Bug 699209

Summary: Kernel error and no display on monitor kernel 2.6.37.6
Product: [openSUSE] openSUSE 11.4 Reporter: Forgotten User 36kTZSIwRV <forgotten_36kTZSIwRV>
Component: KernelAssignee: E-mail List <kernel-maintainers>
Status: RESOLVED DUPLICATE QA Contact: E-mail List <qa-bugs>
Severity: Critical    
Priority: P5 - None CC: fisiu, jdelvare
Version: Final   
Target Milestone: ---   
Hardware: i686   
OS: openSUSE 11.4   
Whiteboard:
Found By: --- Services Priority:
Business Priority: Blocker: ---
Marketing QA Status: --- IT Deployment: ---

Description Forgotten User 36kTZSIwRV 2011-06-10 06:41:11 UTC
User-Agent:       Opera/9.80 (X11; Linux x86_64; U; OpenSUSE 11.4; pl) Presto/2.8.131 Version/11.11

After updating the kernel 2.6.37.6 there was an error that the graphics card does not start. The older version everything works properly. Repository file, update and results boot.msg.

<6>[    2.833968] [drm] Initialized drm 1.1.0 20060810
<6>[    2.906448] [drm] radeon defaulting to kernel modesetting.
<6>[    2.906453] [drm] radeon kernel modesetting enabled.
<7>[    2.906520] checking generic (d0000000 1000000) vs hw (d0000000 8000000)
<3>[    2.906523] fb: conflicting fb hw usage radeondrmfb vs VESA VGA - removing generic driver
<4>[    2.906567] Console: switching to colour dummy device 80x25
<6>[    2.907197] radeon 0000:01:00.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16
<6>[    2.910349] [drm] initializing kernel modesetting (RV280 0x1002:0x5960).
<6>[    2.911781] [drm] register mmio base: 0xDFDF0000
<6>[    2.911785] [drm] register mmio size: 65536
<6>[    2.912782] agpgart-intel 0000:00:00.0: AGP 2.0 bridge
<6>[    2.912802] agpgart-intel 0000:00:00.0: putting AGP V2 device into 4x mode
<6>[    2.912836] radeon 0000:01:00.0: putting AGP V2 device into 4x mode
<6>[    2.912864] radeon 0000:01:00.0: GTT: 256M 0xE0000000 - 0xEFFFFFFF
<6>[    2.912869] [drm] Generation 2 PCI interface, using max accessible memory
<6>[    2.912874] radeon 0000:01:00.0: VRAM: 128M 0x00000000D0000000 - 0x00000000D7FFFFFF (128M used)
<6>[    2.912913] [drm] radeon: irq initialized.
<6>[    2.913022] [drm] Detected VRAM RAM=128M, BAR=128M
<6>[    2.913027] [drm] RAM width 128bits DDR
<6>[    2.913250] [TTM] Zone  kernel: Available graphics memory: 443166 kiB.
<6>[    2.913285] [TTM] Zone highmem: Available graphics memory: 512770 kiB.
<6>[    2.913288] [TTM] Initializing pool allocator.
<6>[    2.913313] [drm] radeon: 128M of VRAM memory ready
<6>[    2.913317] [drm] radeon: 256M of GTT memory ready.
<6>[    2.914444] radeon 0000:01:00.0: WB disabled
<6>[    2.914538] [drm] Loading R200 Microcode
<6>[    2.918116] [drm] radeon: ring at 0x00000000E0001000
<6>[    2.918138] [drm] ring test succeeded in 1 usecs
<6>[    2.918470] [drm] radeon: ib pool ready.
<6>[    2.918548] [drm] ib test succeeded in 0 usecs
<4>[    2.918589] Radeon i2c bit bus DVI_DDC: bus seems to be busy
<3>[    2.919172] [drm:radeon_i2c_create] *ERROR* Failed to register bit i2c DVI_DDC
<4>[    2.919194] Radeon i2c bit bus VGA_DDC: bus seems to be busy
<3>[    2.919234] [drm:radeon_i2c_create] *ERROR* Failed to register bit i2c VGA_DDC
<4>[    2.919313] Radeon i2c bit bus MONID: bus seems to be busy
<3>[    2.928465] [drm:radeon_i2c_create] *ERROR* Failed to register bit i2c MONID
<4>[    2.928486] Radeon i2c bit bus CRT2_DDC: bus seems to be busy
<3>[    2.928527] [drm:radeon_i2c_create] *ERROR* Failed to register bit i2c CRT2_DDC
<6>[    2.928641] [drm] Radeon Display Connectors
<6>[    2.928644] [drm] Connector 0:
<6>[    2.928646] [drm]   S-video
<6>[    2.928647] [drm]   Encoders:
<6>[    2.928649] [drm]     TV1: INTERNAL_DAC2
<6>[    2.931662] No connectors reported connected with modes
<6>[    2.931666] [drm] Cannot find any crtc or sizes - going 1024x768
<6>[    2.948850] [drm] fb mappable at 0xD0040000
<6>[    2.948853] [drm] vram apper at 0xD0000000
<6>[    2.948855] [drm] size 3145728
<6>[    2.948857] [drm] fb depth is 24
<6>[    2.948858] [drm]    pitch is 4096
<4>[    2.949065] Console: switching to colour frame buffer device 128x48
<6>[    3.071550] fb0: radeondrmfb frame buffer device
<6>[    3.071554] drm: registered panic notifier
<6>[    3.071568] [drm] Initialized radeon 2.7.0 20080528 for 0000:01:00.0 on minor 0

