[PATCH 28/29] drm/amdkfd: Refactor migrate init to support partition switch

[Alex Deucher]

On Mon, Aug 7, 2023 at 12:04 PM Michel Dänzer <michel at daenzer.net> wrote:
>
> On 7/28/23 19:20, Alex Deucher wrote:
> > On Fri, Jul 28, 2023 at 1:19 PM Michel Dänzer <michel at daenzer.net> wrote:
> >> On 7/28/23 18:43, Alex Deucher wrote:
> >>> On Fri, Jul 28, 2023 at 10:25 AM Michel Dänzer <michel at daenzer.net> wrote:
> >>>> On 7/28/23 11:30, Michel Dänzer wrote:
> >>>>> On 7/28/23 03:38, Zhang, Jesse(Jie) wrote:
> >>>>>>
> >>>>>> Could you try the patch again ?  That work for me.
> >>>>>>
> >>>>>> https://patchwork.freedesktop.org/patch/549605/ <https://patchwork.freedesktop.org/patch/549605/>
> >>>>>
> >>>>> This patch fixes the symptoms described in https://gitlab.freedesktop.org/drm/amd/-/issues/2659 for me as well.
> >>>>>
> >>>>> However, it does not fix the IOMMU page faults[0] or the IB test failures on the compute rings. Should I try amdgpu.ignore_crat=1 for these symptoms as well?
> >>>>
> >>>> I tried ignore_crat=1, it avoids the remaining symptoms as well.
> >>>
> >>> The first 3 patches of this set may also fix it:
> >>> https://patchwork.freedesktop.org/series/121538/
> >>
> >> They do fix all symptoms I'm seeing.
> >>
> >> Is this a feasible solution for the final 6.5 release?
> >
> > I think the first3 are.
>
> Even with Felix's feedback on patch 2?
>
> If so, will you merge these for 6.5 final?

Yes, just sent out a new patch set.  The changes for Felix' feedback
was pretty trivial.  Once they are reviewed, I'll land them for 6.5.
Hopefully this week.

Alex

>
> Or should we revert 84b4dd3f84de ("drm/amdkfd: Refactor migrate init to support partition switch") for that?
>
>
> --
> Earthling Michel Dänzer            |                  https://redhat.com
> Libre software enthusiast          |         Mesa and Xwayland developer
>