Regression 2.6.34->2.6.35-rc4: radeaon KMS an RS690 broken

[Torsten Kaiser]

On Thu, Jul 15, 2010 at 12:16 AM, Alex Deucher <alexdeucher at gmail.com> wrote:
> I discussed this with Jerome and I think we found the root cause.
> Does this patch help?
(patch 0001-drm-radeon-kms-fix-gtt-MC-base-alignment-on-rs4xx-rs.patch)

Yes:
[    0.426978] Linux agpgart interface v0.103
[    0.427141] [drm] Initialized drm 1.1.0 20060810
[    0.427242] [drm] radeon defaulting to kernel modesetting.
[    0.427300] [drm] radeon kernel modesetting enabled.
[    0.427417] radeon 0000:01:05.0: PCI INT A -> GSI 18 (level, low) -> IRQ 18
[    0.429732] [drm] initializing kernel modesetting (RS690 0x1002:0x791E).
[    0.429890] [drm] register mmio base: 0xFE9F0000
[    0.429948] [drm] register mmio size: 65536
[    0.430537] ATOM BIOS: ATI
[    0.430612] radeon 0000:01:05.0: VRAM: 32M 0xDE000000 - 0xDFFFFFFF (32M used)
[    0.430672] radeon 0000:01:05.0: GTT: 512M 0xA0000000 - 0xBFFFFFFF
[    0.430754] [drm] radeon: irq initialized.
[    0.430817] mtrr: type mismatch for fc000000,2000000 old:
write-back new: write-combining
[    0.430890] [drm] Detected VRAM RAM=32M, BAR=32M
[    0.430947] [drm] RAM width 128bits DDR
[    0.431090] [TTM] Zone  kernel: Available graphics memory: 2010234 kiB.
[    0.431149] [TTM] Initializing pool allocator.
[    0.431224] [drm] radeon: 32M of VRAM memory ready
[    0.431283] [drm] radeon: 512M of GTT memory ready.
[    0.431343] [drm] GART: num cpu pages 131072, num gpu pages 131072
[    0.434732] [drm] radeon: 1 quad pipes, 1 z pipes initialized.
[    0.441796] [drm] Loading RS690/RS740 Microcode
[    0.442006] [drm] radeon: ring at 0x00000000A0000000
[    0.442080] [drm] ring test succeeded in 1 usecs
[    0.442223] [drm] radeon: ib pool ready.
[    0.442289] [drm] ib test succeeded in 0 usecs
[    0.442370] [drm] Default TV standard: NTSC
[    0.442587] [drm] Default TV standard: NTSC
[    0.442866] [drm] Radeon Display Connectors
[    0.442924] [drm] Connector 0:
[    0.442981] [drm]   VGA
[    0.443039] [drm]   DDC: 0x7e50 0x7e40 0x7e54 0x7e44 0x7e58 0x7e48
0x7e5c 0x7e4c
[    0.443111] [drm]   Encoders:
[    0.443169] [drm]     CRT1: INTERNAL_KLDSCP_DAC1
[    0.443227] [drm] Connector 1:
[    0.443284] [drm]   S-video
[    0.443340] [drm]   Encoders:
[    0.443398] [drm]     TV1: INTERNAL_KLDSCP_DAC1
[    0.443455] [drm] Connector 2:
[    0.443512] [drm]   HDMI-A
[    0.443570] [drm]   DDC: 0x7e40 0x7e50 0x7e44 0x7e54 0x7e48 0x7e58
0x7e4c 0x7e5c
[    0.443642] [drm]   Encoders:
[    0.443700] [drm]     DFP3: INTERNAL_LVTM1
[    0.643372] [drm] fb mappable at 0xFC040000
[    0.643432] [drm] vram apper at 0xFC000000
[    0.643489] [drm] size 786432
[    0.643546] [drm] fb depth is 8
[    0.643603] [drm]    pitch is 1024
[    0.643742] fbcon: radeondrmfb (fb0) is primary device
[    0.663232] Console: switching to colour frame buffer device 128x48
[    0.664818] fb0: radeondrmfb frame buffer device
[    0.664837] drm: registered panic notifier
[    0.664856] Slow work thread pool: Starting up
[    0.664919] Slow work thread pool: Ready
[    0.664940] [drm] Initialized radeon 2.5.0 20080528 for
0000:01:05.0 on minor 0

Please note, that I'm only looking for "ring test succeeded" / "ring
test failed", as I'm only using the KMS fb console on that system.

Thanks for the fix,
                            Torsten