[Piglit] [PATCH] Demonstrate a bug with fast clears on MS FBOs with missing comps
Neil Roberts
neil at linux.intel.com
Wed Nov 4 09:09:56 PST 2015
There is currently a bug with the i965 driver in Mesa with the fast
clear optimisation when there are missing components in the
framebuffer. If the texture that was rendered to is later sampled it
will use the fast clear color without overriding the missing
components to 0 or 1. This patch adds a test to demonstrate that.
---
tests/all.py | 1 +
.../ext_framebuffer_multisample/CMakeLists.gl.txt | 1 +
.../ext_framebuffer_multisample/fast-clear-rgb.c | 131 +++++++++++++++++++++
3 files changed, 133 insertions(+)
create mode 100644 tests/spec/ext_framebuffer_multisample/fast-clear-rgb.c
diff --git a/tests/all.py b/tests/all.py
index bd3e79d..4386c60 100644
--- a/tests/all.py
+++ b/tests/all.py
@@ -2752,6 +2752,7 @@ with profile.group_manager(
g(['ext_framebuffer_multisample-alpha-blending'], 'alpha-blending')
g(['ext_framebuffer_multisample-alpha-blending', 'slow_cc'],
'alpha-blending slow_cc')
+ g(['ext_framebuffer_multisample-fast-clear-rgb'], 'fast-clear-rgb')
for num_samples in MSAA_SAMPLE_COUNTS:
if num_samples % 2 != 0:
diff --git a/tests/spec/ext_framebuffer_multisample/CMakeLists.gl.txt b/tests/spec/ext_framebuffer_multisample/CMakeLists.gl.txt
index 1733f57..5c04f90 100644
--- a/tests/spec/ext_framebuffer_multisample/CMakeLists.gl.txt
+++ b/tests/spec/ext_framebuffer_multisample/CMakeLists.gl.txt
@@ -65,3 +65,4 @@ piglit_add_executable (ext_framebuffer_multisample-sample-alpha-to-one common.cp
piglit_add_executable (ext_framebuffer_multisample-turn-on-off common.cpp turn-on-off.cpp)
piglit_add_executable (ext_framebuffer_multisample-unaligned-blit common.cpp unaligned-blit.cpp)
piglit_add_executable (ext_framebuffer_multisample-upsample common.cpp upsample.cpp)
+piglit_add_executable (ext_framebuffer_multisample-fast-clear-rgb fast-clear-rgb.c)
diff --git a/tests/spec/ext_framebuffer_multisample/fast-clear-rgb.c b/tests/spec/ext_framebuffer_multisample/fast-clear-rgb.c
new file mode 100644
index 0000000..0883711
--- /dev/null
+++ b/tests/spec/ext_framebuffer_multisample/fast-clear-rgb.c
@@ -0,0 +1,131 @@
+/*
+ * Copyright © 2015 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 fast-clear-rgb.c
+ *
+ * Clears a multisample RGB texture to 1,1,1,0 and then samples from
+ * it in a shader to ensure that the alpha channel gets overriden to
+ * 1. The reason for this test is that there is a bug in the i965
+ * implementation of fast clears where the fast clear color doesn't
+ * take into account the available components in the render target.
+ * When it is later sampled it will read the fast clear color value
+ * for the alpha instead of overriding it to 1.
+ */
+
+#include "piglit-util-gl.h"
+
+PIGLIT_GL_TEST_CONFIG_BEGIN
+
+ config.supports_gl_core_version = 31;
+ config.window_visual = PIGLIT_GL_VISUAL_RGBA | PIGLIT_GL_VISUAL_DOUBLE;
+
+PIGLIT_GL_TEST_CONFIG_END
+
+static const char
+vertex_source[] =
+ "#version 140\n"
+ "\n"
+ "in vec4 piglit_vertex;\n"
+ "\n"
+ "void\n"
+ "main()\n"
+ "{\n"
+ " gl_Position = piglit_vertex;\n"
+ "}\n";
+
+static const char
+fragment_source[] =
+ "#version 140\n"
+ "#extension GL_ARB_texture_multisample : require\n"
+ "\n"
+ "uniform sampler2DMS tex;\n"
+ "out vec4 color_out;\n"
+ "\n"
+ "void\n"
+ "main()\n"
+ "{\n"
+ " color_out = texelFetch(tex, ivec2(0), 0);\n"
+ "}\n";
+
+enum piglit_result
+piglit_display()
+{
+ /* Although we cleared the texture to 1,1,1,0, the alpha
+ * channel should end up as 1.0 when sampled.
+ */
+ static const GLfloat expected_color[] = {
+ 1.0, 1.0, 1.0, 1.0
+ };
+ bool pass = true;
+
+ glClearColor(0.0, 0.0, 0.0, 0.0);
+ glClear(GL_COLOR_BUFFER_BIT);
+ piglit_draw_rect(-1, -1, 2, 2);
+ pass = piglit_probe_rect_rgba(0, 0, piglit_width, piglit_height,
+ expected_color) && pass;
+
+ piglit_present_results();
+
+ return pass ? PIGLIT_PASS : PIGLIT_FAIL;
+}
+
+void
+piglit_init(int argc, char **argv)
+{
+ GLuint prog;
+ GLint tex_location;
+ GLuint tex;
+ GLuint fbo;
+
+ piglit_require_extension("GL_ARB_texture_multisample");
+
+ prog = piglit_build_simple_program(vertex_source, fragment_source);
+ glUseProgram(prog);
+ tex_location = glGetUniformLocation(prog, "tex");
+ glUniform1i(tex_location, 0);
+
+ glGenTextures(1, &tex);
+ glBindTexture(GL_TEXTURE_2D_MULTISAMPLE, tex);
+ glTexImage2DMultisample(GL_TEXTURE_2D_MULTISAMPLE,
+ 1, /* samples */
+ GL_RGB,
+ 1, 1, /* width/height */
+ GL_FALSE /* fixed sample locations */);
+ glGenFramebuffers(1, &fbo);
+ glBindFramebuffer(GL_FRAMEBUFFER, fbo);
+ glFramebufferTexture(GL_FRAMEBUFFER,
+ GL_COLOR_ATTACHMENT0,
+ tex,
+ 0);
+ if (glCheckFramebufferStatus(GL_FRAMEBUFFER) !=
+ GL_FRAMEBUFFER_COMPLETE) {
+ fprintf(stderr, "texture FBO not complete\n");
+ piglit_report_result(PIGLIT_SKIP);
+ }
+
+ glClearColor(1.0, 1.0, 1.0, 0.0);
+ glClear(GL_COLOR_BUFFER_BIT);
+ glBindFramebuffer(GL_FRAMEBUFFER, piglit_winsys_fbo);
+ glDeleteFramebuffers(1, &fbo);
+}
--
1.9.3
More information about the Piglit
mailing list