ACK: [SRU][X][PULL v2] bcache: fix hung task timeout in bch_bucket_alloc()

[Kleber Souza]

On 8/7/19 3:18 PM, Andrea Righi wrote:
> BugLink: https://bugs.launchpad.net/bugs/1784665
> 
> [Impact]
> 
> bcache_allocator can call the following:
> 
>  bch_allocator_thread()
>   -> bch_prio_write()
>      -> bch_bucket_alloc()
>         -> wait on &ca->set->bucket_wait
> 
> But the wake up event on bucket_wait is supposed to come from
> bch_allocator_thread() itself causing a deadlock.
> 
> [Test Case]
> 
> This is a simple script that can easily trigger the deadlock condition:
> https://launchpadlibrarian.net/381282009/bcache-basic-repro.sh
> 
> A better test case has been also provided in LP: #1796292:
> https://bugs.launchpad.net/curtin/+bug/1796292/+attachment/5280353/+files/curtin-nvme.sh
> 
> [Fix]
> 
> Fix by making the call to bch_prio_write() non-blocking, so that
> bch_allocator_thread() never waits on itself. Moreover, make sure to
> wake up the garbage collector thread when bch_prio_write() is failing to
> allocate buckets to increase the chance of freeing up more buckets.
> 
> In addition to that it would be safer to also import other upstream
> bcache fixes (all clean cherry picks):
> 
> ecb37ce9baac653cc09e2b631393dde3df82979f bcache: Move couple of functions to sysfs.c
> 20d3a518713e394efa5a899c84574b4b79ec5098 bcache: Reduce the number of sparse complaints about lock imbalances
> 42361469ae84c851e40cb1f94c8c9a14cdd94039 bcache: Suppress more warnings about set-but-not-used variables
> f0d3814090ac77de94c42b7124c37ece23629197 bcache: Remove an unused variable
> 47344e330eabc1515cbe6061eb337100a3ab6d37 bcache: Fix kernel-doc warnings
> 9dfbdec7b7fea1ff1b7b5d5d12980dbc7dca46c7 bcache: Annotate switch fall-through
> 4a4e443835a43a79113cc237c472c0d268eb1e1c bcache: Add __printf annotation to __bch_check_keys()
> fd01991d5c20098c5c1ffc4dca6c821cc60a2f74 bcache: Fix indentation
> ca71df31661a0518ed58a1a59cf1993962153ebb bcache: fix using of loop variable in memory shrink
> f3641c3abd1da978ee969b0203b71b86ec1bfa93 bcache: fix error return value in memory shrink
> 688892b3bc05e25da94866e32210e5f503f16f69 bcache: fix incorrect sysfs output value of strip size
> c4dc2497d50d9c6fb16aa0d07b6a14f3b2adb1e0 bcache: fix high CPU occupancy during journal
> a728eacbbdd229d1d903e46261c57d5206f87a4a bcache: add journal statistic
> eb8cbb6df38f6e5124a3d5f1f8a3dbf519537c60 bcache: improve bcache_reboot()
> 
> [Regression Potential]
> 
> The upstream fixes are all clean cherry picks from stable (most of them
> are small cleanups), so regression potential is minimal.
> 
> The only special patch is "UBUNTU: SAUCE: bcache: fix deadlock in
> bcache_allocator" that is addressing the main deadlock bug. Even if this
> patch is not yet applied upstream (posted to the LKML), it seems to
> reliably fix/prevent the specific deadlock problem reported in this bug,
> so it should be considered safe to apply it as it is for now, to prevent
> potential hung task timeout conditions.
> 
> Changes in v2:
>  - fix potential buckets leak in "UBUNTU: SAUCE: bcache: fix deadlock in
>    bcache_allocator"
> 
> ----------------------------------------------------------------
> The following changes since commit 4a916e0ee66c44ecdf473dc96a5ff2e4ad4ae196:
> 
>   floppy: fix out-of-bounds read in copy_buffer (2019-08-07 00:54:59 -0400)
> 
> are available in the Git repository at:
> 
>   git://git.launchpad.net/~arighi/+git/xenial-linux bcache-fix-v2
> 
> for you to fetch changes up to 4f40d1b13ac6288a8a842c1fa6fcd49ccc8de0b9:
> 
>   UBUNTU: SAUCE: bcache: fix deadlock in bcache_allocator (2019-08-07 14:19:12 +0200)
> 
> ----------------------------------------------------------------
> Andrea Righi (1):
>       UBUNTU: SAUCE: bcache: fix deadlock in bcache_allocator
> 
> Andy Shevchenko (1):
>       bcache: Move couple of functions to sysfs.c
> 
> Bart Van Assche (7):
>       bcache: Fix indentation
>       bcache: Add __printf annotation to __bch_check_keys()
>       bcache: Annotate switch fall-through
>       bcache: Fix kernel-doc warnings
>       bcache: Remove an unused variable
>       bcache: Suppress more warnings about set-but-not-used variables
>       bcache: Reduce the number of sparse complaints about lock imbalances
> 
> Coly Li (1):
>       bcache: improve bcache_reboot()
> 
> Tang Junhui (5):
>       bcache: add journal statistic
>       bcache: fix high CPU occupancy during journal
>       bcache: fix incorrect sysfs output value of strip size
>       bcache: fix error return value in memory shrink
>       bcache: fix using of loop variable in memory shrink
> 
>  drivers/md/bcache/alloc.c     |  5 +++-
>  drivers/md/bcache/bcache.h    |  8 +++++-
>  drivers/md/bcache/bset.c      |  4 +--
>  drivers/md/bcache/bset.h      |  5 ++--
>  drivers/md/bcache/btree.c     | 15 ++++++-----
>  drivers/md/bcache/closure.c   |  8 +++---
>  drivers/md/bcache/extents.c   |  2 --
>  drivers/md/bcache/journal.c   | 56 ++++++++++++++++++++++++++++------------
>  drivers/md/bcache/request.c   |  1 +
>  drivers/md/bcache/super.c     | 39 ++++++++++++++++++++++------
>  drivers/md/bcache/sysfs.c     | 52 ++++++++++++++++++++++++++++++++++++-
>  drivers/md/bcache/util.c      | 60 +++++++++++--------------------------------
>  drivers/md/bcache/util.h      |  7 ++---
>  drivers/md/bcache/writeback.h |  2 +-
>  14 files changed, 170 insertions(+), 94 deletions(-)
> 

Acked-by: Kleber Sacilotto de Souza <kleber.souza at canonical.com>