Making it easier to work with shaders https://git.yummers.dev/slang.git/atom?h=master 2018-02-02T16:49:04+00:00 2018-02-02T16:49:04+00:00 Tim Foley tfoleyNV@users.noreply.github.com 2018-02-02T16:49:04+00:00 urn:sha1:0360f81b9741ece65768a65731bd23455a3b7a96 * Remove support for the -no-checking flag Fixes #381 Fixes #383 Work on #382 - No longer expose flag through API (`SLANG_COMPILE_FLAG_NO_CHECKING`) and command-line (`-no-checking`) options - Remove all logic in `check.cpp` that was withholding diagnostics (including errors) when the no-checking mode was enabled - Remove `HiddenImplicitCastExpr`, which was only created to support no-checking mode (it represented an implicit cast that our checking through was needed, but couldn't emit because it might be wrong) - Remove logic for storing function bodies as raw token lists when checking is turned off. I'm leaving in the `UnparsedStmt` AST node in case we ever need/want to lazily parse and check function bodies down the line. - Remove a few of the code-generation paths we had to contend with, but keep the comment about them in place. - Remove GLSL-based tests that can't meaningfully work with the new approach. - Fix other tests that used a GLSL baseline so that their GLSL compiles with `-pass-through glslang` instead of invoking `slang` with the `-no-checking` flag. - Remove tests that were explicitly added to test the "rewriter + IR" path, since that is no longer supported. There is more cleanup that can be done here, now that we know that AST-based rewrite and IR will never co-exist, but it is probably easier to deal with that as part of removing the AST-based rewrite path. We've lost some test coverage here, but actually not too much if we consider that we are dropping GLSL input anyway. * Fixup: test runner was mis-counting ignored tests * Fixup: turn on dumping on test failure under Travis * Fixup: enable extensions in Linux build of glslang 2017-06-14T15:35:43+00:00 Tim Foley tfoley@nvidia.com 2017-06-14T15:35:43+00:00 urn:sha1:34c7cdc68fd5b7b8132250c4de0d562c9ba0adbb Most of the Vulkan GLSL shaders in the `sascha-willems` corpus use completely regular parameter bindings, e.g.: ``` layout (binding = 0) uniform sampler2D samplerPositionDepth; layout (binding = 1) uniform sampler2D samplerNormal; layout (binding = 2) uniform sampler2D ssaoNoise; ``` With the Slang compiler, we can write this kind of stuff more compactly as: ``` uniform sampler2D samplerPositionDepth; uniform sampler2D samplerNormal; uniform sampler2D ssaoNoise; ``` and get the exact same result (in fact, we will generate output GLSL matching the first example). There are a few spot tests for this in the HLSL case, but I hadn't done anything for GLSL yet. Now that we have the ability to specify multiple tests to run on each input file, it was easy enough to go in and tweak some of these files to be usable as binding-generation tests. I didn't ammend all of the GLSL shaders for two reasons: 1. Not all of the shaders will work with completely automatic binding generation done on a per-file basis. In some cases, the parameter layout needs to consider multiple files (and the order in which the files are supplied could matter). This is probably best handled with more directed tests. 2. There is going to be a ton of duplication if I just have 100s of tests that all confirm that, yes, the Slang compiler can count vertex inputs starting from zero. These shaders aren't really presenting a whole lot of unique cases to work with. That said, a level of confidence greater than zero is always better than zero confidence. 2017-06-09T20:44:59+00:00 Tim Foley tfoley@nvidia.com 2017-06-09T18:34:21+00:00 urn:sha1:fcf83dbf9effab3bd98bad2b83b2468b7eb05cfd