[PATCH v2 2/2] drm/i915: Handle validation of relocation object with size = 2GB

Chris Wilson chris at chris-wilson.co.uk
Fri Apr 7 11:59:30 UTC 2017


On Fri, Apr 07, 2017 at 05:13:56PM +0530, Kamble, Sagar A wrote:
> 
> 
> On 4/7/2017 4:48 PM, Chris Wilson wrote:
> >On Fri, Apr 07, 2017 at 04:32:59PM +0530, Sagar Arun Kamble wrote:
> >>With prefaulting extended to support 2GB size, i915 can handle
> >>relocation objects of 2GB.
> >>
> >>v2: Removed relocs_maxs and relocs_count type change.
> >>
> >>Testcase: igt/gem_exec_reloc/wc-31
> >>Cc: Daniel Vetter <daniel.vetter at ffwll.ch>
> >>Cc: Chris Wilson <chris at chris-wilson.co.uk>
> >>Cc: Joonas Lahtinen <joonas.lahtinen at linux.intel.com>
> >>Signed-off-by: Sagar Arun Kamble <sagar.a.kamble at intel.com>
> >The testcases were written to exercise patches that were already on the
> >list.
> >-Chris
> Hi Chris,
> 
> This testcase is enumerated during yocto testing and is failing due
> to i915 not handling 2GB relocations.
> Do we need to change to skip this subtest ? Or add this change to
> handle 2GB relocations?

Ask why wc-32 and friends are excluded from the test list? What's the
point in a partial fix that conflicts with an older complete fix? And
having demonstrated a competency with the code, perhaps you might like
to critique https://patchwork.freedesktop.org/series/22099/
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre


More information about the Intel-gfx-trybot mailing list