summaryrefslogtreecommitdiffstats
path: root/drivers/md
Commit message (Collapse)AuthorAgeFilesLines
* dm cache: add cache block invalidation supportJoe Thornber2013-11-111-3/+222
| | | | | | | | | | | | | | Cache block invalidation is removing an entry from the cache without writing it back. Cache blocks can be invalidated via the 'invalidate_cblocks' message, which takes an arbitrary number of cblock ranges: invalidate_cblocks [<cblock>|<cblock begin>-<cblock end>]* E.g. dmsetup message my_cache 0 invalidate_cblocks 2345 3456-4567 5678-6789 Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache: add remove_cblock method to policy interfaceJoe Thornber2013-11-113-4/+57
| | | | | | | | | | | | Implement policy_remove_cblock() and add remove_cblock method to the mq policy. These methods will be used by the following cache block invalidation patch which adds the 'invalidate_cblocks' message to the cache core. Also, update some comments in dm-cache-policy.h Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache policy mq: reduce memory requirementsJoe Thornber2013-11-111-312/+231
| | | | | | | | | | | | | | Rather than storing the cblock in each cache entry, we allocate all entries in an array and infer the cblock from the entry position. Saves 4 bytes of memory per cache block. In addition, this gives us an easy way of looking up cache entries by cblock. We no longer need to keep an explicit bitset to track which cblocks have been allocated. And no searching is needed to find free cblocks. Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache metadata: check the metadata version when reading the superblockJoe Thornber2013-11-111-3/+21
| | | | | | | Need to check the version to verify on-disk metadata is supported. Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache: add passthrough modeJoe Thornber2013-11-113-35/+184
| | | | | | | | | | | | | | | | | | | | | | | | | | | | "Passthrough" is a dm-cache operating mode (like writethrough or writeback) which is intended to be used when the cache contents are not known to be coherent with the origin device. It behaves as follows: * All reads are served from the origin device (all reads miss the cache) * All writes are forwarded to the origin device; additionally, write hits cause cache block invalidates This mode decouples cache coherency checks from cache device creation, largely to avoid having to perform coherency checks while booting. Boot scripts can create cache devices in passthrough mode and put them into service (mount cached filesystems, for example) without having to worry about coherency. Coherency that exists is maintained, although the cache will gradually cool as writes take place. Later, applications can perform coherency checks, the nature of which will depend on the type of the underlying storage. If coherency can be verified, the cache device can be transitioned to writethrough or writeback mode while still warm; otherwise, the cache contents can be discarded prior to transitioning to the desired operating mode. Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Heinz Mauelshagen <heinzm@redhat.com> Signed-off-by: Morgan Mears <Morgan.Mears@netapp.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache: cache shrinking supportJoe Thornber2013-11-112-9/+120
| | | | | | | | Allow a cache to shrink if the blocks being removed from the cache are not dirty. Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache: promotion optimisation for writesJoe Thornber2013-11-091-6/+87
| | | | | | | | | | | | If a write block triggers promotion and covers a whole block we can avoid a copy. Introduce dm_{hook,unhook}_bio to simplify saving and restoring bio fields (bi_private is now used by overwrite). Switch writethrough support over to using these helpers too. Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache: be much more aggressive about promoting writes to discarded blocksJoe Thornber2013-11-091-21/+63
| | | | | | | | | | | | Previously these promotions only got priority if there were unused cache blocks. Now we give them priority if there are any clean blocks in the cache. The fio_soak_test in the device-mapper-test-suite now gives uniform performance across subvolumes (~16 seconds). Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache policy mq: implement writeback_work() and mq_{set,clear}_dirty()Joe Thornber2013-11-091-19/+128
