Lines Matching refs:pass
110 // How many regions to render to in multi-pass tests
183 bool clearBeforeRenderPass; //!< Whether loadOp=CLEAR should be used, or clear is done before render pass and loadOp=LOAD is used
516 VK_STENCIL_OP_INCREMENT_AND_CLAMP, // depth & stencil pass
933 // The attachments are either cleared already or should be cleared now. If an attachment was used in a previous render pass,
1101 // The attachments are either cleared already or should be cleared now. If an attachment was used in a previous render pass,
2299 // If not rendering to the whole framebuffer and the images were cleared before the render pass, verify that the area outside the render pass is untouched.
2485 return tcu::TestStatus::pass("Pass");
2989 // Create a render pass and a framebuffer
3179 // Always clear before render pass so outside render area can be verified.
3314 // Create a render pass and a framebuffer
3674 // Create a render pass and a framebuffer
3789 // Create a render pass and a framebuffer
3950 // Generate a random number of passes (either subpass or render pass)
3960 // Decide which attachments will be used in which pass. This is a bit mask.
3980 // Set the sample count for attachments. Multisampled attachments that are used in the same pass will get the same number of samples.
4000 // If all attachments are single-sampled, choose a random number of samples for the render pass.
4017 // Make all multisampled attachments used in the pass have the same number of samples. Additionally, make all the multisampled attachments
4090 // Make every multisampled attachment resolve in the last pass it's used.
4105 DBG(" - Resolved 0x%x in pass %u\n", unresolvedAttachments, passNdx - 1);
4108 // Decide whether clear should be done as part of the render pass. Tests loadOp=CLEAR vs loadOp=LOAD.
4140 // Integer resolve may choose any sample, so we modify only one channel per pass (hence the maximum of 4 passes). This way, the verification
4169 // Second draw call overwrites only the samples that pass the depth test (which is GREATER).
4195 // Note that integer formats don't blend, so always take the last value that's written. Each pass writes to only one channel, and color mask is used
4220 // does not average), so the test makes sure every channel is written to in only one pass, which again means there's no need to perform a resolve in between passes.
4236 // If depth/stencil is single-sampled, prepare the data for the next pass. If multisampled, it will no longer be used after the resolve.
4441 // Note that integer formats don't blend, so clear values are discarded, except for channels that are never written to. Each pass
4501 // Create a render pass and a framebuffer
4726 // Subpass 0 renders to color1 and depth/stencil only. They are resolved at the end of the pass.
4749 // Always clear before render pass so outside render area can be verified.
4931 return tcu::TestStatus::pass("Pass");
5142 // Test 3: Tests with a single render pass, potentially with multiple subpasses.
5146 // Single render pass with multiple subpasses