Reproducible: Always

Actual Results:  
I downgrade to 2.6.37.1
Comment 1 Forgotten User 36kTZSIwRV 2011-06-10 06:43:20 UTC
I have radeon 9250
Comment 2 Forgotten User 36kTZSIwRV 2011-06-10 21:10:53 UTC
In kernel 2.6.39.1 from repositories openssue stable kernel:

linux-0j0c:/var/log # grep drm boot.msg
<6>[    2.025181] [drm] Initialized drm 1.1.0 20060810
<6>[    2.293939] [drm] radeon defaulting to kernel modesetting.
<6>[    2.293944] [drm] radeon kernel modesetting enabled.
<6>[    2.294018] fb: conflicting fb hw usage radeondrmfb vs VESA VGA - removing generic driver
<6>[    2.294887] [drm] initializing kernel modesetting (RV280 0x1002:0x5960).
<6>[    2.294915] [drm] register mmio base: 0xDFDF0000
<6>[    2.294918] [drm] register mmio size: 65536
<6>[    2.295962] [drm] Generation 2 PCI interface, using max accessible memory
<6>[    2.295984] [drm] Supports vblank timestamp caching Rev 1 (10.10.2010).
<6>[    2.295987] [drm] Driver supports precise vblank timestamp query.
<6>[    2.296045] [drm] radeon: irq initialized.
<6>[    2.296191] [drm] Detected VRAM RAM=128M, BAR=128M
<6>[    2.296196] [drm] RAM width 128bits DDR
<6>[    2.304195] [drm] radeon: 128M of VRAM memory ready
<6>[    2.304198] [drm] radeon: 256M of GTT memory ready.
<6>[    2.305718] [drm] Loading R200 Microcode
<6>[    2.322424] [drm] radeon: ring at 0x00000000E0001000
<6>[    2.322446] [drm] ring test succeeded in 1 usecs
<6>[    2.322816] [drm] radeon: ib pool ready.
<6>[    2.322901] [drm] ib test succeeded in 0 usecs
<3>[    2.323312] [drm:radeon_i2c_create] *ERROR* Failed to register bit i2c DVI_DDC
<3>[    2.323370] [drm:radeon_i2c_create] *ERROR* Failed to register bit i2c VGA_DDC
<3>[    2.332049] [drm:radeon_i2c_create] *ERROR* Failed to register bit i2c MONID
<3>[    2.332125] [drm:radeon_i2c_create] *ERROR* Failed to register bit i2c CRT2_DDC
<3>[    2.332159] [drm:radeon_add_legacy_connector] *ERROR* VGA: Failed to assign ddc bus! Check dmesg for i2c errors.
<3>[    2.332237] [drm:radeon_add_legacy_connector] *ERROR* DVI: Failed to assign ddc bus! Check dmesg for i2c errors.
<6>[    2.332315] [drm] Radeon Display Connectors
<6>[    2.332318] [drm] Connector 0:
<6>[    2.332320] [drm]   VGA
<6>[    2.332322] [drm]   DDC: no ddc bus - possible BIOS bug - please report to xorg-driver-ati@lists.x.org
<6>[    2.332325] [drm]   Encoders:
<6>[    2.332327] [drm]     CRT1: INTERNAL_DAC1
<6>[    2.332329] [drm] Connector 1:
<6>[    2.332331] [drm]   DVI-I
<6>[    2.332333] [drm]   HPD1
<6>[    2.332334] [drm]   DDC: no ddc bus - possible BIOS bug - please report to xorg-driver-ati@lists.x.org
<6>[    2.332337] [drm]   Encoders:
<6>[    2.332339] [drm]     CRT2: INTERNAL_DAC2
<6>[    2.332341] [drm]     DFP1: INTERNAL_TMDS1
<6>[    2.332344] [drm] Connector 2:
<6>[    2.332345] [drm]   S-video
<6>[    2.332347] [drm]   Encoders:
<6>[    2.332349] [drm]     TV1: INTERNAL_DAC2
<6>[    2.347298] [drm] Cannot find any crtc or sizes - going 1024x768
<6>[    2.364570] [drm] fb mappable at 0xD0040000
<6>[    2.364573] [drm] vram apper at 0xD0000000
<6>[    2.364575] [drm] size 3145728
<6>[    2.364577] [drm] fb depth is 24
<6>[    2.364579] [drm]    pitch is 4096
<6>[    2.364680] fbcon: radeondrmfb (fb0) is primary device
<6>[    2.488163] fb0: radeondrmfb frame buffer device
<6>[    2.488168] drm: registered panic notifier
<6>[    2.488187] [drm] Initialized radeon 2.9.0 20080528 for 0000:01:00.0 on minor 0
Comment 3 Mariusz Fik 2011-06-10 21:58:12 UTC
It looks similar to https://bugzilla.novell.com/show_bug.cgi?id=691052
Comment 4 Forgotten User 36kTZSIwRV 2011-06-14 13:41:23 UTC
Udało mi się rozwiązać problem przy pomocy ludzi z freedesktop. Posiadam w komputerze zainstalowaną kartę graficzną Radeon 9250 i Kartę TV. Po aktualizacji jądra do 2.6.37.6-0.5 następował błąd w postaci nie właczającego się monitora. W logach pojawiały się informacje przedstawione powyżej. Przyczyną owego błędu krytycznego jest ustawienie opcji w konfiguracji kernela i2c_algo_bit.bit_test na 1 co powodowało konflikt w automatycznym wyborze monitora wyświetlania. Aby to anulować można wykonać za pomocą dwóch sposobów. PIerwszy to dodadnie do grub'a podczas startu opcji nomodeset, która wyłacza testowanie albo dopisanie do gruba i2c_algo_bit.bit_test=0 co spowoduje wyłaczenie opcji, która powodowała błąd. Poniżej zamieszczam treść maila oraz linka do patcha, którego warto dodać do paczek jądra w openSUSE. Jest także link do podobnego problemu co mój. (Po Polsku)

