[Suggestion] kernel: cgroup: mount failure in LTP cgroup_regression_test.sh

[Suggestion] kernel: cgroup: mount failure in LTP cgroup_regression_test.sh

[Chen Gang]

Hello Maintainers:

On Fedora16 with defconfig for next-20131107 kernel, use latest LTP
(Linux Test Project) version, test_5() in cgroup_regression_test.sh
will be fail. The related LTP output:

  <<<test_start>>>
  tag=cgroup stime=1383562810
  cmdline="       cgroup_regression_test.sh"
  contacts=""
  analysis=exit
  <<<test_output>>>
  cgroup_regression_test    1  TPASS  :  no kernel bug was found
  /opt/ltp/testcases/bin/cgroup_regression_test.sh: line 118: 13283 Terminated              ./fork_processes
  cgroup_regression_test    2  TPASS  :  notify_on_release is inherited
  cgroup_regression_test    3  TCONF  :  CONFIG_SCHED_DEBUG is not enabled
  cgroup_regression_test    4  TCONF  :  CONFIG_LOCKDEP is not enabled
  mount: xxx already mounted or cgroup/ busy
  cgroup_regression_test    5  TFAIL  :  mount freezer and cpuacct failed
  cgroup_regression_test    6  TCONF  :  CONFIG_CGROUP_NS
  /opt/ltp/testcases/bin/cgroup_regression_test.sh: line 360: 24428 Terminated              sleep 100 < cgroup/0
  /opt/ltp/testcases/bin/cgroup_regression_test.sh: line 381: 24447 Terminated              sleep 100 < cgroup/0
  cgroup_regression_test    7  TPASS  :  no kernel bug was found
  cgroup_regression_test    8  TPASS  :  no kernel bug was found
  cgroup_regression_test    9  TPASS  :  no kernel warning was found
  rmdir: failed to remove `cgroup/0': No such file or directory
  umount: /sys/fs/cgroup/systemd: device is busy.
          (In some cases useful info about processes that use
           the device is found by lsof(8) or fuser(1))
  cgroup_regression_test   10  TPASS  :  no kernel warning was found
  <<<execution_status>>>
  initiation_status="ok"
  duration=62 termination_type=exited termination_id=1 corefile=no
  cutime=3341 cstime=6353
  <<<test_end>>>

After simplify, the related operation, environments and output are:

  [root@gchenlinux tmp]# df -Th | grep cgroup
  tmpfs                             tmpfs    1001M     0 1001M   0% /sys/fs/cgroup
  [root@gchenlinux tmp]# lsof | grep cgroup | grep -v grep
  systemd       1          root    6r      DIR               0,18         0       5998 /sys/fs/cgroup/systemd/system
  [root@gchenlinux tmp]# cat /proc/cgroups 
  #subsys_name	hierarchy	num_cgroups	enabled
  cpuset	3	4	1
  cpu	4	35	1
  cpuacct	4	35	1
  freezer	5	4	1
  [root@gchenlinux tmp]# mkdir cgroup
  [root@gchenlinux tmp]# mount -t cgroup -o freezer,cpuacct xxx cgroup/
  mount: xxx already mounted or cgroup/ busy

Is it real issue of cgroup? If it is, I will/should continue analyzing.


Welcome any additional suggestions and completions.

Thanks.
--
Chen Gang

Re: [Suggestion] kernel: cgroup: mount failure in LTP cgroup_regression_test.sh

[Tejun Heo]

Hello,

Sorry about the delay.

On Fri, Nov 08, 2013 at 04:15:23PM +0800, Chen Gang wrote:
> After simplify, the related operation, environments and output are:
> 
>   [root@gchenlinux tmp]# df -Th | grep cgroup
>   tmpfs                             tmpfs    1001M     0 1001M   0% /sys/fs/cgroup
>   [root@gchenlinux tmp]# lsof | grep cgroup | grep -v grep
>   systemd       1          root    6r      DIR               0,18         0       5998 /sys/fs/cgroup/systemd/system
>   [root@gchenlinux tmp]# cat /proc/cgroups 
>   #subsys_name	hierarchy	num_cgroups	enabled
>   cpuset	3	4	1
>   cpu	4	35	1
>   cpuacct	4	35	1
>   freezer	5	4	1
>   [root@gchenlinux tmp]# mkdir cgroup
>   [root@gchenlinux tmp]# mount -t cgroup -o freezer,cpuacct xxx cgroup/
>   mount: xxx already mounted or cgroup/ busy
> 
> Is it real issue of cgroup? If it is, I will/should continue analyzing.

