Bug 52252

Bisect shows:bce58e155db7202a98642c10e6132dee4e08162b is the first bad commit.
commit bce58e155db7202a98642c10e6132dee4e08162b
Author:     Eric Anholt <eric@anholt.net>
AuthorDate: Wed Jul 4 10:52:32 2012 -0700
Commit:     Chad Versace <chad.versace@linux.intel.com>
CommitDate: Tue Jul 17 08:17:56 2012 -0700

    intel: Convert to using private depth/stencil buffers (v2)

    This means that GLX buffer sharing of these no longer works.  On the
    other hand, just *look* at this code reduction.

    v2:
      - [chad] Fix intelCreateBuffer for gen < 6. When the branch for
        !screen->hw_has_separate_stencil was taken,
        intel_create_private_renderbuffer was incorrectly not used.

      - [chad] Remove all code in intel_process_dri2_buffer for processing
        depth, stencil, and hiz buffers. That code is now dead.

    CC: Eric Anholt <eric@anholt.net>
    Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
    Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>

Assigning to self (chadv). The assertion failure occurs in HiZ code.

I am unable to reproduce this on Ivybridge with either Mesa master-996ff1c or the bisected commit. In my attempt to reproduce the bug, I used the following ~/.xinitrc.

Lu, can you still reproduce this bug? If so, please provide a backtrace.

--------
# file: ~/.xinitrc
startkde &
for ((x=0;x<1024;++x))
do
    if ! ./glx-visuals-depth -auto
    then
        echo "fail: $x"
        exit 1
     fi
done
--------

After googling I found this bz.  I'm also experiencing this issue with mesa-8.1_rc1 (Gentoo Linux)  and Totem on SandyBridge Mobile.  Specifically when I attempt to resize or fullscreen the totem window it happens 100% of the time.

    DRM_IOCTL_I915_GEM_CONTEXT_CREATE failed: Invalid argument
    DRM_IOCTL_I915_GEM_CONTEXT_CREATE failed: Invalid argument
    totem: intel_resolve_map.c:46: intel_resolve_map_set: Assertion `(*tail)->need == need' failed.
    Aborted

Backtrace:
#0  0x00000037d4036415 in raise () from /lib64/libc.so.6
#1  0x00000037d4037d2b in abort () from /lib64/libc.so.6
#2  0x00000037d402ed0d in __assert_fail () from /lib64/libc.so.6
#3  0x00007ffff4ed3055 in intel_resolve_map_set (head=<optimized out>, level=0,
    layer=0, need=GEN6_HIZ_OP_DEPTH_RESOLVE) at intel_resolve_map.c:46
#4  0x00007ffff4ee5f2a in brw_postdraw_set_buffers_need_resolve (brw=0x64b880)
    at brw_draw.c:350
#5  brw_try_draw_prims (max_index=<optimized out>, min_index=<optimized out>, ib=0x0,
    nr_prims=1, prim=0x7fffffffdf80, arrays=0x995b98, ctx=0x64b880) at brw_draw.c:539
#6  brw_draw_prims (ctx=0x64b880, prim=0x7fffffffdf80, nr_prims=1, ib=0x0,
    index_bounds_valid=<optimized out>, min_index=0, max_index=3, tfb_vertcount=0x0)
    at brw_draw.c:583
#7  0x00007ffff4a46709 in vbo_draw_arrays (ctx=0x64b880, mode=7, start=0,
    count=<optimized out>, numInstances=1, baseInstance=<optimized out>)
    at ./../vbo/vbo_exec_array.c:645
#8  0x00007ffff7d4e0bf in piglit_draw_rect (x=0, y=0, w=10, h=20)
    at /GFX/Test/Piglit/piglit/tests/util/piglit-util-gl.c:817
#9  0x0000000000400cd4 in draw (dpy=0x602010, config=0x624cb0)
    at /GFX/Test/Piglit/piglit/tests/glx/glx-visuals-stencil.c:64
#10 0x00007ffff7d4f92a in piglit_iterate_visuals_event_loop (dpy=0x602010,
    draw=0x400bd4 <draw>, config=0x624cb0)
    at /GFX/Test/Piglit/piglit/tests/util/piglit-glx-util.c:224
#11 0x00007ffff7d4fd76 in piglit_glx_iterate_visuals (draw=0x400bd4 <draw>)
    at /GFX/Test/Piglit/piglit/tests/util/piglit-glx-util.c:362
#12 0x0000000000400eca in main (argc=2, argv=0x7fffffffe308)
    at /GFX/Test/Piglit/piglit/tests/glx/glx-visuals-stencil.c:107

Patch is submitted to mesa-dev.

Author: Chad Versace <chad.versace@linux.intel.com>
Date:   Tue Aug 14 11:23:21 2012 -0700

    i965: Move hiz resolve to after renderbuffer resizing
    
    Do all pre-draw hiz resolves *after* the renderbuffers are resized by
    intel_prepare_render. Otherwise, we may resolve buffers that are
    immediately discarded afterwards.
    
    Fixes the assertion failure below when resizing windows in KDE and under
    some unknown circumstance in Chrome OS:
        intel_resolve_map.c:46: intel_resolve_map_set: Assertion
        `(*tail)->need == need' failed.
    
    Also, remove the comment that "resolves must occur [...] before setting up
    any hardware state". That was true when resolves were implemented with
    meta-ops, but no longer with blorp.
    
    CC: Stephane Marchesin <marcheu@chromium.org>
    Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=52252
    Reported-by: Lu Hua <huax.lu@intel.com>
    Signed-off-by: Chad Versace <chad.versace@linux.intel.com>

Can't find "i965: Move hiz resolve to after renderbuffer resizing" on mesa master commit 1597176f7090eea73f41b3114ae2a02a50ac7a12.
Issue still exists.

(In reply to comment #7)
> Can't find "i965: Move hiz resolve to after renderbuffer resizing" on mesa
> master commit 1597176f7090eea73f41b3114ae2a02a50ac7a12.
> Issue still exists.

The patch has not yet been committed to master; it is still under reviewon the mailing list. In comment #6, I stated that the "Patch is submitted to mesa-dev".

*** Bug 53495 has been marked as a duplicate of this bug. ***

Fixed by

commit 88edbdf9f0c79e05593a71929f791fa4efb08ffb
Author: Chad Versace <chad.versace@linux.intel.com>
Date:   Tue Aug 14 11:23:21 2012 -0700

    i965: Move hiz resolve to after renderbuffer resizing (v2)

Verified.Fixed on master branch commit 87a34131c427b40a561cfef1513b446a0eeabc39.