Any SYSRQ key to force mode switch to console mode when using KMS?

Any SYSRQ key to force mode switch to console mode when using KMS?

[Uwe Bonnes]

Hello,

is there any SYSRQ key that forces the video system to console mode,
especially when using KMS and allows to work on. 

Since going to KMS on a Thinkpad R50 with ATI Radeon RV250 via going from
Opensuse-11.2 to Opensuse-11.3 I experience frequent situations where I am
left with a black screen, with only ALT-SYSREQ-C switching the screen back
on and hard reboot as only chance to escape that situation.

Situations that result in a black screen are:

* Going with the mouse over the NOSCRIPT buttom in firefox. Load goes up and
screen goes black after some seconds. xorg.log ends with "Backtrace", but no
backtrace output...

* Calling pm-hibernate on a system that has hibernated before. First
  pm-hibernate/wake-up-per power buttom sequence is successfull.

Disabling KMS results in very slow "kaffeine" full screen output...

pm-hibernate and firefox/noscript worked pre-KMS on opensuse-11.2.

Some traces of other people hitting the NOSCRIPT xorg crash and others
hitting the black unresponsive screen with KMS on RV250 are
found on the net, but I didn't find a conclusive resolution.

Opensuse default kernel and the recent Opensuse 2.6.36-rc3 kernel-of-the day
were tested with no difference

Thanks

PS: I only follow linux-kernel via gmane.org, so if you need fast feedback,
please CC me. 
-- 
Uwe Bonnes                bon@elektron.ikp.physik.tu-darmstadt.de

Institut fuer Kernphysik  Schlossgartenstrasse 9  64289 Darmstadt
--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

Re: Any SYSRQ key to force mode switch to console mode when using KMS?

[Randy Dunlap]

On Wed, 15 Sep 2010 19:40:37 +0200 Uwe Bonnes wrote:

> Hello,
> 
> is there any SYSRQ key that forces the video system to console mode,
> especially when using KMS and allows to work on. 

sysrq-v help message says:

force-fb(V) = "Restore framebuffer console"

> Since going to KMS on a Thinkpad R50 with ATI Radeon RV250 via going from
> Opensuse-11.2 to Opensuse-11.3 I experience frequent situations where I am
> left with a black screen, with only ALT-SYSREQ-C switching the screen back
> on and hard reboot as only chance to escape that situation.
> 
> Situations that result in a black screen are:
> 
> * Going with the mouse over the NOSCRIPT buttom in firefox. Load goes up and
> screen goes black after some seconds. xorg.log ends with "Backtrace", but no
> backtrace output...
> 
> * Calling pm-hibernate on a system that has hibernated before. First
>   pm-hibernate/wake-up-per power buttom sequence is successfull.
> 
> Disabling KMS results in very slow "kaffeine" full screen output...
> 
> pm-hibernate and firefox/noscript worked pre-KMS on opensuse-11.2.
> 
> Some traces of other people hitting the NOSCRIPT xorg crash and others
> hitting the black unresponsive screen with KMS on RV250 are
> found on the net, but I didn't find a conclusive resolution.
> 
> Opensuse default kernel and the recent Opensuse 2.6.36-rc3 kernel-of-the day
> were tested with no difference
> 
> Thanks
> 
> PS: I only follow linux-kernel via gmane.org, so if you need fast feedback,
> please CC me. 

---
~Randy
*** Remember to use Documentation/SubmitChecklist when testing your code ***

Re: Any SYSRQ key to force mode switch to console mode when using KMS?

[Uwe Bonnes]

>>>>> "Randy" == Randy Dunlap <randy.dunlap@oracle.com> writes:

    Randy> On Wed, 15 Sep 2010 19:40:37 +0200 Uwe Bonnes wrote:
    >> Hello,
    >> 
    >> is there any SYSRQ key that forces the video system to console mode,
    >> especially when using KMS and allows to work on.

    Randy> sysrq-v help message says:

    Randy> force-fb(V) = "Restore framebuffer console"

How is that documented outside the sysrq itself?

