&dc->timer_lock);
+ mutex_lock(&dc->timer_lock);
if (!timer_pending(&dc->delay_timer) || expires < dc->delay_timer.expires)
mod_timer(&dc->delay_timer, expires);
- up(&dc->timer_lock);
+ mutex_unlock(&dc->timer_lock);
}
static void flush_bios(struct bio *bio)
@@ -80,7 +80,7 @@ static void flush_bios(struct bio *bio)
static struct bio *flush_delayed_bios(struct delay_c *dc, int flush_all)
{
- struct delay_info *delayed, *next;
+ struct dm_delay_info *delayed, *next;
unsigned long next_expires = 0;
int start_timer = 0;
BIO_LIST(flush_bios);
@@ -193,13 +193,11 @@ out:
goto bad;
}
- init_timer(&dc->delay_timer);
- dc->delay_timer.function = handle_delayed_timer;
- dc->delay_timer.data = (unsigned long)dc;
+ setup_timer(&dc->delay_timer, handle_delayed_timer, (unsigned long)dc);
INIT_WORK(&dc->flush_expired_bios, flush_expired_bios);
INIT_LIST_HEAD(&dc->delayed_bios);
- init_MUTEX(&dc->timer_lock);
+ mutex_init(&dc->timer_lock);
atomic_set(&dc->may_delay, 1);
ti->private = dc;
@@ -227,7 +225,7 @@ static void delay_dtr(struct dm_target *ti)
static int delay_bio(struct delay_c *dc, int delay, struct bio *bio)
{
- struct delay_info *delayed;
+ struct dm_delay_info *delayed;
unsigned long expires = 0;
if (!delay || !atomic_read(&dc->may_delay))
@@ -338,10 +336,7 @@ static int __init dm_delay_init(void)
goto bad_queue;
}
- delayed_cache = kmem_cache_create("dm-delay",
- sizeof(struct delay_info),
- __alignof__(struct delay_info),
- 0, NULL, NULL);
+ delayed_cache = KMEM_CACHE(dm_delay_info, 0);
if (!delayed_cache) {
DMERR("Couldn't create delayed bio cache.");
goto bad_memcache;
diff --git a/drivers/md/dm-exception-store.c b/drivers/md/dm-exception-store.c
index 07e0a0c84f6..3d65917a1bb 100644
--- a/drivers/md/dm-exception-store.c
+++ b/drivers/md/dm-exception-store.c
@@ -125,9 +125,11 @@ struct pstore {
uint32_t callback_count;
struct commit_callback *callbacks;
struct dm_io_client *io_client;
+
+ struct workqueue_struct *metadata_wq;
};
-static inline unsigned int sectors_to_pages(unsigned int sectors)
+static unsigned sectors_to_pages(unsigned sectors)
{
return sectors / (PAGE_SIZE >> 9);
}
@@ -156,10 +158,24 @@ static void free_area(struct pstore *ps)
ps->area = NULL;
}
+struct mdata_req {
+ struct io_region *where;
+ struct dm_io_request *io_req;
+ struct work_struct work;
+ int result;
+};
+
+static void do_metadata(struct work_struct *work)
+{
+ struct mdata_req *req = container_of(work, struct mdata_req, work);
+
+ req->result = dm_io(req->io_req, 1, req->where, NULL);
+}
+
/*
* 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)
+static int chunk_io(struct pstore *ps, uint32_t chunk, int rw, int metadata)
{
struct io_region where = {
.bdev = ps->snap->cow->bdev,
@@ -173,8 +189,23 @@ static int chunk_io(struct pstore *ps, uint32_t chunk, int rw)
.client = ps->io_client,
.notify.fn = NULL,
};
+ struct mdata_req req;
+
+ if (!metadata)
+ return dm_io(&io_req, 1, &where, NULL);
+
+ req.where = &where;
+ req.io_req = &io_req;
- return dm_io(&io_req, 1, &where, NULL);
+ /*
+ * Issue the synchronous I/O from a different thread
+ * to avoid generic_make_request recursion.
