Bug 11146

Summary: randr-1.2 does not work with an ATI Radeon Mobility 9200 (M9+) 5C63 (AGP) on powerpc
Product: xorg Reporter: Gallien Matthieu <matthieu_gallien>
Component: Driver/RadeonAssignee: xf86-video-ati maintainers <xorg-driver-ati>
Status: RESOLVED FIXED QA Contact: Xorg Project Team <xorg-team>
Severity: normal    
Priority: medium CC: matthieu_gallien
Version: git   
Hardware: PowerPC   
OS: Linux (All)   
Whiteboard:
i915 platform: i915 features:
Attachments:
Description Flags
server log with a working version using mergedfb and with the external screen plugged on
none
server log with a driver compiled from gir branch randr-1.2
none
In case it is usefull, server log with the same setup than the working one with mergedfb and only the internal screen
none
a log with the last git master version
none
my xorg.conf none

Description Gallien Matthieu 2007-06-05 02:04:08 UTC
Hi,
Thanks for all your work.
I just give a try to the branch of the driver with support of randr-1.2.
It does not work on my iBook. I have a debian unstable with xserver-xorg 1.3.
Exactly, when I start the xserver with no external display connected, the display is messed up. It seems like some vertical lines are not at their right place. If you want, I can send you a picture made with my phone.

Can I do anything usefull to help you ?
Feel free to ask me any informations that can be usefull.

Best regards

PS: sorry for my English and thanks for all your great work.
Comment 1 Dave Airlie 2007-06-05 02:51:56 UTC
can you attach an Xorg.0.log? and one from a working master driver as well?
Comment 2 Gallien Matthieu 2007-06-05 06:34:00 UTC
Created attachment 10193 [details]
server log with a working version using mergedfb and with the external screen plugged on
Comment 3 Gallien Matthieu 2007-06-05 06:34:59 UTC
Created attachment 10194 [details]
server log with a driver compiled from gir branch randr-1.2
Comment 4 Gallien Matthieu 2007-06-05 06:36:33 UTC
Created attachment 10195 [details]
In case it is usefull, server log with the same setup than the working one with mergedfb and only the internal screen
Comment 5 Erik Andren 2007-06-10 11:41:31 UTC
Please attach (not pasting) your xorg.conf to this bug.
Comment 6 Gallien Matthieu 2007-06-13 04:12:12 UTC
Created attachment 10294 [details]
a log with the last git master version
Comment 7 Gallien Matthieu 2007-06-13 04:13:01 UTC
Created attachment 10295 [details]
my xorg.conf
Comment 8 Andrew Randrianasulu 2007-06-21 17:39:46 UTC
....i'm on x86, and my video card is:

01:00.0 VGA compatible controller: ATI Technologies Inc RV280 [Radeon 9200 SE] (rev 01) (prog-if 00 [VGA])

i'm using randr-1.2 aranch of ati driver (up to and include ec1fd2c388ec2c4033644776ee8588405a3d25f1)

i have strange screen corruption after server startup. But only if i have also ColorTiling enabled (default, as i can see from logs). If I disable it - i get normal picture (but slower 3d). 


And should i fill another bug about non-working VT switch?  (again, but here is more serious problem - i can't see anything on text consoles, only blank screen. but Xserver itself working, i can return to graphics mode.)
Comment 9 Andrew Randrianasulu 2007-06-21 20:35:02 UTC
....issue with not restored text mode resolved!

in radeon_driver.c , in function RADEONRestore i enable again call to RADEONUnblank(pScrn); [around line 5103] and all working again! But i have only one monitor (VGA connector, CRT monitor) - may be this 'fix' broke things for another people?
Comment 10 Gallien Matthieu 2007-06-22 07:31:52 UTC
I got a correct display on the LVDS if I disable ColorTiling AND the hardware cursor, otherwise, with commit 3bb46c03fddd3bd79bdab887366aeeced0290a3a, I got the messed display.
Comment 11 Alex Deucher 2007-06-25 12:55:21 UTC
Should be fixed now:
52cc1dc1491559a9055f3ba6dd54064bb382ad86

Default size is 1600x1200; use a virtual line in your config to make it bigger.

cursor and depth 16 palette set up still need fixes for big endian.
Comment 12 Alex Deucher 2007-06-25 15:42:20 UTC
these should be resolved now.
Comment 13 Gallien Matthieu 2007-07-02 09:02:37 UTC
Hi,
I have tested the driver today with git commits up to bdcae62...
The problem is still there.
You can ask for informations or testing or anything else that does not require too much time.

Thanks in advance
Best regards
Comment 14 Alex Deucher 2007-07-02 09:17:30 UTC
(In reply to comment #13)
> Hi,
> I have tested the driver today with git commits up to bdcae62...
> The problem is still there.

Which problem?  The stride issues or the HW cursor issues?  Depth 16 is still broken and HW cursors are still broken on ppc.  The stride issues should be fixed however.  What size desktop are you using?
Comment 15 Gallien Matthieu 2007-07-02 09:34:27 UTC
(In reply to comment #14)
> (In reply to comment #13)
> > Hi,
> > I have tested the driver today with git commits up to bdcae62...
> > The problem is still there.
> 
> Which problem?  The stride issues or the HW cursor issues?  Depth 16 is still
> broken and HW cursors are still broken on ppc.  The stride issues should be
> fixed however.  What size desktop are you using?
> 
I have checked and I am using depth 24 with a desktop of 1024x768 according to the xorg log. It is the native display of my iBook.
I am still experiencing the stride issue.

PS : sorry for the delay since you closed the bug, but I had no time to test before today.
Comment 16 Alex Deucher 2007-07-02 10:05:32 UTC
(In reply to comment #15)
> I have checked and I am using depth 24 with a desktop of 1024x768 according to
> the xorg log. It is the native display of my iBook.
> I am still experiencing the stride issue.
> 

1024x768 is the size of the mode, but the max desktop size is different.  the default is 1600x1200 unless you add a virtual line to your config.  Does changing the virtual size help?  try 1024 or 1664 for the virtual width:
e.g.,

Virtual 1024 768
Virtual 1664 1200
Comment 17 Michel Dänzer 2007-08-01 10:36:45 UTC
See bug 11796 for the HW cursor issue.
Comment 18 Gallien Matthieu 2007-08-01 12:38:56 UTC
The software Tiling bug is now closed (see https://bugs.freedesktop.org/show_bug.cgi?id=11357 for reference) and since bug 11796 is taking care of the hardware cursor, I am closing the bug.

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.