/usr/src/linux-2.6.36-rc4-19/Documentation/sysrq.txt:

'v'     - Dumps Voyager SMP processor info to your console

Bye
-- 
Uwe Bonnes                bon@elektron.ikp.physik.tu-darmstadt.de

Institut fuer Kernphysik  Schlossgartenstrasse 9  64289 Darmstadt
--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

Re: Any SYSRQ key to force mode switch to console mode when using KMS?

[Randy Dunlap]

On 09/22/10 02:44, Uwe Bonnes wrote:
>>>>>> "Randy" == Randy Dunlap <randy.dunlap@oracle.com> writes:
> 
>     Randy> On Wed, 15 Sep 2010 19:40:37 +0200 Uwe Bonnes wrote:
>     >> Hello,
>     >> 
>     >> is there any SYSRQ key that forces the video system to console mode,
>     >> especially when using KMS and allows to work on.
> 
>     Randy> sysrq-v help message says:
> 
>     Randy> force-fb(V) = "Restore framebuffer console"
> 
> How is that documented outside the sysrq itself?
> 
> /usr/src/linux-2.6.36-rc4-19/Documentation/sysrq.txt:
> 
> 'v'     - Dumps Voyager SMP processor info to your console

Yes, I noticed that and I posted a patch for it yesterday.

-- 
~Randy
*** Remember to use Documentation/SubmitChecklist when testing your code ***

Re: Any SYSRQ key to force mode switch to console mode when using KMS?

[Uwe Bonnes]

>>>>> "Randy" == Randy Dunlap <randy.dunlap@oracle.com> writes:

    Randy> On 09/22/10 02:44, Uwe Bonnes wrote:
    >> >> is there any SYSRQ key that forces the video system to console
    >> mode, >> especially when using KMS and allows to work on.
    >> 
    Randy> sysrq-v help message says:
    >>
    Randy> force-fb(V) = "Restore framebuffer console"
    >>  How is that documented outside the sysrq itself?
    >> 
    >> /usr/src/linux-2.6.36-rc4-19/Documentation/sysrq.txt:
    >> 
    >> 'v' - Dumps Voyager SMP processor info to your console

    Randy> Yes, I noticed that and I posted a patch for it yesterday.

Thanks for caring!

However 2.6.36-rc4 also fixed my hibernate problem:

commit 84e1d836ef0759a152578a961894824bde89596f
Merge: 20f4cad 6715045
Author: Linus Torvalds <torvalds@linux-foundation.org>
Date:   Sat Sep 11 15:50:53 2010 -0700

    Merge branch 'pm-fixes' of
    git://git.kernel.org/pub/scm/linux/kernel/git/rafael/suspend-2.6
    
    * 'pm-fixes' of
      ...
      PM / Hibernate: Avoid hitting OOM during preallocation of memory
      ...

so I no longer hit the problem with the dark screen...

Bye
-- 
Uwe Bonnes                bon@elektron.ikp.physik.tu-darmstadt.de

Institut fuer Kernphysik  Schlossgartenstrasse 9  64289 Darmstadt
--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

Re: Any SYSRQ key to force mode switch to console mode when using KMS?

[Uwe Bonnes]

>>>>> "Uwe" == Uwe Bonnes <bon@elektron.ikp.physik.tu-darmstadt.de> writes:

>>>>> "Randy" == Randy Dunlap <randy.dunlap@oracle.com> writes:
...
    Uwe> However 2.6.36-rc4 also fixed my hibernate problem:

    Uwe> commit 84e1d836ef0759a152578a961894824bde89596f Merge: 20f4cad
    Uwe> 6715045 Author: Linus Torvalds <torvalds@linux-foundation.org>
    Uwe> Date: Sat Sep 11 15:50:53 2010 -0700

    Uwe>     Merge branch 'pm-fixes' of
    Uwe> git://git.kernel.org/pub/scm/linux/kernel/git/rafael/suspend-2.6
    
    Uwe>     * 'pm-fixes' of ...  PM / Hibernate: Avoid hitting OOM during
    Uwe> preallocation of memory ...

