<html> <head> <base href="https://bugs.freedesktop.org/"> </head> <body> <div> <a class="bz_bug_link bz_status_NEW " title="NEW - [SKL] ES31-CTS.core.sepshaderobjs.StateInteraction fails sporadically" href="https://bugs.freedesktop.org/show_bug.cgi?id=102177#c4">Comment # 4</a> on <a class="bz_bug_link bz_status_NEW " title="NEW - [SKL] ES31-CTS.core.sepshaderobjs.StateInteraction fails sporadically" href="https://bugs.freedesktop.org/show_bug.cgi?id=102177">bug 102177</a> from <a class="email" href="mailto:t_arceri@yahoo.com.au" title="Timothy Arceri <t_arceri@yahoo.com.au>"> Timothy Arceri</a> <pre>I took a quick look at this after talking to Ken. There seems to be a bunch of scenarios but I'm wondering if in at least some cases we are marking things as dirty when they shouldn't be. For example: ==30669== Invalid read of size 4 ==30669== at 0x9532B94: gen6_upload_push_constants (gen6_constant_state.c:85) ==30669== by 0x9529250: check_and_emit_atom (brw_state_upload.c:456) ==30669== by 0x9529250: brw_upload_pipeline_state (brw_state_upload.c:570) ==30669== by 0x9529250: brw_upload_render_state (brw_state_upload.c:592) ==30669== by 0x951BAAC: brw_try_draw_prims (brw_draw.c:770) ==30669== by 0x951BAAC: brw_draw_prims (brw_draw.c:862) ==30669== by 0x931F0BD: vbo_validated_drawrangeelements (vbo_exec_array.c:925) ==30669== by 0x931F84E: vbo_exec_DrawElements (vbo_exec_array.c:1075) ==30669== by 0xFC744A: glcts::StateInteractionCase::iterate() (in /home/timothy/data/git_mesa/build-es-cts-new/cts/glcts) ==30669== by 0xE19E75: es31cts::TestCaseWrapper::iterate(tcu::TestCase*) (in /home/timothy/data/git_mesa/build-es-cts-new/cts/glcts) ==30669== by 0x27DD23F: tcu::TestSessionExecutor::iterateTestCase(tcu::TestCase*) (in /home/timothy/data/git_mesa/build-es-cts-new/cts/glcts) ==30669== by 0x27DC794: tcu::TestSessionExecutor::iterate() (in /home/timothy/data/git_mesa/build-es-cts-new/cts/glcts) ==30669== by 0x27B1F3E: tcu::App::iterate() (in /home/timothy/data/git_mesa/build-es-cts-new/cts/glcts) ==30669== by 0xE14313: main (in /home/timothy/data/git_mesa/build-es-cts-new/cts/glcts) ==30669== Address 0xa744580 is 48 bytes inside a block of size 64 free'd ==30669== at 0x4C2ED4A: free (vg_replace_malloc.c:530) ==30669== by 0x944C14C: unsafe_free (ralloc.c:269) ==30669== by 0x92B8EEF: _mesa_clear_shader_program_data (shaderobj.c:334) ==30669== by 0x93A21BD: _mesa_glsl_link_shader (ir_to_mesa.cpp:3088) ==30669== by 0x92B7158: link_program (shaderapi.c:1163) ==30669== by 0x92B7158: link_program_error (shaderapi.c:1241) Here DrawElements ends up calling gen6_upload_push_constants() (which accesses the now freed uniform storage) but I would have presumed constants would still be in the correct state.</pre> </div> <hr> You are receiving this mail because: <ul> <li>You are the QA Contact for the bug.</li> </ul> </body> </html>