Bug 31383

Summary:	[r600g] crash when using xeglthreads
Product:	Mesa	Reporter:	Kevin DeKorte <kdekorte>
Component:	Drivers/Gallium/r600	Assignee:	Default DRI bug account <dri-devel>
Status:	RESOLVED FIXED	QA Contact:
Severity:	normal
Priority:	medium
Version:	git
Hardware:	Other
OS:	All
Whiteboard:
i915 platform:		i915 features:

Description Kevin DeKorte 2010-11-04 09:37:34 UTC

Using mesa git e7f5d19a1106351f2db8f62f59f51be86eaa93df
kernel 2.6.36 (kernel-2.6.35.6-37.fc13.x86_64)
libdrm git
xf86-video-ati git 

I'm able to duplicate a crash in r600g pretty reliably with this command 
./xeglthreads -n 6 -t -l

it gives an error:  r600_priv.h:184: radeon_bo_unmap: Assertion `bo->map_count >= 0' failed.

works ok with -n values < 6, but real performance drop between 4 & 5

Comment 1 Kevin DeKorte 2010-11-04 09:45:01 UTC

Seems to also happen with this command

./glthreads -n 3 -t
glthreads: No explict locking.
glthreads: Single display connection.
XInitThreads() returned 1 (success)
glthreads: creating windows
glthreads: creating threads
glthreads: Created thread 0x7f83360d3700
glthreads: Created thread 0x7f83358d2700
glthreads: Created thread 0x7f8327fff700
glthreads: 1: GL_RENDERER = Gallium 0.4 on RV635
glthreads: 0: GL_RENDERER = Gallium 0.4 on RV635
glthreads: 2: GL_RENDERER = Gallium 0.4 on RV635
glthreads: r600_priv.h:184: radeon_bo_unmap: Assertion `bo->map_count >= 0' failed.

Comment 2 Kevin DeKorte 2010-11-04 09:46:17 UTC

with glthreads you need to press the arrow keys to make it happen

Comment 3 Kevin DeKorte 2010-11-05 12:59:49 UTC

backtrace as requested by tilman

run -n 6 -t -l
Starting program: /home/kdekorte/git/mesa/progs/egl/xeglthreads -n 6 -t -l
[Thread debugging using libthread_db enabled]
xeglthreads: Using explicit locks around Xlib calls.
xeglthreads: Single display connection.
xeglthreads: creating windows
xeglthreads: creating threads
[New Thread 0x7ffff5ae0700 (LWP 29394)]
xeglthreads: Created thread 0x7ffff5ae0700
[New Thread 0x7ffff52df700 (LWP 29395)]
xeglthreads: Created thread 0x7ffff52df700
[New Thread 0x7ffff4ade700 (LWP 29396)]
xeglthreads: Created thread 0x7ffff4ade700
[New Thread 0x7fffeffff700 (LWP 29397)]
xeglthreads: Created thread 0x7fffeffff700
[New Thread 0x7fffef7fe700 (LWP 29398)]
xeglthreads: Created thread 0x7fffef7fe700
[New Thread 0x7fffeeffd700 (LWP 29399)]
xeglthreads: Created thread 0x7fffeeffd700
xeglthreads: 0: GL_RENDERER = Gallium 0.4 on RV635
xeglthreads: 4: GL_RENDERER = Gallium 0.4 on RV635
xeglthreads: 3: GL_RENDERER = Gallium 0.4 on RV635
xeglthreads: 2: GL_RENDERER = Gallium 0.4 on RV635
xeglthreads: 1: GL_RENDERER = Gallium 0.4 on RV635
xeglthreads: 5: GL_RENDERER = Gallium 0.4 on RV635
xeglthreads: r600_priv.h:184: radeon_bo_unmap: Assertion `bo->map_count >= 0' failed.

