Re: [PATCH 01/26] Revert "drm/i915/gem: Async GPU relocations only"

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Quoting Chris Wilson (2020-06-23 16:23:31)
> Quoting Maarten Lankhorst (2020-06-23 15:28:18)
> > This reverts commit 9e0f9464e2ab36b864359a59b0e9058fdef0ce47,
> > and related commit 7ac2d2536dfa7 ("drm/i915/gem: Delete unused code").
> 
> Regardless that you haven't adapted the series...
> 
> This still prevents concurrent submission between clients, and does not
> remove the allocation mutexes. That latter we can do in a couple of
> patches that preserve the status quo with just a name switch and the
> forced removal of nestable shrinkers.

To be clear, I think the first patch has to be
https://patchwork.freedesktop.org/series/78772/ (provided the rebasing
went ok) to fully commit ourselves to removing obj->mm.mutex. Empirically,
we cannot use trylock inside the mmu_notifier as that often fails during
normal unmap.
-Chris
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx



[Index of Archives]     [AMD Graphics]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux