From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
stable@vger.kernel.org, David Rientjes <rientjes@google.com>,
Guenter Roeck <linux@roeck-us.net>,
Douglas Anderson <dianders@chromium.org>,
Mike Snitzer <snitzer@redhat.com>
Subject: [PATCH 4.4 39/47] dm bufio: avoid sleeping while holding the dm_bufio lock
Date: Tue, 10 Jul 2018 20:25:03 +0200 [thread overview]
Message-ID: <20180710182338.807810204@linuxfoundation.org> (raw)
In-Reply-To: <20180710182337.047502999@linuxfoundation.org>
4.4-stable review patch. If anyone has any objections, please let me know.
------------------
From: Douglas Anderson <dianders@chromium.org>
commit 9ea61cac0b1ad0c09022f39fd97e9b99a2cfc2dc upstream.
We've seen in-field reports showing _lots_ (18 in one case, 41 in
another) of tasks all sitting there blocked on:
mutex_lock+0x4c/0x68
dm_bufio_shrink_count+0x38/0x78
shrink_slab.part.54.constprop.65+0x100/0x464
shrink_zone+0xa8/0x198
In the two cases analyzed, we see one task that looks like this:
Workqueue: kverityd verity_prefetch_io
__switch_to+0x9c/0xa8
__schedule+0x440/0x6d8
schedule+0x94/0xb4
schedule_timeout+0x204/0x27c
schedule_timeout_uninterruptible+0x44/0x50
wait_iff_congested+0x9c/0x1f0
shrink_inactive_list+0x3a0/0x4cc
shrink_lruvec+0x418/0x5cc
shrink_zone+0x88/0x198
try_to_free_pages+0x51c/0x588
__alloc_pages_nodemask+0x648/0xa88
__get_free_pages+0x34/0x7c
alloc_buffer+0xa4/0x144
__bufio_new+0x84/0x278
dm_bufio_prefetch+0x9c/0x154
verity_prefetch_io+0xe8/0x10c
process_one_work+0x240/0x424
worker_thread+0x2fc/0x424
kthread+0x10c/0x114
...and that looks to be the one holding the mutex.
The problem has been reproduced on fairly easily:
0. Be running Chrome OS w/ verity enabled on the root filesystem
1. Pick test patch: http://crosreview.com/412360
2. Install launchBalloons.sh and balloon.arm from
http://crbug.com/468342
...that's just a memory stress test app.
3. On a 4GB rk3399 machine, run
nice ./launchBalloons.sh 4 900 100000
...that tries to eat 4 * 900 MB of memory and keep accessing.
4. Login to the Chrome web browser and restore many tabs
With that, I've seen printouts like:
DOUG: long bufio 90758 ms
...and stack trace always show's we're in dm_bufio_prefetch().
The problem is that we try to allocate memory with GFP_NOIO while
we're holding the dm_bufio lock. Instead we should be using
GFP_NOWAIT. Using GFP_NOIO can cause us to sleep while holding the
lock and that causes the above problems.
The current behavior explained by David Rientjes:
It will still try reclaim initially because __GFP_WAIT (or
__GFP_KSWAPD_RECLAIM) is set by GFP_NOIO. This is the cause of
contention on dm_bufio_lock() that the thread holds. You want to
pass GFP_NOWAIT instead of GFP_NOIO to alloc_buffer() when holding a
mutex that can be contended by a concurrent slab shrinker (if
count_objects didn't use a trylock, this pattern would trivially
deadlock).
This change significantly increases responsiveness of the system while
in this state. It makes a real difference because it unblocks kswapd.
