Re: 2.6.34-rc1: pci 0000:00:00.0: address space collision / spontaenous reboots [now 2.6.34-rc1]

[Yinghai Lu]

On 03/12/2010 01:52 PM, Justin Piszcz wrote:
> 
> 
> On Fri, 12 Mar 2010, Yinghai Lu wrote:
> 
>> On 03/12/2010 01:30 PM, Justin Piszcz wrote:
>>>
>>>
>>> On Fri, 12 Mar 2010, Justin Piszcz wrote:
>>>
>>>> On Fri, 12 Mar 2010, Justin Piszcz wrote:
>>>>
>>>
>>> # Disable Bootmem code (NO_BOOTMEM) [Y/n/?] (NEW)
>>>
>>> I see this new option ^
>>>
>>> Ok, booting 2.6.34-rc1:
>>>
>>> [    0.132248] PCI: pci_cache_line_size set to 64 bytes
>>> [    0.132248] pci 0000:00:00.0: BAR 3: reserving [mem
>>> 0xe0000000-0xffffffff flags 0x120204] (d=0, p=0)
>>> [    0.132248] pci 0000:00:00.0: no compatible bridge window for [mem
>>> 0xe0000000-0xffffffff 64bit]
>>> [    0.132248] pci 0000:00:00.0: can't reserve [mem
>>> 0xe0000000-0xffffffff 64bit]
>>>
>>> Full dmesg:
>>> http://home.comcast.net/~jpiszcz/20100312/dmesg-2.6.34-rc1.txt
>>>
>> [    0.097651] node 0 link 0: io port [b000, ffff]
>> [    0.097654] TOM: 0000000040000000 aka 1024M
>> [    0.097657] Fam 10h mmconf [e0000000, e00fffff]
>> [    0.097659] node 0 link 0: mmio [a0000, bffff]
>> [    0.097661] node 0 link 0: mmio [40000000, dfffffff]
>> [    0.097663] node 0 link 0: mmio [f0000000, fe02ffff]
>> [    0.097665] node 0 link 0: mmio [e0000000, e04fffff] ==> [e0100000,
>> e04fffff]
>> [    0.097668] bus: [00, 04] on node 0 link 0
>> [    0.097670] bus: 00 index 0 [io  0x0000-0xffff]
>> [    0.097672] bus: 00 index 1 [mem 0x000a0000-0x000bffff]
>> [    0.097674] bus: 00 index 2 [mem 0x40000000-0xdfffffff]
>> [    0.097676] bus: 00 index 3 [mem 0xe0500000-0xffffffff]
>> [    0.097677] bus: 00 index 4 [mem 0xe0100000-0xe04fffff]
>>
>> that looks like setting MMCONFIG through ATI chipset.
>>
>> and cpu northbridge set 1M for accessing 1M...
>>
>> BIOS really should only use northbridge MSR to set that to cover bus
>> [0, 255]
>>
>> and just disable the ATI pci 00:00.0 BAR3
> 
> Ok, so the BIOS is broken, is there a pci= option to workaround this in
> the interim, or does a quirk need to be implemented?

should be ok, just some warning.

YH
--
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/