Skip to contentSkip to conversation

e52ddd946a: xfstests generic/211 fail

aio Fengguang Wu 2 participants 5 messages Mar 23, 2014 VFS

lore
Share & tools
lore
#0 lore raw

Hi Benjamin,

We noticed xfstests fail and the first bad commit is

commit e52ddd946ab1def55c8282c8b3d0e80403abae12

Author:     Benjamin LaHaise <[email protected]>
AuthorDate: Fri Mar 21 14:26:43 2014 -0400
    aio: ensure access to ctx->ring_pages is correctly serialised

# good: [dcb99fd9b08cfe1afe426af4d8d3cbc429190f15] Linux 3.14-rc7
git bisect good dcb99fd9b08cfe1afe426af4d8d3cbc429190f15
# bad: [61fabf948117a9f91e679323b49e6a0cf56d25a6] 0day head guard for 'devel-hourly-2014032214'
git bisect bad 61fabf948117a9f91e679323b49e6a0cf56d25a6
# good: [56ba91913d9f2c108be0e33eeb4a473863d91600] Merge 'regulator/topic/bcm590xx' into devel-hourly-2014032214 git bisect good 56ba91913d9f2c108be0e33eeb4a473863d91600
# good: [a3b8ced9d666d7e4da8e4867d58767e8e5e0a7c9] Merge 'kees/nak_fw-relative' into devel-hourly-2014032214 git bisect good a3b8ced9d666d7e4da8e4867d58767e8e5e0a7c9
# bad: [7a74b3c608ebb48fd99791f570c24387104dd425] Merge 'kees/nak_recv-leak' into devel-hourly-2014032214
git bisect bad 7a74b3c608ebb48fd99791f570c24387104dd425
# good: [4b518d474c67c909d82f4ed6242dee38a681daf8] Merge 'ath6kl/ath-next-test' into devel-hourly-2014032214 git bisect good 4b518d474c67c909d82f4ed6242dee38a681daf8
# good: [37e32762b783f435265c1a8a86dd074c81ad87f2] Merge 'spi/for-next' into devel-hourly-2014032214
git bisect good 37e32762b783f435265c1a8a86dd074c81ad87f2
# bad: [655a90c342c5878de4a3fef60b1a7a0cd18bc08c] Merge 'aio-next/master' into devel-hourly-2014032214
git bisect bad 655a90c342c5878de4a3fef60b1a7a0cd18bc08c
# good: [v3.14-rc6] Linux 3.14-rc6
git bisect good v3.14-rc6
# bad: [e52ddd946ab1def55c8282c8b3d0e80403abae12] aio: ensure access to ctx->ring_pages is correctly serialised git bisect bad e52ddd946ab1def55c8282c8b3d0e80403abae12
# good: [v3.14-rc6] Linux 3.14-rc6
git bisect good v3.14-rc6
# bad: [e52ddd946ab1def55c8282c8b3d0e80403abae12] aio: ensure access to ctx->ring_pages is correctly serialised git bisect bad e52ddd946ab1def55c8282c8b3d0e80403abae12
# good: [e6a4b6f5eaa8478b7a0b9a17e40c51463631db1a] Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
git bisect good e6a4b6f5eaa8478b7a0b9a17e40c51463631db1a
# good: [8712a00514e50aafa7c9bf5cd3955fa60758e53b] Merge branch 'akpm' (patches from Andrew Morton)
git bisect good 8712a00514e50aafa7c9bf5cd3955fa60758e53b

# first bad commit: [e52ddd946ab1def55c8282c8b3d0e80403abae12] aio: ensure access to ctx->ring_pages is correctly serialised

Thanks,
Fengguang

#1 lore raw
Quoted 3 lines · Fengguang Wu

Hi Benjamin,

We noticed xfstests fail and the first bad commit is

What is the failure? This is awfully vague for a bug report.

-ben

Signature · 1 line
"Thought is the essence of where you are now."
#2 lore raw
Quoted 6 lines · Benjamin LaHaise
Quoted 3 lines · Fengguang Wu

Hi Benjamin,

We noticed xfstests fail and the first bad commit is

What is the failure? This is awfully vague for a bug report.

In the email title: the generic/211 case of xfstests.

Here are the rough steps to reproduce the problem in KVM:

mkfs -t ext4 -q /dev/vda mkfs -t ext4 -q /dev/vdd mount -t ext4 /dev/vda /fs/vda mount -t ext4 /dev/vdd /fs/vdd export TEST_DIR=/fs/vda export TEST_DEV=/dev/vda export SCRATCH_DEV=/dev/vdd export SCRATCH_MNT=/fs/scratch mkdir /fs/scratch -p ./check generic/211

Thanks,
Fengguang

#3 lore raw
Quoted 12 lines · Fengguang Wu
Quoted 6 lines · Benjamin LaHaise
Quoted 3 lines · Fengguang Wu

Hi Benjamin,

We noticed xfstests fail and the first bad commit is

What is the failure? This is awfully vague for a bug report.

In the email title: the generic/211 case of xfstests.

Here are the rough steps to reproduce the problem in KVM:

Quoted 10 lines

mkfs -t ext4 -q /dev/vda mkfs -t ext4 -q /dev/vdd mount -t ext4 /dev/vda /fs/vda mount -t ext4 /dev/vdd /fs/vdd export TEST_DIR=/fs/vda export TEST_DEV=/dev/vda export SCRATCH_DEV=/dev/vdd export SCRATCH_MNT=/fs/scratch mkdir /fs/scratch -p ./check generic/211

generic/211 [failed, exit status 1] - output mismatch (see /lkp/benchmarks/xfstests/results//generic/211.out.bad)

    --- tests/generic/211.out   2013-09-16 21:19:49.000000000 +0800
    +++ /lkp/benchmarks/xfstests/results//generic/211.out.bad   2014-03-23 02:55:53.390625000 +0800
    @@ -1,2 +1,2 @@
     QA output created by 211
    -aio-free-ring-with-bogus-nr-pages: Success!
    +aio-free-ring-with-bogus-nr-pages: Error: io_setup returned -11, expected -ENOMEM
     ...
     (Run 'diff -u tests/generic/211.out /lkp/benchmarks/xfstests/results//generic/211.out.bad' to see the entire diff)
#4 lore raw
Quoted 4 lines · Fengguang Wu

@@ -1,2 +1,2 @@
QA output created by 211
-aio-free-ring-with-bogus-nr-pages: Success! +aio-free-ring-with-bogus-nr-pages: Error: io_setup returned -11, expected -ENOMEM

This isn't anything serious, as it is just a change in the error code being reported. I'll look into this on Monday.

-ben

Signature · 1 line
"Thought is the essence of where you are now."