In the bug report analyzed, kswapd was hung:
kswapd0 D ffffffc000204fd8 0 72 2 0x00000000
Call trace:
[<ffffffc000204fd8>] __switch_to+0x9c/0xa8
[<ffffffc00090b794>] __schedule+0x440/0x6d8
[<ffffffc00090bac0>] schedule+0x94/0xb4
[<ffffffc00090be44>] schedule_preempt_disabled+0x28/0x44
[<ffffffc00090d900>] __mutex_lock_slowpath+0x120/0x1ac
[<ffffffc00090d9d8>] mutex_lock+0x4c/0x68
[<ffffffc000708e7c>] dm_bufio_shrink_count+0x38/0x78
[<ffffffc00030b268>] shrink_slab.part.54.constprop.65+0x100/0x464
[<ffffffc00030dbd8>] shrink_zone+0xa8/0x198
[<ffffffc00030e578>] balance_pgdat+0x328/0x508
[<ffffffc00030eb7c>] kswapd+0x424/0x51c
[<ffffffc00023f06c>] kthread+0x10c/0x114
[<ffffffc000203dd0>] ret_from_fork+0x10/0x40
By unblocking kswapd memory pressure should be reduced.
Suggested-by: David Rientjes <rientjes@google.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
drivers/md/dm-bufio.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
--- a/drivers/md/dm-bufio.c
+++ b/drivers/md/dm-bufio.c
@@ -818,7 +818,8 @@ static struct dm_buffer *__alloc_buffer_
* dm-bufio is resistant to allocation failures (it just keeps
* one buffer reserved in cases all the allocations fail).
* So set flags to not try too hard:
- * GFP_NOIO: don't recurse into the I/O layer
+ * GFP_NOWAIT: don't wait; if we need to sleep we'll release our
+ * mutex and wait ourselves.
* __GFP_NORETRY: don't retry and rather return failure
* __GFP_NOMEMALLOC: don't use emergency reserves
* __GFP_NOWARN: don't print a warning in case of failure
@@ -828,7 +829,7 @@ static struct dm_buffer *__alloc_buffer_
*/
while (1) {
if (dm_bufio_cache_size_latch != 1) {
- b = alloc_buffer(c, GFP_NOIO | __GFP_NORETRY | __GFP_NOMEMALLOC | __GFP_NOWARN);
+ b = alloc_buffer(c, GFP_NOWAIT | __GFP_NORETRY | __GFP_NOMEMALLOC | __GFP_NOWARN);
if (b)
return b;
}
next prev parent reply other threads:[~2018-07-10 18:28 UTC|newest]
Thread overview: 56+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-07-10 18:24 [PATCH 4.4 00/47] 4.4.140-stable review Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 01/47] usb: cdc_acm: Add quirk for Uniden UBC125 scanner Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 02/47] USB: serial: cp210x: add CESINEL device ids Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 03/47] USB: serial: cp210x: add Silicon Labs IDs for Windows Update Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 04/47] n_tty: Fix stall at n_tty_receive_char_special() Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 05/47] staging: android: ion: Return an ERR_PTR in ion_map_kernel Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 06/47] n_tty: Access echo_* variables carefully Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 07/47] x86/boot: Fix early command-line parsing when matching at end Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 08/47] ath10k: fix rfc1042 header retrieval in QCA4019 with eth decap mode Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 09/47] i2c: rcar: fix resume by always initializing registers before transfer Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 10/47] ipv4: Fix error return value in fib_convert_metrics() Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 11/47] kprobes/x86: Do not modify singlestep buffer while resuming Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 12/47] nvme-pci: initialize queue memory before interrupts Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 13/47] netfilter: nf_tables: use WARN_ON_ONCE instead of BUG_ON in nft_do_chain() Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 14/47] ARM: dts: imx6q: Use correct SDMA script for SPI5 core Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 15/47] ubi: fastmap: Correctly handle interrupted erasures in EBA Greg Kroah-Hartman
2018-07-26 2:12 ` Ben Hutchings
2018-07-26 6:25 ` Richard Weinberger
2018-07-28 1:28 ` Ben Hutchings
2018-07-28 5:56 ` Richard Weinberger
