[Openchrome-users] [openChrome] #309: P4M890 trunk regression: lock up on resolution changing (0.2.903 OK)

[OpenChrome Trac]

#309: P4M890 trunk regression: lock up on resolution changing (0.2.903 OK)
----------------------+-----------------------------------------------------
 Reporter:  Zajec     |        Owner:  somebody   
     Type:  defect    |       Status:  new        
 Priority:  critical  |    Component:  modesetting
  Version:  trunk     |   Resolution:             
 Keywords:            |     Blocking:  261        
Blockedby:            |  
----------------------+-----------------------------------------------------

Comment(by Zajec):

 Witam :)

 That's weird. Bug I experience is really simple one: changing resolution
 (by game or by xrandr) locks up card. End.

 In your case simple changing resolution with xrandr seems to work, but
 when Diablo does it (change resolution) it causes problem. My only guess
 is that after changing resolution your card is stable enought to display
 desktop, but something more advanced (like Diablo graphics) is not
 possible in that state. Does it make sense for you? Do you have other
 explaination?

 There are probably more tests we could make (for ex. starting X in 640x480
 and then start Diablo, so it doesn't change resolution - all without patch
 of course) but not sure if this is really important to understand that
 problem so well. Personally I can just accept: "set 0x1A or card won't
 work stable".

 If this little patch can fix so unpredictable bugs (on more than 1
 chipset) and shouldn't cause problems, I agree we should use this for all
 chipsets. Who knows what other bugs (hard to detect) can be triggered by
 ignoring 0x1A register.

-- 
Ticket URL: <http://www.openchrome.org/trac/ticket/309#comment:12>
openChrome <http://www.openchrome.org/>
The openChrome project