+ */
+ INIT_WORK(&req.work, do_metadata);
+ queue_work(ps->metadata_wq, &req.work);
+ flush_workqueue(ps->metadata_wq);
+
+ return req.result;
}
/*
@@ -189,7 +220,7 @@ static int area_io(struct pstore *ps, uint32_t area, int rw)
/* convert a metadata area index to a chunk index */
chunk = 1 + ((ps->exceptions_per_area + 1) * area);
- r = chunk_io(ps, chunk, rw);
+ r = chunk_io(ps, chunk, rw, 0);
if (r)
return r;
@@ -230,7 +261,7 @@ static int read_header(struct pstore *ps, int *new_snapshot)
if (r)
return r;
- r = chunk_io(ps, 0, READ);
+ r = chunk_io(ps, 0, READ, 1);
if (r)
goto bad;
@@ -292,7 +323,7 @@ static int write_header(struct pstore *ps)
dh->version = cpu_to_le32(ps->version);
dh->chunk_size = cpu_to_le32(ps->snap->chunk_size);
- return chunk_io(ps, 0, WRITE);
+ return chunk_io(ps, 0, WRITE, 1);
}
/*
@@ -393,7 +424,7 @@ static int read_exceptions(struct pstore *ps)
return 0;
}
-static inline struct pstore *get_info(struct exception_store *store)
+static struct pstore *get_info(struct exception_store *store)
{
return (struct pstore *) store->context;
}
@@ -409,6 +440,7 @@ static void persistent_destroy(struct exception_store *store)
{
struct pstore *ps = get_info(store);
+ destroy_workqueue(ps->metadata_wq);
dm_io_client_destroy(ps->io_client);
vfree(ps->callbacks);
free_area(ps);
@@ -457,11 +489,6 @@ static int persistent_read_metadata(struct exception_store *store)
/*
* 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);
@@ -469,6 +496,12 @@ static int persistent_read_metadata(struct exception_store *store)
}
/*
+ * Metadata are valid, but snapshot is invalidated
+ */
+ if (!ps->valid)
+ return 1;
+
+ /*
* Read the metadata.
*/
r = read_exceptions(ps);
@@ -480,7 +513,7 @@ static int persistent_read_metadata(struct exception_store *store)
}
static int persistent_prepare(struct exception_store *store,
- struct exception *e)
+ struct dm_snap_exception *e)
{
struct pstore *ps = get_info(store);
uint32_t stride;
@@ -505,7 +538,7 @@ static int persistent_prepare(struct exception_store *store,
}
static void persistent_commit(struct exception_store *store,
- struct exception *e,
+ struct dm_snap_exception *e,
void (*callback) (void *, int success),
void *callback_context)
{
@@ -588,6 +621,12 @@ int dm_create_persistent(struct exception_store *store)
atomic_set(&ps->pending_count, 0);
ps->callbacks = NULL;
+ ps->metadata_wq = create_singlethread_workqueue("ksnaphd");
+ if (!ps->metadata_wq) {
+ DMERR("couldn't start header metadata update thread");
+ return -ENOMEM;
+ }
+
store->destroy = persistent_destroy;
store->read_metadata = persistent_read_metadata;
store->prepare_exception = persistent_prepare;
@@ -616,7 +655,8 @@ static int transient_read_metadata(struct exception_store *store)
return 0;
}
-static int transient_prepare(struct exception_store *store, struct exception *e)
+static int transient_prepare(struct exception_store *store,
+ struct dm_snap_exception *e)
{
struct transient_c *tc = (struct transient_c *) store->context;
sector_t size = get_dev_size(store->snap->cow->bdev);
@@ -631,9 +671,9 @@ static int transient_prepare(struct exception_store *store, struct exception *e)
}
static void transient_commit(struct exception_store *store,
- struct exception *e,
- void (*callback) (void *, int success),
- void *callback_context)
+ struct dm_snap_exception *e,
+ void (*callback) (void *, int success),
+ void *callback_context)
{
/* Just succeed */
callback(callback_context, 1);
diff --git a/drivers/md/dm-io.c b/drivers/md/dm-io.c
index 352c6fbeac5..f3a77248643 100644
--- a/drivers/md/dm-io.c
+++ b/drivers/md/dm-io.c
@@ -293,7 +293,10 @@ static void do_region(int rw, unsigned int region, struct io_region *where,
* bvec for bio_get/set_region() and decrement bi_max_vecs
* to hide it from bio_add_page().
