- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
System:
Linux ubuntu64 4.4.0-34-generic #53-Ubuntu SMP Wed Jul 27 16:06:39 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
Intel(R) Core(TM) i5-4570 CPU @ 3.20GHz
libgl1-mesa-dri:amd64 11.2.0-1ubuntu2.1
Driver:
GL_VENDOR Intel Open Source Technology Center
GL_RENDERER Mesa DRI Intel(R) Haswell Desktop
GL_VERSION 3.3 (Core Profile) Mesa 11.2.0
GL_SHADING_LANGUAGE_VERSION 3.30
Steps to reproduce:
#include <fenv.h>
...feenableexcept( FE_DIVBYZERO | FE_INVALID | FE_OVERFLOW );
...
glClearColor(0.5,0,0,1);
glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT);
Expected results: running program.
Actual results: floating point exception, aborted process.
Thread 1 "game" received signal SIGFPE, Arithmetic exception.
The crash goes away using a different clear color like 1,1,1,1 or 0,0,0,1
Why on earth does the clear color matter?
The driver should not do any bad FP operations. Never divide by zero, not overflow or underflow. Intel QA should test their driver with FP exceptions enabled, so this would get caught.
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Some additional info:
stacktrace:
(gdb) where
#0 0x00007fffeed83459 in ?? () from /usr/lib/x86_64-linux-gnu/dri/i965_dri.so
#1 0x00007fffeed6ba41 in ?? () from /usr/lib/x86_64-linux-gnu/dri/i965_dri.so
#2 0x00007fffeed52818 in ?? () from /usr/lib/x86_64-linux-gnu/dri/i965_dri.so
#3 0x00007fffeed5ae5b in ?? () from /usr/lib/x86_64-linux-gnu/dri/i965_dri.so
#4 0x00007fffeed5baa7 in ?? () from /usr/lib/x86_64-linux-gnu/dri/i965_dri.so
#5 0x00007fffeed42a18 in ?? () from /usr/lib/x86_64-linux-gnu/dri/i965_dri.so
#6 0x0000000000451ed2 in Gfx::TargetClear (r=0.0100228256, g=0.0100228256, b=0.039681904, a=1, depth=1, stencil=0) at ../graphics/graphics.cpp:559
#7 0x000000000040d578 in Vis::Update (this=0x7ffffffeba90, elapsed=2.33999998e-07, numsteps=1, dt=0.00416666688, wld=0x7ffffffeeef8) at vis.cpp:535
#8 0x000000000042b9ca in main (argc=1, argv=0x7fffffffddb8) at main.cpp:124
(crash is in glClear())
And some HW is affected, others not.
Affected: GL_RENDERER Mesa DRI Intel(R) Haswell Desktop
Not affected: GL_RENDERER Mesa DRI Intel(R) Sandybridge Desktop
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi,
Unfortunately I can't help with Linux driver questions/issues. We do have some places you can go for help though.
For questions and issues with the Mesa driver, your best options are:
- File a bug in Mesa's public bug tracker: https://bugs.freedesktop.org/enter_bug.cgi?product=Mesa . In the component field, select Drivers/DRI./i965.
- Send a message to Mesa's development mailing list http://lists.freedesktop.org/mailman/listinfo/mesa-dev .
- Most of the Intel Mesa developers are on IRC in irc://irc.freenode.net/#intel-gfx or irc://irc.freenode.net/#dri-devel. People there are very helpful.
For LVDS issues, your best option is to contact one of Intel's kernel graphics developers. You can find them on IRC atirc://irc.freenode.net/#intel-gfx and on the mailing list http://lists.freedesktop.org/mailman/listinfo/intel-gfx
-Michael
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
>Unfortunately I can't help with Linux driver questions/issues. We do have some places you can go for help though.
Thanks Michael,
Full analysis continued here: https://lists.freedesktop.org/archives/mesa-dev/2016-September/127960.html
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page