Hmmm... I'm a bit confused.  What is it testing?  "cat /proc/cgroup"
is showing that freezer is already mounted and the kernel seems to
have correctly refused to mount it in a different hierarchy.  What am
I missing here?

Thanks.

-- 
tejun

Re: [Suggestion] kernel: cgroup: mount failure in LTP cgroup_regression_test.sh

[Chen Gang]

On 11/21/2013 01:35 AM, Tejun Heo wrote:
> Hello,
> 
> Sorry about the delay.
> 

Oh, No problem, every members time resources are expensive, thank you
for your reply.


> On Fri, Nov 08, 2013 at 04:15:23PM +0800, Chen Gang wrote:
>> After simplify, the related operation, environments and output are:
>>
>>   [root@gchenlinux tmp]# df -Th | grep cgroup
>>   tmpfs                             tmpfs    1001M     0 1001M   0% /sys/fs/cgroup
>>   [root@gchenlinux tmp]# lsof | grep cgroup | grep -v grep
>>   systemd       1          root    6r      DIR               0,18         0       5998 /sys/fs/cgroup/systemd/system
>>   [root@gchenlinux tmp]# cat /proc/cgroups 
>>   #subsys_name	hierarchy	num_cgroups	enabled
>>   cpuset	3	4	1
>>   cpu	4	35	1
>>   cpuacct	4	35	1
>>   freezer	5	4	1
>>   [root@gchenlinux tmp]# mkdir cgroup
>>   [root@gchenlinux tmp]# mount -t cgroup -o freezer,cpuacct xxx cgroup/
>>   mount: xxx already mounted or cgroup/ busy
>>
>> Is it real issue of cgroup? If it is, I will/should continue analyzing.
> 
> Hmmm... I'm a bit confused.  What is it testing?  "cat /proc/cgroup"
> is showing that freezer is already mounted and the kernel seems to
> have correctly refused to mount it in a different hierarchy.  What am
> I missing here?
> 

Hmm... I am not quit familiar about it, either, but at least I can
contact related LTP members (I guess they can make sure about it).


Excuse me, I am changing/finding new job, so I should not use Asianux
Corporation mail again.  And it seems I can not be focus on upstream
kernel as much as before.  :-(

But at least, I will/should continue providing some contributions (more
or less) to Public Open Source as volunteers. And for upstream kernel,
I plan to make 1-3 patches per month (mainly for cross compiling).

 - if can not get companies support, I can not provide enough time
   resource on upstream kernel (I have to get pay check for family).

 - New company may mainly focus on open source tool chains (not only
   kernel), so I will/should also provide some contributions to them
   (e.g. KVM, GCC ...).


At last, sorry again for I have to give up 10 patches per month for
upstream kernel.


Thanks.
-- 
Chen Gang

Fwd: [Suggestion] kernel: cgroup: mount failure in LTP cgroup_regression_test.sh

[Chen Gang]

Hello LTP maintainers:

When I use ltp to test cgroup, I find one failure, the related forward
mail is below.

Related LTP git commit is "75df824 Add cgroup regression tests to LTP:
Those testcases can reproduce bugs we found in cgroups, and is useful
to track future regressions." (lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org and subrata_modak)

Please help check when you have time.

Thanks.