| | | | | | | | | | | | | | | | There are now two multiqueues for in cache blocks. A clean one and a dirty one. writeback_work comes from the dirty one. Demotions come from the clean one. There are two benefits: - Performance improvement, since demoting a clean block is a noop. - The cache cleans itself when io load is light. Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Heinz Mauelshagen <heinzm@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache: optimize commit_if_neededHeinz Mauelshagen2013-11-091-5/+7
| | | | | | | | | | | | Check commit_requested flag _before_ calling dm_cache_changed_this_transaction() superfluously. Also, be sure to set last_commit_jiffies _after_ dm_cache_commit() completes. Signed-off-by: Heinz Mauelshagen <heinzm@redhat.com> Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm space map disk: optimise sm_disk_dec_blockJoe Thornber2013-11-091-17/+1
| | | | | | | | | | Don't waste time spotting blocks that have been allocated and then freed in the same transaction. The extra lookup is expensive, and I don't think it really gives us much. Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm: fix Kconfig menu indentationMikulas Patocka2013-11-091-11/+11
| | | | | | | | | The option DM_LOG_USERSPACE is sub-option of DM_MIRROR, so place it right after DM_MIRROR. Doing so fixes various other Device mapper targets/features to be properly nested under "Device mapper support". Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm: allow remove to be deferredMikulas Patocka2013-11-093-10/+86
| | | | | | | | | | | | | | | | | | | | | | | | | This patch allows the removal of an open device to be deferred until it is closed. (Previously such a removal attempt would fail.) The deferred remove functionality is enabled by setting the flag DM_DEFERRED_REMOVE in the ioctl structure on DM_DEV_REMOVE or DM_REMOVE_ALL ioctl. On return from DM_DEV_REMOVE, the flag DM_DEFERRED_REMOVE indicates if the device was removed immediately or flagged to be removed on close - if the flag is clear, the device was removed. On return from DM_DEV_STATUS and other ioctls, the flag DM_DEFERRED_REMOVE is set if the device is scheduled to be removed on closure. A device that is scheduled to be deleted can be revived using the message "@cancel_deferred_remove". This message clears the DMF_DEFERRED_REMOVE flag so that the device won't be deleted on close. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Alasdair G Kergon <agk@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm table: print error on preresume failureMike Snitzer2013-11-091-1/+4
| | | | | | | | | | | | | If preresume fails it is worth logging an error given that a device is left suspended due to the failure. This change was motivated by local preresume error logging that was added to the cache target ("preresume failed"). Elevating this target-agnostic context for the where the target-specific error occurred relative to the DM core's callouts makes sense. Signed-off-by: Mike Snitzer <snitzer@redhat.com> Signed-off-by: Joe Thornber <ejt@redhat.com>
* dm crypt: add TCW IV mode for old CBC TCRYPT containersMilan Broz2013-11-091-2/+183
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | dm-crypt can already activate TCRYPT (TrueCrypt compatible) containers in LRW or XTS block encryption mode. TCRYPT containers prior to version 4.1 use CBC mode with some additional tweaks, this patch adds support for these containers. This new mode is implemented using special IV generator named TCW (TrueCrypt IV with whitening). TCW IV only supports containers that are encrypted with one cipher (Tested with AES, Twofish, Serpent, CAST5 and TripleDES). While this mode is legacy and is known to be vulnerable to some watermarking attacks (e.g. revealing of hidden disk existence) it can still be useful to activate old containers without using 3rd party software or for independent forensic analysis of such containers. (Both the userspace and kernel code is an independent implementation based on the format documentation and it completely avoids use of original source code.) The TCW IV generator uses two additional keys: Kw (whitening seed, size is always 16 bytes - TCW_WHITENING_SIZE) and Kiv (IV seed, size is always the IV size of the selected cipher). These keys are concatenated at the