5.11-rc1 TTM list corruption

[Christian König]

Am 05.01.21 um 13:20 schrieb Huang Rui:
> On Tue, Jan 05, 2021 at 07:43:51PM +0800, Borislav Petkov wrote:
>> On Tue, Jan 05, 2021 at 07:08:52PM +0800, Huang Rui wrote:
>>> Ah, this asic is a bit old and still use radeon driver. So we didn't
>>> reproduce it on amdgpu driver. I don't have such the old asic in my hand.
>>> May we know whether this issue can be duplicated after SI which is used
>>> amdgpu module (not sure whether you have recent APU or GPU)?
>> The latest I have (I think it is the latest) is:
>>
>> [    1.826102] [drm] initializing kernel modesetting (RENOIR 0x1002:0x1636 0x17AA:0x5099 0xD1).
>>
>> and so far that hasn't triggered it. Which makes sense because that
>> thing uses amdgpu:
>>
>> [    1.810260] [drm] amdgpu kernel modesetting enabled.
> Yes! Renoir is late enough for amdgpu kernel module. :-)
> Please let us know if you still encounter the issue.

Thanks for the hints guys. You need a rather specific configuration, but 
I can reproduce this now.

Let's see what the problem is here.

Thanks,
Christian.

>
> Thanks,
> Ray