FYI, we noticed the below changes on
git://internal_merge_and_test_tree revert-e594571852b03a5503ba48edd718462e99bf238e-e594571852b03a5503ba48edd718462e99bf238e
commit e594571852b03a5503ba48edd718462e99bf238e ("direct-io: Fix negative return from dio read beyond eof")
2015-11-22 15:56:21 export TEST_DIR=/fs/sda5
2015-11-22 15:56:21 export TEST_DEV=/dev/sda5
2015-11-22 15:56:21 export FSTYP=btrfs
2015-11-22 15:56:21 export SCRATCH_MNT=/fs/scratch
2015-11-22 15:56:21 mkdir /fs/scratch -p
2015-11-22 15:56:21 export SCRATCH_DEV_POOL="/dev/sda6 /dev/sda7 /dev/sda8"
2015-11-22 15:56:21 ./check generic/027 generic/074 generic/076 generic/083 generic/091 generic/094 generic/114 generic/133 generic/208 generic/225 generic/256 generic/269 generic/299 generic/300 generic/311 generic/323
FSTYP -- btrfs
PLATFORM -- Linux/x86_64 snb-drag 4.4.0-rc1-00001-ge594571
MKFS_OPTIONS -- /dev/sda6
MOUNT_OPTIONS -- /dev/sda6 /fs/scratch
generic/027 793s
generic/074 96s
generic/076 9s
generic/083 59s
generic/091 22s
generic/094 13s
generic/114 8s
generic/133 56s
generic/208 201s
generic/225 6s
generic/256 98s
generic/269 41s
generic/299 [failed, exit status 1] - output mismatch (see /lkp/benchmarks/xfstests/results//generic/299.out.bad)
--- tests/generic/299.out 2015-11-12 16:56:51.000000000 +0800
+++ /lkp/benchmarks/xfstests/results//generic/299.out.bad 2015-11-22 16:19:50.892344014 +0800
@@ -3,3 +3,5 @@
Run fio with random aio-dio pattern
Start fallocate/truncate loop
+failed: '/usr/bin/fio /tmp/425.fio'
+(see /lkp/benchmarks/xfstests/results//generic/299.full for details)
...
(Run 'diff -u tests/generic/299.out /lkp/benchmarks/xfstests/results//generic/299.out.bad' to see the entire diff)
generic/300 17s
generic/311 114s
generic/323 120s
Ran: generic/027 generic/074 generic/076 generic/083 generic/091 generic/094 generic/114 generic/133 generic/208 generic/225 generic/256 generic/269 generic/299 generic/300 generic/311 generic/323
Failures: generic/299
Failed 1 of 16 tests
To reproduce:
git clone git://git.kernel.org/pub/scm/linux/kernel/git/wfg/lkp-tests.git
cd lkp-tests
bin/lkp install job.yaml # job file is attached in this email
bin/lkp run job.yaml
Thanks,
Ying Huang
Hello,
On Mon 23-11-15 16:29:25, kernel test robot wrote:
> FYI, we noticed the below changes on
>
> git://internal_merge_and_test_tree revert-e594571852b03a5503ba48edd718462e99bf238e-e594571852b03a5503ba48edd718462e99bf238e
> commit e594571852b03a5503ba48edd718462e99bf238e ("direct-io: Fix negative return from dio read beyond eof")
Do you have the file:
/lkp/benchmarks/xfstests/results//generic/299.full
produced by the failed test? Thanks!
Honza
--
Jan Kara <[email protected]>
SUSE Labs, CR
Jan Kara <[email protected]> writes:
> Hello,
>
> On Mon 23-11-15 16:29:25, kernel test robot wrote:
>> FYI, we noticed the below changes on
>>
>> git://internal_merge_and_test_tree
>> revert-e594571852b03a5503ba48edd718462e99bf238e-e594571852b03a5503ba48edd718462e99bf238e
>> commit e594571852b03a5503ba48edd718462e99bf238e ("direct-io: Fix negative return from dio read beyond eof")
>
> Do you have the file:
>
> /lkp/benchmarks/xfstests/results//generic/299.full
>
> produced by the failed test? Thanks!
Sorry, we don't collect that data during regular regression test. Is
that needed all the time? We will change our test scripts to collect
that data. But that need some time.
Best Regards,
Huang, Ying
Jan Kara <[email protected]> writes:
> Hello,
>
> On Mon 23-11-15 16:29:25, kernel test robot wrote:
>> FYI, we noticed the below changes on
>>
>> git://internal_merge_and_test_tree
>> revert-e594571852b03a5503ba48edd718462e99bf238e-e594571852b03a5503ba48edd718462e99bf238e
>> commit e594571852b03a5503ba48edd718462e99bf238e ("direct-io: Fix negative return from dio read beyond eof")
>
> Do you have the file:
>
> /lkp/benchmarks/xfstests/results//generic/299.full
>
> produced by the failed test? Thanks!
I found we have this feature in our test scripts already, just need to
fix a bug in script, but that is just done yesterday! Please check the
299.full attached.
Best Regards,
Huang, Ying
On Tue 24-11-15 12:45:54, Huang, Ying wrote:
> Jan Kara <[email protected]> writes:
>
> > Hello,
> >
> > On Mon 23-11-15 16:29:25, kernel test robot wrote:
> >> FYI, we noticed the below changes on
> >>
> >> git://internal_merge_and_test_tree
> >> revert-e594571852b03a5503ba48edd718462e99bf238e-e594571852b03a5503ba48edd718462e99bf238e
> >> commit e594571852b03a5503ba48edd718462e99bf238e ("direct-io: Fix negative return from dio read beyond eof")
> >
> > Do you have the file:
> >
> > /lkp/benchmarks/xfstests/results//generic/299.full
> >
> > produced by the failed test? Thanks!
>
> I found we have this feature in our test scripts already, just need to
> fix a bug in script, but that is just done yesterday! Please check the
> 299.full attached.
Thanks for the file. The cause of failure is ENOMEM error aio-dio-verifier
got - so it is very likely unrelated to my patch:
fio: io_u.c:1315: __get_io_u: Assertion `io_u->flags & IO_U_F_FREE' failed.
fio: pid=607, got signal=6
fio: io_u error on file /fs/scratch/aio-dio-verifier: Cannot allocate memory
read offset=8446353408, buflen=4096
fio: io_u error on file /fs/scratch/aio-dio-verifier: Cannot allocate memory
read offset=71839776768, buflen=4096
aio-dio-verifier: (groupid=0, jobs=1): err=12 (file:io_u.c:1548, func=io_u
error, error=Cannot allocate memory): pid=605: Tue Nov 24 11:18:39 2015
cpu : usr=0.00%, sys=0.10%, ctx=42, majf=0, minf=14
IO depths : 1=0.8%, 2=1.6%, 4=3.1%, 8=6.2%, 16=12.5%, 32=25.0%, >=64=50.8%
submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
issued : total=r=64/w=64/d=0, short=r=0/w=0/d=0
errors : total=0, first_error=0/<Success>
latency : target=0, window=0, percentile=100.00%, depth=128
If you have time, you can check why ENOMEM actually happens in your test
setup...
Honza
--
Jan Kara <[email protected]>
SUSE Labs, CR