On 1/28/20 8:27 PM, Michal Koutný wrote:
> I agree with your changes, however, I have some suggestions:
>
> 1) squash [PATCH 2/2] "cgroup_procs_next should increase position index"
> and [PATCH] "__cgroup_procs_start cleanup"
> - make it clear in the commit message that it's fixing the small
> buffer listing, it's not a mere cleanup(!)
> 2) for completeness, I propose squashing also this change (IOW,
> cgroup_procs_start should only initialize the iterator, not move it):
got it
> 3) I was not able to reproduce the corrupted listing into small buffer
> on v1 hierarchy, i.e. the [PATCH 1/2] "cgroup_pidlist_next should update
> position index" log message should just explain the change is to
> satisfy seq_file iterator requirements.
[root@localhost ~]# uname -a
Linux localhost.localdomain 5.4.7-200.fc31.x86_64 #1 SMP Tue Dec 31 22:25:12 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux
[root@localhost ~]# mount | grep cgroup
cgroup2 on /sys/fs/cgroup type cgroup2 (rw,nosuid,nodev,noexec,relatime,seclabel,nsdelegate)
cgroup on /mnt type cgroup (rw,relatime,seclabel,cpu)
[root@localhost ~]# dd if=/mnt/cgroup.procs bs=1 # normal output
...
1294
1295
1296
1304
1382
584+0 records in
584+0 records out
584 bytes copied, 0.00281916 s, 207 kB/s
[root@localhost ~]# dd if=/mnt/cgroup.procs bs=581 skip=1 # read after lseek in middle of last line
dd: /mnt/cgroup.procs: cannot skip to specified offset
83 <<< generates end of last line
1383 <<< ... and whole last line once again
0+1 records in
0+1 records out
8 bytes copied, 0.000171759 s, 46.6 kB/s
[root@localhost ~]# dd if=/mnt/cgroup.procs bs=1000 skip=1 # read after lseek beyond end of file
dd: /mnt/cgroup.procs: cannot skip to specified offset
1386 <<< generates last line anyway
0+1 records in
0+1 records out
5 bytes copied, 0.000198117 s, 25.2 kB/s
> I can send my complete diffs if the suggestions are unclear.
>
> Michal
>
> P.S. I really recommend using `git send-email` for sending out the
> patches, it makes mail threading more readable.
thank you, will resent v2
Vasily Averin
