From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
stable@vger.kernel.org, Greg Thelen <gthelen@google.com>,
Michal Hocko <mhocko@suse.cz>,
"Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>,
Johannes Weiner <hannes@cmpxchg.org>,
Andrew Morton <akpm@linux-foundation.org>,
Linus Torvalds <torvalds@linux-foundation.org>
Subject: [ 33/40] memcg: fix multiple large threshold notifications
Date: Tue, 24 Sep 2013 17:12:05 -0700 [thread overview]
Message-ID: <20130925001045.542629293@linuxfoundation.org> (raw)
In-Reply-To: <20130925001041.939335518@linuxfoundation.org>
3.4-stable review patch. If anyone has any objections, please let me know.
------------------
From: Greg Thelen <gthelen@google.com>
commit 2bff24a3707093c435ab3241c47dcdb5f16e432b upstream.
A memory cgroup with (1) multiple threshold notifications and (2) at least
one threshold >=2G was not reliable. Specifically the notifications would
either not fire or would not fire in the proper order.
The __mem_cgroup_threshold() signaling logic depends on keeping 64 bit
thresholds in sorted order. mem_cgroup_usage_register_event() sorts them
with compare_thresholds(), which returns the difference of two 64 bit
thresholds as an int. If the difference is positive but has bit[31] set,
then sort() treats the difference as negative and breaks sort order.
This fix compares the two arbitrary 64 bit thresholds returning the
classic -1, 0, 1 result.
The test below sets two notifications (at 0x1000 and 0x81001000):
cd /sys/fs/cgroup/memory
mkdir x
for x in 4096 2164264960; do
cgroup_event_listener x/memory.usage_in_bytes $x | sed "s/^/$x listener:/" &
done
echo $$ > x/cgroup.procs
anon_leaker 500M
v3.11-rc7 fails to signal the 4096 event listener:
Leaking...
Done leaking pages.
Patched v3.11-rc7 properly notifies:
Leaking...
4096 listener:2013:8:31:14:13:36
Done leaking pages.
The fixed bug is old. It appears to date back to the introduction of
memcg threshold notifications in v2.6.34-rc1-116-g2e72b6347c94 "memcg:
implement memory thresholds"
Signed-off-by: Greg Thelen <gthelen@google.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
mm/memcontrol.c | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -4349,7 +4349,13 @@ static int compare_thresholds(const void
const struct mem_cgroup_threshold *_a = a;
const struct mem_cgroup_threshold *_b = b;
- return _a->threshold - _b->threshold;
+ if (_a->threshold > _b->threshold)
+ return 1;
+
+ if (_a->threshold < _b->threshold)
+ return -1;
+
+ return 0;
}
static int mem_cgroup_oom_notify_cb(struct mem_cgroup *memcg)
next prev parent reply other threads:[~2013-09-25 0:15 UTC|newest]
Thread overview: 44+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-09-25 0:11 [ 00/40] 3.4.63-stable review Greg Kroah-Hartman
2013-09-25 0:11 ` [ 01/40] SCSI: sd: Fix potential out-of-bounds access Greg Kroah-Hartman
2013-09-25 0:11 ` [ 02/40] crypto: api - Fix race condition in larval lookup Greg Kroah-Hartman
2013-09-25 0:11 ` [ 03/40] powerpc: Handle unaligned ldbrx/stdbrx Greg Kroah-Hartman
2013-09-25 0:11 ` [ 04/40] xen-gnt: prevent adding duplicate gnt callbacks Greg Kroah-Hartman
2013-09-25 0:11 ` [ 05/40] ARM: PCI: versatile: Fix SMAP register offsets Greg Kroah-Hartman
2013-09-25 0:11 ` [ 06/40] xhci-plat: Dont enable legacy PCI interrupts Greg Kroah-Hartman
2013-09-25 0:11 ` [ 07/40] usb: xhci: Disable runtime PM suspend for quirky controllers Greg Kroah-Hartman
2013-09-25 0:11 ` [ 08/40] cifs: ensure that srv_mutex is held when dealing with ssocket pointer Greg Kroah-Hartman
