Created attachment 72358 [details] X Start up log messages The Radeon driver will not load with the Radeon kernel drivers in Linux 3.8.0 The following is an excerpt from the startup: [ 109.112] (II) Loading extension DRI2 [ 109.112] (II) LoadModule: "radeon" [ 109.112] (II) Loading /usr/lib/xorg/modules/drivers/radeon_drv.so [ 109.112] (II) Module radeon: vendor="X.Org Foundation" [ 109.112] compiled for 1.12.3, module version = 7.0.99 [ 109.112] Module class: X.Org Video Driver [ 109.112] ABI class: X.Org Video Driver, version 12.0 [ 109.112] (II) RADEON: Driver for ATI Radeon chipsets: ATI Radeon Mobility X600 (M24) 3150 (PCIE), ATI FireMV 2400 (PCI), (many cards snipped) [ 109.114] (--) using VT number 8 [ 109.119] (II) [KMS] Kernel modesetting enabled. [ 109.120] (II) RADEON(0): Creating default Display subsection in Screen section "Default Screen Section" for depth/fbbpp 24/32 [ 109.120] (==) RADEON(0): Depth 24, (--) framebuffer bpp 32 [ 109.120] (II) RADEON(0): Pixel depth = 24 bits stored in 4 bytes (32 bpp pixmaps) [ 109.120] (==) RADEON(0): Default visual is TrueColor [ 109.120] (==) RADEON(0): RGB weight 888 [ 109.120] (II) RADEON(0): Using 8 bits per RGB (8 bit DAC) [ 109.120] (--) RADEON(0): Chipset: "ATI Radeon HD 4350" (ChipID = 0x954f) [ 109.120] drmOpenDevice: node name is /dev/dri/card0 [ 109.120] drmOpenDevice: open result is 9, (OK) [ 109.120] drmOpenByBusid: Searching for BusID pci:0000:04:00.0 [ 109.120] drmOpenDevice: node name is /dev/dri/card0 [ 109.120] drmOpenDevice: open result is 9, (OK) [ 109.120] drmOpenByBusid: drmOpenMinor returns 9 [ 109.120] drmOpenByBusid: Searching for BusID pci:0000:04:00.0 [ 109.120] drmOpenDevice: node name is /dev/dri/card0 [ 109.120] drmOpenDevice: open result is 9, (OK) [ 109.120] drmOpenByBusid: drmOpenMinor returns 9 [ 109.120] drmOpenByBusid: Interface 1.4 failed, trying 1.1 [ 109.120] drmOpenByBusid: drmGetBusid reports That last line is not a copy/paste issue - it literally is "[ 109.120] drmOpenByBusid: drmGetBusid reports " (as in, there is no bus ID given) Then the X Radeon driver looks for any other DRI node, fails, and dies. [ 109.120] drmOpenDevice: node name is /dev/dri/card1 [ 109.125] drmOpenByBusid: drmOpenMinor returns -1 (snip many other searches) [ 109.170] drmOpenDevice: node name is /dev/dri/card14 [ 109.174] drmOpenByBusid: drmOpenMinor returns -1 [ 109.174] drmOpenDevice: node name is /dev/dri/card15 [ 109.178] drmOpenByBusid: drmOpenMinor returns -1 [ 109.178] drmOpenDevice: node name is /dev/dri/card0 [ 109.178] drmOpenDevice: open result is 9, (OK) [ 109.178] drmOpenDevice: node name is /dev/dri/card0 [ 109.178] drmOpenDevice: open result is 9, (OK) [ 109.178] drmGetBusid returned '' [ 109.178] (EE) RADEON(0): [drm] failed to set drm interface version. [ 109.178] (EE) RADEON(0): Kernel modesetting setup failed [ 109.178] (II) UnloadModule: "radeon" [ 109.178] (EE) Screen(s) found, but none have a usable configuration. I have tried the following: 1) adding the following parameters in /etc/modprobe.d/radeon.conf options radeon modeset=1 dynclks=1 gartsize=256 audio=1 pcie_gen2=-1 msi=1 2) Adding the following to /etc/initramfs-tools/modules radeon modeset=1 dynclks=1 gartsize=256 audio=1 pcie_gen2=-1 msi=1 and rebuilding the ramdisk. The kernel has the following configuration with respect to the Radeon drivers: CONFIG_DRM_RADEON=m CONFIG_DRM_RADEON_KMS=y # CONFIG_FB_RADEON is not set
Please attach the corresponding dmesg output.