    Uwe> so I no longer hit the problem with the dark screen...

I burried the problem too early. 

On another hibernate attempt on 2.6.36-rc4, now the image again was not
written to disk, the screen turn black and no ALT-CTL-SYSRQ combination
helped. Especially I tried  ALT-CTL-SYSRQ ad 'v','V', ALT-CTL 'v' and
ALT-CTL 'V'...
 
-- 
Uwe Bonnes                bon@elektron.ikp.physik.tu-darmstadt.de

Institut fuer Kernphysik  Schlossgartenstrasse 9  64289 Darmstadt
--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

Re: Any SYSRQ key to force mode switch to console mode when using KMS?

[Randy Dunlap]

On Thu, 23 Sep 2010 15:01:39 +0200 Uwe Bonnes wrote:

> >>>>> "Uwe" == Uwe Bonnes <bon@elektron.ikp.physik.tu-darmstadt.de> writes:
> 
> >>>>> "Randy" == Randy Dunlap <randy.dunlap@oracle.com> writes:
> ...
>     Uwe> However 2.6.36-rc4 also fixed my hibernate problem:
> 
>     Uwe> commit 84e1d836ef0759a152578a961894824bde89596f Merge: 20f4cad
>     Uwe> 6715045 Author: Linus Torvalds <torvalds@linux-foundation.org>
>     Uwe> Date: Sat Sep 11 15:50:53 2010 -0700
> 
>     Uwe>     Merge branch 'pm-fixes' of
>     Uwe> git://git.kernel.org/pub/scm/linux/kernel/git/rafael/suspend-2.6
>     
>     Uwe>     * 'pm-fixes' of ...  PM / Hibernate: Avoid hitting OOM during
>     Uwe> preallocation of memory ...
> 
>     Uwe> so I no longer hit the problem with the dark screen...
> 
> I burried the problem too early. 
> 
> On another hibernate attempt on 2.6.36-rc4, now the image again was not
> written to disk, the screen turn black and no ALT-CTL-SYSRQ combination
> helped. Especially I tried  ALT-CTL-SYSRQ ad 'v','V', ALT-CTL 'v' and
> ALT-CTL 'V'...

Is CONFIG_DRM_KMS_HELPER enabled in your kernel .config file?
If not, we need to find out how to enable it (there is no kconfig
prompt for it).

Can you check if Alt-Sysrq-H (help) lists V for "force-fb(V)"?

---
~Randy
*** Remember to use Documentation/SubmitChecklist when testing your code ***

Re: Any SYSRQ key to force mode switch to console mode when using KMS?

[Uwe Bonnes]

>>>>> "Randy" == Randy Dunlap <rdunlap@xenotime.net> writes:


    Randy> Is CONFIG_DRM_KMS_HELPER enabled in your kernel .config file?  If
    Randy> not, we need to find out how to enable it (there is no kconfig
    Randy> prompt for it).

Yes, it is enabled

    Randy> Can you check if Alt-Sysrq-H (help) lists V for "force-fb(V)"?
Yes, it is listed.

And yes, my last experiments where with ALT-CTL-sysreq, so doomed for non
function due to operator error.

However:
When running X, X.Org X Server 1.9.0 on Thinkpad R50 RV250, Alt-Sysrq-v
switches to an all-black display. By console switching with ALT-CLT-F1 I
get back to the booting console and ALT-CLT-F7 brings me back to X.  Running
on the console and calling Alt-Sysrq-v prints
 SysRq : Restore framebuffer console

Is there some omission by the distribution or my setup that Alt-Sysrq-v when
running X results in an all-black screen?

And when I hit the hibernate problem, the screen stays black even with a
console switch after Alt-Sysrq-v

Bye
-- 
Uwe Bonnes                bon@elektron.ikp.physik.tu-darmstadt.de

Institut fuer Kernphysik  Schlossgartenstrasse 9  64289 Darmstadt
--------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

Re: Any SYSRQ key to force mode switch to console mode when using KMS?