-------- Original Message --------
Subject: [Suggestion] kernel: cgroup: mount failure in LTP cgroup_regression_test.sh
Date: Fri, 08 Nov 2013 16:15:23 +0800
From: Chen Gang <gang.chen-bOixZGp5f+dBDgjK7y7TUQ@public.gmane.org>
To: Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>, Li Zefan <lizefan-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
CC: containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org, cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,  "linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org" <linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>

Hello Maintainers:

On Fedora16 with defconfig for next-20131107 kernel, use latest LTP
(Linux Test Project) version, test_5() in cgroup_regression_test.sh
will be fail. The related LTP output:

  <<<test_start>>>
  tag=cgroup stime=1383562810
  cmdline="       cgroup_regression_test.sh"
  contacts=""
  analysis=exit
  <<<test_output>>>
  cgroup_regression_test    1  TPASS  :  no kernel bug was found
  /opt/ltp/testcases/bin/cgroup_regression_test.sh: line 118: 13283 Terminated              ./fork_processes
  cgroup_regression_test    2  TPASS  :  notify_on_release is inherited
  cgroup_regression_test    3  TCONF  :  CONFIG_SCHED_DEBUG is not enabled
  cgroup_regression_test    4  TCONF  :  CONFIG_LOCKDEP is not enabled
  mount: xxx already mounted or cgroup/ busy
  cgroup_regression_test    5  TFAIL  :  mount freezer and cpuacct failed
  cgroup_regression_test    6  TCONF  :  CONFIG_CGROUP_NS
  /opt/ltp/testcases/bin/cgroup_regression_test.sh: line 360: 24428 Terminated              sleep 100 < cgroup/0
  /opt/ltp/testcases/bin/cgroup_regression_test.sh: line 381: 24447 Terminated              sleep 100 < cgroup/0
  cgroup_regression_test    7  TPASS  :  no kernel bug was found
  cgroup_regression_test    8  TPASS  :  no kernel bug was found
  cgroup_regression_test    9  TPASS  :  no kernel warning was found
  rmdir: failed to remove `cgroup/0': No such file or directory
  umount: /sys/fs/cgroup/systemd: device is busy.
          (In some cases useful info about processes that use
           the device is found by lsof(8) or fuser(1))
  cgroup_regression_test   10  TPASS  :  no kernel warning was found
  <<<execution_status>>>
  initiation_status="ok"
  duration=62 termination_type=exited termination_id=1 corefile=no
  cutime=3341 cstime=6353
  <<<test_end>>>

After simplify, the related operation, environments and output are:

  [root@gchenlinux tmp]# df -Th | grep cgroup
  tmpfs                             tmpfs    1001M     0 1001M   0% /sys/fs/cgroup
  [root@gchenlinux tmp]# lsof | grep cgroup | grep -v grep
  systemd       1          root    6r      DIR               0,18         0       5998 /sys/fs/cgroup/systemd/system
  [root@gchenlinux tmp]# cat /proc/cgroups
  #subsys_name	hierarchy	num_cgroups	enabled
  cpuset	3	4	1
  cpu	4	35	1
  cpuacct	4	35	1
  freezer	5	4	1
  [root@gchenlinux tmp]# mkdir cgroup
  [root@gchenlinux tmp]# mount -t cgroup -o freezer,cpuacct xxx cgroup/
  mount: xxx already mounted or cgroup/ busy

Is it real issue of cgroup? If it is, I will/should continue analyzing.


Welcome any additional suggestions and completions.

Thanks.
--
Chen Gang

Re: [Suggestion] kernel: cgroup: mount failure in LTP cgroup_regression_test.sh

[Jan Stancek]

Hi,

it looks like those 2 subsystems are already mounted by systemd,
each in different hierarchy:

>   cpuacct	4	35	1
>   freezer	5	4	1

from Documentation/cgroups/cgroups.txt:
"If an active hierarchy with exactly the same set of subsystems already
exists, it will be reused for the new mount. If no existing hierarchy
matches, and any of the requested subsystems are in use in an existing
hierarchy, the mount will fail with -EBUSY. Otherwise, a new hierarchy
is activated, associated with the requested subsystems."