[ 1.298538] [drm] radeon kernel modesetting enabled. [ 1.298806] [drm] initializing kernel modesetting (RV710 0x1002:0x954F 0x1002:0x2143). [ 1.298938] [drm] register mmio base: 0xF7BE0000 [ 1.299008] [drm] register mmio size: 65536 [ 1.299140] ATOM BIOS: 0 [ 1.299228] radeon 0000:04:00.0: VRAM: 512M 0x0000000000000000 - 0x000000001FFFFFFF (512M used) [ 1.299344] radeon 0000:04:00.0: GTT: 256M 0x0000000020000000 - 0x000000002FFFFFFF [ 1.299465] mtrr: type mismatch for d0000000,10000000 old: write-back new: write-combining [ 1.299581] [drm] Detected VRAM RAM=512M, BAR=256M [ 1.299654] [drm] RAM width 64bits DDR [ 1.299766] [TTM] Zone kernel: Available graphics memory: 3085752 kiB [ 1.299843] [TTM] Zone dma32: Available graphics memory: 2097152 kiB [ 1.299920] [TTM] Initializing pool allocator [ 1.299993] [TTM] Initializing DMA pool allocator [ 1.300080] [drm] radeon: 512M of VRAM memory ready [ 1.300148] [drm] radeon: 256M of GTT memory ready. [ 1.300222] [drm] Supports vblank timestamp caching Rev 1 (10.10.2010). [ 1.300292] [drm] Driver supports precise vblank timestamp query. [ 1.300393] radeon 0000:04:00.0: irq 71 for MSI/MSI-X [ 1.300401] radeon 0000:04:00.0: radeon: using MSI. [ 1.300531] [drm] radeon: irq initialized. [ 1.300639] [drm] GART: num cpu pages 65536, num gpu pages 65536 [ 1.300997] [drm] probing gen 2 caps for device 8086:340a = 2/0 [ 1.301067] [drm] enabling PCIE gen 2 link speeds, disable with radeon.pcie_gen2=0 [ 1.301228] [drm] Loading RV710 Microcode 1.324868] [drm] PCIE GART of 256M enabled (table at 0x0000000000040000). [ 1.325019] radeon 0000:04:00.0: WB enabled [ 1.325083] radeon 0000:04:00.0: fence driver on ring 0 use gpu addr 0x0000000020000c00 and cpu addr 0xffff8801b4bb6c00 [ 1.325187] radeon 0000:04:00.0: fence driver on ring 3 use gpu addr 0x0000000020000c0c and cpu addr 0xffff8801b4bb6c0c [ 1.374327] [drm] radeon: power management initialized [ 1.465494] [drm] fb mappable at 0xD00C2000 [ 1.465556] [drm] vram apper at 0xD0000000 [ 1.465617] [drm] size 8294400 [ 1.465677] [drm] fb depth is 24 [ 1.465736] [drm] pitch is 7680 [ 1.465995] radeon 0000:04:00.0: fb0: radeondrmfb frame buffer device [ 1.466069] radeon 0000:04:00.0: registered panic notifier [ 1.466142] [drm] Initialized radeon 2.27.0 20080528 for 0000:04:00.0 on minor 0
Hmm, can't see anything wrong there... I assume it works with older kernels? Can you bisect?
I cannot say with rigor that it did work with older kernels, as prior to upgrading to Ubuntu 12.4 I was running the non-Free fglrx drivers (forgive me RMS for I have sinned....) Bisecting is hard, because I am using many of the features of the newer kernel now, and backtracking would be difficult. But if you can suggest a version to test I can start. I have seen other people running the Intel drivers reporting this same sort of error (the missing busID in the drmGetBusid call), and they thought it was a race condition (hence putting the driver into the initrd, which the Intel driver folks said fixed their problem).
(In reply to comment #4) > But if you can suggest a version to test I can start. I haven't seen any other similar reports, so if it happens with 3.7 kernels, it's probably not a (recent) kernel regression. > I have seen other people running the Intel drivers reporting this same sort > of error (the missing busID in the drmGetBusid call), and they thought it > was a race condition That seems unlikely in your case, given the timestamps in the log snippets you've provided.
-- GitLab Migration Automatic Message -- This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity. You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/drm/amd/issues/317.
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.