[PATCH 2/5] dma-buf: remove kmap_atomic interface

Daniel Vetter daniel at ffwll.ch
Thu Jun 21 06:30:03 UTC 2018


On Thu, Jun 21, 2018 at 8:26 AM, Daniel Vetter <daniel at ffwll.ch> wrote:
> On Wed, Jun 20, 2018 at 4:21 PM, Christian König
> <christian.koenig at amd.com> wrote:
>> Am 20.06.2018 um 16:04 schrieb Christian König:
>>>
>>> Am 20.06.2018 um 14:52 schrieb Daniel Vetter:
>>>>
>>>> On Wed, Jun 20, 2018 at 2:46 PM, Christian König
>>>> <ckoenig.leichtzumerken at gmail.com> wrote:
>>>>>
>>>>> [SNIP]
>>>>>>
>>>>>> Go ahead, that's the point of commit rights. dim might complain if you
>>>>>> cherry picked them and didn't pick them up using dim apply though ...
>>>>>
>>>>>
>>>>> I've fixed up the Link tags, but when I try "dim push-branch
>>>>> drm-misc-next"
>>>>> I only get the error message "error: dst ref refs/heads/drm-misc-next
>>>>> receives from more than one src."
>>>>>
>>>>> Any idea what is going wrong here?
>>>>
>>>> Sounds like multiple upstreams for your local drm-misc-next branch,
>>>> and git then can't decide which one to pick. If you delete the branch
>>>> and create it using dim checkout drm-misc-next this shouldn't happen.
>>>> We're trying to fit into existing check-outs and branches, but if you
>>>> set things up slightly different than dim would have you're off script
>>>> and there's limited support for that.
>>>>
>>>> Alternative check out your .git/config and remove the other upstreams.
>>>> Or attach your git config if this isn't the issue (I'm just doing some
>>>> guessing here).
>>>
>>>
>>> I've tried to delete my drm-misc-next branch and recreate it, but that
>>> doesn't seem to help.
>>>
>>> Attached is my .git/config, but at least on first glance it looks ok as
>>> well.
>>>
>>> Any ideas?
>>
>>
>> Ok that seems to be a bug in dim.
>>
>> "bash -x dim push drm-misc-next" looks like it tries to push the branch
>> drm-misc-next twice to the drm-misc remote: git push drm-misc drm-misc-next
>> drm-misc-next
>>
>> When I try that manually I get the same result, but "git push drm-misc
>> drm-misc-next" just seemed to work fine.
>>
>> Let's hope that I haven't messed things up totally on the server now.
>
> Tree looks all intact, except for some build fail. For drm-misc please
> use the 3 defconfings (for x86, arm and arm64) to check that not too
> much broke, they should enable all the gfx relevant stuff. Not full
> combinatorials, but at least the most obvious of things.
>
> Wrt the issue, it's indeed a bug in dim. The shorthand for pushing the
> current branch works:
>
> $ dim push
>
> But the explicit form somehow gained a bug.
>
> $ dim push drm-misc-next
>
> I'll look into why this is.

Should have waited for coffee to kick in. This is a feature :-)

$ dim push [git push arguments]

is the short-hand. If you want to specify the branch explicitly, you need to use

$ dim push-branch drm-misc-next [git push arguments]

Usually the only thing is adding an -f for a non-fast-forward push
(when maintainers have rebased the -fixes tree, shouldn't be done
anywhere else). Docs seem accurate, so not sure where you copypasted
this from? And not sure we should catch this in the script, since it's
kinda ambiguous.
-Daniel


> -Daniel
>
>>
>> Christian.
>>
>>>
>>> Thanks,
>>> Christian.
>>>
>>>> -Daniel
>>>>
>>>>
>>>
>>
>
>
>
> --
> Daniel Vetter
> Software Engineer, Intel Corporation
> +41 (0) 79 365 57 48 - http://blog.ffwll.ch



-- 
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch


More information about the amd-gfx mailing list