build failures on drm-next

Russell, Kent Kent.Russell at
Mon Oct 2 15:05:37 UTC 2017

Agreed. I did the same change locally as Tom did and it seems to be fine. It was just an issue of not rebasing and re-testing before pushing. It builds again now as well, so it should hopefully be alright Harry. Reverting did not go well locally.


-----Original Message-----
From: amd-gfx [mailto:amd-gfx-bounces at] On Behalf Of Tom St Denis
Sent: Monday, October 02, 2017 10:52 AM
To: Wentland, Harry; amd-gfx mailing list
Cc: Zhu, Rex
Subject: Re: build failures on drm-next

I already pushed Rb'ed commits that patch them up :-)

The commits add functionality but they were written before Rex had ported smumgr stuff up into hwmgr and when they were merged nobody compile tested them.

So reverting is probably not the best idea.

On 02/10/17 10:47 AM, Harry Wentland wrote:
> Does anyone have a strong opinion for or against reverting the 
> offending commits?
> Harry
> On 2017-10-02 09:49 AM, Tom St Denis wrote:
>> The lastest good commit is
>> commit f5a16c561fb8a15e94a4384e880b9d84a2657098
>> Author: Colin Ian King <colin.king at>
>> Date:   Thu Sep 28 14:46:17 2017 +0100
>>      drm/radeon: make functions alloc_pasid and free_pasid static
>>      The functions alloc_pasid  and free_pasid are local to the
>>      source and do not need to be in global scope, so make them static.
>>      Cleans up sparse warnings:
>>      warning: symbol 'alloc_pasid' was not declared. Should it be static?
>>      warning: symbol 'free_pasid' was not declared. Should it be static?
>>      Signed-off-by: Colin Ian King <colin.king at>
>>      Signed-off-by: Alex Deucher <alexander.deucher at>
>> I see errors (e.g., attached) when building from the tip of drm-next.
>> Cheers,
>> Tom
>> _______________________________________________
>> amd-gfx mailing list
>> amd-gfx at

amd-gfx mailing list
amd-gfx at

More information about the amd-gfx mailing list