end of the main encryption key provided in mapping table. While whitening is completely independent from IV, it is implemented inside IV generator for simplification. The whitening value is always 16 bytes long and is calculated per sector from provided Kw as initial seed, xored with sector number and mixed with CRC32 algorithm. Resulting value is xored with ciphertext sector content. IV is calculated from the provided Kiv as initial IV seed and xored with sector number. Detailed calculation can be found in the Truecrypt documentation for version < 4.1 and will also be described on dm-crypt site, see: http://code.google.com/p/cryptsetup/wiki/DMCrypt The experimental support for activation of these containers is already present in git devel brach of cryptsetup. Signed-off-by: Milan Broz <gmazyland@gmail.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm crypt: properly handle extra key string in initializationMilan Broz2013-11-091-11/+18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Some encryption modes use extra keys (e.g. loopAES has IV seed) which are not used in block cipher initialization but are part of key string in table constructor. This patch adds an additional field which describes the length of the extra key(s) and substracts it before real key encryption setting. The key_size always includes the size, in bytes, of the key provided in mapping table. The key_parts describes how many parts (usually keys) are contained in the whole key buffer. And key_extra_size contains size in bytes of additional keys part (this number of bytes must be subtracted because it is processed by the IV generator). | K1 | K2 | .... | K64 | Kiv | |----------- key_size ----------------- | | |-key_extra_size-| | [64 keys] | [1 key] | => key_parts = 65 Example where key string contains main key K, whitening key Kw and IV seed Kiv: | K | Kiv | Kw | |--------------- key_size --------------| | |-----key_extra_size------| | [1 key] | [1 key] | [1 key] | => key_parts = 3 Because key_extra_size is calculated during IV mode setting, key initialization is moved after this step. For now, this change has no effect to supported modes (thanks to ilog2 rounding) but it is required by the following patch. Also, fix a sparse warning in crypt_iv_lmk_one(). Signed-off-by: Milan Broz <gmazyland@gmail.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache: log error message if dm_kcopyd_copy() failsHeinz Mauelshagen2013-11-091-1/+3
| | | | | | | | A migration failure should be logged (albeit limited). Signed-off-by: Heinz Mauelshagen <heinzm@redhat.com> Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache: use cell_defer() boolean argument consistentlyHeinz Mauelshagen2013-11-091-4/+4
| | | | | | | | Fix a few cell_defer() calls that weren't passing a bool. Signed-off-by: Heinz Mauelshagen <heinzm@redhat.com> Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache: return -EINVAL if the user specifies unknown cache policyMikulas Patocka2013-11-092-8/+9
| | | | | | | | Return -EINVAL when the specified cache policy is unknown rather than returning -ENOMEM. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache metadata: return bool from __superblock_all_zeroesJoe Thornber2013-11-091-4/+5
| | | | | Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache policy mq: a few small fixesJoe Thornber2013-11-091-10/+12
| | | | | | | | | | | | Rename takeout_queue to concat_queue. Fix a harmless bug in mq policies pop() function. Currently pop() always succeeds, with up coming changes this wont be the case. Fix typo in comment above pre_cache_to_cache prototype. Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache policy: remove return from void policy_remove_mappingJoe Thornber2013-11-091-1/+1
| | | | | | | No need to return from a void function. Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache: improve efficiency of quiescing flag managementJoe Thornber2013-11-091-22/+5
| | | | | | | | Make the quiescing flag an atomic_t and stop protecting it with a spin lock. Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache: fix a race condition between queuing new migrations and quiescing ↵Joe Thornber2013-11-091-14/+40
