Bug 26821

Summary:	x1300 kms failure
Product:	xorg	Reporter:	kowalski marcin <yoshi314>
Component:	Driver/Radeon	Assignee:	xf86-video-ati maintainers <xorg-driver-ati>
Status:	RESOLVED INVALID	QA Contact:	Xorg Project Team <xorg-team>
Severity:	normal
Priority:	medium
Version:	7.4 (2008.09)
Hardware:	x86 (IA32)
OS:	Linux (All)
Whiteboard:
i915 platform:		i915 features:

Description kowalski marcin 2010-03-01 07:55:08 UTC

there are multiple problems rolled into one. 

to sum it up:

booting 2.6.32.* and 2.6.33 kernel with radeon.modeset=1 causes it to stall for ~33 seconds. finally kms kicks in . on 2.3.32 it set an incorrect mode (too high refresh rate), on 2.6.33 it works correctly. 

2.6.31 was the last properly working kernel where kms had no delays.

however, i found in dmesg that delay was (probably) caused by missing firmware, namely radeon/R520_cp.bin . 

i included that in my initrd and when booting with KMS display goes entirely white. and it seems to lock up completely.

lspci:
05:00.0 VGA compatible controller: ATI Technologies Inc RV515 PRO [Radeon X1300/X1550 Series]
05:00.1 Display controller: ATI Technologies Inc RV515 PRO [Radeon X1300/X1550 Series] (Secondary)

right now i'm using daewoo cmc-1427x old crt display. maybe that's the source of the problem?

i'll provide necessary logs, if needed.

Comment 1 kowalski marcin 2010-03-01 07:57:16 UTC

so, currently to get kms working properly i need to 
- boot with radeon.modeset=1 (or set it in kernel config)
- exclude radeon firmware from initrd
- put up with the delay

if the firmware loads later on (and logs show it does) there are no issues.

Comment 2 kowalski marcin 2010-03-09 09:16:19 UTC

as suggested by jglisse on #radeon i tried to modprobe radeon modeset=1 and try to access dmesg after that.

unfortunately system completely locks up and ssh session dies. with or without firmware in place. i can see vertical black and white stripes, and few seconds after screen becomes white and locks up.

fedora 13 alpha kernel *works* and KMS starts correctly , although there is a non-fatal segfault after requesting firmware somewhere in dmesg, maybe it doesn't load at all?. 

also, there is no ~30second delay. X also starts up (i assume that's what graphical installer uses).

i'm attaching dmesg + dmidecode from fedora13 live media.

http://pastebin.ca/1830262 - dmesg
http://pastebin.ca/1830263 - dmidecode (maybe it'l be useful)

Comment 3 kowalski marcin 2010-03-11 00:13:52 UTC

ok i discovered that this boils down to a dodgy d-sub connector on my oooold crt.

still, old x.org mode setting was way more robust than KMS - at least i didn't get gpu lockups like now.

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.