Bug 48422 - Radeon KMS fails on Radeon X850XT (R480) graphics card
Summary: Radeon KMS fails on Radeon X850XT (R480) graphics card
Status: RESOLVED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Radeon (show other bugs)
Version: XOrg 6.7.0
Hardware: x86-64 (AMD64) All
: medium normal
Assignee: Default DRI bug account
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2012-04-07 10:50 UTC by Mike B
Modified: 2012-04-18 15:13 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
Dmesg output from boot (123.04 KB, text/plain)
2012-04-07 10:50 UTC, Mike B
no flags Details
Dmesg output from boot (no drm debug) (75.27 KB, text/plain)
2012-04-07 17:07 UTC, Mike B
no flags Details

Description Mike B 2012-04-07 10:50:17 UTC
Created attachment 59628 [details]
Dmesg output from boot

Video output fails completely (apparently due to a KMS regression) on all kernels >= 3.2, rendering the system unusable as a desktop. Access via SSH over network is unaffected.

Steps to reproduce: Boot any recent kernel. This has been encountered with Debian, and the kernel versions in question are:-

linux-image-3.0.0-2-amd64         3.0.0-5               functions as expected
linux-image-3.2.0-1-amd64         3.2.7-1               fails
linux-image-3.2.0-2-amd64         3.2.13-1              fails
linux-image-3.3.0-trunk-amd64     3.3-1~experimental.1  fails


Expected behaviour:-

At system boot, initial kernel output appears on the VGA console, until the drm driver is loaded, when the kernel switches the output font and resolution. Boot continues until a subsequent switch (I believe when the kernel attempts to modeswitch the card), at which point the font visibly changes at this point and boot continues, with further console output until the X desktop appears.

Actual behaviour with failing kernels:-

Boot starts normally until the modeswitch happens, at which point video output ceases, and the monitor shows no signal on the DVI input.

Graphics hardware details from lspci is:-

01:00.0 VGA compatible controller: Advanced Micro Devices [AMD] nee ATI R480 [Radeon X850XT (PCIE)] (Primary)
01:00.1 Display controller: Advanced Micro Devices [AMD] nee ATI R480 [Radeon X850XT (PCIE)] (Secondary)

Attached file is dmesg output from booting the 3.3 kernel with drm.debug=0x6 passed as a kernel parameter.
Comment 1 Jonathan Nieder 2012-04-07 12:16:24 UTC
Context: http://bugs.debian.org/667799 which mentions that the system is still usable using ssh after the failed modeset.

Looks like the log with drm.debug=0x6 overflowed the kernel buffer so it's cut off at the beginning.  Does /var/log/dmesg have more?
Comment 2 Mike B 2012-04-07 17:07:13 UTC
Created attachment 59633 [details]
Dmesg output from boot (no drm debug)

Attached file is the dmesg output from booting 3.3 without the drm debug parameter set.
Comment 3 Alex Deucher 2012-04-09 05:41:43 UTC
Should be fixed with this patch:
http://lists.freedesktop.org/archives/dri-devel/2012-April/020963.html
Comment 4 Jonathan Nieder 2012-04-18 15:13:54 UTC
Mike Brodbelt wrote[1]:

> I have done basic testing. Pulling a kernel source tree for 3.3.y and
> applying the patch results in a new kernel which now boots
> successfully.
>
> Currently running kernel is :-
>
> $ uname -a
> Linux mordor 3.3.1+ #5 SMP Tue Apr 10 01:53:45 BST 2012 x86_64 GNU/Linux
>
> Which is working.... I've not tested this tree without the patch yes,
> so can't be 100% certain, but that patch is looking pretty convincing
> so far.

So it looks like this is fixed by

commit afceb9319f21b18ee3bc15ee9a5f92e18ef8a8c9
Author: Alex Deucher <alexander.deucher@amd.com>
Date:   Tue Apr 3 17:05:41 2012 -0400

    drm/radeon/kms: fix DVO setup on some r4xx chips
    
    Some r4xx chips have the wrong frev in the
    DVOEncoderControl table.  It should always be 1
    on r4xx.  Fixes modesetting on DVO on r4xx chips
    with the bad frev.
    
    Reported by twied on #radeon.
    
    Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
    Cc: stable@vger.kernel.org
    Signed-off-by: Dave Airlie <airlied@redhat.com>

which is part of v3.4-rc3.

[1] http://bugs.debian.org/667799#36


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.