Created attachment 21419 [details] log System Environment: -------------------------- --Platform: G45 --Architecture(32-bit,64-bit,compatiblity): 64-bit --2D driver: (xf86-video-intel-2.6-branch)83f3c376b5942e134047a220e6e5f2432ffc492c --3D driver: (intel-2008-q4)f96baeaac3ef41260ac3975750627ece073fdce0 --DRM: shipped with kernel 2.6.28-rc8-for-airlied --libdrm: (master)0243c9f801a35de3465a0321c02f18a4d07ce5b8 --Xserver: (server-1.6-branch)32e81074b967716865aef08b66ec29caf0fec2c5 --Kernel: 2.6.28-rc8-for-airlied Bug detailed description: -------------------------- start X and run SpecViewPerf, after it run the viewset grabs/test11full.png, it quit with a segmentation fault. following is the output: Writing PNG file '../../viewperf/viewperf10.0/results/ugnx-01/grabs/test10full.png'...done. libGL: XF86DRIGetClientDriverName: 1.9.0 i965 (screen 0) libGL: OpenDriver: trying /opt/X11R7/lib/dri/i965_dri.so drmOpenDevice: node name is /dev/dri/card0 drmOpenDevice: open result is 6, (OK) drmOpenByBusid: Searching for BusID pci:0000:00:02.0 drmOpenDevice: node name is /dev/dri/card0 drmOpenDevice: open result is 6, (OK) drmOpenByBusid: drmOpenMinor returns 6 drmOpenByBusid: drmGetBusid reports pci:0000:00:02.0 GetXEnvironInfo: no window set vpGrabScreen Writing PNG file '../../viewperf/viewperf10.0/results/ugnx-01/grabs/test11small.png'...done. vpGrabScreen Writing PNG file '../../viewperf/viewperf10.0/results/ugnx-01/grabs/test11full.png'...done. GetXEnvironInfo: no window set Segmentation fault Reproduce steps: ---------------- 1. xinit & 2. ./Run_All.csh
Created attachment 21420 [details] conf
Actually this is an old issue, and it's case bug. See bug#10658.
(In reply to comment #2) > Actually this is an old issue, and it's case bug. See bug#10658. > It seems not. It works well until the last line " $viewperf -generateScore -viewset "ugnx-01" -versionpath $1" is ran. Following is the run script. ################ $viewperf -viewset "ugnx-01" -versionpath $1 -testnum 7 -weight 9.0 -dl -mh EngineCooling-shaded+6lights $commonOptions -grab grab.scr -alternateShareFile 2 -alternatePackFile 1 -ignoreExtraLists $viewperf -viewset "ugnx-01" -versionpath $1 -testnum 8 -weight 9.0 -dl -mh bodyparts-wireframe+linesmooth $commonOptions -grab grab.scr -alternateShareFile 1 $viewperf -viewset "ugnx-01" -versionpath $1 -testnum 9 -weight 9.0 -dl -mh bodyparts-shaded $commonOptions -grab grab.scr -alternateShareFile 1 $viewperf -viewset "ugnx-01" -versionpath $1 -testnum 10 -weight 9.0 -dl -mh bodyparts-shaded+edges $commonOptions -grab grab.scr -alternateShareFile 1 $viewperf -viewset "ugnx-01" -versionpath $1 -testnum 11 -weight 10.0 -dl -mh bodyparts-shaded+6lights $commonOptions -grab grab.scr -alternateShareFile 1 $viewperf -generateScore -viewset "ugnx-01" -versionpath $1 ################ And the bt info: Current language: auto; currently asm (gdb) bt #0 glPixelStorei () at ../../../src/mesa/x86-64/glapi_x86-64.S:8947 #1 0x000000000040942d in myCreateFont () #2 0x0000000000418b44 in main () ã
Case issue. totalNumVertices is 4 and SUB_STROKEMAX is 20 in this case. It results in invalid write (please see vierperf.c lines 11324-11375).
Shuang, please follow up this case issue.
(In reply to comment #4) > Case issue. > totalNumVertices is 4 and SUB_STROKEMAX is 20 in this case. It results in > invalid write (please see vierperf.c lines 11324-11375). > Ok, I had a look, found the symptom is exactly like the reporter stated. but seems this segment fault is from SpecViewPerf10 itself. at this line in file viewsets/ugnx-01/ugnx-01.csh (It seems it's used to generate a score for this case): $viewperf -generateScore -viewset "ugnx-01" -versionpath $1 So I will agree to mark it as not our bug, verified.
Oh, just debug a bit. seems it's still a issue with mesa, see the backtrace: (gdb) bt #0 glPixelStorei () at ../../../src/mesa/x86-64/glapi_x86-64.S:8947 #1 0x000000000040945d in myCreateFont () #2 0x0000000000418b24 in main () I'd have to reopen it.
Shuang, does this issue still exist? It should be a case issue, the stack is destroyed by viewperf.
(In reply to comment #8) > Shuang, does this issue still exist? > It should be a case issue, the stack is destroyed by viewperf. > It still happen. I agree with you that it's a case issue. so close it.
verified
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.