> > I don't see anything which would generate this warning in 2.6.33.4's
> > fs/fs-writeback.c:597. ?RH must have mucked with it.
> Already fixed in mainstream by
> commit 7c8a3554c683f512dbcee26faedb42e4c05f12fa
> Author: Jens Axboe <[email protected]>
> Date: ? Tue May 18 14:29:29 2010 +0200
Do you send it to stable ? I am hitting the same bug and count is high
on kerneloops (http://www.kerneloops.org/searchweek.php?search=writeback_inodes_wb)
Bastien
On Thu, 10 Jun 2010 11:06:17 +0200
Bastien ROUCARIES <[email protected]> wrote:
> > > I don't see anything which would generate this warning in 2.6.33.4's
> > > fs/fs-writeback.c:597. __RH must have mucked with it.
> > Already fixed in mainstream by
> > commit 7c8a3554c683f512dbcee26faedb42e4c05f12fa
> > Author: Jens Axboe <[email protected]>
> > Date: __ Tue May 18 14:29:29 2010 +0200
>
> Do you send it to stable ? I am hitting the same bug and count is high
> on kerneloops (http://www.kerneloops.org/searchweek.php?search=writeback_inodes_wb)
>
The commit didn't have a cc:stable in the changelog so no, it probably
wasn't backported.
But that commit was later reverted anyway. Hopefully when the real fix
gets merged, it will be backported.
On Mon, Jun 14, 2010 at 12:57:16PM -0700, Andrew Morton wrote:
> > Do you send it to stable ? I am hitting the same bug and count is high
> > on kerneloops (http://www.kerneloops.org/searchweek.php?search=writeback_inodes_wb)
> >
>
> The commit didn't have a cc:stable in the changelog so no, it probably
> wasn't backported.
>
> But that commit was later reverted anyway. Hopefully when the real fix
> gets merged, it will be backported.
>
Yes, we backported it when it hit mainline, and have since pulled them
until the real fix is sorted out.
thanks, Kyle.
On Mon, Jun 14, 2010 at 9:57 PM, Andrew Morton
<[email protected]> wrote:
> On Thu, 10 Jun 2010 11:06:17 +0200
> Bastien ROUCARIES <[email protected]> wrote:
>
>> > > I don't see anything which would generate this warning in 2.6.33.4's
>> > > fs/fs-writeback.c:597. __RH must have mucked with it.
>> > Already fixed in mainstream by
>> > commit 7c8a3554c683f512dbcee26faedb42e4c05f12fa
>> > Author: Jens Axboe <[email protected]>
>> > Date: __ Tue May 18 14:29:29 2010 +0200
>>
>> Do you send it to stable ? I am hitting the same bug and count is high
>> on kerneloops (http://www.kerneloops.org/searchweek.php?search=writeback_inodes_wb)
>>
>
> The commit didn't have a cc:stable in the changelog so no, it probably
> wasn't backported.
>
> But that commit was later reverted anyway. ?Hopefully when the real fix
> gets merged, it will be backported.
Thank Andrew. I am hitting this warning at each boot, without anything
fancy on my machine
(I am hitting it just after mounting the root partition, i suppose
reiserfs user_xattr, acl help but I was expecting that automatic
testing could catch this kind of bug)
Regards
Bastien
On Tue, Jun 15, 2010 at 10:25 AM, Bastien ROUCARIES
<[email protected]> wrote:
> On Mon, Jun 14, 2010 at 9:57 PM, Andrew Morton
> <[email protected]> wrote:
>> On Thu, 10 Jun 2010 11:06:17 +0200
>> Bastien ROUCARIES <[email protected]> wrote:
>>
>>> > > I don't see anything which would generate this warning in 2.6.33.4's
>>> > > fs/fs-writeback.c:597. __RH must have mucked with it.