| | | | | | | | | | | | | | | | for a shutdown The code that was trying to do this was inadequate. The postsuspend method (in ioctl context), needs to wait for the worker thread to acknowledge the request to quiesce. Otherwise the migration count may drop to zero temporarily before the worker thread realises we're quiescing. In this case the target will be taken down, but the worker thread may have issued a new migration, which will cause an oops when it completes. Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org # 3.9+
* dm cache: io destined for the cache device can now serve as tick biosJoe Thornber2013-11-091-0/+1
| | | | | | | | | | | | Previously only origin bios could trigger ticks, which meant if all the io was destined for the cache no ticks were generated. If no ticks are generated then multiple hits, and movements in general, are attributed to the same tick. Only a stop gap fix, we need a better solution. Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm cache policy mq: protect residency method with existing mutexJoe Thornber2013-11-091-2/+6
| | | | | | | | | It is safe to use a mutex in mq_residency() at this point since it is only called from ioctl context. But future-proof mq_residency() by using might_sleep() to catch new contexts that cannot sleep. Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm array: fix bug in growing arrayJoe Thornber2013-11-051-1/+4
| | | | | | | | Entries would be lost if the old tail block was partially filled. Signed-off-by: Joe Thornber <ejt@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org # 3.9+
* dm mpath: requeue I/O during pg_initHannes Reinecke2013-11-051-4/+12
| | | | | | | | | | | | | | | | | | | | | | | | When pg_init is running no I/O can be submitted to the underlying devices, as the path priority etc might change. When using queue_io for this, requests will be piling up within multipath as the block I/O scheduler just sees a _very fast_ device. All of this queued I/O has to be resubmitted from within multipathing once pg_init is done. This approach has the problem that it's virtually impossible to abort I/O when pg_init is running, and we're adding heavy load to the devices after pg_init since all of the queued I/O needs to be resubmitted _before_ any requests can be pulled off of the request queue and normal operation continues. This patch will requeue the I/O that triggers the pg_init call, and return 'busy' when pg_init is in progress. With these changes the block I/O scheduler will stop submitting I/O during pg_init, resulting in a quicker path switch and less I/O pressure (and memory consumption) after pg_init. Signed-off-by: Hannes Reinecke <hare@suse.de> [patch header edited for clarity and typos by Mike Snitzer] Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* dm mpath: fix race condition between multipath_dtr and pg_init_doneShiva Krishna Merla2013-10-311-3/+15
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Whenever multipath_dtr() is happening we must prevent queueing any further path activation work. Implement this by adding a new 'pg_init_disabled' flag to the multipath structure that denotes future path activation work should be skipped if it is set. By disabling pg_init and then re-enabling in flush_multipath_work() we also avoid the potential for pg_init to be initiated while suspending an mpath device. Without this patch a race condition exists that may result in a kernel panic: 1) If after pg_init_done() decrements pg_init_in_progress to 0, a call to wait_for_pg_init_completion() assumes there are no more pending path management commands. 2) If pg_init_required is set by pg_init_done(), due to retryable mode_select errors, then process_queued_ios() will again queue the path activation work. 3) If free_multipath() completes before activate_path() work is called a NULL pointer dereference like the following can be seen when accessing members of the recently destructed multipath: BUG: unable to handle kernel NULL pointer dereference at 0000000000000090 RIP: 0010:[<ffffffffa003db1b>] [<ffffffffa003db1b>] activate_path+0x1b/0x30 [dm_multipath] [<ffffffff81090ac0>] worker_thread+0x170/0x2a0 [<ffffffff81096c80>] ? autoremove_wake_function+0x0/0x40 [switch to disabling pg_init in flush_multipath_work & header edits by Mike Snitzer] Signed-off-by: Shiva Krishna Merla <shivakrishna.merla@netapp.com> Reviewed-by: Krishnasamy Somasundaram <somasundaram.krishnasamy@netapp.com> Tested-by: Speagle Andy <Andy.Speagle@netapp.com> Acked-by: Junichi Nomura <j-nomura@ce.jp.nec.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