*/
- num_bvecs = (remaining / (PAGE_SIZE >> SECTOR_SHIFT)) + 2;
+ num_bvecs = dm_sector_div_up(remaining,
+ (PAGE_SIZE >> SECTOR_SHIFT));
+ num_bvecs = 1 + min_t(int, bio_get_nr_vecs(where->bdev),
+ num_bvecs);
bio = bio_alloc_bioset(GFP_NOIO, num_bvecs, io->client->bios);
bio->bi_sector = where->sector + (where->count - remaining);
bio->bi_bdev = where->bdev;
diff --git a/drivers/md/dm-mpath-rdac.c b/drivers/md/dm-mpath-rdac.c
new file mode 100644
index 00000000000..8b776b8cb7f
--- /dev/null
+++ b/drivers/md/dm-mpath-rdac.c
@@ -0,0 +1,700 @@
+/*
+ * Engenio/LSI RDAC DM HW handler
+ *
+ * Copyright (C) 2005 Mike Christie. All rights reserved.
+ * Copyright (C) Chandra Seetharaman, IBM Corp. 2007
+ *
+ * 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 <scsi/scsi.h>
+#include <scsi/scsi_cmnd.h>
+#include <scsi/scsi_eh.h>
+
+#define DM_MSG_PREFIX "multipath rdac"
+
+#include "dm.h"
+#include "dm-hw-handler.h"
+
+#define RDAC_DM_HWH_NAME "rdac"
+#define RDAC_DM_HWH_VER "0.4"
+
+/*
+ * LSI mode page stuff
+ *
+ * These struct definitions and the forming of the
+ * mode page were taken from the LSI RDAC 2.4 GPL'd
+ * driver, and then converted to Linux conventions.
+ */
+#define RDAC_QUIESCENCE_TIME 20;
+/*
+ * Page Codes
+ */
+#define RDAC_PAGE_CODE_REDUNDANT_CONTROLLER 0x2c
+
+/*
+ * Controller modes definitions
+ */
+#define RDAC_MODE_TRANSFER_ALL_LUNS 0x01
+#define RDAC_MODE_TRANSFER_SPECIFIED_LUNS 0x02
+
+/*
+ * RDAC Options field
+ */
+#define RDAC_FORCED_QUIESENCE 0x02
+
+#define RDAC_FAILOVER_TIMEOUT (60 * HZ)
+
+struct rdac_mode_6_hdr {
+ u8 data_len;
+ u8 medium_type;
+ u8 device_params;
+ u8 block_desc_len;
+};
+
+struct rdac_mode_10_hdr {
+ u16 data_len;
+ u8 medium_type;
+ u8 device_params;
+ u16 reserved;
+ u16 block_desc_len;
+};
+
+struct rdac_mode_common {
+ u8 controller_serial[16];
+ u8 alt_controller_serial[16];
+ u8 rdac_mode[2];
+ u8 alt_rdac_mode[2];
+ u8 quiescence_timeout;
+ u8 rdac_options;
+};
+
+struct rdac_pg_legacy {
+ struct rdac_mode_6_hdr hdr;
+ u8 page_code;
+ u8 page_len;
+ struct rdac_mode_common common;
+#define MODE6_MAX_LUN 32
+ u8 lun_table[MODE6_MAX_LUN];
+ u8 reserved2[32];
+ u8 reserved3;
+ u8 reserved4;
+};
+
+struct rdac_pg_expanded {
+ struct rdac_mode_10_hdr hdr;
+ u8 page_code;
+ u8 subpage_code;
+ u8 page_len[2];
+ struct rdac_mode_common common;
+ u8 lun_table[256];
+ u8 reserved3;
+ u8 reserved4;
+};
+
+struct c9_inquiry {
+ u8 peripheral_info;
+ u8 page_code; /* 0xC9 */
+ u8 reserved1;
+ u8 page_len;
+ u8 page_id[4]; /* "vace" */
