DisplayPort and radeon

[89c51]

I have my monitor connected to a 6850 through DisplayPort.

On occasion i end up with a blank screen. I am quite sure this is due to the radeon module because it happens when it loads.

Anything i can do or any info you might need?

[89c51]

Sorry to bump but could anyone tell me if there is a solution or how to file a bug on that (or if its needed to file one). Its not reproduceable and seems to happen once in a while.

[89c51]

dmesg | grep radeon
[    0.739761] [drm] radeon defaulting to kernel modesetting.
[    0.739763] [drm] radeon kernel modesetting enabled.
[    0.739783] fb: conflicting fb hw usage radeondrmfb vs EFI VGA - removing generic driver
[    0.740452] radeon 0000:01:00.0: VRAM: 1024M 0x0000000000000000 - 0x000000003FFFFFFF (1024M used)
[    0.740454] radeon 0000:01:00.0: GTT: 512M 0x0000000040000000 - 0x000000005FFFFFFF
[    0.740505] [drm] radeon: 1024M of VRAM memory ready
[    0.740506] [drm] radeon: 512M of GTT memory ready.
[    0.740547] radeon 0000:01:00.0: irq 42 for MSI/MSI-X
[    0.740553] radeon 0000:01:00.0: radeon: using MSI.
[    0.740586] [drm] radeon: irq initialized.
[    0.740803] [drm] radeon: ib pool ready.
[    0.744288] radeon 0000:01:00.0: WB enabled
[    0.763028] [drm] radeon: power management initialized
[    0.828106] fbcon: radeondrmfb (fb0) is primary device
[    1.244321] [drm:radeon_dp_link_train_cr] *ERROR* clock recovery tried 5 times
[    1.244322] [drm:radeon_dp_link_train_cr] *ERROR* clock recovery failed
[    1.249800] fb0: radeondrmfb frame buffer device
[    1.249804] [drm] Initialized radeon 2.13.0 20080528 for 0000:01:00.0 on minor 0
[  247.395379] [drm:radeon_dp_get_link_status] *ERROR* displayport link status failed
[  247.395381] [drm:radeon_dp_link_train_cr] *ERROR* clock recovery failed
[  247.397604] [drm:radeon_dp_get_link_status] *ERROR* displayport link status failed
[  247.397605] [drm:radeon_dp_link_train_cr] *ERROR* clock recovery failed

any ideas ???

[Wilfred]

As someone who has done helpdesk are you sure the cables are plugged in correctly?

[89c51]

I unlpugged and plugged the cable after it happened the first time so yes. I assume it wouldn't be happening randomly in case it was just a loose cable.

As i said before it happens when the module loads. I get the normal bios screen and when the kernel loads it goes blank.

I updated the kernel today to 3.4.2 so i ll see how this goes.

[89c51]

Because i know there are radeon people around and they might be interested the error continues in the 3.4.x kernel but its less frequent.

 dmesg | grep radeon
[    0.575232] [drm] radeon defaulting to kernel modesetting.
[    0.575233] [drm] radeon kernel modesetting enabled.
[    0.575253] fb: conflicting fb hw usage radeondrmfb vs EFI VGA - removing generic driver
[    0.575943] radeon 0000:01:00.0: VRAM: 1024M 0x0000000000000000 - 0x000000003FFFFFFF (1024M used)
[    0.575945] radeon 0000:01:00.0: GTT: 512M 0x0000000040000000 - 0x000000005FFFFFFF
[    0.575993] [drm] radeon: 1024M of VRAM memory ready
[    0.575994] [drm] radeon: 512M of GTT memory ready.
[    0.576037] radeon 0000:01:00.0: irq 42 for MSI/MSI-X
[    0.576043] radeon 0000:01:00.0: radeon: using MSI.
[    0.576077] [drm] radeon: irq initialized.
[    0.576295] [drm] radeon: ib pool ready.
[    0.580220] radeon 0000:01:00.0: WB enabled
[    0.598839] [drm] radeon: power management initialized
[    0.660999] fbcon: radeondrmfb (fb0) is primary device
[    1.074768] [drm:radeon_dp_link_train_cr] *ERROR* clock recovery tried 5 times
[    1.074769] [drm:radeon_dp_link_train_cr] *ERROR* clock recovery failed
[    1.080283] fb0: radeondrmfb frame buffer device
[    1.080286] [drm] Initialized radeon 2.16.0 20080528 for 0000:01:00.0 on minor 0

[chithanh]

I would check if maybe something goes wrong during fb handover, by disabling efifb and testing whether the problem persists.

[89c51]

I would check if maybe something goes wrong during fb handover, by disabling efifb and testing whether the problem persists.

The problem is that it doesn't happen all the time. With 3.3.x kernel it happened once every 5-6 boots with 3.4.x happened only twice in the last 30 days.

If i switch the cable to DVI it works.

[chithanh]

I understand that, but it doesn't exclude fb handover being the problem.