2013-09-25 0:11 ` [ 09/40] staging: comedi: dt282x: dt282x_ai_insn_read() always fails Greg Kroah-Hartman
2013-09-25 0:11 ` [ 10/40] USB: mos7720: use GFP_ATOMIC under spinlock Greg Kroah-Hartman
2013-09-25 0:11 ` [ 11/40] USB: mos7720: fix big-endian control requests Greg Kroah-Hartman
2013-09-25 0:11 ` [ 12/40] usb: ehci-mxc: check for pdata before dereferencing Greg Kroah-Hartman
2013-09-25 0:11 ` [ 13/40] USB: cdc-wdm: fix race between interrupt handler and tasklet Greg Kroah-Hartman
2013-09-25 0:11 ` [ 14/40] usb: config->desc.bLength may not exceed amount of data returned by the device Greg Kroah-Hartman
2013-09-25 0:11 ` [ 15/40] rculist: list_first_or_null_rcu() should use list_entry_rcu() Greg Kroah-Hartman
2013-09-25 0:11 ` [ 16/40] ASoC: wm8960: Fix PLL register writes Greg Kroah-Hartman
2013-09-25 0:11 ` [ 17/40] ALSA: hda - Add Toshiba Satellite C870 to MSI blacklist Greg Kroah-Hartman
2013-09-25 0:11 ` [ 18/40] brcmsmac: Fix WARNING caused by lack of calls to dma_mapping_error() Greg Kroah-Hartman
2013-09-25 0:11 ` [ 19/40] ath9k: always clear ps filter bit on new assoc Greg Kroah-Hartman
2013-09-25 0:11 ` [ 20/40] ath9k: fix rx descriptor related race condition Greg Kroah-Hartman
2013-09-25 0:11 ` [ 21/40] ath9k: avoid accessing MRC registers on single-chain devices Greg Kroah-Hartman
2013-09-25 0:11 ` [ 22/40] HID: pantherlord: validate output report details Greg Kroah-Hartman
2013-09-25 0:11 ` [ 23/40] HID: Fix Speedlink VAD Cezanne support for some devices Greg Kroah-Hartman
2013-09-25 0:11 ` [ 24/40] HID: validate HID report id size Greg Kroah-Hartman
2013-09-25 0:11 ` [ 25/40] HID: ntrig: validate feature report details Greg Kroah-Hartman
2013-09-25 0:11 ` [ 26/40] HID: battery: dont do DMA from stack Greg Kroah-Hartman
2013-09-25 0:11 ` [ 27/40] HID: check for NULL field when setting values Greg Kroah-Hartman
2013-09-25 0:12 ` [ 28/40] HID: usbhid: quirk for N-Trig DuoSense Touch Screen Greg Kroah-Hartman
2013-09-25 0:12 ` [ 29/40] media: v4l2: added missing mutex.h include to v4l2-ctrls.h Greg Kroah-Hartman
2013-09-25 0:12 ` [ 30/40] MIPS: ath79: Fix ar933x watchdog clock Greg Kroah-Hartman
2013-09-25 0:12 ` [ 31/40] intel-iommu: Fix leaks in pagetable freeing Greg Kroah-Hartman
2013-09-25 0:12 ` [ 32/40] ocfs2: fix the end cluster offset of FIEMAP Greg Kroah-Hartman
2013-09-25 0:12 ` Greg Kroah-Hartman [this message]
2013-09-25 0:12 ` [ 34/40] mm/huge_memory.c: fix potential NULL pointer dereference Greg Kroah-Hartman
2013-09-25 0:12 ` [ 35/40] isofs: Refuse RW mount of the filesystem instead of making it RO Greg Kroah-Hartman
2013-09-25 0:12 ` [ 36/40] drm/edid: add quirk for Medion MD30217PG Greg Kroah-Hartman
2013-09-25 0:12 ` [ 37/40] mmc: tmio_mmc_dma: fix PIO fallback on SDHI Greg Kroah-Hartman
2013-09-25 0:12 ` [ 38/40] of: Fix missing memory initialization on FDT unflattening Greg Kroah-Hartman
2013-09-25 0:12 ` [ 39/40] fuse: postpone end_page_writeback() in fuse_writepage_locked() Greg Kroah-Hartman
2013-09-25 0:12 ` [ 40/40] fuse: invalidate inode attributes on xattr modification Greg Kroah-Hartman
2013-09-25 4:35 ` [ 00/40] 3.4.63-stable review Guenter Roeck
2013-09-26 1:09 ` Greg Kroah-Hartman
2013-09-26 2:24 ` 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=20130925001045.542629293@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=akpm@linux-foundation.org \
--cc=gthelen@google.com \
--cc=hannes@cmpxchg.org \
--cc=kirill.shutemov@linux.intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mhocko@suse.cz \
--cc=stable@vger.kernel.org \
--cc=torvalds@linux-foundation.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