[LTP] [PATCH 5.4 00/92] 5.4.241-rc1 review

[Yang Xu (Fujitsu)]

on 2023/04/21 16:04, Petr Vorel wrote:
>> Hi!
>>>> This is the start of the stable review cycle for the 5.4.241 release.
>>>> There are 92 patches in this series, all will be posted as a response
>>>> to this one.  If anyone has any issues with these being applied, please
>>>> let me know.
> 
>>>> Responses should be made by Thu, 20 Apr 2023 12:02:44 +0000.
>>>> Anything received after that time might be too late.
> 
>>>> The whole patch series can be found in one patch at:
>>>>          https://www.kernel.org/pub/linux/kernel/v5.x/stable-review/patch-5.4.241-rc1.gz
>>>> or in the git tree and branch at:
>>>>          git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-5.4.y
>>>> and the diffstat can be found below.
> 
>>>> thanks,
> 
>>>> greg k-h
> 
> 
>>> Recently we have upgraded the LTP test suite version and started noticing
>>> these test failures on 5.4.
>>> Test getting skipped on 4.19 and 4.14 as not supported features.
> 
>>> Need to investigate test case issues or kernel issues.
> 
>>> Reported-by: Linux Kernel Functional Testing <lkft@linaro.org>
> 
>>> NOTE:
> 
>>> ---
>>> creat09.c:73: TINFO: User nobody: uid = 65534, gid = 65534
>>> creat09.c:75: TINFO: Found unused GID 11: SUCCESS (0)
>>> creat09.c:120: TINFO: File created with umask(0)
>>> creat09.c:106: TPASS: mntpoint/testdir/creat.tmp: Owned by correct group
>>> creat09.c:112: TPASS: mntpoint/testdir/creat.tmp: Setgid bit not set
>>> creat09.c:106: TPASS: mntpoint/testdir/open.tmp: Owned by correct group
>>> creat09.c:112: TPASS: mntpoint/testdir/open.tmp: Setgid bit not set
>>> creat09.c:120: TINFO: File created with umask(S_IXGRP)
>>> creat09.c:106: TPASS: mntpoint/testdir/creat.tmp: Owned by correct group
>>> creat09.c:110: TFAIL: mntpoint/testdir/creat.tmp: Setgid bit is set
>>> creat09.c:106: TPASS: mntpoint/testdir/open.tmp: Owned by correct group
>>> creat09.c:110: TFAIL: mntpoint/testdir/open.tmp: Setgid bit is set
> 
>>> Test history links,
>>>   - https://qa-reports.linaro.org/lkft/linux-stable-rc-linux-5.4.y/build/v5.4.238-199-g230f1bde44b6/testrun/16338751/suite/ltp-syscalls/test/creat09/history/
>>>   - https://qa-reports.linaro.org/lkft/linux-stable-rc-linux-5.4.y/build/v5.4.238-199-g230f1bde44b6/testrun/16337895/suite/ltp-cve/test/cve-2018-13405/history/
>>>   - https://qa-reports.linaro.org/lkft/linux-stable-rc-linux-5.4.y/build/v5.4.238-199-g230f1bde44b6/testrun/16338751/suite/ltp-syscalls/test/creat09/log
> 
>> That's likely a missing kernel patch, as this is a regression test there
>> should have been links to the patches and CVE referencies in the test
>> output as the test is tagged with kernel commits and CVE numbers:
> 
>>          .tags = (const struct tst_tag[]) {
>>                  {"linux-git", "0fa3ecd87848"},
>>                  {"CVE", "2018-13405"},
>>                  {"CVE", "2021-4037"},
>>                  {"linux-git", "01ea173e103e"},
> Only this one has been backported (as
> e76bd6da51235ce86f5a8017dd6c056c76da64f9), the other two are missing.
>>                  {"linux-git", "1639a49ccdce"},
>>                  {"linux-git", "426b4ca2d6a5"},
> The last one is merge tag, I wonder if it's correct:
> 426b4ca2d6a5 ("Merge tag 'fs.setgid.v6.0' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux")
> Maybe just 1639a49ccdce would be ok.
> 
> @Yang Xu
> 1) why 1639a49ccdce has not been merged to stable tree? It does not apply now,
> was that the only reason? Or is it not applicable?

In fact, I don't know the stable kernel tree details.

> 
> @Yang Xu is really 426b4ca2d6a5 needed? Was it easier to list merge commit than
> particular fixes? Merge commit contains:
> 
> 5fadbd992996 ("ceph: rely on vfs for setgid stripping")
> 1639a49ccdce ("fs: move S_ISGID stripping into the vfs_*() helpers")
> ac6800e279a2 ("fs: Add missing umask strip in vfs_tmpfile")
> 2b3416ceff5e ("fs: add mode_strip_sgid() helper")

We just need 1639a49ccdce commit is ok and this commit will depend on 
2b3416ceff5e because the previous commit needs to use  mode_strip_sgid api.

For the merged commit, we have a disscussion for 5.19 or 6.0 with cyril 
on last year
see url 
https://lore.kernel.org/ltp/1663143142-2283-1-git-send-email-xuyang2018.jy@fujitsu.com/T/#t

> 
> They have not been backported to 5.4 stable, nor to the older releases.
> Again, they don't apply.
> 

I don't have attention to stable kernel tree, maybe we can ask 5.14 
stable maintainer?

Best Regards
Yang Xu
> 
>>                  {}
>>          },
> 
>>> ---
> 
>>> fanotify14.c:161: TCONF: FAN_REPORT_TARGET_FID not supported in kernel?
>>> fanotify14.c:157: TINFO: Test case 7: fanotify_init(FAN_CLASS_NOTIF |
>>> FAN_REPORT_TARGET_FID | FAN_REPORT_DFID_FID, O_RDONLY)
>>> fanotify14.c:161: TCONF: FAN_REPORT_TARGET_FID not supported in kernel?
>>> [  377.081993] EXT4-fs (loop0): mounting ext3 file system using the
>>> ext4 subsystem
>>> fanotify14.c:157: TINFO: Test case 8: fanotify_init(FAN_CLASS_NOTIF |
>>> FAN_REPORT_DFID_FID, O_RDONLY)
>>> [  377.099137] EXT4-fs (loop0): mounted filesystem with ordered data
>>> mode. Opts: (null)
>>> fanotify14.c:175: TFAIL: fanotify_init(tc->init.flags, O_RDONLY)
>>> failed: EINVAL (22)
> 
>> Possibly like the test may be missing check for a FAN_REPORT_DFID_FID
>> support.
> 
> @Amir could you please look at this fanotify14.c failure on 5.4.241-rc1?
> 
> Kind regards,
> Petr