* dm: allocate buffer for messages with small number of arguments using GFP_NOIOMikulas Patocka2013-10-311-2/+16
| | | | | | | | | | | | | | | | | dm-mpath and dm-thin must process messages even if some device is suspended, so we allocate argv buffer with GFP_NOIO. These messages have a small fixed number of arguments. On the other hand, dm-switch needs to process bulk data using messages so excessive use of GFP_NOIO could cause trouble. The patch also lowers the default number of arguments from 64 to 8, so that there is smaller load on GFP_NOIO allocations. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Cc: stable@vger.kernel.org Acked-by: Alasdair G Kergon <agk@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* bcache: Fix a null ptr deref regressionKent Overstreet2013-10-101-2/+1
| | | | | | | | | | | Commit c0f04d88e46d ("bcache: Fix flushes in writeback mode") was fixing a reported data corruption bug, but it seems some last minute refactoring or rebasing introduced a null pointer deref. Signed-off-by: Kent Overstreet <kmo@daterainc.com> Cc: linux-stable <stable@vger.kernel.org> # >= v3.10 Reported-by: Gabriel de Perthuis <g2p.code@gmail.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* Merge tag 'dm-3.12-fixes' of ↵Linus Torvalds2013-09-258-25/+118
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm Pull device-mapper fixes from Mike Snitzer: "A few fixes for dm-snapshot, a 32 bit fix for dm-stats, a couple error handling fixes for dm-multipath. A fix for the thin provisioning target to not expose non-zero discard limits if discards are disabled. Lastly, add two DM module parameters which allow users to tune the emergency memory reserves that DM mainatins per device -- this helps fix a long-standing issue for dm-multipath. The conservative default reserve for request-based dm-multipath devices (256) has proven problematic for users with many multipathed SCSI devices but relatively little memory. To responsibly select a smaller value users should use the new nr_bios tracepoint info (via commit 75afb352 "block: Add nr_bios to block_rq_remap tracepoint") to determine the peak number of bios their workloads create" * tag 'dm-3.12-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm: dm: add reserved_bio_based_ios module parameter dm: add reserved_rq_based_ios module parameter dm: lower bio-based mempool reservation dm thin: do not expose non-zero discard limits if discards disabled dm mpath: disable WRITE SAME if it fails dm-snapshot: fix performance degradation due to small hash size dm snapshot: workaround for a false positive lockdep warning dm stats: fix possible counter corruption on 32-bit systems dm mpath: do not fail path on -ENOSPC
| * dm: add reserved_bio_based_ios module parameterMike Snitzer2013-09-233-5/+20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Allow user to change the number of IOs that are reserved by bio-based DM's mempools by writing to this file: /sys/module/dm_mod/parameters/reserved_bio_based_ios The default value is RESERVED_BIO_BASED_IOS (16). The maximum allowed value is RESERVED_MAX_IOS (1024). Export dm_get_reserved_bio_based_ios() for use by DM targets and core code. Switch to sizing dm-io's mempool and bioset using DM core's configurable 'reserved_bio_based_ios'. Signed-off-by: Mike Snitzer <snitzer@redhat.com> Signed-off-by: Frank Mayhar <fmayhar@google.com>
| * dm: add reserved_rq_based_ios module parameterMike Snitzer2013-09-233-4/+42
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Allow user to change the number of IOs that are reserved by request-based DM's mempools by writing to this file: /sys/module/dm_mod/parameters/reserved_rq_based_ios The default value is RESERVED_REQUEST_BASED_IOS (256). The maximum allowed value is RESERVED_MAX_IOS (1024). Export dm_get_reserved_rq_based_ios() for use by DM targets and core code. Switch to sizing dm-mpath's mempool using DM core's configurable 'reserved_rq_based_ios'. Signed-off-by: Mike Snitzer <snitzer@redhat.com> Signed-off-by: Frank Mayhar <fmayhar@google.com> Acked-by: Mikulas Patocka <mpatocka@redhat.com>
