<html> <head> <base href="https://bugs.freedesktop.org/" /> </head> <body> <div> <a class="bz_bug_link bz_status_ASSIGNED " title="ASSIGNED - [SKL bisected] texsubimage pbo intermittent failures" href="https://bugs.freedesktop.org/show_bug.cgi?id=91926#c32">Comment # 32</a> on <a class="bz_bug_link bz_status_ASSIGNED " title="ASSIGNED - [SKL bisected] texsubimage pbo intermittent failures" href="https://bugs.freedesktop.org/show_bug.cgi?id=91926">bug 91926</a> from <a class="email" href="mailto:topi.pohjolainen@intel.com" title="Topi Pohjolainen <topi.pohjolainen@intel.com>"> Topi Pohjolainen</a> <pre>(In reply to Ben Widawsky from <a href="show_bug.cgi?id=91926#c31">comment #31</a>) > I wonder why the blit is failing always on SKL. I can't think of anything > IVB can do that SKL cannot do. And you are correct. Neither actually fail the blit when doing the texture upload, the fail comes from elsewhere. I made the same mistake I did before when interpreting my logs - only this time in both IVB and SKL. I didn't bother creating special driver hook for only the texture upload path to use but simply used the general. I assumed the test is simple enough not require blitting elsewhere - this is not true unfortunately. When I disabled the glsl-blit option in the driver hook some of these non-tex-upload blits started to fall to the software path because the blitter couldn't do a format conversion. Furthermore, the use of blitter engine helps the special of the test I've been studying on IVB. Full test (original version of the piglit test) fails on both IVB and SKL with blitter engine. In fact, it looks to fail in more cases than the glsl blit. I need to look into this some more.</pre> </div> <hr> You are receiving this mail because: <ul> <li>You are the QA Contact for the bug.</li> </ul> </body> </html>