* [LTP] statx04.c:53: FAIL: STATX_ATTR_COMPRESSED flag is not set - Testing on ext2
@ 2018-09-15 9:53 Naresh Kamboju
2018-09-16 16:26 ` Rafael David Tinoco
2018-09-17 8:27 ` Cyril Hrubis
0 siblings, 2 replies; 4+ messages in thread
From: Naresh Kamboju @ 2018-09-15 9:53 UTC (permalink / raw)
To: ltp
statx04 test failed on Hikey (arm64) while testing for ext2 which is
running 4.18 stable kernel and LTP master branch. Same test case got
PASS for ext3 and ext4.
Is this test case need any special prerequirements ?
Test log.
tst_test.c:1124: INFO: Testing on ext2
tst_mkfs.c:90: INFO: Formatting /dev/loop0 with ext2 opts='' extra opts=''
mke2fs 1.43.5 (04-Aug-2017)
tst_test.c:1063: INFO: Timeout per run is 0h 15m 00s
statx04.c:45: PASS: sys_statx(AT_FDCWD, mntpoint/test_dir1, 0, 0, &buf)
statx04.c:53: FAIL: STATX_ATTR_COMPRESSED flag is not set
statx04.c:58: FAIL: STATX_ATTR_APPEND flag is not set
statx04.c:63: FAIL: STATX_ATTR_IMMUTABLE flag is not set
statx04.c:68: FAIL: STATX_ATTR_NODUMP flag is not set
statx04.c:79: PASS: sys_statx(AT_FDCWD, mntpoint/test_dir2, 0, 0, &buf)
statx04.c:86: PASS: STATX_ATTR_COMPRESSED flag is not set
statx04.c:91: PASS: STATX_ATTR_APPEND flag is not set
statx04.c:96: PASS: STATX_ATTR_IMMUTABLE flag is not set
statx04.c:101: PASS: STATX_ATTR_NODUMP flag is not set
Full test log,
https://lkft.validation.linaro.org/scheduler/job/420924#L13367
Please investigate this failure.
P.S. I did not tested on any other architecture and board.
- Naresh
^ permalink raw reply [flat|nested] 4+ messages in thread
* [LTP] statx04.c:53: FAIL: STATX_ATTR_COMPRESSED flag is not set - Testing on ext2
2018-09-15 9:53 [LTP] statx04.c:53: FAIL: STATX_ATTR_COMPRESSED flag is not set - Testing on ext2 Naresh Kamboju
@ 2018-09-16 16:26 ` Rafael David Tinoco
2018-09-17 8:36 ` Cyril Hrubis
2018-09-17 8:27 ` Cyril Hrubis
1 sibling, 1 reply; 4+ messages in thread
From: Rafael David Tinoco @ 2018-09-16 16:26 UTC (permalink / raw)
To: ltp
Naresh,
I could not reproduce this result, after 200 iterations, in KVM ARM64,
nor in Hikey960, using v4.18.8-0-gdfba61ec760e and latest LTP. I have
also compiled the same kernel as this LKFT test (94710cac0e) and
couldn't reproduce this issue. My config file is a merge in between
LKFT hikey config and some Debian options.
I have downloaded and booted my Hikey960 with kernel from:
http://snapshots.linaro.org/openembedded/ltp/rocko/hikey/300/rpb/
And noticed that:
inaddy@hikey:~$ zcat /proc/config.gz | grep EXT
...
CONFIG_EXT2_FS=y
# CONFIG_EXT2_FS_XATTR is not set
...
And just to confirm I tried:
tst_test.c:1124: INFO: Testing on ext2
tst_mkfs.c:90: INFO: Formatting /dev/loop1 with ext2 opts='' extra opts=''
mke2fs 1.44.3 (10-July-2018)
tst_test.c:1063: INFO: Timeout per run is 0h 05m 00s
fsetxattr01.c:186: PASS: fsetxattr(2) failed: EINVAL
fsetxattr01.c:150: CONF: fsetxattr(2) not supported
Then I tried the same xattr04 test and:
tst_test.c:1124: INFO: Testing on ext2
tst_mkfs.c:90: INFO: Formatting /dev/loop1 with ext2 opts='' extra opts=''
mke2fs 1.44.3 (10-July-2018)
tst_test.c:1063: INFO: Timeout per run is 0h 05m 00s
statx04.c:45: PASS: sys_statx(AT_FDCWD, mntpoint/test_dir1, 0, 0, &buf)
statx04.c:53: FAIL: STATX_ATTR_COMPRESSED flag is not set
statx04.c:58: FAIL: STATX_ATTR_APPEND flag is not set
statx04.c:63: FAIL: STATX_ATTR_IMMUTABLE flag is not set
statx04.c:68: FAIL: STATX_ATTR_NODUMP flag is not set
and it failed like yours.
