[Bug 110825] valgrind invalid reads when spilling

bugzilla-daemon at freedesktop.org bugzilla-daemon at freedesktop.org
Fri Jun 7 06:04:46 UTC 2019


https://bugs.freedesktop.org/show_bug.cgi?id=110825

--- Comment #5 from Tapani Pälli <lemody at gmail.com> ---
FYI now that I tested with KBL, I see following error/warning. I do see it
before the offending commit though so this is something that has existed
before. For some reason I did not spot these in HSW. That test eventually
crashes on HSW but passes on KBL with this single warning.

--- 8< ---
==12320== Memcheck, a memory error detector
==12320== Copyright (C) 2002-2017, and GNU GPL'd, by Julian Seward et al.
==12320== Using Valgrind-3.14.0 and LibVEX; rerun with -h for copyright info
==12320== Command: bin/shader_runner /home/tpalli/slow.shader_test -auto
==12320== 
==12320== Conditional jump or move depends on uninitialised value(s)
==12320==    at 0xC2A36FC: ra_get_best_spill_node (register_allocate.c:960)
==12320==    by 0xBF99499: fs_reg_alloc::choose_spill_reg()
(brw_fs_reg_allocate.cpp:937)
==12320==    by 0xBF9A0A8: fs_reg_alloc::assign_regs(bool, bool)
(brw_fs_reg_allocate.cpp:1161)
==12320==    by 0xBF9A3AA: fs_visitor::assign_regs(bool, bool)
(brw_fs_reg_allocate.cpp:1201)
==12320==    by 0xBF619A7: fs_visitor::allocate_registers(unsigned int, bool)
(brw_fs.cpp:7282)
==12320==    by 0xBF62D20: fs_visitor::run_fs(bool, bool) (brw_fs.cpp:7694)
==12320==    by 0xBF63CC8: brw_compile_fs (brw_fs.cpp:8028)
==12320==    by 0xB7BCE62: brw_codegen_wm_prog (brw_wm.c:123)
==12320==    by 0xB7BE25D: brw_fs_precompile (brw_wm.c:592)
==12320==    by 0xB7A423C: brw_shader_precompile(gl_context*,
gl_shader_program*) (brw_link.cpp:56)
==12320==    by 0xB7A541F: brw_link_shader (brw_link.cpp:374)
==12320==    by 0xBAC983D: _mesa_glsl_link_shader (ir_to_mesa.cpp:3170)
==12320==

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-3d-bugs/attachments/20190607/41fe642e/attachment.html>


More information about the intel-3d-bugs mailing list