Program received signal SIGABRT, Aborted.
[Switching to Thread 0x7fffef7fe700 (LWP 29398)]
0x0000003f0e634065 in raise () from /lib64/libc.so.6
Missing separate debuginfos, use: debuginfo-install expat-2.0.1-10.fc13.x86_64 glibc-2.12.90-18.x86_64 libX11-1.3.4-3.fc14.x86_64 libXau-1.0.6-1.fc14.x86_64 libXdamage-1.1.3-1.fc14.x86_64 libXext-1.1.2-2.fc14.x86_64 libXfixes-4.0.5-1.fc14.x86_64 libXxf86vm-1.1.0-1.fc13.x86_64 libdrm-2.4.23-0.1.20101019.fc14.x86_64 libgcc-4.5.1-4.fc14.x86_64 libstdc++-4.5.1-4.fc14.x86_64 libtalloc-2.0.1-1.fc13.x86_64 libv4l-0.8.1-1.fc14.x86_64 libxcb-1.7-1.fc14.x86_64
(gdb) bt
#0  0x0000003f0e634065 in raise () from /lib64/libc.so.6
#1  0x0000003f0e635a16 in abort () from /lib64/libc.so.6
#2  0x0000003f0e62c8a5 in __assert_fail () from /lib64/libc.so.6
#3  0x00007ffff76e4653 in radeon_bo_unmap (_mgr=<value optimized out>) at r600_priv.h:184
#4  radeon_bo_pbmgr_flush_maps (_mgr=<value optimized out>) at radeon_bo_pb.c:264
#5  0x00007ffff7730b91 in r600_context_flush (ctx=0x1bc6f30) at r600_hw_context.c:1085
#6  0x00007ffff6f96468 in st_context_flush (stctxi=<value optimized out>, flags=9, 
    fence=<value optimized out>) at state_tracker/st_manager.c:508
#7  0x00007ffff7968971 in egl_g3d_swap_buffers (drv=<value optimized out>, 
    dpy=<value optimized out>, surf=0x2156820) at common/egl_g3d_api.c:602
#8  0x00007ffff7bb789c in eglSwapBuffers (dpy=0x612ba0, surface=0x2156820) at eglapi.c:681
#9  0x000000000040310f in draw_loop (p=0x604280) at xeglthreads.c:301
#10 thread_function (p=0x604280) at xeglthreads.c:567
#11 0x0000003f0f206d5b in start_thread () from /lib64/libpthread.so.0
#12 0x0000003f0e6e427d in clone () from /lib64/libc.so.6

Comment 4 Kevin DeKorte 2010-11-05 13:00:30 UTC

gdb> thread apply all bt

Thread 7 (Thread 0x7fffeeffd700 (LWP 29399)):
#0  0x0000003f0f20e1ac in __lll_lock_wait () from /lib64/libpthread.so.0
#1  0x0000003f0f2094a4 in _L_lock_997 () from /lib64/libpthread.so.0
#2  0x0000003f0f2092ba in pthread_mutex_lock () from /lib64/libpthread.so.0
#3  0x00000000004032cc in draw_loop (p=0x6042c8) at xeglthreads.c:249
#4  thread_function (p=0x6042c8) at xeglthreads.c:567
#5  0x0000003f0f206d5b in start_thread () from /lib64/libpthread.so.0
#6  0x0000003f0e6e427d in clone () from /lib64/libc.so.6

Thread 6 (Thread 0x7fffef7fe700 (LWP 29398)):
#0  0x0000003f0e634065 in raise () from /lib64/libc.so.6
#1  0x0000003f0e635a16 in abort () from /lib64/libc.so.6
#2  0x0000003f0e62c8a5 in __assert_fail () from /lib64/libc.so.6
#3  0x00007ffff76e4653 in radeon_bo_unmap (_mgr=<value optimized out>) at r600_priv.h:184
#4  radeon_bo_pbmgr_flush_maps (_mgr=<value optimized out>) at radeon_bo_pb.c:264
#5  0x00007ffff7730b91 in r600_context_flush (ctx=0x1bc6f30) at r600_hw_context.c:1085
#6  0x00007ffff6f96468 in st_context_flush (stctxi=<value optimized out>, flags=9, 
    fence=<value optimized out>) at state_tracker/st_manager.c:508
#7  0x00007ffff7968971 in egl_g3d_swap_buffers (drv=<value optimized out>, 
    dpy=<value optimized out>, surf=0x2156820) at common/egl_g3d_api.c:602
#8  0x00007ffff7bb789c in eglSwapBuffers (dpy=0x612ba0, surface=0x2156820) at eglapi.c:681
#9  0x000000000040310f in draw_loop (p=0x604280) at xeglthreads.c:301
#10 thread_function (p=0x604280) at xeglthreads.c:567
#11 0x0000003f0f206d5b in start_thread () from /lib64/libpthread.so.0
#12 0x0000003f0e6e427d in clone () from /lib64/libc.so.6