Regards,
Jan

----- Original Message -----
> From: "Chen Gang" <gang.chen.5i5j-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
> To: "Tejun Heo" <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>, "Garrett Cooper" <yanegomi-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>, gaowanlong-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org, chris-8NJIiSa5LzA@public.gmane.org,
> jstancek-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org, mitani-Czb31pHLG9l3+QwDJ9on6Q@public.gmane.org, lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org
> Cc: "Li Zefan" <lizefan-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>, containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org, cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> Sent: Thursday, 21 November, 2013 5:18:50 AM
> Subject: Fwd: [Suggestion] kernel: cgroup: mount failure in LTP cgroup_regression_test.sh
> 
> Hello LTP maintainers:
> 
> When I use ltp to test cgroup, I find one failure, the related forward
> mail is below.
> 
> Related LTP git commit is "75df824 Add cgroup regression tests to LTP:
> Those testcases can reproduce bugs we found in cgroups, and is useful
> to track future regressions." (lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org and subrata_modak)
> 
> Please help check when you have time.
> 
> Thanks.
> 
> -------- Original Message --------
> Subject: [Suggestion] kernel: cgroup: mount failure in LTP
> cgroup_regression_test.sh
> Date: Fri, 08 Nov 2013 16:15:23 +0800
> From: Chen Gang <gang.chen-bOixZGp5f+dBDgjK7y7TUQ@public.gmane.org>
> To: Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>, Li Zefan <lizefan-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
> CC: containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org, cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
> "linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org" <linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>
> 
> Hello Maintainers:
> 
> On Fedora16 with defconfig for next-20131107 kernel, use latest LTP
> (Linux Test Project) version, test_5() in cgroup_regression_test.sh
> will be fail. The related LTP output:
> 
>   <<<test_start>>>
>   tag=cgroup stime=1383562810
>   cmdline="       cgroup_regression_test.sh"
>   contacts=""
>   analysis=exit
>   <<<test_output>>>
>   cgroup_regression_test    1  TPASS  :  no kernel bug was found
>   /opt/ltp/testcases/bin/cgroup_regression_test.sh: line 118: 13283
>   Terminated              ./fork_processes
>   cgroup_regression_test    2  TPASS  :  notify_on_release is inherited
>   cgroup_regression_test    3  TCONF  :  CONFIG_SCHED_DEBUG is not enabled
>   cgroup_regression_test    4  TCONF  :  CONFIG_LOCKDEP is not enabled
>   mount: xxx already mounted or cgroup/ busy
>   cgroup_regression_test    5  TFAIL  :  mount freezer and cpuacct failed
>   cgroup_regression_test    6  TCONF  :  CONFIG_CGROUP_NS
>   /opt/ltp/testcases/bin/cgroup_regression_test.sh: line 360: 24428
>   Terminated              sleep 100 < cgroup/0
>   /opt/ltp/testcases/bin/cgroup_regression_test.sh: line 381: 24447
>   Terminated              sleep 100 < cgroup/0
>   cgroup_regression_test    7  TPASS  :  no kernel bug was found
>   cgroup_regression_test    8  TPASS  :  no kernel bug was found
>   cgroup_regression_test    9  TPASS  :  no kernel warning was found
>   rmdir: failed to remove `cgroup/0': No such file or directory
>   umount: /sys/fs/cgroup/systemd: device is busy.
>           (In some cases useful info about processes that use
>            the device is found by lsof(8) or fuser(1))
>   cgroup_regression_test   10  TPASS  :  no kernel warning was found
>   <<<execution_status>>>
>   initiation_status="ok"
>   duration=62 termination_type=exited termination_id=1 corefile=no
>   cutime=3341 cstime=6353
>   <<<test_end>>>
> 
> After simplify, the related operation, environments and output are:
> 
>   [root@gchenlinux tmp]# df -Th | grep cgroup
>   tmpfs                             tmpfs    1001M     0 1001M   0%
>   /sys/fs/cgroup
>   [root@gchenlinux tmp]# lsof | grep cgroup | grep -v grep
>   systemd       1          root    6r      DIR               0,18         0
>   5998 /sys/fs/cgroup/systemd/system
>   [root@gchenlinux tmp]# cat /proc/cgroups
>   #subsys_name	hierarchy	num_cgroups	enabled
>   cpuset	3	4	1
>   cpu	4	35	1
>   cpuacct	4	35	1
>   freezer	5	4	1
>   [root@gchenlinux tmp]# mkdir cgroup
>   [root@gchenlinux tmp]# mount -t cgroup -o freezer,cpuacct xxx cgroup/
>   mount: xxx already mounted or cgroup/ busy
> 
> Is it real issue of cgroup? If it is, I will/should continue analyzing.
> 
> 
> Welcome any additional suggestions and completions.
> 
> Thanks.
> --
> Chen Gang
> 
> 
> 