+ u8 avte_cvp;
+ u8 path_prio;
+ u8 reserved2[38];
+};
+
+#define SUBSYS_ID_LEN 16
+#define SLOT_ID_LEN 2
+
+struct c4_inquiry {
+ u8 peripheral_info;
+ u8 page_code; /* 0xC4 */
+ u8 reserved1;
+ u8 page_len;
+ u8 page_id[4]; /* "subs" */
+ u8 subsys_id[SUBSYS_ID_LEN];
+ u8 revision[4];
+ u8 slot_id[SLOT_ID_LEN];
+ u8 reserved[2];
+};
+
+struct rdac_controller {
+ u8 subsys_id[SUBSYS_ID_LEN];
+ u8 slot_id[SLOT_ID_LEN];
+ int use_10_ms;
+ struct kref kref;
+ struct list_head node; /* list of all controllers */
+ spinlock_t lock;
+ int submitted;
+ struct list_head cmd_list; /* list of commands to be submitted */
+ union {
+ struct rdac_pg_legacy legacy;
+ struct rdac_pg_expanded expanded;
+ } mode_select;
+};
+struct c8_inquiry {
+ u8 peripheral_info;
+ u8 page_code; /* 0xC8 */
+ u8 reserved1;
+ u8 page_len;
+ u8 page_id[4]; /* "edid" */
+ u8 reserved2[3];
+ u8 vol_uniq_id_len;
+ u8 vol_uniq_id[16];
+ u8 vol_user_label_len;
+ u8 vol_user_label[60];
+ u8 array_uniq_id_len;
+ u8 array_unique_id[16];
+ u8 array_user_label_len;
+ u8 array_user_label[60];
+ u8 lun[8];
+};
+
+struct c2_inquiry {
+ u8 peripheral_info;
+ u8 page_code; /* 0xC2 */
+ u8 reserved1;
+ u8 page_len;
+ u8 page_id[4]; /* "swr4" */
+ u8 sw_version[3];
+ u8 sw_date[3];
+ u8 features_enabled;
+ u8 max_lun_supported;
+ u8 partitions[239]; /* Total allocation length should be 0xFF */
+};
+
+struct rdac_handler {
+ struct list_head entry; /* list waiting to submit MODE SELECT */
+ unsigned timeout;
+ struct rdac_controller *ctlr;
+#define UNINITIALIZED_LUN (1 << 8)
+ unsigned lun;
+ unsigned char sense[SCSI_SENSE_BUFFERSIZE];
+ struct dm_path *path;
+ struct work_struct work;
+#define SEND_C2_INQUIRY 1
+#define SEND_C4_INQUIRY 2
+#define SEND_C8_INQUIRY 3
+#define SEND_C9_INQUIRY 4
+#define SEND_MODE_SELECT 5
+ int cmd_to_send;
+ union {
+ struct c2_inquiry c2;
+ struct c4_inquiry c4;
+ struct c8_inquiry c8;
+ struct c9_inquiry c9;
+ } inq;
+};
+
+static LIST_HEAD(ctlr_list);
+static DEFINE_SPINLOCK(list_lock);
+static struct workqueue_struct *rdac_wkqd;
+
+static inline int had_failures(struct request *req, int error)
+{
+ return (error || host_byte(req->errors) != DID_OK ||
+ msg_byte(req->errors) != COMMAND_COMPLETE);
+}
+
+static void rdac_resubmit_all(struct rdac_handler *h)
+{
+ struct rdac_controller *ctlr = h->ctlr;
+ struct rdac_handler *tmp, *h1;
+
+ spin_lock(&ctlr->lock);
+ list_for_each_entry_safe(h1, tmp, &ctlr->cmd_list, entry) {
+ h1->cmd_to_send = SEND_C9_INQUIRY;
+ queue_work(rdac_wkqd, &h1->work);