| * dm: lower bio-based mempool reservationMike Snitzer2013-09-231-4/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Bio-based device mapper processing doesn't need larger mempools (like request-based DM does), so lower the number of reserved entries for bio-based operation. 16 was already used for bio-based DM's bioset but mistakenly wasn't used for it's _io_cache. Formalize difference between bio-based and request-based defaults by introducing RESERVED_BIO_BASED_IOS and RESERVED_REQUEST_BASED_IOS. (based on older code from Mikulas Patocka) Signed-off-by: Mike Snitzer <snitzer@redhat.com> Signed-off-by: Frank Mayhar <fmayhar@google.com> Acked-by: Mikulas Patocka <mpatocka@redhat.com>
| * dm thin: do not expose non-zero discard limits if discards disabledMike Snitzer2013-09-231-3/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fix issue where the block layer would stack the discard limits of the pool's data device even if the "ignore_discard" pool feature was specified. The pool and thin device(s) still had discards disabled because the QUEUE_FLAG_DISCARD request_queue flag wasn't set. But to avoid user confusion when "ignore_discard" is used: both the pool device and the thin device(s) have zeroes for all discard limits. Also, always set discard_zeroes_data_unsupported in targets because they should never advertise the 'discard_zeroes_data' capability (even if the pool's data device supports it). Signed-off-by: Mike Snitzer <snitzer@redhat.com> Acked-by: Joe Thornber <ejt@redhat.com>
| * dm mpath: disable WRITE SAME if it failsMike Snitzer2013-09-202-1/+21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Workaround the SCSI layer's problematic WRITE SAME heuristics by disabling WRITE SAME in the DM multipath device's queue_limits if an underlying device disabled it. The WRITE SAME heuristics, with both the original commit 5db44863b6eb ("[SCSI] sd: Implement support for WRITE SAME") and the updated commit 66c28f971 ("[SCSI] sd: Update WRITE SAME heuristics"), default to enabling WRITE SAME(10) even without successfully determining it is supported. After the first failed WRITE SAME the SCSI layer will disable WRITE SAME for the device (by setting sdkp->device->no_write_same which results in 'max_write_same_sectors' in device's queue_limits to be set to 0). When a device is stacked ontop of such a SCSI device any changes to that SCSI device's queue_limits do not automatically propagate up the stack. As such, a DM multipath device will not have its WRITE SAME support disabled. This causes the block layer to continue to issue WRITE SAME requests to the mpath device which causes paths to fail and (if mpath IO isn't configured to queue when no paths are available) it will result in actual IO errors to the upper layers. This fix doesn't help configurations that have additional devices stacked ontop of the mpath device (e.g. LVM created linear DM devices ontop). A proper fix that restacks all the queue_limits from the bottom of the device stack up will need to be explored if SCSI will continue to use this model of optimistically allowing op codes and then disabling them after they fail for the first time. Before this patch: EXT4-fs (dm-6): mounted filesystem with ordered data mode. Opts: (null) device-mapper: multipath: XXX snitm debugging: got -EREMOTEIO (-121) device-mapper: multipath: XXX snitm debugging: failing WRITE SAME IO with error=-121 end_request: critical target error, dev dm-6, sector 528 dm-6: WRITE SAME failed. Manually zeroing. device-mapper: multipath: Failing path 8:112. end_request: I/O error, dev dm-6, sector 4616 dm-6: WRITE SAME failed. Manually zeroing. end_request: I/O error, dev dm-6, sector 4616 end_request: I/O error, dev dm-6, sector 5640 end_request: I/O error, dev dm-6, sector 6664 end_request: I/O error, dev dm-6, sector 7688 end_request: I/O error, dev dm-6, sector 524288 Buffer I/O error on device dm-6, logical block 65536 lost page write due to I/O error on dm-6 JBD2: Error -5 detected when updating journal superblock for dm-6-8. end_request: I/O error, dev dm-6, sector 524296 Aborting journal on device dm-6-8. end_request: I/O error, dev dm-6, sector 524288 Buffer I/O error on device dm-6, logical block 65536 lost page write due to I/O error on dm-6 JBD2: Error -5 detected when updating journal superblock for dm-6-8. # cat /sys/block/sdh/queue/write_same_max_bytes 0 # cat /sys/block/dm-6/queue/write_same_max_bytes 33553920 After this patch: EXT4-fs (dm-6): mounted filesystem with ordered data mode. Opts: (null) device-mapper: multipath: XXX snitm debugging: got -EREMOTEIO (-121) device-mapper: multipath: XXX snitm debugging: WRITE SAME I/O failed with error=-121 end_request: critical target error, dev dm-6, sector 528 dm-6: WRITE SAME failed. Manually zeroing. # cat /sys/block/sdh/queue/write_same_max_bytes 0 # cat /sys/block/dm-6/queue/write_same_max_bytes 0 It should be noted that WRITE SAME support wasn't enabled in DM multipath until v3.10. Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: Martin K. Petersen <martin.petersen@oracle.com> Cc: Hannes Reinecke <hare@suse.de> Cc: stable@vger.kernel.org # 3.10+
