<html> <head> <base href="https://bugs.freedesktop.org/"> </head> <body> <div> <a class="bz_bug_link bz_status_REOPENED " title="REOPENED - Invalid imageAtomicExchange() writes." href="https://bugs.freedesktop.org/show_bug.cgi?id=106390#c10">Comment # 10</a> on <a class="bz_bug_link bz_status_REOPENED " title="REOPENED - Invalid imageAtomicExchange() writes." href="https://bugs.freedesktop.org/show_bug.cgi?id=106390">bug 106390</a> from <a class="email" href="mailto:denys.kostin@globallogic.com" title="Denis <denys.kostin@globallogic.com>"> Denis</a> <pre>Created <a href="attachment.cgi?id=142211" name="attach_142211" title="denys_expected">attachment 142211</a> <a href="attachment.cgi?id=142211&action=edit" title="denys_expected">[details]</a> denys_expected hi guys. I tested last attachment from Alexander on radeon and intel. Found out, that on radeon test looks without visible squads (on the right part of the picture). See attachment "denys_expected" with my "expected" output. Then I checked old mesa versions and found out, that the same picture I can see on 13.0.6 mesa. Below there is a result of bisection (tested on SKL and KBL): 9919542f1cfff70524bc6117d19bf88e59159caa is the first bad commit commit 9919542f1cfff70524bc6117d19bf88e59159caa Author: Kenneth Graunke <<a href="mailto:kenneth@whitecape.org">kenneth@whitecape.org</a>> Date: Sat Jan 14 23:32:12 2017 -0800 i965: Make DCE set null destinations on messages with side effects. (Co-authored by Matt Turner.) Image atomics, for example, return a value - but the shader may not want to use it. We assigned a useless VGRF destination. This seemed harmless, but it can actually be quite harmful. The register allocator has to assign that VGRF to a real register. It may assign the same actual GRF to the destination of an instruction that follows soon after. This results in a write-after-write (WAW) dependency, and stall. A number of "Deus Ex: Mankind Divided" shaders use image atomics, but don't use the return value. Several of these were hitting WAW stalls for nearly 14,000 (poorly estimated) cycles a pop. Making dead code elimination null out the destination avoids this issue. This patch cuts one shader's estimated cycles by -98.39%! Removing the message response should also help with data cluster bandwidth. On Skylake: (instruction counts remain identical) total cycles in shared programs: 255413890 -> 248081010 (-2.87%) cycles in affected programs: 12019948 -> 4687068 (-61.01%) helped: 24 HURT: 10 v2: Make can_omit_write independent of can_eliminate (Curro). Signed-off-by: Kenneth Graunke <<a href="mailto:kenneth@whitecape.org">kenneth@whitecape.org</a>> Reviewed-by: Francisco Jerez <<a href="mailto:currojerez@riseup.net">currojerez@riseup.net</a>> Reviewed-by: Matt Turner <<a href="mailto:mattst88@gmail.com">mattst88@gmail.com</a>> :040000 040000 d3ac50f716bb55e2919ea2a8098656be67dffdc4 9763545331f08faf330b432997836bbcdce7e8c1 M src</pre> </div> <hr> You are receiving this mail because: <ul> <li>You are the QA Contact for the bug.</li> <li>You are the assignee for the bug.</li> </ul> </body> </html>