+ list_del(&h1->entry);
+ }
+ ctlr->submitted = 0;
+ spin_unlock(&ctlr->lock);
+}
+
+static void mode_select_endio(struct request *req, int error)
+{
+ struct rdac_handler *h = req->end_io_data;
+ struct scsi_sense_hdr sense_hdr;
+ int sense = 0, fail = 0;
+
+ if (had_failures(req, error)) {
+ fail = 1;
+ goto failed;
+ }
+
+ if (status_byte(req->errors) == CHECK_CONDITION) {
+ scsi_normalize_sense(req->sense, SCSI_SENSE_BUFFERSIZE,
+ &sense_hdr);
+ sense = (sense_hdr.sense_key << 16) | (sense_hdr.asc << 8) |
+ sense_hdr.ascq;
+ /* If it is retryable failure, submit the c9 inquiry again */
+ if (sense == 0x59136 || sense == 0x68b02 || sense == 0xb8b02 ||
+ sense == 0x62900) {
+ /* 0x59136 - Command lock contention
+ * 0x[6b]8b02 - Quiesense in progress or achieved
+ * 0x62900 - Power On, Reset, or Bus Device Reset
+ */
+ h->cmd_to_send = SEND_C9_INQUIRY;
+ queue_work(rdac_wkqd, &h->work);
+ goto done;
+ }
+ if (sense)
+ DMINFO("MODE_SELECT failed on %s with sense 0x%x",
+ h->path->dev->name, sense);
+ }
+failed:
+ if (fail || sense)
+ dm_pg_init_complete(h->path, MP_FAIL_PATH);
+ else
+ dm_pg_init_complete(h->path, 0);
+
+done:
+ rdac_resubmit_all(h);
+ __blk_put_request(req->q, req);
+}
+
+static struct request *get_rdac_req(struct rdac_handler *h,
+ void *buffer, unsigned buflen, int rw)
+{
+ struct request *rq;
+ struct request_queue *q = bdev_get_queue(h->path->dev->bdev);
+
+ rq = blk_get_request(q, rw, GFP_KERNEL);
+
+ if (!rq) {
+ DMINFO("get_rdac_req: blk_get_request failed");
+ return NULL;
+ }
+
+ if (buflen && blk_rq_map_kern(q, rq, buffer, buflen, GFP_KERNEL)) {
+ blk_put_request(rq);
+ DMINFO("get_rdac_req: blk_rq_map_kern failed");
+ return NULL;
+ }
+
+ memset(&rq->cmd, 0, BLK_MAX_CDB);
+ rq->sense = h->sense;
+ memset(rq->sense, 0, SCSI_SENSE_BUFFERSIZE);
+ rq->sense_len = 0;
+
+ rq->end_io_data = h;
+ rq->timeout = h->timeout;
+ rq->cmd_type = REQ_TYPE_BLOCK_PC;
+ rq->cmd_flags = REQ_FAILFAST | REQ_NOMERGE;
+ return rq;
+}
+
+static struct request *rdac_failover_get(struct rdac_handler *h)
+{
+ struct request *rq;
+ struct rdac_mode_common *common;
+ unsigned data_size;
+
+ if (h->ctlr->use_10_ms) {
+ struct rdac_pg_expanded *rdac_pg;
+
+ data_size = sizeof(struct rdac_pg_expanded);
+ rdac_pg = &h->ctlr->mode_select.expanded;
+ memset(rdac_pg, 0, data_size);
+ common = &rdac_pg->common;
+ rdac_pg->page_code = RDAC_PAGE_CODE_REDUNDANT_CONTROLLER + 0x40;
+ rdac_pg->subpage_code = 0x1;
+ rdac_pg->page_len[0] = 0x01;
+ rdac_pg->page_len[1] = 0x28;
+ rdac_pg->lun_table[h->lun] = 0x81;
