I have a Dell OptiPlex 380 box, this has the following videocard: 01:00.0 VGA compatible controller [0300]: Advanced Micro Devices [AMD] nee ATI RV710 [Radeon HD 4350/4550] [1002:954f] Two monitors attached, one with 1920x1080 resolution, the other with 1280x1024. With kernel 3.9.11 this card works perfectly (with included external firmware: radeon/R700_rlc.bin). Automatically changes to graphics mode without kernel parameters before root filesystem comes up, works with X, and can switch from X to console and back without problems. This kernel boots up to console login without the firmware too, but corruption can be seen on non-used area on the larger screen. With almost the same setup, 3.10.7 hangs the hardware where modesetting should occur, no kernel panic. The card stays in text mode, and the last written text before mode setting can be seen. The same happens without the firmware included into the binary. With almost the same setup, 3.11-rc7 hangs too, but a bit different way, the hang occurs a bit later, and the screens turn off a few seconds after the hang. The same happens with or without the firmware included. I'll attach configs and hang screens to the report.
Created attachment 107317 [details] hang screen with 3.10.9, both with and without firmware
Created attachment 107318 [details] hang screen with 3.11-rc7, before screens turn to blank, both with and without firmware
Created attachment 107319 [details] Working config for 3.9.11
Created attachment 107320 [details] Config for 3.10.9, this one with firmware included
Created attachment 107321 [details] Config for 3.11-rc7, this one without firmware
One more thing, I don't know if it's relevant or not, all three kernels are throwing a warning during configuration: 3.9.11 throws this: warning: (DRM_RADEON && DRM_I915 && DRM_TILCDC && FB_BACKLIGHT && USB_APPLEDISPLAY && FB_OLPC_DCON && ASUS_LAPTOP && SONY_LAPTOP && THINKPAD_ACPI && EEEPC_LAPTOP && ACPI_CMPC && SAMSUNG_Q10) selects BACKLIGHT_CLASS_DEVICE which has unmet direct dependencies (HAS_IOMEM && BACKLIGHT_LCD_SUPPORT) 3.10.9 throws this: warning: (DRM_RADEON && DRM_I915 && DRM_GMA500 && DRM_TILCDC && FB_BACKLIGHT && USB_APPLEDISPLAY && FB_OLPC_DCON && ASUS_LAPTOP && SONY_LAPTOP && THINKPAD_ACPI && EEEPC_LAPTOP && ACPI_CMPC && SAMSUNG_Q10) selects BACKLIGHT_CLASS_DEVICE which has unmet direct dependencies (HAS_IOMEM && BACKLIGHT_LCD_SUPPORT) 3.11-rc7 throws this: warning: (DRM_RADEON && DRM_NOUVEAU && DRM_I915 && DRM_GMA500 && DRM_TILCDC && FB_BACKLIGHT && USB_APPLEDISPLAY && FB_OLPC_DCON && ASUS_LAPTOP && SONY_LAPTOP && THINKPAD_ACPI && EEEPC_LAPTOP && ACPI_CMPC && SAMSUNG_Q10) selects BACKLIGHT_CLASS_DEVICE which has unmet direct dependencies (HAS_IOMEM && BACKLIGHT_LCD_SUPPORT)
git bisect gave me the following result: Bisecting: a merge base must be tested [c1be5a5b1b355d40e6cf79cc979eb66dafa24ad1] Linux 3.9 Bisecting: 60 revisions left to test after this (roughly 6 steps) [f8e6bfc2ce162855fa4f9822a45659f4b542c960] drm/radeon: fix possible segfault when parsing pm tables Bisecting: 30 revisions left to test after this (roughly 5 steps) [bea5497bfc1067620c8c8e9d37a42e0bb6d7d7fa] drm/radeon: update wait_for_vblank for r5xx-r7xx Bisecting: 14 revisions left to test after this (roughly 4 steps) [ef0e6e657cfe6e80036b5263887c6ec102c4bae9] drm/radeon: add set_uvd_clocks callback for r7xx v3 Bisecting: 7 revisions left to test after this (roughly 3 steps) [574490401d8fd553ac4baa33ea22fa315a2b4294] drm/radeon: UVD doesn't needs VM on SI v2 Bisecting: 3 revisions left to test after this (roughly 2 steps) [7062ab67d4c6568ec423da39321423721b925fdc] drm/radeon: add radeon_atom_get_clock_dividers helper Bisecting: 1 revision left to test after this (roughly 1 step) [f2ba57b5eab8817d86d0f108fdf1878e51dc0a37] drm/radeon: UVD bringup v8 Bisecting: 0 revisions left to test after this (roughly 0 steps) [4474f3a91f95e3fcc62d97e36f1e8e3392c96ee0] drm/radeon: rework fallback handling v2 f2ba57b5eab8817d86d0f108fdf1878e51dc0a37 is the first bad commit commit f2ba57b5eab8817d86d0f108fdf1878e51dc0a37 Author: Christian König <deathsimple@vodafone.de> Date: Mon Apr 8 12:41:29 2013 +0200 drm/radeon: UVD bringup v8 Just everything needed to decode videos using UVD. v6: just all the bugfixes and support for R7xx-SI merged in one patch v7: UVD_CGC_GATE is a write only register, lockup detection fix v8: split out VRAM fallback changes, remove support for RV770, add support for HEMLOCK, add buffer sizes checks Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Jerome Glisse <jglisse@redhat.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> :040000 040000 1bfd368f291a2cfd7d0ebeb70c9109dfd6ebb97d e4307ff6aa51cf9831815528b7dcf802dd045022 M drivers :040000 040000 1e16513d7c6ac3bde842a2ef9e9993ddfeb3faab 8a79c88d63cce4341a3885effcb69024d6d36917 M include
It seems I'm not alone here: https://lkml.org/lkml/2013/6/16/55
Well have you updated your firmware to include the rlc, pfp, me and uvd ucode?
I see it now, my card needs different firmware set for newer kernels. Conclusion: 3.11 works with the new firmware set if everything is loaded correctly. 3.10 still freezes, but it does not matter now. It's fixed in 3.11. I'm closing this as invalid.