| * dm-snapshot: fix performance degradation due to small hash sizeMikulas Patocka2013-09-201-3/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | LVM2, since version 2.02.96, creates origin with zero size, then loads the snapshot driver and then loads the origin. Consequently, the snapshot driver sees the origin size zero and sets the hash size to the lower bound 64. Such small hash table causes performance degradation. This patch changes it so that the hash size is determined by the size of snapshot volume, not minimum of origin and snapshot size. It doesn't make sense to set the snapshot size significantly larger than the origin size, so we do not need to take origin size into account when calculating the hash size. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org
| * dm snapshot: workaround for a false positive lockdep warningMikulas Patocka2013-09-201-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The kernel reports a lockdep warning if a snapshot is invalidated because it runs out of space. The lockdep warning was triggered by commit 0976dfc1d0cd80a4e9dfaf87bd87 ("workqueue: Catch more locking problems with flush_work()") in v3.5. The warning is false positive. The real cause for the warning is that the lockdep engine treats different instances of md->lock as a single lock. This patch is a workaround - we use flush_workqueue instead of flush_work. This code path is not performance sensitive (it is called only on initialization or invalidation), thus it doesn't matter that we flush the whole workqueue. The real fix for the problem would be to teach the lockdep engine to treat different instances of md->lock as separate locks. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Acked-by: Alasdair G Kergon <agk@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Cc: stable@vger.kernel.org # 3.5+
| * dm stats: fix possible counter corruption on 32-bit systemsMikulas Patocka2013-09-181-6/+17
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | There was a deliberate race condition in dm_stat_for_entry() to avoid the overhead of disabling and enabling interrupts. The race could result in some events not being counted on 64-bit architectures. However, on 32-bit architectures, operations on long long variables are not atomic, so the race condition could cause the counter to jump by 2^32. Such jumps could be disruptive, so we need to do proper locking on 32-bit architectures. Signed-off-by: Mikulas Patocka <mpatocka@redhat.com> Cc: Alasdair G. Kergon <agk@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
| * dm mpath: do not fail path on -ENOSPCJun'ichi Nomura2013-09-181-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Since ENOSPC is a target-side error, dm-mpath should just pass the error information to upper layer instead of retrying itself with path failover. Otherwise it will end up failing all paths down while path checkers find all paths ok. ENOSPC can now be returned from SCSI device after commit a9d6ceb8 ("[SCSI] return ENOSPC on thin provisioning failure"). Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com> Acked-by: Hannes Reinecke <hare@suse.de> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
* | bcache: Fix flushes in writeback modeKent Overstreet2013-09-241-6/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In writeback mode, when we get a cache flush we need to make sure we issue a flush to the backing device. The code for sending down an extra flush was wrong - by cloning the bio we were probably getting flags that didn't make sense for a bare flush, and also the old code was firing for FUA bios, for which we don't need to send a flush to the backing device. This was causing data corruption somehow - the mechanism was never determined, but this patch fixes it for the users that were seeing it. Signed-off-by: Kent Overstreet <kmo@daterainc.com> Cc: linux-stable <stable@vger.kernel.org> # >= v3.10 Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* | bcache: Fix for handling overlapping extents when reading in a btree nodeKent Overstreet2013-09-241-11/+28
