summaryrefslogtreecommitdiff
path: root/drivers/md/raid5.c
AgeCommit message (Expand)AuthorFilesLines
2015-02-25raid5: check faulty flag for array status during recovery.Eric Mei1-4/+9
2015-02-18md/raid5: Fix livelock when array is both resyncing and degraded.NeilBrown1-1/+2
2015-02-06md: make reconfig_mutex optional for writes to md sysfs files.NeilBrown1-58/+76
2015-02-06md/raid5: use ->lock to protect accessing raid5 sysfs attributes.NeilBrown1-16/+28
2015-02-04md: rename ->stop to ->freeNeilBrown1-7/+5
2015-02-04md: split detach operation out from ->stop.NeilBrown1-1/+0
2015-02-04md: make merge_bvec_fn more robust in face of personality changes.NeilBrown1-4/+4
2015-02-04md: make ->congested robust against personality changes.NeilBrown1-15/+4
2015-02-04md/raid5: need_this_block: tidy/fix last condition.NeilBrown1-10/+32
2015-02-04md/raid5: need_this_block: start simplifying the last two conditions.NeilBrown1-5/+23
2015-02-04md/raid5: separate out the easy conditions in need_this_block.NeilBrown1-9/+29
2015-02-04md/raid5: separate large if clause out of fetch_block().NeilBrown1-4/+15
2015-02-04md: do_release_stripe(): No need to call md_wakeup_thread() twiceJes Sorensen1-5/+2
2015-02-02md/raid5: fix another livelock caused by non-aligned writes.NeilBrown1-0/+5
2014-12-03md/raid5: fetch_block must fetch all the blocks handle_stripe_dirtying wants.NeilBrown1-2/+5
2014-10-14md: remove unwanted white space from md.cNeilBrown1-22/+1
2014-10-14md/raid5: fix init_stripe() inconsistenciesMarkus Stockhausen1-3/+1
2014-10-09md: use set_bit/clear_bit instead of shift/mask for bi_flags changes.NeilBrown1-1/+1
2014-10-02md/raid5: disable 'DISCARD' by default due to safety concerns.NeilBrown1-1/+17
2014-08-18md/raid6: avoid data corruption during recovery of double-degraded RAID6NeilBrown1-0/+2
2014-08-18md/raid5: avoid livelock caused by non-aligned writes.NeilBrown1-1/+1
2014-06-11Merge tag 'md/3.16' of git://neil.brown.name/mdLinus Torvalds1-35/+123
2014-06-10raid5: speedup sync_request processingEivind Sarto1-1/+1
2014-06-05md/raid5: deadlock between retry_aligned_read with barrier iohui jiao1-1/+1
2014-05-29raid5: add an option to avoid copy data from bio to stripe cacheShaohua Li1-21/+98
2014-05-29raid5: avoid release list until last reference of the stripeEivind Sarto1-0/+5
2014-05-29md/raid56: Don't perform reads to support writes until stripe is ready.NeilBrown1-12/+18
2014-04-18arch: Mass conversion of smp_mb__*()Peter Zijlstra1-1/+1
2014-04-17raid5: fix a race of stripe count checkShaohua Li1-2/+1
2014-04-09raid5: get_active_stripe avoids device_lockShaohua Li1-7/+2
2014-04-09raid5: make_request does less prepare waitShaohua Li1-5/+14
2014-02-13md/raid5: Fix CPU hotplug callback registrationOleg Nesterov1-46/+44
2014-01-30Merge branch 'for-3.14/drivers' of git://git.kernel.dk/linux-blockLinus Torvalds1-0/+1
2014-01-30Merge branch 'for-3.14/core' of git://git.kernel.dk/linux-blockLinus Torvalds1-41/+43
2014-01-22md/raid5: close recently introduced race in stripe_head management.NeilBrown1-4/+4
2014-01-16md/raid5: fix long-standing problem with bitmap handling on write failure.NeilBrown1-0/+1
2014-01-14md/raid5: fix a recently broken BUG_ON().NeilBrown1-1/+2
2014-01-14md/raid5: Fix possible confusion when multiple write errors occur.NeilBrown1-2/+2
2014-01-09bcache/md: Use raid stripe sizeKent Overstreet1-0/+1
2013-12-31Merge tag 'v3.13-rc6' into for-3.14/coreJens Axboe1-8/+5
2013-11-28md/raid5: fix newly-broken locking in get_active_stripe.NeilBrown1-7/+4
2013-11-28md/raid5: fix new memory-reference bug in alloc_thread_groups.NeilBrown1-1/+1
2013-11-24block: Convert bio_for_each_segment() to bvec_iterKent Overstreet1-6/+6
2013-11-24block: Abstract out bvec iteratorKent Overstreet1-35/+37
2013-11-21Merge tag 'md/3.13' of git://neil.brown.name/mdLinus Torvalds1-101/+319
2013-11-19md/raid5: Use conf->device_lock protect changing of multi-thread resources.majianpeng1-24/+39
2013-11-19md/raid5: Before freeing old multi-thread worker, it should flush them.majianpeng1-0/+3
2013-11-19md/raid5: For stripe with R5_ReadNoMerge, we replace REQ_FLUSH with REQ_NOMERGE.majianpeng1-1/+1
2013-11-19md: use MD_RECOVERY_INTR instead of kthread_should_stop in resync thread.NeilBrown1-4/+15
2013-11-19raid5: Retry R5_ReadNoMerge flag when hit a read error.Bian Yu1-0/+3