2015-12-28 01:39:52

by kernel test robot

[permalink] [raw]

Subject: [lkp] [f2fs] 75cd4e098d: xfstests.generic.042.fail

---

FYI, we noticed the below changes on

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
commit 75cd4e098d178433436abce08146a21647bb4585 ("f2fs: support FALLOC_FL_ZERO_RANGE")

We found after your commit, the xfstests test case 009/042 failed.

=========================================================================================
compiler/disk/fs/kconfig/rootfs/tbox_group/test/testcase:
gcc-4.9/4HDD/f2fs/x86_64-rhel/debian-x86_64-2015-02-07.cgz/vm-kbuild-4G/generic-quick/xfstests

commit:
b4ace33703243fed56f8bfc80a001533acb9decb
75cd4e098d178433436abce08146a21647bb4585

b4ace33703243fed 75cd4e098d178433436abce081
---------------- --------------------------
fail:runs %reproduction fail:runs
| | |
:18 94% 17:18 xfstests.generic.009.fail
:18 94% 17:18 xfstests.generic.042.fail


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

---

Attachments:

(No filename) (1.15 kB)
config-4.1.0-rc2-00138-g75cd4e0 (134.00 kB)
dmesg.xz (27.69 kB)
job.yaml (2.92 kB)
reproduce.sh (2.44 kB)
Download all attachments

2015-12-31 10:17:25

by Chao Yu

[permalink] [raw]

Subject: RE: [lkp] [f2fs] 75cd4e098d: xfstests.generic.042.fail

---

Hi Ying,

> -----Original Message-----
> From: kernel test robot [mailto:[email protected]]
> Sent: Monday, December 28, 2015 9:40 AM
> To: Chao Yu
> Cc: [email protected]; LKML; Jaegeuk Kim
> Subject: [lkp] [f2fs] 75cd4e098d: xfstests.generic.042.fail
>
> FYI, we noticed the below changes on
>
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
> commit 75cd4e098d178433436abce08146a21647bb4585 ("f2fs: support FALLOC_FL_ZERO_RANGE")
>
> We found after your commit, the xfstests test case 009/042 failed.

Thank you for the report, 042 is a known issue, IIRC, the reason is f2fs
will flush inode page with last i_size in ioctl(F2FS_IOC_SHUTDOWN, 0x01),
but 042 expects an zero-sized inode. So I supposed it happened due to the
different semantics of metadata flushing between f2fs and xfs.

For 009, I haven't saw it failed for long time, did you test with last f2fs?

Thanks,

>
> =========================================================================================
> compiler/disk/fs/kconfig/rootfs/tbox_group/test/testcase:
>
> gcc-4.9/4HDD/f2fs/x86_64-rhel/debian-x86_64-2015-02-07.cgz/vm-kbuild-4G/generic-quick/xfst
> ests
>
> commit:
> b4ace33703243fed56f8bfc80a001533acb9decb
> 75cd4e098d178433436abce08146a21647bb4585
>
> b4ace33703243fed 75cd4e098d178433436abce081
> ---------------- --------------------------
> fail:runs %reproduction fail:runs
> | | |
> :18 94% 17:18 xfstests.generic.009.fail
> :18 94% 17:18 xfstests.generic.042.fail
>
>
> 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