<div dir="ltr">Thanks for the replies. <div><br></div><div>I tried to upgrade LLVM from 3.3 to 3.4 after I sent the e-mail but ended up having "an internal compiler error" from gcc linaro toolchain compiler. Nice. I could try a more up-to-date version of Mesa to see if it helps. Although I have learned that I can not take the most up-to-date version since something has gotten broken related to SW rendering/wayland/egl.</div><div><br></div><div>Marko</div><div><br></div><div> </div></div><div class="gmail_extra"><br><div class="gmail_quote">On Wed, Apr 15, 2015 at 6:45 PM, Jan Vesely <span dir="ltr"><<a href="mailto:jan.vesely@rutgers.edu" target="_blank">jan.vesely@rutgers.edu</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hi,<br>
I can run arm llvmpipe glxgears on IFC6410 (Snapdragon 600 APQ8064,<br>
Krait 300) without problems.<br>
llvm 3.5<br>
mesa 10.4.3<br>
both are fedora 21 standard packages without modifications.<br>
I haven't tried anything more complex than glxgears<br>
<br>
jan<br>
<br>
On Wed, 2015-04-15 at 16:30 +0200, Roland Scheidegger wrote:<br>
> There were reports with varying success of running llvmpipe with arm in<br>
> the past but I haven't heard anything for a while. In theory it should<br>
> work (and crashing in the pixel shader is an indication the vertex<br>
> shader worked or at least didn't crash).<br>
> But it's not really tested and hence might easily break. My suggestion<br>
> would be to try a newer Mesa version since 9.2.5 is very old, but I<br>
> really don't know if it will help.<br>
> If that doesn't help, you could try debugging it, you can get the<br>
> crashing instruction in gdb with x/i <address> and see if that was an<br>
> unaligned access which should have been aligned or something else.<br>
> GALLIVM_DEBUG=ir,asm will get you the IR and compiled code (with debug<br>
> builds).<br>
><br>
> Roland<br>
><br>
><br>
> Am 15.04.2015 um 14:37 schrieb Marko Moberg:<br>
> > Hi,<br>
> ><br>
> > I have been trying to get OpenGL sw rastering to work on Linux based ARM<br>
> > CortexA15 device but for some reason gallium llvmpipe gives me a<br>
> > segmentation fault. Has anybody managed to run sw rastering using<br>
> > gallium llvmpipe on ARM? Softpipe seems to be working ok.<br>
> ><br>
> > I am using the following versions:<br>
> > - Mesa 9.2.5<br>
> > - LLVM3.3<br>
> > - Wayland/Weston 1.5<br>
> ><br>
> > The issue is related to opengl/gles2 rendering i.e. weston-simple-egl<br>
> > demo and Qt openGL demos give seg faults when executed with llvmpipe<br>
> > enabled.<br>
> ><br>
> > Backtrace from weston-simple-egl looks something like this:<br>
> > #0 0x6d7de620 in ?? ()<br>
> > #1 0xb354004c in fs17_variant0_partial ()<br>
> > #2 0xb6841740 in lp_rast_shade_quads_mask (task=task@entry=0x1e360,<br>
> > inputs=inputs@entry=0x47a00, x=128, y=80, mask=4369) at lp_rast.c:466<br>
> > #3 0xb6842d04 in do_block_4_1 (c=<optimized out>, y=<optimized out>,<br>
> > x=<optimized out>, plane=<optimized out>, tri=<optimized out>,<br>
> > task=<optimized out>) at lp_rast_tri_tmp.h:61<br>
> > #4 do_block_16_1 (c=<synthetic pointer>, y=<optimized out>,<br>
> > x=<optimized out>, plane=0xb4f4ccf0, tri=<optimized out>,<br>
> > task=<optimized out>) at lp_rast_tri_tmp.h:130<br>
> > #5 lp_rast_triangle_1 (task=0x1e360, arg=...) at lp_rast_tri_tmp.h:232<br>
> > #6 0xb6840420 in do_rasterize_bin (bin=<optimized out>, task=0x1e360,<br>
> > x=<optimized out>, y=<optimized out>) at lp_rast.c:607<br>
> > #7 rasterize_bin (y=<optimized out>, x=<optimized out>, bin=<optimized<br>
> > out>,<br>
> > task=0x1e360) at lp_rast.c:626<br>
> > #8 rasterize_scene (task=task@entry=0x1e360, scene=0xb3646008)<br>
> > at lp_rast.c:675<br>
> > #9 0xb6840cb0 in thread_function (init_data=0x1e360) at lp_rast.c:788<br>
> > #10 0xb6d98ed2 in start_thread () from /lib/libpthread.so.0<br>
> > #11 0xb6c91058 in ?? () from /lib/libc.so.6<br>
> > #12 0xb6c91058 in ?? () from /lib/libc.so.6<br>
> ><br>
> > regards,<br>
> > Marko<br>
> ><br>
> > <a href="mailto:marko.s.moberg@gmail.com">marko.s.moberg@gmail.com</a> <mailto:<a href="mailto:marko.s.moberg@gmail.com">marko.s.moberg@gmail.com</a>><br>
> ><br>
> ><br>
> ><br>
> > _______________________________________________<br>
> > mesa-dev mailing list<br>
> > <a href="mailto:mesa-dev@lists.freedesktop.org">mesa-dev@lists.freedesktop.org</a><br>
> > <a href="https://urldefense.proofpoint.com/v2/url?u=http-3A__lists.freedesktop.org_mailman_listinfo_mesa-2Ddev&d=AwIGaQ&c=Sqcl0Ez6M0X8aeM67LKIiDJAXVeAw-YihVMNtXt-uEs&r=Vjtt0vs_iqoI31UfJxBl7yv9I2FeiaeAYgMTLKRBc_I&m=5qZ4KpVhu_wXA8-dTn8xYiul5G713jYwztPr8kfcUZM&s=wBNyDvf5DiNBjxXXl7QvyHOvNuTVjZAaw0n7WaQM-bA&e=" target="_blank">https://urldefense.proofpoint.com/v2/url?u=http-3A__lists.freedesktop.org_mailman_listinfo_mesa-2Ddev&d=AwIGaQ&c=Sqcl0Ez6M0X8aeM67LKIiDJAXVeAw-YihVMNtXt-uEs&r=Vjtt0vs_iqoI31UfJxBl7yv9I2FeiaeAYgMTLKRBc_I&m=5qZ4KpVhu_wXA8-dTn8xYiul5G713jYwztPr8kfcUZM&s=wBNyDvf5DiNBjxXXl7QvyHOvNuTVjZAaw0n7WaQM-bA&e=</a><br>
> ><br>
><br>
> _______________________________________________<br>
> mesa-dev mailing list<br>
> <a href="mailto:mesa-dev@lists.freedesktop.org">mesa-dev@lists.freedesktop.org</a><br>
> <a href="http://lists.freedesktop.org/mailman/listinfo/mesa-dev" target="_blank">http://lists.freedesktop.org/mailman/listinfo/mesa-dev</a><br>
<span class="HOEnZb"><font color="#888888"><br>
--<br>
Jan Vesely <<a href="mailto:jan.vesely@rutgers.edu">jan.vesely@rutgers.edu</a>><br>
</font></span></blockquote></div><br></div>