+ } else {
+ struct rdac_pg_legacy *rdac_pg;
+
+ data_size = sizeof(struct rdac_pg_legacy);
+ rdac_pg = &h->ctlr->mode_select.legacy;
+ memset(rdac_pg, 0, data_size);
+ common = &rdac_pg->common;
+ rdac_pg->page_code = RDAC_PAGE_CODE_REDUNDANT_CONTROLLER;
+ rdac_pg->page_len = 0x68;
+ rdac_pg->lun_table[h->lun] = 0x81;
+ }
+ common->rdac_mode[1] = RDAC_MODE_TRANSFER_SPECIFIED_LUNS;
+ common->quiescence_timeout = RDAC_QUIESCENCE_TIME;
+ common->rdac_options = RDAC_FORCED_QUIESENCE;
+
+ /* get request for block layer packet command */
+ rq = get_rdac_req(h, &h->ctlr->mode_select, data_size, WRITE);
+ if (!rq) {
+ DMERR("rdac_failover_get: no rq");
+ return NULL;
+ }
+
+ /* Prepare the command. */
+ if (h->ctlr->use_10_ms) {
+ rq->cmd[0] = MODE_SELECT_10;
+ rq->cmd[7] = data_size >> 8;
+ rq->cmd[8] = data_size & 0xff;
+ } else {
+ rq->cmd[0] = MODE_SELECT;
+ rq->cmd[4] = data_size;
+ }
+ rq->cmd_len = COMMAND_SIZE(rq->cmd[0]);
+
+ return rq;
+}
+
+/* Acquires h->ctlr->lock */
+static void submit_mode_select(struct rdac_handler *h)
+{
+ struct request *rq;
+ struct request_queue *q = bdev_get_queue(h->path->dev->bdev);
+
+ spin_lock(&h->ctlr->lock);
+ if (h->ctlr->submitted) {
+ list_add(&h->entry, &h->ctlr->cmd_list);
+ goto drop_lock;
+ }
+
+ if (!q) {
+ DMINFO("submit_mode_select: no queue");
+ goto fail_path;
+ }
+
+ rq = rdac_failover_get(h);
+ if (!rq) {
+ DMERR("submit_mode_select: no rq");
+ goto fail_path;
+ }
+
+ DMINFO("queueing MODE_SELECT command on %s", h->path->dev->name);
+
+ blk_execute_rq_nowait(q, NULL, rq, 1, mode_select_endio);
+ h->ctlr->submitted = 1;
+ goto drop_lock;
+fail_path:
+ dm_pg_init_complete(h->path, MP_FAIL_PATH);
+drop_lock:
+ spin_unlock(&h->ctlr->lock);
+}
+
+static void release_ctlr(struct kref *kref)
+{
+ struct rdac_controller *ctlr;
+ ctlr = container_of(kref, struct rdac_controller, kref);
+
+ spin_lock(&list_lock);
+ list_del(&ctlr->node);
+ spin_unlock(&list_lock);
+ kfree(ctlr);
+}
+
+static struct rdac_controller *get_controller(u8 *subsys_id, u8 *slot_id)
+{
+ struct rdac_controller *ctlr, *tmp;
+
+ spin_lock(&list_lock);
+
+ list_for_each_entry(tmp, &ctlr_list, node) {
+ if ((memcmp(tmp->subsys_id, subsys_id, SUBSYS_ID_LEN) == 0) &&
+ (memcmp(tmp->slot_id, slot_id, SLOT_ID_LEN) == 0)) {
+ kref_get(&tmp->kref);
+ spin_unlock(&list_lock);
+ return tmp;
+ }
+ }
+ ctlr = kmalloc(sizeof(*ctlr), GFP_ATOMIC);
+ if (!ctlr)
+ goto done;
+
+ /* initialize fields of controller */
+ memcpy(ctlr->subsys_id, subsys_id, SUBSYS_ID_LEN);
