Bug 105140

Summary: clearing related crash in Dying Light
Product: Mesa Reporter: Grazvydas Ignotas <notasas>
Component: Drivers/Gallium/radeonsiAssignee: Default DRI bug account <dri-devel>
Status: RESOLVED MOVED QA Contact: Default DRI bug account <dri-devel>
Severity: normal    
Priority: medium    
Version: git   
Hardware: Other   
OS: All   
Whiteboard:
i915 platform: i915 features:

Description Grazvydas Ignotas 2018-02-17 01:17:01 UTC
util_format_write_4f() gets called with a pipe_format PIPE_FORMAT_Z24_UNORM_S8_UINT that has no pack_rgba_float()

Thread 9 "gallium_drv:0" received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7f78b1397700 (LWP 30748)]
#0  0x0000000000000000 in ?? ()
#1  0x00007f78baf0e547 in util_format_write_4f (format=PIPE_FORMAT_Z24_UNORM_S8_UINT, src=0x1eb422c, src_stride=0, dst=0x7f78b13963a0, 
    dst_stride=0, x=0, y=0, w=1, h=1) at util/u_format.c:337
#2  0x00007f78bafd5853 in util_pack_color (rgba=0x1eb422c, format=PIPE_FORMAT_Z24_UNORM_S8_UINT, uc=0x7f78b13963a0)
    at ../../../../src/gallium/auxiliary/util/u_pack_color.h:433
#3  0x00007f78bafd5ad0 in si_set_clear_color (rtex=0x141f56e0, surface_format=PIPE_FORMAT_Z24_UNORM_S8_UINT, color=0x1eb422c)
    at si_clear.c:89
#4  0x00007f78bafd69f3 in si_do_fast_color_clear (sctx=0x1d60690, buffers=0x7f78b1396474, color=0x1eb422c) at si_clear.c:508
#5  0x00007f78bafd6b29 in si_clear (ctx=0x1d60690, buffers=4, color=0x1eb422c, depth=0, stencil=0) at si_clear.c:527
#6  0x00007f78baf344e7 in tc_batch_execute (job=job@entry=0x1eb4180, thread_index=thread_index@entry=0) at util/u_threaded_context.c:96
#7  0x00007f78badbb03f in util_queue_thread_func (input=input@entry=0x1e81550) at u_queue.c:271
#8  0x00007f78badbabc7 in impl_thrd_routine (p=<optimized out>) at ../../include/c11/threads_posix.h:87
#9  0x00007f78ce8e06ba in start_thread (arg=0x7f78b1397700) at pthread_create.c:333
#10 0x00007f78c734241d in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:109
Comment 1 Roland Scheidegger 2018-02-17 03:15:56 UTC
Looks like pretty much the same issue as 105139 to me.
Somehow a depth/stencil format texture gets attached as a color attachment (I can't think of any reason why anyone would even try that...), and if that gets through all bets are off.
Though the framebuffer completeness code does have code for checking color/ds format mismatch, so no idea how it could happen.
Comment 2 GitLab Migration User 2019-09-25 18:02:40 UTC
-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/mesa/mesa/issues/1303.

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.