Re: [Suggestion] kernel: cgroup: mount failure in LTP cgroup_regression_test.sh

[Chen Gang]

On 11/21/2013 04:04 PM, Jan Stancek wrote:
> Hi,
> 
> it looks like those 2 subsystems are already mounted by systemd,
> each in different hierarchy:
> 
>>   cpuacct	4	35	1
>>   freezer	5	4	1
> 
> from Documentation/cgroups/cgroups.txt:
> "If an active hierarchy with exactly the same set of subsystems already
> exists, it will be reused for the new mount. If no existing hierarchy
> matches, and any of the requested subsystems are in use in an existing
> hierarchy, the mount will fail with -EBUSY. Otherwise, a new hierarchy
> is activated, associated with the requested subsystems."
> 

OK, for me, these information above are valuable for analyzing.

Hmm... do we have members who both the related author of LTP cgroup and
kernel cgroup? If we have, I guess this issue is very simple to them,
and don't need any additional analyzing by other members.

If we really have no related members for both LTP cgroup and kernel
cgroup, at least, I will/should continue analyzing.


Thanks.

> Regards,
> Jan
> 
> ----- Original Message -----
>> From: "Chen Gang" <gang.chen.5i5j-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
>> To: "Tejun Heo" <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>, "Garrett Cooper" <yanegomi-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>, gaowanlong-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org, chris-8NJIiSa5LzA@public.gmane.org,
>> jstancek-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org, mitani-Czb31pHLG9l3+QwDJ9on6Q@public.gmane.org, lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org
>> Cc: "Li Zefan" <lizefan-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>, containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org, cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
>> Sent: Thursday, 21 November, 2013 5:18:50 AM
>> Subject: Fwd: [Suggestion] kernel: cgroup: mount failure in LTP cgroup_regression_test.sh
>>
>> Hello LTP maintainers:
>>
>> When I use ltp to test cgroup, I find one failure, the related forward
>> mail is below.
>>
>> Related LTP git commit is "75df824 Add cgroup regression tests to LTP:
>> Those testcases can reproduce bugs we found in cgroups, and is useful
>> to track future regressions." (lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org and subrata_modak)
>>
>> Please help check when you have time.
>>
>> Thanks.
>>
>> -------- Original Message --------
>> Subject: [Suggestion] kernel: cgroup: mount failure in LTP
>> cgroup_regression_test.sh
>> Date: Fri, 08 Nov 2013 16:15:23 +0800
>> From: Chen Gang <gang.chen-bOixZGp5f+dBDgjK7y7TUQ@public.gmane.org>
>> To: Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>, Li Zefan <lizefan-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
>> CC: containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org, cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
>> "linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org" <linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org>
>>
>> Hello Maintainers:
>>
>> On Fedora16 with defconfig for next-20131107 kernel, use latest LTP
>> (Linux Test Project) version, test_5() in cgroup_regression_test.sh
>> will be fail. The related LTP output:
>>
>>   <<<test_start>>>
>>   tag=cgroup stime=1383562810
>>   cmdline="       cgroup_regression_test.sh"
>>   contacts=""
>>   analysis=exit
>>   <<<test_output>>>
>>   cgroup_regression_test    1  TPASS  :  no kernel bug was found
>>   /opt/ltp/testcases/bin/cgroup_regression_test.sh: line 118: 13283
>>   Terminated              ./fork_processes
>>   cgroup_regression_test    2  TPASS  :  notify_on_release is inherited
>>   cgroup_regression_test    3  TCONF  :  CONFIG_SCHED_DEBUG is not enabled
>>   cgroup_regression_test    4  TCONF  :  CONFIG_LOCKDEP is not enabled
>>   mount: xxx already mounted or cgroup/ busy
>>   cgroup_regression_test    5  TFAIL  :  mount freezer and cpuacct failed
>>   cgroup_regression_test    6  TCONF  :  CONFIG_CGROUP_NS
>>   /opt/ltp/testcases/bin/cgroup_regression_test.sh: line 360: 24428
>>   Terminated              sleep 100 < cgroup/0
>>   /opt/ltp/testcases/bin/cgroup_regression_test.sh: line 381: 24447
>>   Terminated              sleep 100 < cgroup/0
>>   cgroup_regression_test    7  TPASS  :  no kernel bug was found
>>   cgroup_regression_test    8  TPASS  :  no kernel bug was found
>>   cgroup_regression_test    9  TPASS  :  no kernel warning was found
>>   rmdir: failed to remove `cgroup/0': No such file or directory
>>   umount: /sys/fs/cgroup/systemd: device is busy.
>>           (In some cases useful info about processes that use
>>            the device is found by lsof(8) or fuser(1))
>>   cgroup_regression_test   10  TPASS  :  no kernel warning was found
>>   <<<execution_status>>>
>>   initiation_status="ok"
>>   duration=62 termination_type=exited termination_id=1 corefile=no
>>   cutime=3341 cstime=6353
>>   <<<test_end>>>
>>
>> After simplify, the related operation, environments and output are:
>>
>>   [root@gchenlinux tmp]# df -Th | grep cgroup
>>   tmpfs                             tmpfs    1001M     0 1001M   0%
>>   /sys/fs/cgroup
>>   [root@gchenlinux tmp]# lsof | grep cgroup | grep -v grep
>>   systemd       1          root    6r      DIR               0,18         0
>>   5998 /sys/fs/cgroup/systemd/system
>>   [root@gchenlinux tmp]# cat /proc/cgroups
>>   #subsys_name	hierarchy	num_cgroups	enabled
>>   cpuset	3	4	1
>>   cpu	4	35	1
>>   cpuacct	4	35	1
>>   freezer	5	4	1
>>   [root@gchenlinux tmp]# mkdir cgroup
>>   [root@gchenlinux tmp]# mount -t cgroup -o freezer,cpuacct xxx cgroup/
>>   mount: xxx already mounted or cgroup/ busy
>>
>> Is it real issue of cgroup? If it is, I will/should continue analyzing.
>>
>>
>> Welcome any additional suggestions and completions.
>>
>> Thanks.
>> --
>> Chen Gang
>>
>>
>>


