Re: Linux 2.4.20-rc1

[Carl-Daniel Hailfinger <c-d.hailfinger.kernel.2002-Q4@gmx.net>]

[-- Attachment #1: Type: text/plain, Size: 1481 bytes --]

Carl-Daniel Hailfinger wrote:
> [CC: linux-fbdev-devel to find out what was wrong with the patch]
> Marcelo Tosatti wrote:
>> On Sat, 2 Nov 2002, Carl-Daniel Hailfinger wrote:
>>> Marcelo Tosatti wrote:
>>>
>>>> Hi,
>>>>
>>>> Finally, rc1.
>>>> [snipped]
>>>>
>>>> Please stress test it.
>>>>
>>>
>>> My system comes up with a blank console after hardware suspend and 
>>> resume.
>>> The cursor is still visible, but no text is there. Switching to another
>>> console and back fixes it. Vesafb is enabled with vga=791.
>>> Hardware is a Toshiba Satellite 4100XCDT notebook with Trident 
>>> Cyber9525DVD
>>> graphics chipset, but this also can be reproduced with Dell notebooks.
>>>
>>> I just verified the problem exists still with 2.4.20-rc1.
>>> A binary search turned up 2.4.18-pre7 as the kernel which broke,
>>> specifically the changes made to apm.c back then.
>>
>>
>>
>> Have you tried to revert 2.4.18-pre7's changes to apm.c to make sure 
>> it is
>> the cause?>
> 
> 
> I tried it and found out that my results were incorrect. The problem was 
> introduced in 2.4.18-pre1 by the changes to drivers/video/fbcon.c
> 
> Reverting the attached patch fixes my problem. However, I am not exactly 
> sure why a patch intended to fix a PM problem introduced another one.

I managed to trim down the offending patch further. Reverting the new 
attached patch is enough to fix my problem. Can someone of the framebuffer 
experts please comment on this one?

Regards
Carl-Daniel

[-- Attachment #2: patch-fbdev-5 --]
[-- Type: application/x-java-vm, Size: 450 bytes --]