+ memcpy(ctlr->slot_id, slot_id, SLOT_ID_LEN);
+ kref_init(&ctlr->kref);
+ spin_lock_init(&ctlr->lock);
+ ctlr->submitted = 0;
+ ctlr->use_10_ms = -1;
+ INIT_LIST_HEAD(&ctlr->cmd_list);
+ list_add(&ctlr->node, &ctlr_list);
+done:
+ spin_unlock(&list_lock);
+ return ctlr;
+}
+
+static void c4_endio(struct request *req, int error)
+{
+ struct rdac_handler *h = req->end_io_data;
+ struct c4_inquiry *sp;
+
+ if (had_failures(req, error)) {
+ dm_pg_init_complete(h->path, MP_FAIL_PATH);
+ goto done;
+ }
+
+ sp = &h->inq.c4;
+
+ h->ctlr = get_controller(sp->subsys_id, sp->slot_id);
+
+ if (h->ctlr) {
+ h->cmd_to_send = SEND_C9_INQUIRY;
+ queue_work(rdac_wkqd, &h->work);
+ } else
+ dm_pg_init_complete(h->path, MP_FAIL_PATH);
+done:
+ __blk_put_request(req->q, req);
+}
+
+static void c2_endio(struct request *req, int error)
+{
+ struct rdac_handler *h = req->end_io_data;
+ struct c2_inquiry *sp;
+
+ if (had_failures(req, error)) {
+ dm_pg_init_complete(h->path, MP_FAIL_PATH);
+ goto done;
+ }
+
+ sp = &h->inq.c2;
+
+ /* If more than MODE6_MAX_LUN luns are supported, use mode select 10 */
+ if (sp->max_lun_supported >= MODE6_MAX_LUN)
+ h->ctlr->use_10_ms = 1;
+ else
+ h->ctlr->use_10_ms = 0;
+
+ h->cmd_to_send = SEND_MODE_SELECT;
+ queue_work(rdac_wkqd, &h->work);
+done:
+ __blk_put_request(req->q, req);
+}
+
+static void c9_endio(struct request *req, int error)
+{
+ struct rdac_handler *h = req->end_io_data;
+ struct c9_inquiry *sp;
+
+ if (had_failures(req, error)) {
+ dm_pg_init_complete(h->path, MP_FAIL_PATH);
+ goto done;
+ }
+
+ /* We need to look at the sense keys here to take clear action.
+ * For now simple logic: If the host is in AVT mode or if controller
+ * owns the lun, return dm_pg_init_complete(), otherwise submit
+ * MODE SELECT.
+ */
+ sp = &h->inq.c9;
+
+ /* If in AVT mode, return success */
+ if ((sp->avte_cvp >> 7) == 0x1) {
+ dm_pg_init_complete(h->path, 0);
+ goto done;
+ }
+
+ /* If the controller on this path owns the LUN, return success */
+ if (sp->avte_cvp & 0x1) {
+ dm_pg_init_complete(h->path, 0);
+ goto done;
+ }
+
+ if (h->ctlr) {
+ if (h->ctlr->use_10_ms == -1)
+ h->cmd_to_send = SEND_C2_INQUIRY;
+ else
+ h->cmd_to_send = SEND_MODE_SELECT;
+ } else
+ h->cmd_to_send = SEND_C4_INQUIRY;
+ queue_work(rdac_wkqd, &h->work);
+done:
+ __blk_put_request(req->q, req);
+}
+
+static void c8_endio(struct request *req, int error)
+{
+ struct rdac_handler *h = req->end_io_data;
+ struct c8_inquiry *sp;
+
+ if (had_failures(req, error)) {
+ dm_pg_init_complete(h->path, MP_FAIL