>>> > Already fixed in mainstream by
>>> > commit 7c8a3554c683f512dbcee26faedb42e4c05f12fa
>>> > Author: Jens Axboe <[email protected]>
>>> > Date: __ Tue May 18 14:29:29 2010 +0200
>>>
>>> Do you send it to stable ? I am hitting the same bug and count is high
>>> on kerneloops (http://www.kerneloops.org/searchweek.php?search=writeback_inodes_wb)
>>>
>>
>> The commit didn't have a cc:stable in the changelog so no, it probably
>> wasn't backported.
>>
>> But that commit was later reverted anyway. ?Hopefully when the real fix
>> gets merged, it will be backported.
>
Any news of this bug ?
Regard
Bastien
On Thu, 1 Jul 2010 09:54:07 +0200
Bastien ROUCARIES <[email protected]> wrote:
> On Tue, Jun 15, 2010 at 10:25 AM, Bastien ROUCARIES
> <[email protected]> wrote:
> > On Mon, Jun 14, 2010 at 9:57 PM, Andrew Morton
> > <[email protected]> wrote:
> >> On Thu, 10 Jun 2010 11:06:17 +0200
> >> Bastien ROUCARIES <[email protected]> wrote:
> >>
> >>> > > I don't see anything which would generate this warning in 2.6.33.4's
> >>> > > fs/fs-writeback.c:597. __RH must have mucked with it.
> >>> > Already fixed in mainstream by
> >>> > commit 7c8a3554c683f512dbcee26faedb42e4c05f12fa
> >>> > Author: Jens Axboe <[email protected]>
> >>> > Date: __ Tue May 18 14:29:29 2010 +0200
> >>>
> >>> Do you send it to stable ? I am hitting the same bug and count is high
> >>> on kerneloops (http://www.kerneloops.org/searchweek.php?search=writeback_inodes_wb)
> >>>
> >>
> >> The commit didn't have a cc:stable in the changelog so no, it probably
> >> wasn't backported.
> >>
> >> But that commit was later reverted anyway. __Hopefully when the real fix
> >> gets merged, it will be backported.
> >
>
> Any news of this bug ?
>
There have been various dramas in the writeback code recently.
Christoph has been working on it.
Christoph, is this oops known about, and are those patches expected to
fix it?
Thanks.
On Wed, Jul 07, 2010 at 01:51:18PM -0700, Andrew Morton wrote:
> There have been various dramas in the writeback code recently.
> Christoph has been working on it.
>
> Christoph, is this oops known about, and are those patches expected to
> fix it?
I have no idea what oops you mean. But the writeback code is very
different now compared to 2.6.33, and with the latest fixes in Jens'
tree there are no known issues left. (know to me anyway)
On Wed, 7 Jul 2010 23:02:09 +0200
Christoph Hellwig <[email protected]> wrote:
> On Wed, Jul 07, 2010 at 01:51:18PM -0700, Andrew Morton wrote:
> > There have been various dramas in the writeback code recently.
> > Christoph has been working on it.
> >
> > Christoph, is this oops known about, and are those patches expected to
> > fix it?
>
> I have no idea what oops you mean.
http://lmgtfy.com/?q=fs%2Ffs-writeback+oopses+in+kernel+2.6.33.4-95.fc13.i686.PAE
> But the writeback code is very
> different now compared to 2.6.33, and with the latest fixes in Jens'
> tree there are no known issues left. (know to me anyway)
Well the issue is that 2.6.33.x goes oops and we'd like to fix that.
if it's fixed in current mainline then which patch did it?
On Wed, Jul 07, 2010 at 02:14:31PM -0700, Andrew Morton wrote:
> >
> > I have no idea what oops you mean.
>
> http://lmgtfy.com/?q=fs%2Ffs-writeback+oopses+in+kernel+2.6.33.4-95.fc13.i686.PAE
Not particularly useful. Line numbers and function have changed
completely since .33. Given that 2.6.33-stable is at the end of line I really
can't be bothered to go back, figure out what the hell was going on back
then and backport things. In fact we'd probably end up backporting all
writeback changes since then anyway, as the changes to sort out the
per-sb syncs were very large and interdependent.