On Tue, Oct 2, 2012 at 7:19 AM, Robert Bragg wrote:
> On Thu, Sep 27, 2012 at 6:48 PM, Matt Turner wrote:
>> On Thu, Sep 20, 2012 at 8:59 AM, Robert Bragg wrote:
>>> This adds an egl_probe_front_pixel_rgb function that is analogous to
>>> piglit_probe_pixel_rgba except it probes the front buffer
Zero-sized tex images and sub images should not generate an error.
Full images which are not multiples of the block size should also be legal.
Verified w/ NVIDIA's driver.
---
tests/texturing/s3tc-errors.c | 38 ++
1 files changed, 38 insertions(+), 0 deletio
Hi,
Rob Bradford was discussing a bug in Mesa where the fast path code for
uploading texture data to BGRA textures on Sandybridge does not work
with a subregion of the buffer using GL_UNPACK_SKIP_*. Here is a test
for Piglit that exposes this bug on GLES2. It is based on a previous
test called ext
On Thu, Sep 27, 2012 at 6:48 PM, Matt Turner wrote:
> On Thu, Sep 20, 2012 at 8:59 AM, Robert Bragg wrote:
>> This adds an egl_probe_front_pixel_rgb function that is analogous to
>> piglit_probe_pixel_rgba except it probes the front buffer instead of
>> probing the back buffer.
>> ---
>> tests/e