Created attachment 114970 [details] Xorg.log Test Environment -------------------- Failed Platform: Broadwell Failed GFX SW stack Libdrm: (master)libdrm-2.4.60-18-gbc6b5174d362a01f6d638a4db6aaf2fd355db69c Mesa: (master)8f7338f284cdb1fef64c85e3293d2200d0cc6387 Xserver: (master)xorg-server-1.17.0-72-gf1da6bf5d94911e78d2e27e6accf0c6e3aefb331 Xf86_video_intel: (master)2.99.917-219-g745c21fb1aa1fcd54e9a9aeaa7fe6877af39f2f1 Cairo: (master)71392653803277cec6deea9e7aa4696f8b22dd73 Kernel: (drm-intel-nightly)d72ff1ab1499711c941831400629c14493313b3a Mesa Regression: Yes First bad commit is: commit 47c4b3854076adfe5a27b537f36262ac4ec4530d Author: Matt Turner <mattst88@gmail.com> AuthorDate: Tue Jan 27 19:18:46 2015 -0800 Commit: Matt Turner <mattst88@gmail.com> CommitDate: Tue Mar 31 14:14:36 2015 -0700 i965/fs: Allow CSE to handle MULs with negated arguments. mul x, -y is equivalent to mul -x, y; and mul x, y is the negation of mul x, -y. With NIR: total instructions in shared programs: 6167779 -> 6161193 (-0.11%) instructions in affected programs: 983511 -> 976925 (-0.67%) helped: 4106 HURT: 16 GAINED: 18 LOST: 7 Without NIR: total instructions in shared programs: 6192323 -> 6185299 (-0.11%) instructions in affected programs: 987875 -> 980851 (-0.71%) helped: 4146 HURT: 16 GAINED: 16 LOST: 0 Bug Description 4 Synmark2_v6 cases have been impacted by the bad commit: Image validation will fail when running case, we cannot get the performance data for these 4 cases on BDW now. OglDrvRes OglDrvShComp OglDrvState OglPSPom root@x-bdw21:/usr/local/games/OpenGL/SynMark2_6# ./synmark2 OglDrvRes -- SynMark2 6.0.0 -- 10:04:57 04/09/15 -- Test name: OglDrvRes Config name: Default Validation: Failed - different (1.219 %), different (0.915 %) OglDrvRes: Image validation failed - different (1.219 %), different (0.915 %) Reproduce step: 1. cd /usr/local/games/OpenGL/SynMark2_6 2. ./synmark2 OglDrvRes
Created attachment 114971 [details] dmesg log file
Interesting. Thanks. I'll have a look.
This happens also on BSW i.e. it's GEN8 specific. The main failing test is PSPom, the other two are composite tests, containing pieces of multiple other tests, including PSPom.
Patch sent.
Fixed with commit 3ca17e75e4fe129511a4dcad47e139485beef880 Author: Matt Turner <mattst88@gmail.com> Date: Mon Apr 13 11:29:14 2015 -0700 i965/fs: Correct mistake in determining whether a MUL is negated. a * b is equivalent to -a * -b, and the previous code was failing at that. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89961 Reviewed-by: Anuj Phogat <anuj.phogat@gmail.com> Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> Thanks for the report!
Verified on BDW with 2015-04-15 GFX SW stack, this issue been fixed. Libdrm: (master)libdrm-2.4.60-33-g32258e4dbd32466b4286ab2ad7883b7cb90c105f Mesa: (master)3cbefe3cf4c745c7c681cfc18a1e47461fec91db Xserver: (master)xorg-server-1.17.0-74-g41932dfbc841a1adc6512d41085ea3f8ebecb42c Xf86_video_intel: (master)2.99.917-266-g66f3db6789727b99ec34afe3ceaecc9e549e55ee Cairo: (master)2cf2d8e340a325adb205baf8e4bd64e1d1858008 Libva: (master)062a63932c0f1439aa587aa986bbcfb758ff38f2 Libva_intel_driver: (master)ed03aebc6e702dab65204cc1469eef0da73e2372 Kernel: (drm-intel-nightly)5ea91de4ff45adb60031853d64314c3405378fbd
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.