Re: linux-next: Tree for Sept 28 (ext4|vfs related issues?)

Re: linux-next: Tree for Sept 28 (ext4|vfs related issues?)

[Sedat Dilek]

On Fri, Sep 28, 2012 at 1:22 PM, Sedat Dilek <sedat.dilek@gmail.com> wrote:
> On Fri, Sep 28, 2012 at 8:51 AM, Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>> Hi all,
>>
>> Changes since 201209027:
>>
>> The vfs tree lost its conflicts.
>>
>> The infiniband tree gained a conflict against Linus' tree.
>>
>> The net-next tree gained a conflict against the net tree and a build
>> failure for which I reverted a commit.
>>
>> The block tree lost its build failure.
>>
>> The tip tree gained conflicts against the rr tree.
>>
>> The signal tree gained a conflict against the vfs tree.
>>
>> The akpm tree lost several patches that turned up elsewhere and lost its
>> build failure but gained another for which I applied a merge fix patch.
>>
>> ----------------------------------------------------------------------------
>>
>
> [ Added some more maintainers of ext4|vfs|block ]
>
> Hi,
>
> after solving the thermal-next problem (patch still pending in thermal
> tree) today's Linux-Next is still not running fine here (I reported
> that already).
>
> After switching from SLUB to SLAB I could take finally a screenshot of
> a call-trace (sorry, it's 600KiB).
> It seems to be a problem of Ext4-FS or in combination with VFS (can't
> say as I am not an FS-expert).
> Symptoms:
> Harddisk I/O is fu**ing slow, apps like firefox start slowly as used
> to and stall often, a dmesg-output in a console takes sometimes ages,
> etc.
> IOW, you cannot work on such a system :-)!
>
> Hope the screenshot helps to dig down the problem.
>

I pulled in latest ext4.git#dev [0] and [1] seems to be the pending fix.
Still compiling... (will report later on).

- Sedat -

[0] http://git.kernel.org/?p=linux/kernel/git/tytso/ext4.git;a=shortlog;h=refs/heads/dev
[1] http://git.kernel.org/?p=linux/kernel/git/tytso/ext4.git;a=commitdiff;h=9bebbf9b6266e63a68845ade39e368f2d162ea1c

> Regards,
> - Sedat -
>
> P.S.: I also attached the latest kernel-config.

Re: linux-next: Tree for Sept 28 (ext4|vfs related issues?)

[Sedat Dilek]

On Fri, Sep 28, 2012 at 1:51 PM, Sedat Dilek <sedat.dilek@gmail.com> wrote:
> On Fri, Sep 28, 2012 at 1:22 PM, Sedat Dilek <sedat.dilek@gmail.com> wrote:
>> On Fri, Sep 28, 2012 at 8:51 AM, Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>>> Hi all,
>>>
>>> Changes since 201209027:
>>>
>>> The vfs tree lost its conflicts.
>>>
>>> The infiniband tree gained a conflict against Linus' tree.
>>>
>>> The net-next tree gained a conflict against the net tree and a build
>>> failure for which I reverted a commit.
>>>
>>> The block tree lost its build failure.
>>>
>>> The tip tree gained conflicts against the rr tree.
>>>
>>> The signal tree gained a conflict against the vfs tree.
>>>
>>> The akpm tree lost several patches that turned up elsewhere and lost its
>>> build failure but gained another for which I applied a merge fix patch.
>>>
>>> ----------------------------------------------------------------------------
>>>
>>
>> [ Added some more maintainers of ext4|vfs|block ]
>>
>> Hi,
>>
>> after solving the thermal-next problem (patch still pending in thermal
>> tree) today's Linux-Next is still not running fine here (I reported
>> that already).
>>
>> After switching from SLUB to SLAB I could take finally a screenshot of
>> a call-trace (sorry, it's 600KiB).
>> It seems to be a problem of Ext4-FS or in combination with VFS (can't
>> say as I am not an FS-expert).
>> Symptoms:
>> Harddisk I/O is fu**ing slow, apps like firefox start slowly as used
>> to and stall often, a dmesg-output in a console takes sometimes ages,
>> etc.
>> IOW, you cannot work on such a system :-)!
>>
>> Hope the screenshot helps to dig down the problem.
>>
>
> I pulled in latest ext4.git#dev [0] and [1] seems to be the pending fix.
> Still compiling... (will report later on).
>

Unfortunately, this did not help!
Any hints/help appreciated.

- Sedat -

> - Sedat -
>
> [0] http://git.kernel.org/?p=linux/kernel/git/tytso/ext4.git;a=shortlog;h=refs/heads/dev
> [1] http://git.kernel.org/?p=linux/kernel/git/tytso/ext4.git;a=commitdiff;h=9bebbf9b6266e63a68845ade39e368f2d162ea1c
>
>> Regards,
>> - Sedat -
>>
>> P.S.: I also attached the latest kernel-config.

Re: linux-next: Tree for Sept 28 (ext4|vfs related issues?)

[Theodore Ts'o]

On Fri, Sep 28, 2012 at 01:22:45PM +0200, Sedat Dilek wrote:
> 
> After switching from SLUB to SLAB I could take finally a screenshot of
> a call-trace (sorry, it's 600KiB).
> It seems to be a problem of Ext4-FS or in combination with VFS (can't
> say as I am not an FS-expert).

This is strictly speaking an ext4 bug, and it's a known problem.  It
will be fixed in the next linux-next release; my apologies for letting
this slip through....

	   	       		     	     	       - Ted