Created attachment 1535974 [details]
coredump

xfwm4 --vblank=glx causes an assertion fail in i965_dri.so. Without the vblank argument everything works. ABRT won't let me report the issue automatically because it says that the coredump has "low informational value". Here's the text backtrace:

#0  0x00007fed689bd53f in __GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:50
#1  0x00007fed689a7895 in __GI_abort () at abort.c:79
#2  0x00007fed689a7769 in __assert_fail_base
    (fmt=0x7fed68b0ee90 "%s%s%s:%u: %s%sAssertion `%s' failed.\n%n", assertion=0x7fed663240a0 "image->TexObject->Target == mt->target", file=0x7fed66323f30 "../src/mesa/drivers/dri/i965/intel_mipmap_tree.c", line=1287, function=<optimized out>) at assert.c:92
#3  0x00007fed689b59f6 in __GI___assert_fail
    (assertion=assertion@entry=0x7fed663240a0 "image->TexObject->Target == mt->target", file=file@entry=0x7fed66323f30 "../src/mesa/drivers/dri/i965/intel_mipmap_tree.c", line=line@entry=1287, function=function@entry=0x7fed66325790 <__PRETTY_FUNCTION__.38805> "intel_miptree_match_image") at assert.c:101
#4  0x00007fed65cf4f43 in intel_miptree_match_image (mt=<optimized out>, image=image@entry=0x55c4804e0150) at ../src/mesa/main/macros.h:666
#5  0x00007fed65d0470a in intel_finalize_mipmap_tree (brw=brw@entry=0x55c48027a770, tObj=0x55c4804539c0) at ../src/mesa/drivers/dri/i965/intel_tex_validate.c:111
#6  0x00007fed65d04a5c in brw_validate_textures (brw=brw@entry=0x55c48027a770) at ../src/mesa/drivers/dri/i965/intel_tex_validate.c:195
#7  0x00007fed65cc82ce in brw_prepare_drawing (max_index=3, min_index=0, index_bounds_valid=<optimized out>, ib=0x0, ctx=0x55c48027a770) at ../src/mesa/drivers/dri/i965/brw_draw.c:811
#8  0x00007fed65cc82ce in brw_draw_prims
    (ctx=0x55c48027a770, prims=0x55c4802b3f48, nr_prims=1, ib=0x0, index_bounds_valid=<optimized out>, min_index=<optimized out>, max_index=<optimized out>, gl_xfb_obj=0x0, stream=0, indirect=0x0) at ../src/mesa/drivers/dri/i965/brw_draw.c:1070
#9  0x00007fed65e76f3c in vbo_exec_vtx_flush (exec=exec@entry=0x55c4802b3750, keepUnmapped=keepUnmapped@entry=1 '\001') at ../src/mesa/vbo/vbo_exec_draw.c:393
#10 0x00007fed65e767db in vbo_exec_FlushVertices_internal (unmap=1 '\001', exec=<optimized out>) at ../src/mesa/vbo/vbo_exec_api.c:1255
#11 0x00007fed65e767db in vbo_exec_FlushVertices (ctx=ctx@entry=0x55c48027a770, flags=flags@entry=1) at ../src/mesa/vbo/vbo_exec_api.c:1255
#12 0x00007fed65dee155 in _mesa_PopMatrix () at ../src/mesa/main/matrix.c:287
#13 0x000055c47e388da4 in paint_all (screen_info=<optimized out>, region=<optimized out>, buffer=<optimized out>) at compositor.c:1575
#14 0x000055c47e3891ec in compositor_timeout_cb (data=<optimized out>) at compositor.c:2346
#15 0x00007fed68eabb31 in  () at /lib64/libglib-2.0.so.0
#16 0x00007fed68eab06d in g_main_context_dispatch () at /lib64/libglib-2.0.so.0
#17 0x00007fed68eab438 in  () at /lib64/libglib-2.0.so.0
#18 0x00007fed68eab762 in g_main_loop_run () at /lib64/libglib-2.0.so.0
#19 0x00007fed696ddb6d in gtk_main () at /lib64/libgtk-3.so.0
#20 0x000055c47e37e0b7 in main (argc=<optimized out>, argv=<optimized out>) at main.c:762

Coredump is attached.

The fail happens with mesa-dri-drivers-18.3.3-1.fc29.x86_64 but used to work with mesa-dri-drivers-18.2.8-1.fc29.x86_64. xfwm4 is from xfwm4-4.13.1-4.fc29.x86_64

The only reference I could find is this one https://bugs.freedesktop.org/show_bug.cgi?id=107117 , it's resolved as fixed but evidently it's not, or Fedora is not receiving the upstream patch somehow.