Bug 104703

I've reproduced this with Chrome on SKL and HSW. I took a brief look and there might be some issues in the testsuite itself. I've dumped the shaders but I haven't figured out which shaders are used by 'shader06' test.


I can see at least following errors from apitrace that indicates browser/app issues:

--- 8< ---
54120: message: major api error 80: GL_INVALID_OPERATION in glGetNamedBufferPointerv(non-existent buffer object 21)
54120 @0 glDeleteBuffers(n = 1, buffers = &21)
54120: warning: glGetError(glDeleteBuffers) = GL_INVALID_OPERATION
54696: message: major api error 80: GL_INVALID_OPERATION in glGetNamedBufferPointerv(non-existent buffer object 16)
54696 @0 glDeleteBuffers(n = 1, buffers = &16)
54696: warning: glGetError(glDeleteBuffers) = GL_INVALID_OPERATION


...

92498: message: api issue 4: FBO incomplete: Unsupported HW texture/renderbuffer format attached: MESA_FORMAT_RGB_FLOAT32
92498: message: api issue 5: FBO incomplete: driver marked FBO as incomplete [-1]

...

112229: message: major shader compiler issue 120: 0:77(11): warning: `webgl_dfd20e96eeb1f964' used uninitialized
112229 @0 glCompileShader(shader = 74)
112229: warning: 0:77(11): warning: `webgl_dfd20e96eeb1f964' used uninitialized

Issue is reproducible on my laptop:

OS: Ubuntu 17.10 64-bit
CPU: Intel® Core™ i7-7500U CPU @ 2.70GHz × 4
GPU: Intel® HD Graphics 620 (Kaby Lake GT2)
mesa: 18.1.0-devel (git-7d4007d58a)
kernel: 4.13.0-31-generic

In dmesg I noticed this message after firefox crash:

[23838.361098] traps: Web Content[30658] trap divide error ip:7f6dee287e9f sp:7ffc7d493f20 error:0 in i965_dri.so[7f6dedf27000+8dd000]


Additionally I run firefox with gdb and got following error message:

Thread 1 "Web Content" received signal SIGFPE, Arithmetic exception.

0x00007f18ffe87e9f in evaluate_idiv (num_components=1, bit_size=<optimized out>, _src=_src@entry=0x7ffc0c2fb620) at nir/nir_constant_expressions.c:8791
8791	            int32_t dst = src0 / src1;
(gdb) 
Continuing.

At this point src1 equals to zero.

seems like something has changed, for me tests 4, 6, 9, 11 fail

Bisected:

40e9f2f13847ddd94e1216088aa00456d7b02d2b is the first bad commit
commit 40e9f2f13847ddd94e1216088aa00456d7b02d2b
Author: Timothy Arceri <timothy.arceri@collabora.com>
Date:   Tue Dec 13 11:37:25 2016 +1100

    i965: disable loop unrolling in GLSL IR
    
    There is a single regression in loop unrolling which is:
    
    loops HURT:   shaders/orbital_explorer.shader_test GS SIMD8:    0 -> 1
    
    However the loop is huge so it seems reasonable not to unroll it. It's
    surprising that GLSL IR does unroll it.
    
    shader-db results BDW:
    
    total instructions in shared programs: 13037455 -> 13036947 (-0.00%)
    instructions in affected programs: 17982 -> 17474 (-2.83%)
    helped: 63
    HURT: 25
    
    total cycles in shared programs: 262217870 -> 262227990 (0.00%)
    cycles in affected programs: 2287046 -> 2297166 (0.44%)
    helped: 969
    HURT: 844
    
    total loops in shared programs: 2951 -> 2952 (0.03%)
    loops in affected programs: 0 -> 1
    helped: 0
    HURT: 1
    
    LOST:   0
    GAINED: 1
    
    Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>

Issue is already fixed in https://bugs.freedesktop.org/show_bug.cgi?id=105271 by the following commit:

commit 0c1f37cc2d8555223ade73b244a3ee374be8d9cd
Author: Timothy Arceri <tarceri@itsqueeze.com>
Date:   Wed Feb 28 14:33:55 2018 +1100

    nir: fix interger divide by zero crash during constant folding
    
    From the GLSL 4.60 spec Section 5.9 (Expressions):
    
       "Dividing by zero does not cause an exception but does result in
        an unspecified value."
    
    Fixes: 89285e4d47a6 "nir: add new constant folding infrastructure"
    
    Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
    Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105271

So this can be marked as duplicate.

*** This bug has been marked as a duplicate of bug 105271 ***