Thread 5 (Thread 0x7fffeffff700 (LWP 29397)):
#0  0x0000003f0e6ad5ed in nanosleep () from /lib64/libc.so.6
#1  0x0000003f0e6dce14 in usleep () from /lib64/libc.so.6
#2  0x0000000000402bba in draw_loop (p=0x604238) at xeglthreads.c:307
#3  thread_function (p=0x604238) at xeglthreads.c:567
#4  0x0000003f0f206d5b in start_thread () from /lib64/libpthread.so.0
#5  0x0000003f0e6e427d in clone () from /lib64/libc.so.6

Thread 4 (Thread 0x7ffff4ade700 (LWP 29396)):
#0  0x0000003f0f20e1ac in __lll_lock_wait () from /lib64/libpthread.so.0
#1  0x0000003f0f2094a4 in _L_lock_997 () from /lib64/libpthread.so.0
#2  0x0000003f0f2092ba in pthread_mutex_lock () from /lib64/libpthread.so.0
---Type <return> to continue, or q <return> to quit--- 
#3  0x000000000040319c in draw_loop (p=0x6041f0) at xeglthreads.c:299
#4  thread_function (p=0x6041f0) at xeglthreads.c:567
#5  0x0000003f0f206d5b in start_thread () from /lib64/libpthread.so.0
#6  0x0000003f0e6e427d in clone () from /lib64/libc.so.6

Thread 3 (Thread 0x7ffff52df700 (LWP 29395)):
#0  0x0000003f0f20e1ac in __lll_lock_wait () from /lib64/libpthread.so.0
#1  0x0000003f0f2094a4 in _L_lock_997 () from /lib64/libpthread.so.0
#2  0x0000003f0f2092ba in pthread_mutex_lock () from /lib64/libpthread.so.0
#3  0x000000000040319c in draw_loop (p=0x6041a8) at xeglthreads.c:299
#4  thread_function (p=0x6041a8) at xeglthreads.c:567
#5  0x0000003f0f206d5b in start_thread () from /lib64/libpthread.so.0
#6  0x0000003f0e6e427d in clone () from /lib64/libc.so.6

Thread 2 (Thread 0x7ffff5ae0700 (LWP 29394)):
#0  0x0000003f0f20e1ac in __lll_lock_wait () from /lib64/libpthread.so.0
#1  0x0000003f0f2094a4 in _L_lock_997 () from /lib64/libpthread.so.0
#2  0x0000003f0f2092ba in pthread_mutex_lock () from /lib64/libpthread.so.0
#3  0x00000000004032cc in draw_loop (p=0x604160) at xeglthreads.c:249
#4  thread_function (p=0x604160) at xeglthreads.c:567
#5  0x0000003f0f206d5b in start_thread () from /lib64/libpthread.so.0
#6  0x0000003f0e6e427d in clone () from /lib64/libc.so.6

Thread 1 (Thread 0x7ffff7ba7740 (LWP 29391)):
#0  0x0000003f0e6ad5ed in nanosleep () from /lib64/libc.so.6
#1  0x0000003f0e6dce14 in usleep () from /lib64/libc.so.6
#2  0x0000000000402596 in event_loop (argc=<value optimized out>, argv=<value optimized out>)
    at xeglthreads.c:389
#3  main (argc=<value optimized out>, argv=<value optimized out>) at xeglthreads.c:743

Comment 5 Kevin DeKorte 2010-12-21 10:01:49 UTC

Appears to be fixed with mesa git as of 12/21/10 and kernel 2.6.37-rc6

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.