-- 
Chen Gang

Re: [Suggestion] kernel: cgroup: mount failure in LTP cgroup_regression_test.sh

[Tejun Heo]

Hello, Jan.

On Thu, Nov 21, 2013 at 03:04:54AM -0500, Jan Stancek wrote:
> it looks like those 2 subsystems are already mounted by systemd,
> each in different hierarchy:
> 
> >   cpuacct	4	35	1
> >   freezer	5	4	1
> 
> from Documentation/cgroups/cgroups.txt:
> "If an active hierarchy with exactly the same set of subsystems already
> exists, it will be reused for the new mount. If no existing hierarchy
> matches, and any of the requested subsystems are in use in an existing
> hierarchy, the mount will fail with -EBUSY. Otherwise, a new hierarchy
> is activated, associated with the requested subsystems."

Ah, cgroup now requires that those recycled mounts to have matching
options.  This is an intended behavior change because previously we
could end up with mounts which have actual options which differ from
what was specified by mount, which can be misleading at the very
least.  Would it be possible to update LTP accordingly?

Thanks.

-- 
tejun

Re: [Suggestion] kernel: cgroup: mount failure in LTP cgroup_regression_test.sh

[Jan Stancek]

----- Original Message -----
> From: "Tejun Heo" <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
> To: "Jan Stancek" <jstancek-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
> Cc: "Chen Gang" <gang.chen.5i5j-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>, "Garrett Cooper" <yanegomi-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>, gaowanlong-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org,
> chris-8NJIiSa5LzA@public.gmane.org, mitani-Czb31pHLG9l3+QwDJ9on6Q@public.gmane.org, lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org, "Li Zefan" <lizefan-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>,
> containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org, cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> Sent: Thursday, 21 November, 2013 4:26:45 PM
> Subject: Re: [Suggestion] kernel: cgroup: mount failure in LTP cgroup_regression_test.sh
> 
> Hello, Jan.
> 
> On Thu, Nov 21, 2013 at 03:04:54AM -0500, Jan Stancek wrote:
> > it looks like those 2 subsystems are already mounted by systemd,
> > each in different hierarchy:
> > 
> > >   cpuacct	4	35	1
> > >   freezer	5	4	1
> > 
> > from Documentation/cgroups/cgroups.txt:
> > "If an active hierarchy with exactly the same set of subsystems already
> > exists, it will be reused for the new mount. If no existing hierarchy
> > matches, and any of the requested subsystems are in use in an existing
> > hierarchy, the mount will fail with -EBUSY. Otherwise, a new hierarchy
> > is activated, associated with the requested subsystems."
> 
> Ah, cgroup now requires that those recycled mounts to have matching
> options.  This is an intended behavior change because previously we
> could end up with mounts which have actual options which differ from
> what was specified by mount, which can be misleading at the very
> least.

