Re: [i915] Kernel does not compile when set to use i915 kernel mode-settingper default (CONFIG_DRM_I915_KMS=y)

[Randy Dunlap]

Woody Suwalski wrote:
> Rafael J. Wysocki wrote:
>> On Monday 18 May 2009, Randy Dunlap wrote:
>>  
>>> On Mon, 18 May 2009 21:01:06 +0200 Rafael J. Wysocki wrote:
>>>
>>>    
>>>> Adding CCs.
>>>>
>>>> On Monday 18 May 2009, you wrote:
>>>>      
>>>>> (First post, so please CC to zorael@xxxxxxxxx in case it doesn't
>>>>> automatically.)
>>>>>
>>>>>
>>>>> I'm trying to compile 2.6.30-rc6 to enable kernel mode-setting *per
>>>>> default*, for my Intel 945GME video chipset. Alas, when compiling, it
>>>>> halts with the following tidbit:
>>>>>
>>>>> ...
>>>>>   LD      .tmp_vmlinux1
>>>>> drivers/built-in.o: In function `intel_opregion_free':
>>>>> /usr/src/linux/drivers/gpu/drm/i915/i915_opregion.c:422: undefined
>>>>> reference to `acpi_video_exit'
>>>>> drivers/built-in.o: In function `intel_opregion_init':
>>>>> /usr/src/linux/drivers/gpu/drm/i915/i915_opregion.c:393: undefined
>>>>> reference to `acpi_video_register'
>>>>> make[1]: *** [.tmp_vmlinux1] Error 1
>>>>> make[1]: Leaving directory `/usr/src/linux-2.6.29'
>>>>> make: *** [debian/stamp/build/kernel] Error 2
>>>>>
>>>>>
>>>>> grepping acpi_video_register recursively I see it's defined as an
>>>>> empty function in ./include/acpi/video.h, and as a more proper
>>>>> function in ./drivers/acpi/video.c. Does somehow setting it to use KMS
>>>>> per default via CONFIG_DRM_I915_KMS=y stop it from reading those
>>>>> files?
>>>>>
>>>>> I used to run with KMS by default in the early -30rc*s, but at some
>>>>> point (rc2?), it stopped compiling succesfully. If I don't enable it
>>>>> per default, it *does* compile, though I don't know how to explicitly
>>>>> enable it at boot-time. So I'm living without KMS for the time being.
>>>>> Woe.
>>>>>
>>>>> Anything obvious I'm doing wrong?
>>>>>         
>>> Len posted a patch for this.  It's below.
>>> Len, do you have this patch queued for Linus?
>>>     
>>
>> Ah, it's commit ecb4aed78dcf09e48c8c34c8c2fa7f5c69344be6, which went in
>> slightly after -rc6.  So it should be fixed.
>>
>> Zorael, please test the current Linus' tree.
>>
>>   
> Len's patch idea does not work for me - if ACPI is build as a module, I
> have no way to force acpi-video to be built in.
> And kernel with i915 built-in breaks...
> So to summarize: acpi=m with i915=y is broken...
> 
> For now I am simply commenting-out the i915 acpi-video calls, but that
> is not a good way...
> We need to move the core acpi-video to the main kernel...

Please post your troublesome kernel .config file.

Thanks,
-- 
~Randy
LPC 2009, Sept. 23-25, Portland, Oregon
http://linuxplumbersconf.org/2009/
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/