Doing same thing, by hand, I had:
inaddy@hikey:~$ dd if=/dev/zero of=./teste bs=1M count=100
100+0 records in
100+0 records out
104857600 bytes (105 MB, 100 MiB) copied, 1.30856 s, 80.1 MB/s
inaddy@hikey:~$ sudo losetup -P /dev/loop0 /home/inaddy/teste
inaddy@hikey:~$ sudo mkfs.ext2 /dev/loop0
mke2fs 1.44.3 (10-July-2018)
Discarding device blocks: done
Creating filesystem with 102400 1k blocks and 25688 inodes
...
inaddy@hikey:/mnt$ sudo touch file
inaddy@hikey:/mnt$ sudo mkdir directory
inaddy@hikey:/mnt$ sudo chattr +acid ./file
inaddy@hikey:/mnt$ sudo chattr +acid ./directory
inaddy@hikey:/mnt$ sudo lsattr ./file
----iad-c--------- ./file
inaddy@hikey:/mnt$ sudo lsattr ./directory
inaddy@hikey:/mnt$
which indicates that directory attributes did not work.
Checking differences from my build to LKFT build:
$ zcat /proc/config.gz | grep EXT2
CONFIG_EXT2_FS=y
# CONFIG_EXT2_FS_XATTR is not set
$ cat /boot/config-4.18.0 | grep EXT2
# CONFIG_EXT2_FS is not set
CONFIG_EXT4_USE_FOR_EXT2=y
And checking behaviour in an environment with CONFIG_EXT4_USE_FOR_EXT2:
inaddy@statx04bug:~$ cd /mnt
inaddy@statx04bug:/mnt$ sudo touch file
inaddy@statx04bug:/mnt$ sudo touch directory
inaddy@statx04bug:/mnt$ sudo chattr +acid ./file
inaddy@statx04bug:/mnt$ sudo chattr +acid ./directory
inaddy@statx04bug:/mnt$ sudo lsattr file
----iad-c---------- file
inaddy@statx04bug:/mnt$ sudo lsattr directory
----iad-c---------- directory
I suppose this error is because Hikey builds are actually using EXT2
code and not EXT4 as EXT2 compatible code.
Now, the REAL QUESTION, should EXT2 old code support attributes for
directories ?
Is this a regression or a feature that never existed for EXT2 old code ?
^ permalink raw reply [flat|nested] 4+ messages in thread
* [LTP] statx04.c:53: FAIL: STATX_ATTR_COMPRESSED flag is not set - Testing on ext2
2018-09-16 16:26 ` Rafael David Tinoco
@ 2018-09-17 8:36 ` Cyril Hrubis
0 siblings, 0 replies; 4+ messages in thread
From: Cyril Hrubis @ 2018-09-17 8:36 UTC (permalink / raw)
To: ltp
Hi!
> I suppose this error is because Hikey builds are actually using EXT2
> code and not EXT4 as EXT2 compatible code.
I wouldn't say this to be error, different configuration maybe.
> Now, the REAL QUESTION, should EXT2 old code support attributes for
> directories ?
> Is this a regression or a feature that never existed for EXT2 old code ?
I've talked about that to Jan Kara and he told me that he will get this
fixed so I left the test as it is and noted that it fails on ext2 when
using ext2 driver in the commit message.
--
Cyril Hrubis
chrubis@suse.cz
^ permalink raw reply [flat|nested] 4+ messages in thread
* [LTP] statx04.c:53: FAIL: STATX_ATTR_COMPRESSED flag is not set - Testing on ext2
2018-09-15 9:53 [LTP] statx04.c:53: FAIL: STATX_ATTR_COMPRESSED flag is not set - Testing on ext2 Naresh Kamboju
2018-09-16 16:26 ` Rafael David Tinoco
@ 2018-09-17 8:27 ` Cyril Hrubis
1 sibling, 0 replies; 4+ messages in thread
From: Cyril Hrubis @ 2018-09-17 8:27 UTC (permalink / raw)
To: ltp
Hi!
> statx04 test failed on Hikey (arm64) while testing for ext2 which is
> running 4.18 stable kernel and LTP master branch. Same test case got
> PASS for ext3 and ext4.
>
> Is this test case need any special prerequirements ?
That is a confirmed kernel bug that I've been told by kernel dev will be fixed,
it's written in the test commit message as well:
commit 5ffb23e2c051c718814515527275e92a8d3155e0
Author: vaishnavid <vaishnavi.d@zilogic.com>
Date: Mon Sep 3 12:18:28 2018 +0000
syscalls/statx04: Add new test.
This test will check if statx() reports attribute flags by adding
FS_COMPR_FL, FS_APPEND_FL, FS_IMMUTABLE_FL and FS_NODUMP_FL to a
direcotry and checks that the flags are reported correctly and also
tests a directory with no flags.
The test is executed against all filesystems, since most of the kernel
code that implements this functionality is filesystem specific.
* The ntfs3g fuse driver reports wrong errno for unimplemented ioctls()
EINVAL while every other filesystem reports ENOTTY correctly
* The test fails for ext2 filesystem mounted by kernel ext2 driver,
which is confirmed to be kernel bug that will be fixed later on
^
|
See this part
--
Cyril Hrubis
chrubis@suse.cz
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2018-09-17 8:36 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2018-09-15 9:53 [LTP] statx04.c:53: FAIL: STATX_ATTR_COMPRESSED flag is not set - Testing on ext2 Naresh Kamboju
2018-09-16 16:26 ` Rafael David Tinoco
2018-09-17 8:36 ` Cyril Hrubis
2018-09-17 8:27 ` Cyril Hrubis
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox