Making it easier to work with shaders https://git.yummers.dev/slang.git/atom?h=master 2018-12-07T21:31:06+00:00 2018-12-07T21:31:06+00:00 Tim Foley tfoleyNV@users.noreply.github.com 2018-12-07T21:31:06+00:00 urn:sha1:135eaff6b892fc91a398714ddcf7ef377cd4cccb * Change how buffers are emitted This is a change with a lot of pieces, which can't always be separated out cleanly. I'm going to walk through them in what I hope is a logical order. The main goal of this change was to allow arrays of structured buffers to translate to Vulkan. Consider two declarations of structured buffers in HLSL/Slang: ```hlsl StructuredBuffer<X> single; StructuredBuffer<Y> multiple[10]; ``` The current translation logic was handling `single` by translating it into an *unnamed* GLSL `buffer` block like: ```glsl layout(std430) buffer _S1 { X single[]; }; ``` That syntax allows an expression like `single[i]` in Slang to be translated simply as `single[i]` in GLSL. But that naive translating doesn't work for `multiple`, since we need to declare a array of blocks in GLSL, which requires giving the whole thing a name: ```glsl layout(std430) buffer _S2 { Y _data[]; } multiple[10]; ``` Now a reference to `multiple[i][j]` in Slang needs to become `multiple[i]._data[j]` in GLSL. To avoid having way too many special cases around single structured buffers vs. arrays, it makes sense to allows emit things in the latter form, so that we instead lower `single` as: ```glsl layout(std430) buffer _S1 { X _data[]; } single; ``` So that now a reference to `single[i]` becomes `single._data[i]` in GLSL. Most of that can be handled in the standard library translation of the structured buffer indexing operations. The only wrinkle there is that there were some *old* special-case instructions in the IR intended to handle buffer load/store operations (these were added back when I was trying to keep the "VM" path working). These aren't really needed to have structured-buffer operations work; they can be handled as ordinary functions as far as the stdlib is concerned. I removed the old instructions. Along the way, it became clear that a few other cases follow the same pattern. Byte-addressed buffers are an obvious case. We were lowering HLSL/Slang: ```hlsl ByteAddressBuffer b; ... uint x = b.Load(0); ``` to GLSL like: ```glsl layout(std430) buffer _S1 { uint b[]; }; ... uint x = b[0]; ``` That logic would fail for arrays the same way that the structured buffer case was failing. The fix is the same: use named `buffer` blocks and then introduce an explicit `_data` field: ```glsl layout(std430) buffer _S1 { uint _data[]; } b; ... uint x = b._data[0]; ``` Just like with structured buffers, all of the VK translation for operations on byte-addressed buffers can be implemented directly in teh stdlib, so once the emit logic was changed it was just a matter of adding `._data` to a bunch of VK tranlsations. It turns out that arrays of constant buffers have more or less the same problem, and furthermore we have some problems with any code that directly uses the modern HLSL `ConstantBuffer<T>` type. Note: the emit logic around constant buffers sometimes refers to "parameter groups" because that is being used in the compiler as a catch-all term for constant buffers, texture buffers, and parameter blocks. The existing code was going out of its way to reproduce the way that constant buffer declarations are implicitly referenced in HLSL: ```hlsl cbuffer C { float f; } ... float tmp = f; // No reference to `C` here ``` This can be seen in the emit logic with the `isDerefBaseImplicit` function, which is used to take the internal IR representation for a reference to `f` (which is closer to the expression `(*C).f` or `C->f`) and leave off any reference to `C` so that we emit just `f`. That kind of logic just flat out doesn't work in some important cases. Arrays of constant buffers are a clear one: ```hlsl ConstantBuffer<X> cbArray[3]; ... X x = cbArray[0]; ``` There is no way to translate that to an ordinary `cbuffer` declaration at all. The same problem can be created without arrays, though: ```hlsl ConstantBuffer<X> singleCB; ... X x = singleCB; ``` The current strategy for translating constant buffers was translating `singleCB` into a `cbuffer` declaration that reproduced the fields of `X` as its members, which just wouldn't work: ```hlsl cbuffer singleCB { float f; // field of `X` } ... X x = singleCB; // ERROR: there is nothing named `singleCB` in this HLSL ``` The new strategy is more consistent. We still generate a `cbuffer` declaration for a single constant buffer, but we always give it a single field of the chosen element type: ```hlsl cbuffer singleCB { X singleCB; } ... X x = singleCB; // this works fine! ``` And in the array case we generate code that uses the explicit `ConstantBuffer<T>` type: ```hlsl ConstantBuffer<X> cbArray[3]; ... X x = cbArray[0]; ``` The GLSL output is more complicated because unlike with HLSL there is no implicit conversion from a uniform block to its element type (there is no notion of an element type). The array case thus needs a `_data` field similar to what we do for structured buffers: ```glsl layout(std140) uniform _S3 { X _data; } cbArray[3]; ... X x = cbArray[0]._data; ``` And then the non-array case needs to have a similar `_data` field for consistency: ```glsl layout(std140) uniform _S1 { X _data; } singleCB; ... X x = singleCB._data; ``` This is handled by inserting the necessary reference to `_data` whenever we dereference a constant buffer, either as part of a load instruction (loading from the whole CB as a pointer), or an `IRFieldAddress` instruction which forms a pointer into the CB (e.g., `&(singleCB->f)` becomes `singleCB._data.f`). The current emit logic handles `ParameterBlock<X>` differently from `ConstantBuffer<X>`, but really only to allow parameter blocks to be explicitly named in the output, while constant buffers were left implicit by default. Thus the only difference was a legacy one (from back when trying to exactly reproduce the HLSL text we got as input was considered an important goal), and the new approach to emitting constant buffers would get rid of it. I removed the separate logic for emitting `ParameterBlock<X>` and just let the handling for constant buffers deal with it. Note that any resource types inside of a `ParameterBlock<X>` would have been moved out as part of legalization, so that a parameter block is 100% equivalent to a constant buffer when it comes time to emit code. Unsurprisingly, changing the way we generate HLSL and GLSL output for all these buffer types meant that any tests that were directly comparing the output of `slangc` against `fxc`, `dxc`, or `glslang` broke. The basic approach to fixing the breakage in GLSL tests was to update the GLSL baseline to reflect the new output startegy. In some cases I used macros to name the various `_S<digits>` temporaries so that future renaming will hopefully be easier (it would be great if we auto-generated temporary names with a bit more context). There was one GLSL test (`tests/bugs/vk-structured-buffer-binding`) that was using raw GLSL expected output, and this was changed to use a GLSL baseline to generate SPIR-V for comparison. For HLSL tests we were sometimes running the same input file through `slangc` and `fxc`/`dxc`, and in these cases I macro-ized the various `cbuffer` declarations to generate different declarations depending on the compiler. I completely dropped the tests coming from the D3D SDK because they aren't providing much coverage, and updating them would change them so far from the original code that the purported benefit (using a body of existing shaders) would be lost. I also dropped the explicit matrix layout qualifiers in the `matrix-layout` test because the new output strategy breaks those for GLSL (you can't put matrix layout qualifiers on `struct` fields, and now the body of every constant buffer is inside a `struct`). This isn't as big of a loss as it seems, because our handling of those qualifiers wasn't really right to begin with. Slang users should only be setting the matrix layout mode globally (and we should probably switch to error out on the explicit qualifiers for now). The other thing that got dropped is tests involving `packoffset` modifiers. Slang already warns that it doesn't support these, and the way they were used in the test cases is actually misleading. For the binding/layout-related tests, the goal was to show that Slang reproduces the same layout as fxc, in which case explicitly enforcing a layout via `packoffset` seems like cheating (are we sure we enforced the layout fxc would have produced?). The real reason was that Slang used to emit explicit `packoffset` on *every* field of a `cbuffer` it would output, because of an `fxc` bug where you couldn't use `register` on textures/samplers declared inside a `cbuffer` unless *every* field in the `cbuffer` used a `register` or `packoffset` modifier. Slang hasn't required that behavior in a while because it now splits textures and samplers, and the one test case where we needed `packoffset` to work around the `fxc` bug in the baseline HLSL has been macro-ified even more to work around the bug. The amount of churn in the test cases is unfortunate, but it continues to point at the weakness of any testing strategy that checks for exact equivalent between Slang's output and that of other compilers. We need to keep working to replace these tests with better alternatives. In `check.cpp` there is logic to perform implicit dereferencing, so that if you write `obj.f` where `obj` is a `ConstantBuffer<X>` (or some other "pointer-like" type) and `f` is a field in `X`, then this effectively translates as `(*obj).f`. That is, we dereference the value of type `ConstantBuffer<X>` to get a value of type `X`, and then refer to the field of the `X` value. There was a problem where the logic to insert that kind of implicit dereference operation was using a reference (`auto& type = ...`) for the type of the expression being dereferenced, and then clobbering it. This would mean that an expression of type `ConstantBuffer<X>` would have its type overwritten to be just `X` and then codegen would break later on. I'm not sure how we haven't run into that before. The `array-of-buffers` test case was added to confirm that we now support arrays of constant, structured, and byte-address buffers for both DXIL and SPIR-V output. Okay, so that was a lot of stuff, but hopefully it is clear how this all works to make the output of the compiler more consistent and explicit, while also supporting the required new functionality. * fixup: review feedback 2018-10-29T21:44:39+00:00 Tim Foley tfoleyNV@users.noreply.github.com 2018-10-29T21:44:39+00:00 urn:sha1:725985528f77ba939a5cddc71e5006fee7638465 * Rework command-line options handling for entry points and targets Overview: * The biggest functionality change is that the implicit ordering constraints when multiple `-entry` options are reversed: any `-stage` option affects the `-entry` to its *left* instead of to its *right* as it used to. This is technically a breaking change, but I expect most users aren't using this feature. * The options parsing tries to handle profile versions and stages as distinct data (rather than using the combined `Profile` type all over), and treats a `-profile` option that specifies both a profile version and a stage (e.g., `-profile ps_5_0`) as if it were sugar for both a `-profile` and a `-stage` (e.g., `-profile sm_5_0 -stage fragment`). * We now technically handle multiple `-target` options in one invocation of `-slangc`, but do not advertise that fact in the documentation because it might be confusing for users. Similar to the relationship between `-stage` and `-entry`, any `-profile` option affects the most recent `-target` option unless there is only one `-target`. * The logic for associating `-o` options with corresponding entry points and targets has been beefed up. The rule is that a `-o` option for a compiled kernel binds to the entry point to its left, unless there is only one entry point (just like for `-stage`). The associated target for a `-o` option is found via a search, however, because otherwise it would be impossible to specify `-o` options for both SPIR-V and DXIL in one pass. * The handling of output paths for entry points in the internal compiler structures was changed, because previously it could only handle one output path per entry point (even when there are multiple targets). The new logic builds up a per-target mapping from an entry point to its desired output path (if any). Details: * Support for formatting profile versions, stages, and compile targets (formats) was added to diagnostic printing, so that we can make better error messages. This is fairly ad hoc, and it would be nice to have all of the string<->enum stuff be more data-driven throughout the codebase. * Test cases were added for (almost) all of the error conditions in the current options validation. The main one that is missing is around specifying an `-entry` option before any source file when compiling multiple files. This is because the test runner is putting the source file name first on the command line automatically, so we can't reproduce that case. * Several reflection-related tests now reflect entry points where they didn't before, because the logic for detecting when to infer a default `main` entry point have been made more loose * On the dxc path, beefed up the handling of mapping from Slang `Profile`s to the coresponding string to use when invoking dxc. * A bunch of tests cases were in violation of the newly imposed rules, so those needed to be cleaned up. * There were also a bunch of test cases that had accidentally gotten "disabled" at some point because there were comparing output from `slangc` both with and without a `-pass-through` option, but that meant that any errors in command-line parsing produced the *same* error output in both the Slang and pass-through cases. This change updates `slang-test` to always expect a successful run for these tests, and then manually updates or disables the various test cases that are affected. * When merging the updated test for matrix layout mode, I found that the new command-line logic was failing to propagate a matrix layout mode passed to `render-test` into the compiler. This was because the `-matrix-layout*` options were implemented as per-target, but the target was being set by API while the option came in via command line (passed through the API). It seems like we want matrix layout mode to be a global option anyway (rather than per-target), so I made that change here. * Add missing expected output files * A 64-bit fix * Remove commented-out code noted in review 2017-11-06T18:37:27+00:00 Tim Foley tfoleyNV@users.noreply.github.com 2017-11-06T18:37:27+00:00 urn:sha1:9919c823938ae929b16efac9d507f6d5eb122bf4 * Rename existing ParameterBlock to ParameterGroup We are planning to add a new `ParameterBlock<T>` type, which maps to the notion of a "parameter block" as used in the Spire research work. Unfortunately, the compiler codebase already uses the term `ParameterBlock` as catch-all to encompass all of HLSL `cbuffer`/`tbuffer` and GLSL `uniform`/`buffer`/`in`/`out` blocks (all of which are lexical `{}`-enclosed blocks that define parameters...). This change instead renames all of the existing concepts over to `ParameterGroup`, which isn't an ideal name, but at least doesn't directly overlap the new terminology or any existing terminology. The new `ParameterBlockType` case will probably be a subclass of `ParameterGroupType`, since it is a logical extension of the underlying concept. * Add Shader Model 5.1 profiles The HLSL `register(..., space0)` syntax