| | | | | | | | | | | | | | | | | | | | | | | | | | btree_sort_fixup() was overly clever, because it was trying to avoid pulling a key off the btree iterator in more than one place. This led to a really obscure bug where we'd break early from the loop in btree_sort_fixup() if the current key overlapped with keys in more than one older set, and the next key it overlapped with was zero size. Signed-off-by: Kent Overstreet <kmo@daterainc.com> Cc: linux-stable <stable@vger.kernel.org> # >= v3.10 Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* | bcache: Fix a shrinker deadlockKent Overstreet2013-09-241-1/+1
| | | | | | | | | | | | | | | | | | | | GFP_NOIO means we could be getting called recursively - mca_alloc() -> mca_data_alloc() - definitely can't use mutex_lock(bucket_lock) then. Whoops. Signed-off-by: Kent Overstreet <kmo@daterainc.com> Cc: linux-stable <stable@vger.kernel.org> # >= v3.10 Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* | bcache: Fix a dumb CPU spinning bug in writebackKent Overstreet2013-09-241-2/+1
| | | | | | | | | | | | | | | | schedule_timeout() != schedule_timeout_uninterruptible() Signed-off-by: Kent Overstreet <kmo@daterainc.com> Cc: linux-stable <stable@vger.kernel.org> # >= v3.10 Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* | bcache: Fix a flush/fua performance bugKent Overstreet2013-09-241-0/+1
| | | | | | | | | | | | | | | | | | | | | | bch_journal_meta() was missing the flush to make the journal write actually go down (instead of waiting up to journal_delay_ms)... Whoops Signed-off-by: Kent Overstreet <kmo@daterainc.com> Cc: linux-stable <stable@vger.kernel.org> # >= v3.10 Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* | bcache: Fix a writeback performance regressionKent Overstreet2013-09-244-30/+43
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Background writeback works by scanning the btree for dirty data and adding those keys into a fixed size buffer, then for each dirty key in the keybuf writing it to the backing device. When read_dirty() finishes and it's time to scan for more dirty data, we need to wait for the outstanding writeback IO to finish - they still take up slots in the keybuf (so that foreground writes can check for them to avoid races) - without that wait, we'll continually rescan when we'll be able to add at most a key or two to the keybuf, and that takes locks that starves foreground IO. Doh. Signed-off-by: Kent Overstreet <kmo@daterainc.com> Cc: linux-stable <stable@vger.kernel.org> # >= v3.10 Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* | bcache: Correct printf()-style format length modifierGeert Uytterhoeven2013-09-241-1/+1
| | | | | | | | | | | | | | | | | | | | | | Fix drivers/md/bcache/btree.c: In function ‘bch_btree_node_read’: drivers/md/bcache/btree.c:259: warning: format ‘%lu’ expects type ‘long unsigned int’, but argument 3 has type ‘size_t’ Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> Signed-off-by: Kent Overstreet <kmo@daterainc.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* | bcache: Fix for when no journal entries are foundKent Overstreet2013-09-241-12/+18
| | | | | | | | | | | | | | | | | | The journal replay code didn't handle this case, causing it to go into an infinite loop... Signed-off-by: Kent Overstreet <kmo@daterainc.com> Cc: linux-stable <stable@vger.kernel.org> # >= v3.10 Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* | bcache: Strip endline when writing the label through sysfsGabriel de Perthuis2013-09-241-2/+7
| | | | | | | | | | | | | | | | | | | | | | | | sysfs attributes with unusual characters have crappy failure modes in Squeeze (udev 164); later versions of udev are unaffected. This should make these characters more unusual. Signed-off-by: Gabriel de Perthuis <g2p.code@gmail.com> Signed-off-by: Kent Overstreet <kmo@daterainc.com> Cc: linux-stable <stable@vger.kernel.org> # >= v3.10 Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
OpenPOWER on IntegriCloud