I believe the test always respected this requirement. Problem we see
is because of initial state of the system. Test appears to expect
that it is able to mount cgroups freely, but they are already in use.

> Would it be possible to update LTP accordingly?

I'm not sure the test can do anything (other than not run at all).

Other option is to setup system in a way where cgroups are not in use
prior to running this test, though this may not be possible with systemd.

Regards,
Jan

> 
> Thanks.
> 
> --
> tejun
> 

Re: [Suggestion] kernel: cgroup: mount failure in LTP cgroup_regression_test.sh

[Li Zefan]

> ----- Original Message -----
>> From: "Tejun Heo" <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
>> To: "Jan Stancek" <jstancek-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
>> Cc: "Chen Gang" <gang.chen.5i5j-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>, "Garrett Cooper" <yanegomi-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>, gaowanlong-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org,
>> chris-8NJIiSa5LzA@public.gmane.org, mitani-Czb31pHLG9l3+QwDJ9on6Q@public.gmane.org, lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org, "Li Zefan" <lizefan-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>,
>> containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org, cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
>> Sent: Thursday, 21 November, 2013 4:26:45 PM
>> Subject: Re: [Suggestion] kernel: cgroup: mount failure in LTP cgroup_regression_test.sh
>>
>> Hello, Jan.
>>
>> On Thu, Nov 21, 2013 at 03:04:54AM -0500, Jan Stancek wrote:
>>> it looks like those 2 subsystems are already mounted by systemd,
>>> each in different hierarchy:
>>>
>>>>   cpuacct	4	35	1
>>>>   freezer	5	4	1
>>>
>>> from Documentation/cgroups/cgroups.txt:
>>> "If an active hierarchy with exactly the same set of subsystems already
>>> exists, it will be reused for the new mount. If no existing hierarchy
>>> matches, and any of the requested subsystems are in use in an existing
>>> hierarchy, the mount will fail with -EBUSY. Otherwise, a new hierarchy
>>> is activated, associated with the requested subsystems."
>>
>> Ah, cgroup now requires that those recycled mounts to have matching
>> options.  This is an intended behavior change because previously we
>> could end up with mounts which have actual options which differ from
>> what was specified by mount, which can be misleading at the very
>> least.
> 
> I believe the test always respected this requirement. Problem we see
> is because of initial state of the system. Test appears to expect
> that it is able to mount cgroups freely, but they are already in use.
> 
>> Would it be possible to update LTP accordingly?
> 
> I'm not sure the test can do anything (other than not run at all).
> 
> Other option is to setup system in a way where cgroups are not in use
> prior to running this test, though this may not be possible with systemd.
> 

I believe each test will do cleanup after it finishes (like umount
cgroupfs). I guess the system already mounted cgroupfs after boot
and each cgroup subsystem was mounted separately? And that leads to
this issue.

Re: [Suggestion] kernel: cgroup: mount failure in LTP cgroup_regression_test.sh

[Tejun Heo]

Hello,

On Thu, Nov 21, 2013 at 06:13:01PM -0500, Jan Stancek wrote:
> I believe the test always respected this requirement. Problem we see
> is because of initial state of the system. Test appears to expect
> that it is able to mount cgroups freely, but they are already in use.
> 
> > Would it be possible to update LTP accordingly?
> 
> I'm not sure the test can do anything (other than not run at all).
> 
> Other option is to setup system in a way where cgroups are not in use
> prior to running this test, though this may not be possible with systemd.

You can tell systemd to not mount any controllers.

Thanks.

-- 
tejun