is only allowed on "SM5.1" and later profiles (which is supported by the newer version of `d3dcompiler_47.dll` that comes with the Win10 SDK, but not the older version of `d3dcompiler_47.dll` - good luck figuring out which you have!). This change adds those profiles to our master list of profiles, and nothing else. * First pass at support for `ParameterBlock<T>` - Add the type declaration in stdlib - Add a special case of `ParameterGroupType` for parameter blocks - Handle parameter blocks in type layout (currently handling them identically to constant buffers for now, which isn't going to be right in the long term) - Add an IR pass that basically replaces `ParameterBlock<T>` with `T` - Eventually this should replace it with either `T` or `ConstantBuffer<T>`, depending on whether the layout that was computed required a constant buffer to hold any "free" uniforms - Add first stab at an IR pass to "scalarize" global variables using aggregate types with resources inside. - This currently only applies to global variables, so it won't handle things passed through functions, or used as local variables - It also only supports cases where the references to the original variable are always references to its fields, and not the whole value itself - Add a single test case that technically passes with this level of support, but probably isn't very representative of what we need from the feature * Fold parameter-block desugaring into a more complete "type legalization" pass The basic problem that was arising is that once you desugar `ParameterBlock<T>` into `T`, you then need todeal with splitting `T` into its constituent fields if it contains any resource types. Handling those transformations by following the usual use-def chains wasn't really helping, because you might need systematic rewriting that can really only be handled bottom-up. This change adds a new pass that is intended to perform multiple kinds of type "legalization" at once: - It will turn `ParameterBlock<T>` into `T` - It may at some point also convert `ConstantBuffer<T>` into `T` as well - It will turn an value of an aggregate type that contains resources into N different values (one per field) - As a result of this, it will also deal with AOS-to-SOA conversion of these types Legalization is applied to *every* function/instruction/value, so that it can make large-scale changes that would be tough to manage with a work list. This pass needs to be run *after* generics have been fully specialized, so that we know we are always dealing with fully concrete types, so that their legalization for a given target is completely known. This is still work in progress; there's more to be done to get this working with all our test cases, and finish the remaining `ParameterBlock<T>` work. * Improve binding/layout information when using parameter blocks - When doing type layout for a parameter block, don't include the resources consumed by the element type in the resource usage for the parameter block - Note that this is pretty much identical to how a `ConstantBuffer<T>` does not report any `LayoutResourceKind::Uniform` usage, except that `ParameterBlock<T>` is *also* going to hide underlying texture/sampler reigster usage - The one exception here is that any nested items that use up entire `space`s or `set`s those need to be exposed in the resource usage of the parent (I don't have a test for this) - When type legalization needs to scalarize things, it must propagate layout information down to the new leaf variables. In general, the register/index for a new leaf parameter should be the sum of the offsets for all of the parent variables along the "chain" from the original variable down to the leaf (we aren't dealing with arrays here just yet). - When type legalization decides to eliminate a pointer(-like) type (e.g., desugar `ParameterBlock<T>` over to `T`), actually deal with that in terms of the `LegalVal`s created, so that we can know to turn a `load` into a no-op when applied to a value that got indirection removed. - Hack up the "complex" parameter-block test so that it actually passes (the big hack here is that the HLSL baseline is using names that are generated by the IR, and are unlikely to be stable as we add/remove transformations). - Note: I can't make these be compute tests right now, because regsiter spaces/sets are a feature of D3D12/Vulkan, and our test runner isn't using those APIs. 2017-08-25T22:48:49+00:00 Tim Foley tfoley@nvidia.com 2017-08-25T22:18:10+00:00 urn:sha1:8c68434d6fa6ff1b2e54586637869cceace9d1bb Fixes #171 Fixes #172 These two bugs related to bad logic in handling of splitting resource-containing `cbuffer` declarations. - Issue #171 was the case where a `cbuffer` *only* had resource fields, in which case we crashed whenever referencing any field (some code was assuming there had to be non-resource fields) - Issue #172 was a case where two fields were declared with a single declaration (`Texture2D a, b;`), and the logic we had for tracking resource-type fields was accidentally tagging *both* fields with a single modifier so that field `b` would get confused for `a` in some contexts, and attempts to access `b` would crash. Both issues are now fixed, and regression tests have been added.