Re: [Linux 2.6.27-rc2] suspend to disk, libata problem

Re: [Linux 2.6.27-rc2] suspend to disk, libata problem

[Maciej Rutecki]

I have problem with suspend to disk. On 2.6.26.* works fine.

During s2disk i have this messages (with "no_console_suspend" kernel option):
http://unixy.pl/maciek/download/kernel/2.6.27-rc2/img_0005.jpg
(hibernation fails)

Config, dmesg (normal booting):
http://unixy.pl/maciek/download/kernel/2.6.27-rc2/

-- 
Maciej Rutecki
http://www.maciek.unixy.pl

Re: [Linux 2.6.27-rc2] suspend to disk, libata problem

[Linus Torvalds]

On Fri, 8 Aug 2008, Maciej Rutecki wrote:
>
> I have problem with suspend to disk. On 2.6.26.* works fine.
> 
> During s2disk i have this messages (with "no_console_suspend" kernel option):
> http://unixy.pl/maciek/download/kernel/2.6.27-rc2/img_0005.jpg
> (hibernation fails)
> 
> Config, dmesg (normal booting):
> http://unixy.pl/maciek/download/kernel/2.6.27-rc2/

Any chance to get this bisected? I realize that it's quite a few reboots, 
and _maybe_ Jeff or Bartlomiej can just say "oh, duh, it's because of 
Xyz", but the more likely thing is that without a bisect people will be 
flailing around fairly blindly..

			Linus

Re: [Linux 2.6.27-rc2] suspend to disk, libata problem

[Rafael J. Wysocki]

On Friday, 8 of August 2008, Linus Torvalds wrote:
> 
> On Fri, 8 Aug 2008, Maciej Rutecki wrote:
> >
> > I have problem with suspend to disk. On 2.6.26.* works fine.
> > 
> > During s2disk i have this messages (with "no_console_suspend" kernel option):
> > http://unixy.pl/maciek/download/kernel/2.6.27-rc2/img_0005.jpg
> > (hibernation fails)
> > 
> > Config, dmesg (normal booting):
> > http://unixy.pl/maciek/download/kernel/2.6.27-rc2/
> 
> Any chance to get this bisected? I realize that it's quite a few reboots, 
> and _maybe_ Jeff or Bartlomiej can just say "oh, duh, it's because of 
> Xyz", but the more likely thing is that without a bisect people will be 
> flailing around fairly blindly..

Looks like another incarnation of http://bugzilla.kernel.org/show_bug.cgi?id=11178

Maciej, can you try the patch at
http://bugzilla.kernel.org/attachment.cgi?id=17043&action=view
please?

Thanks,
Rafael

Re: [Linux 2.6.27-rc2] suspend to disk, libata problem

[Jesse Barnes]

On Thursday, August 7, 2008 3:52 pm Rafael J. Wysocki wrote:
> On Friday, 8 of August 2008, Linus Torvalds wrote:
> > On Fri, 8 Aug 2008, Maciej Rutecki wrote:
> > > I have problem with suspend to disk. On 2.6.26.* works fine.
> > >
> > > During s2disk i have this messages (with "no_console_suspend" kernel
> > > option): http://unixy.pl/maciek/download/kernel/2.6.27-rc2/img_0005.jpg
> > > (hibernation fails)
> > >
> > > Config, dmesg (normal booting):
> > > http://unixy.pl/maciek/download/kernel/2.6.27-rc2/
> >
> > Any chance to get this bisected? I realize that it's quite a few reboots,
> > and _maybe_ Jeff or Bartlomiej can just say "oh, duh, it's because of
> > Xyz", but the more likely thing is that without a bisect people will be
> > flailing around fairly blindly..
>
> Looks like another incarnation of
> http://bugzilla.kernel.org/show_bug.cgi?id=11178
>
> Maciej, can you try the patch at
> http://bugzilla.kernel.org/attachment.cgi?id=17043&action=view
> please?

I just pushed this fix into my for-linus tree.  I wanted it (along with the 
rest of my for-linus stuff) to get one day in linux-next before asking Linus 
to pull it though...

Jesse

Re: [Linux 2.6.27-rc2] suspend to disk, libata problem

[Maciej Rutecki]

2008/8/8 Rafael J. Wysocki <rjw@sisk.pl>:
> On Friday, 8 of August 2008, Linus Torvalds wrote:

>>
>> Any chance to get this bisected? I realize that it's quite a few reboots,
>> and _maybe_ Jeff or Bartlomiej can just say "oh, duh, it's because of
>> Xyz", but the more likely thing is that without a bisect people will be
>> flailing around fairly blindly..
>
> Looks like another incarnation of http://bugzilla.kernel.org/show_bug.cgi?id=11178
>
> Maciej, can you try the patch at
> http://bugzilla.kernel.org/attachment.cgi?id=17043&action=view
> please?
>
> Thanks,
> Rafael
>

Thanks
s2disk works, but I notice other problem during resume and normal
booting. Kernel very often stoped booting while detect PATA drive (DVD
drive):
http://unixy.pl/maciek/download/kernel/2.6.27-rc2/msi_path/img_0008.jpg

On 2.6.25.* stable it never happen.
On 2.6.26-rc* and 2.6.26.* stable -  sometimes
On 2.6.27-rc* - very often

So I try disable CONFIG_IDE and test s2disk again. Problem seems has
gone But a I have many warnings during resume (USB related?):
http://unixy.pl/maciek/download/kernel/2.6.27-rc2/without_config_ide/dmesg_resume_ahci.txt

Config kernel without CONFIG_IDE:
http://unixy.pl/maciek/download/kernel/2.6.27-rc2/without_config_ide/config-2.6.27-rc2


@Linus
If anybody tells me how do it, why not. I'm not kernel programmer (or
programmer in  generality). I'm only Linux user - newbie - who found
problem in kernel and think that better report it before stable 2.6.27
will be released than after.

Regards
-- 
Maciej Rutecki
http://www.maciek.unixy.pl

Re: [Linux 2.6.27-rc2] suspend to disk, libata problem

[Alan Jenkins]

Maciej Rutecki wrote:
> Thanks
> s2disk works, but I notice other problem during resume and normal
> booting. Kernel very often stoped booting while detect PATA drive (DVD
> drive):
> http://unixy.pl/maciek/download/kernel/2.6.27-rc2/msi_path/img_0008.jpg
> 
> On 2.6.25.* stable it never happen.
> On 2.6.26-rc* and 2.6.26.* stable -  sometimes
> On 2.6.27-rc* - very often
> 
> So I try disable CONFIG_IDE and test s2disk again. Problem seems has
> gone But a I have many warnings during resume (USB related?):
> http://unixy.pl/maciek/download/kernel/2.6.27-rc2/without_config_ide/dmesg_resume_ahci.txt

That looks like <http://bugzilla.kernel.org/show_bug.cgi?id=11263>.

Comment  #9 From Alan Stern 2008-08-07 07:30:17  [reply] -------  
This is a known problem caused by changes to the driver core.  I haven't had
time yet to update the USB suspend/resume infrastructure to match.

> Config kernel without CONFIG_IDE:
> http://unixy.pl/maciek/download/kernel/2.6.27-rc2/without_config_ide/config-2.6.27-rc2
> 
> 
> @Linus
> If anybody tells me how do it, why not. I'm not kernel programmer (or
> programmer in  generality). I'm only Linux user - newbie - who found
> problem in kernel and think that better report it before stable 2.6.27
> will be released than after.
> 
> Regards