I managed to solve the problem using the freedesktop people. I have installed in your computer graphics card Radeon 9250 and TV card. After updating the kernel to 2.6.37.6-0.5 followed an error in the form of not integrating the monitor. In the logs there were some reports described above. The cause of this fatal error is to set the options in the kernel configuration i2c_algo_bit.bit_test on a conflict which resulted in the automatic selection of the monitor display. To cancel this can be done using two methods. The first is dodadnie to the grub boot options nomodeset, which disables the test or submit a thick i2c_algo_bit.bit_test = 0 which will disable the option, which caused the error. Below is the e-mail text and a link to the patch, which should add to the kernel packages in openSUSE. There is also a link to a similar problem as mine.

Email text:

IRC, openSUSE sets i2c-algo-bit.bit_test=1 by default which causes
the radeon i2c init to fail since the radeon i2c buses require
pre/post i2c setup to work properly which the i2c subsystem didn't
handle by default.  You can either set i2c-algo-bit.bit_test=0 on the
kernel command line in grub, or make sure your kernel has this patch:

link to the patch: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=d3b3e15da14ded61c9654db05863b04a2435f4cc

See this bug for reference:
https://bugs.freedesktop.org/show_bug.cgi?id=36221


Thanks for your help Alex
Comment 5 Forgotten User 36kTZSIwRV 2011-06-14 13:48:45 UTC
Please provide the developers to add the patch to the kernel packages that will avoid many unpleasant surprises.
Comment 6 Jean Delvare 2011-09-30 15:16:56 UTC
I think this bug got misanalysed. The patch in question is what is _causing_ the bug, rather than fixing it. Booting with i2c_algo_bit.bit_test=0 is still the correct workaround though.
Comment 7 Jean Delvare 2011-10-11 12:54:41 UTC
Definitely the same bug as #712023, just for a different openSUSE version.

*** This bug has been marked as a duplicate of bug 712023 ***