<div dir="ltr">On 30 August 2013 01:20, Pohjolainen, Topi <<a href="mailto:topi.pohjolainen@intel.com" target="_blank">topi.pohjolainen@intel.com</a>> wrote: <div class="gmail_extra"><div class="gmail_quote"> <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="HOEnZb"><div class="h5">On Wed, Aug 28, 2013 at 12:12:40PM -0700, Paul Berry wrote: > Verified using the NVIDIA proprietary driver for linux. > --- > tests/all.tests | 6 + > .../glsl-1.50/execution/geometry/CMakeLists.gl.txt | 1 + > .../tri-strip-ordering-with-prim-restart.c | 311 +++++++++++++++++++++ > 3 files changed, 318 insertions(+) > create mode 100644 tests/spec/glsl-1.50/execution/geometry/tri-strip-ordering-with-prim-restart.c > > diff --git a/tests/all.tests b/tests/all.tests > index 7ab841e..3ee7cd2 100644 > --- a/tests/all.tests > +++ b/tests/all.tests > @@ -955,6 +955,12 @@ for prim_type in ['GL_POINTS', 'GL_LINE_LOOP', 'GL_LINE_STRIP', 'GL_LINES', > 'glsl-1.50-geometry-primitive-types {0}'.format( > prim_type)) > > +for prim_type in ['GL_TRIANGLE_STRIP', 'GL_TRIANGLE_STRIP_ADJACENCY']: > + for restart_index in ['ffs', 'other']: > + add_concurrent_test(spec['glsl-1.50'], > + 'glsl-1.50-geometry-tri-strip-ordering-with-prim-restart {0} {1}'.format( > + prim_type, restart_index)) > + > spec['glsl-3.30'] = Group() > import_glsl_parser_tests(spec['glsl-3.30'], > os.path.join(testsDir, 'spec', 'glsl-3.30'), > diff --git a/tests/spec/glsl-1.50/execution/geometry/CMakeLists.gl.txt b/tests/spec/glsl-1.50/execution/geometry/CMakeLists.gl.txt > index 202fcd2..a504412 100644 > --- a/tests/spec/glsl-1.50/execution/geometry/CMakeLists.gl.txt > +++ b/tests/spec/glsl-1.50/execution/geometry/CMakeLists.gl.txt > @@ -12,3 +12,4 @@ ${OPENGL_glu_LIBRARY} > > piglit_add_executable (glsl-1.50-geometry-end-primitive end-primitive.c) > piglit_add_executable (glsl-1.50-geometry-primitive-types primitive-types.c) > +piglit_add_executable (glsl-1.50-geometry-tri-strip-ordering-with-prim-restart tri-strip-ordering-with-prim-restart.c) > diff --git a/tests/spec/glsl-1.50/execution/geometry/tri-strip-ordering-with-prim-restart.c b/tests/spec/glsl-1.50/execution/geometry/tri-strip-ordering-with-prim-restart.c > new file mode 100644 > index 0000000..90f6860 > --- /dev/null > +++ b/tests/spec/glsl-1.50/execution/geometry/tri-strip-ordering-with-prim-restart.c > @@ -0,0 +1,311 @@ > +/* > + * Copyright © 2013 Intel Corporation > + * > + * Permission is hereby granted, free of charge, to any person obtaining a > + * copy of this software and associated documentation files (the "Software"), > + * to deal in the Software without restriction, including without limitation > + * the rights to use, copy, modify, merge, publish, distribute, sublicense, > + * and/or sell copies of the Software, and to permit persons to whom the > + * Software is furnished to do so, subject to the following conditions: > + * > + * The above copyright notice and this permission notice (including the next > + * paragraph) shall be included in all copies or substantial portions of the > + * Software. > + * > + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR > + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, > + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL > + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER > + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING > + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER > + * DEALINGS IN THE SOFTWARE. > + */ > + > +/** > + * \file tri-strip-ordering-with-prim-restart.c > + * > + * Check a subtle corner case that affects the i965/gen7 mesa driver: > + * when the primitive type is either GL_TRIANGLE_STRIP or > + * GL_TRIANGLE_STRIP_ADJACENCY, the hardware delivers the vertices of > + * odd numbered triangles to the geometry shader in the wrong order, > + * so the driver must emit workaround code to re-order them. This > + * test verifies that the workaround code functions correctly in the > + * presence of primitive restart, since the presence of primitive > + * restart can make a triangle "odd numbered" in relation to the > + * current strip even if it is "even numbered" as measured by > + * gl_PrimitiveIDIn. > + * > + * This test works by issuing a single draw call and using primitive > + * restart to split it into a pair of 3-triangle strips (this ensures > + * that triangles in the first strip have the same parity in relation > + * to the strip as they have when measured by gl_PrimitiveIDIn; > + * triangles in the second strip hav opposite parity in relation to > + * the strip from what they have when measured by gl_PrimitiveIDIn). > + * The vertex IDs of all vertices are collected using transform > + * feedback, and checked it C to make sure it matches the expected </div></div> ^ Just checking if this was supposed to be "checked in C" instead? </blockquote><div> </div><div>Oops. Yes, thanks! </div></div></div></div>