[Randy Dunlap]

On Thu, 23 Sep 2010 18:31:32 +0200 Uwe Bonnes wrote:

[just adding cc's]

> >>>>> "Randy" == Randy Dunlap <rdunlap@xenotime.net> writes:
> 
> 
>     Randy> Is CONFIG_DRM_KMS_HELPER enabled in your kernel .config file?  If
>     Randy> not, we need to find out how to enable it (there is no kconfig
>     Randy> prompt for it).
> 
> Yes, it is enabled
> 
>     Randy> Can you check if Alt-Sysrq-H (help) lists V for "force-fb(V)"?
> Yes, it is listed.
> 
> And yes, my last experiments where with ALT-CTL-sysreq, so doomed for non
> function due to operator error.
> 
> However:
> When running X, X.Org X Server 1.9.0 on Thinkpad R50 RV250, Alt-Sysrq-v
> switches to an all-black display. By console switching with ALT-CLT-F1 I
> get back to the booting console and ALT-CLT-F7 brings me back to X.  Running
> on the console and calling Alt-Sysrq-v prints
>  SysRq : Restore framebuffer console
> 
> Is there some omission by the distribution or my setup that Alt-Sysrq-v when
> running X results in an all-black screen?
> 
> And when I hit the hibernate problem, the screen stays black even with a
> console switch after Alt-Sysrq-v
> 
> Bye
> -- 
> Uwe Bonnes                bon@elektron.ikp.physik.tu-darmstadt.de
> 
> Institut fuer Kernphysik  Schlossgartenstrasse 9  64289 Darmstadt
> --------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------


---
~Randy
*** Remember to use Documentation/SubmitChecklist when testing your code ***

Re: Any SYSRQ key to force mode switch to console mode when using KMS?

[Jason Wessel]

On 09/24/2010 10:07 AM, Randy Dunlap wrote:
> On Thu, 23 Sep 2010 18:31:32 +0200 Uwe Bonnes wrote:
>
> [just adding cc's]
>
>   
>>>>>>> "Randy" == Randy Dunlap <rdunlap@xenotime.net> writes:
>>>>>>>               
>>     Randy> Is CONFIG_DRM_KMS_HELPER enabled in your kernel .config file?  If
>>     Randy> not, we need to find out how to enable it (there is no kconfig
>>     Randy> prompt for it).
>>
>> Yes, it is enabled
>>
>>     Randy> Can you check if Alt-Sysrq-H (help) lists V for "force-fb(V)"?
>> Yes, it is listed.
>>
>> And yes, my last experiments where with ALT-CTL-sysreq, so doomed for non
>> function due to operator error.
>>
>> However:
>> When running X, X.Org X Server 1.9.0 on Thinkpad R50 RV250, Alt-Sysrq-v
>> switches to an all-black display. By console switching with ALT-CLT-F1 I
>> get back to the booting console and ALT-CLT-F7 brings me back to X.  Running
>> on the console and calling Alt-Sysrq-v prints
>>  SysRq : Restore framebuffer console
>>
>> Is there some omission by the distribution or my setup that Alt-Sysrq-v when
>> running X results in an all-black screen?
>>
>> And when I hit the hibernate problem, the screen stays black even with a
>> console switch after Alt-Sysrq-v
>>
>>     
I am not exactly sure what you are looking for but alt-sysrq-v might not
tell you a whole lot because an escape back to the text console that was
not actively receiving printk's or crash data is going to in general
provide you with a black screen.

If you were interested in something a bit more invasive such as a
collecting a process listing or examining the dmesg output you might
consider jumping into kdb.  

echo kbd > /sys/module/kgdboc/parameters/kgdboc

Then you would do the "alt-sysrq v" followed by "alt-sysrq g".  From
there you could then run dmesg, and your screen would no longer be blank
in theory.  :-)

We do have some patches for the RV250 for atomic kernel mode setting
plus kdb support but they are not merged to the mainline kernels as of yet.


Jason.