[Mesa-dev] [PATCH] i965: report bound buffer size not underlying buffer size for image size (v2)

Dave Airlie airlied at gmail.com
Tue Aug 23 22:35:15 UTC 2016


On 23 August 2016 at 19:28, Antía Puentes <apuentes at igalia.com> wrote:
> Hi Dave,
>
> The "GL44-CTS.shader_image_size.advanced-nonMS-fs-int" test fails for
> me with current master (which contains this patch), I have tested it
> both in a Broadwell and a Skylake machine.
>
> I recall that you sent to the mailing list a second patch "i965: don't
> fail to shift height images for levels." related to the test. Is this
> patch or a new version of it still needed?. I remember that as it is it
> regressed GL44-CTS.texture_view.gettexparameter, raising the assertion:
> "glcts: intel_mipmap_tree.c:368: intel_miptree_create_layout: Assertion
> `height0 = 1' failed."

Yes still needed.

>
> On a side note, as we are taking about CTS, just remind that Andres has
> reviewed the "mesa/subroutines: start adding per-context subroutine
> index support" series that you sent, which is not pushed yet.

I cleaned up and pushed it yesterday.

Dave.


More information about the mesa-dev mailing list