2018-07-10 18:24 ` [PATCH 4.4 16/47] mm: hugetlb: yield when prepping struct pages Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 17/47] tracing: Fix missing return symbol in function_graph output Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 18/47] scsi: sg: mitigate read/write abuse Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 19/47] s390: Correct register corruption in critical section cleanup Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 20/47] drbd: fix access after free Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 21/47] cifs: Fix infinite loop when using hard mount option Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 22/47] jbd2: dont mark block as modified if the handle is out of credits Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 23/47] ext4: make sure bitmaps and the inode table dont overlap with bg descriptors Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 24/47] ext4: always check block group bounds in ext4_init_block_bitmap() Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 25/47] ext4: only look at the bg_flags field if it is valid Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 26/47] ext4: verify the depth of extent tree in ext4_find_extent() Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 27/47] ext4: include the illegal physical block in the bad map ext4_error msg Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 28/47] ext4: clear i_data in ext4_inode_info when removing inline data Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 29/47] ext4: add more inode number paranoia checks Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 30/47] ext4: add more mount time checks of the superblock Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 31/47] ext4: check superblock mapped prior to committing Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 32/47] HID: i2c-hid: Fix "incomplete report" noise Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 33/47] HID: hiddev: fix potential Spectre v1 Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 34/47] HID: debug: check length before copy_to_user() Greg Kroah-Hartman
2018-07-10 18:24 ` [PATCH 4.4 35/47] x86/mce: Detect local MCEs properly Greg Kroah-Hartman
2018-07-10 18:25 ` [PATCH 4.4 36/47] x86/mce: Fix incorrect "Machine check from unknown source" message Greg Kroah-Hartman
2018-07-10 18:25 ` [PATCH 4.4 37/47] media: cx25840: Use subdev host data for PLL override Greg Kroah-Hartman
2018-07-10 18:25 ` [PATCH 4.4 38/47] mm, page_alloc: do not break __GFP_THISNODE by zonelist reset Greg Kroah-Hartman
2018-07-10 18:25 ` Greg Kroah-Hartman [this message]
2018-07-10 18:25 ` [PATCH 4.4 40/47] dm bufio: drop the lock when doing GFP_NOIO allocation Greg Kroah-Hartman
2018-07-10 18:25 ` [PATCH 4.4 41/47] mtd: rawnand: mxc: set spare area size register explicitly Greg Kroah-Hartman
2018-07-10 18:25 ` [PATCH 4.4 42/47] dm bufio: dont take the lock in dm_bufio_shrink_count Greg Kroah-Hartman
2018-07-10 18:25 ` [PATCH 4.4 43/47] mtd: cfi_cmdset_0002: Change definition naming to retry write operation Greg Kroah-Hartman
2018-07-10 18:25 ` [PATCH 4.4 44/47] mtd: cfi_cmdset_0002: Change erase functions to retry for error Greg Kroah-Hartman
2018-07-10 18:25 ` [PATCH 4.4 45/47] mtd: cfi_cmdset_0002: Change erase functions to check chip good only Greg Kroah-Hartman
2018-07-10 18:25 ` [PATCH 4.4 46/47] netfilter: nf_log: dont hold nf_log_mutex during user access Greg Kroah-Hartman
2018-07-10 18:25 ` [PATCH 4.4 47/47] staging: comedi: quatech_daqp_cs: fix no-op loop daqp_ao_insn_write() Greg Kroah-Hartman
2018-07-10 19:10 ` [PATCH 4.4 00/47] 4.4.140-stable review Nathan Chancellor
2018-07-11 11:21 ` Naresh Kamboju
2018-07-11 13:40 ` Guenter Roeck
2018-07-11 15:10 ` Shuah Khan
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20180710182338.807810204@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=dianders@chromium.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux@roeck-us.net \
--cc=rientjes@google.com \
--cc=snitzer@redhat.com \
--cc=stable@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).