Marking this issue as resolved as the regression is gone and the X.org issue is 
a separate issue that will be resolved as part of http://bts.grml.org/grml/
issue851

Thanks for feedback!

regards,
-mika-

Michael Prokop wrote...

> Michael Prokop <mika@grml.org> added the comment:
> 
> * Christoph Biedl wrote in grml's BTS on 20101028 / 01:16:
> > Michael Prokop wrote...
> 
> > Update: Using grmldailysid as of 101025, some parts of the problem
> > seem to be gone: USB devices are recognized, sound modules are loaded
> > without tons of error messages in the logfile.

Confirmed: This issue has been resolved.

> > The bad news: grml-x seems to make the system unusable. After loading
> > the window manager the system freezes, nothing can be done besides
> > Rebooting via magic sysrq.
> 
> nvidia hardware?

No, Intel:

00:02.0 VGA compatible controller: Intel Corporation 82945G/GZ Integrated Graphics Controller (rev 02) (prog-if 00 [VGA controller])
        Subsystem: Intel Corporation Device 464c
        Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
        Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
        Latency: 0
        Interrupt: pin A routed to IRQ 11
        Region 0: Memory at 88200000 (32-bit, non-prefetchable) [size=512K]
        Region 1: I/O ports at 20e0 [size=8]
        Region 2: Memory at 80000000 (32-bit, prefetchable) [size=128M]
        Region 3: Memory at 88280000 (32-bit, non-prefetchable) [size=128K]
        Capabilities: [90] Message Signalled Interrupts: Mask- 64bit- Queue=0/0 Enable-
                Address: 00000000  Data: 0000
        Capabilities: [d0] Power Management version 2
                Flags: PMEClk- DSI+ D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
                Status: D0 PME-Enable- DSel=0 DScale=0 PME-

> Does startup without xorg.conf? Start e.g. via:
> 
>   grml-x -nostart fluxbox
>   sudo rm -f /etc/X11/xorg.conf
>   /usr/bin/startx

This does the trick. There are minor issues after switching back to
console: Colors flipped, horizontral position on screen shifted.
Nothing to worry about.

    Christoph

* Christoph Biedl wrote in grml's BTS on 20101028 / 01:16:
> Michael Prokop wrote...

> > I'll keep this bugreport open for as long as we don't have new kernel versions 
> > covering this issue so we can track this both for users as well as for 
> > development.

> Update: Using grmldailysid as of 101025, some parts of the problem
> seem to be gone: USB devices are recognized, sound modules are loaded
> without tons of error messages in the logfile.

Thanks for the update, Christoph.

> The bad news: grml-x seems to make the system unusable. After loading
> the window manager the system freezes, nothing can be done besides
> Rebooting via magic sysrq.

nvidia hardware?
Does startup without xorg.conf? Start e.g. via:

  grml-x -nostart fluxbox
  sudo rm -f /etc/X11/xorg.conf
  /usr/bin/startx

> This is somewhat preliminary, I'll do further checks and investigation
> the next days.

Thanks.

regards,
-mika-

Michael Prokop wrote...

> I'll keep this bugreport open for as long as we don't have new kernel versions 
> covering this issue so we can track this both for users as well as for 
> development.

Update: Using grmldailysid as of 101025, some parts of the problem
seem to be gone: USB devices are recognized, sound modules are loaded
without tons of error messages in the logfile.

The bad news: grml-x seems to make the system unusable. After loading
the window manager the system freezes, nothing can be done besides
Rebooting via magic sysrq.

This is somewhat preliminary, I'll do further checks and investigation
the next days.

    Christoph

Hi,

thanks a lot for the detailed information and debugging this issue, Christoph. 
I'll keep this bugreport open for as long as we don't have new kernel versions 
covering this issue so we can track this both for users as well as for 
development. Tagging the issue with "release-stopper" so we make sure our next 
release provides an according bugfix.

thanks && regards,
-mika-

Some additional information, I was a bit too fast sending the previous
message:

Christoph Biedl wrote...

> Providing "agp=off" works around the problem, so does "toram" (tested
> for -small only).

* "agp=off" seems to make X unusable for missing /dev/agpart,
  surprise.  I didn't search for workarounds.
* "toram" works for full-size grml, too, and is also a workaround.-
  Having a lot of RAM is probably a good idea for that.
  However, an USB device like a mouse and sound are unusable.

Matches the expected behaviour :)

    Christoph

Michael Prokop wrote...

> any news regarding this issue from your side?

Thanks for asking.  Since today I have a temporary replacment for the
box so I could do some tests. 

Booting from an USB stick yields strange results, rather late in the
boot process, mostly input/output errors, leaving grml unusable.

Providing "agp=off" works around the problem, so does "toram" (tested
for -small only).

Verified for grml(32)-small and grml64 (full).

    Christoph

Hi Christoph,

any news regarding this issue from your side?

regards,
-mika-

Christoph Biedl wrote...

> grml 2010.04-rc1 (tested: grml64-small via findiso) is affected by the
> regression I have reported in
> https://bugzilla.kernel.org/show_bug.cgi?id=15838
> a few minutes ago.  This means USB and sound are not usable.

I really hate to tell, but:  When I was about to try out your
recommendation for the kernel parameter "agp=off", I was unable to
reproduce that problem at all.  Unfortunately the setup was not
exactly the same, perhaps this had some influence although I cannot
believe this (Booting from sata disk -> booting from CF in pata port;
using grml64-small -> using grml64-medium), or this bus exposes only
after a cold boot.

Unfortunately/2 this computer is in constant use so I don't have much
time for testing.  Will try again tomorrow night.

    Christoph

Fix typo: Name of mainboard was misspelled.

Hi,

to avoid duplications this is rather brief:

grml 2010.04-rc1 (tested: grml64-small via findiso) is affected by the
regression I have reported in
https://bugzilla.kernel.org/show_bug.cgi?id=15838
a few minutes ago.  This means USB and sound are not usable.

Without USB, many of the forensic use cases of grml will be impaired;
and very likely grml will be unable to boot from USB on that hardware.

Reverting the mentioned commit might be a solution but could introduce
other problems I am not aware of.  If you prefer not to patch the
kernel, please drop a warning line in the upcoming release notes.

Thanks,

    Christoph