The __spirv_GenericCastToPtrExplicit_To* builtins and its equivalent OpenCL builtins (to_global, to_local and to_private) were mapped to OpGenericCastToPtr instead of OpGenericCastToPtrExplicit. The patch now uses OpGenericCastToPtrExplicit for these builtins.
This PR is to thoroughly rework duplicate tracker implementation and
tracking of IR entities and types. These are legacy parts of the project
resulting in an extremely bloated intermediate representation and
computational delays due to inefficient data flow and structure choices.
Main results of the rework:
1) Improved compile-time performance. The reference binary LLVM IR used
to measure speed gains in
https://github.com/llvm/llvm-project/pull/120415 shows ~x5 speed up also
after this PR. The timing before this PR is ~42s and after this PR it's
~7.5s. In total this PR and the previous overhaul of the module analysis
in https://github.com/llvm/llvm-project/pull/120415 results in ~x25
speed improvement.
```
$ time llc -O0 -mtriple=spirv64v1.6-unknown-unknown _group_barrier_phi.bc -o 1 --filetype=obj
real 0m7.545s
user 0m6.685s
sys 0m0.859s
```
2) Less bloated intermediate representation of internal translation
steps. Elimination of `spv_track_constant` intrinsic usage for scalar
constants, rework of `spv_assign_name`, removal of the gMIR `GET_XXX`
pseudo code and a smaller number of generated `ASSIGN_TYPE` pseudo codes
substantially decrease volume of data generated during translation.
3) Simpler code and easier maintenance. The duplicate tracker
implementation is simplified, as well as other features.
4) Numerous fixes of issues and logical flaws in different passes. The
main achievement is rework of the duplicate tracker itself that had
never guaranteed a correct caching of LLVM IR entities, rarely and
randomly returning stale/incorrect records (like, remove an instruction
from gMIR but still refer to it). Other fixes comprise consistent
generation of OpConstantNull, assigning types to newly created
registers, creation of integer/bool types, and other minor fixes.
5) Numerous fixes of LIT tests: mainly CHECK-DAG to properly reflect
SPIR-V spec guarantees, `{{$}}` at the end of constants to avoid
matching of substrings, and XFAILS for `SPV_INTEL_long_composites` test
cases, because the feature is not completed in full yet and doesn't
generate a requested by the extension sequence of instructions.
6) New test cases are added.
This PR resolves the issue of the SPIR-V specification, requiring Shader-coupled capabilities to read/write images in the OpenCL SPIR-V environment, from the perspective of the LLVM SPIR-V backend. See https://github.com/KhronosGroup/SPIRV-Headers/issues/487 for details and discussion. Current implementation correctly reproduces requirements of the SPIR-V specification, however, since the requirements are problematic, out current implementation blocks generation of valid SPIR-V code for compute environments. This PR is to implement a solution discussed at the SPIR-V WG to allow proceeding with generation of valid SPIR-V code for the OpenCL environment and do not impact Vulkan environment at the same time.
This PR improves built-in variables and functions support: * extends mapping from an OpenCL C built-in function to the SPIR-V BuiltIn variables as in https://registry.khronos.org/OpenCL/specs/3.0-unified/html/OpenCL_Env.html#_built_in_variables, and * adds SPIR-V friendly builtins for Image Read/Write instructions. Test cases are extended accordingly.
This PR improves the set of rules for type inference by ensuring that a correct pointer type is deduced from the Value argument of OpAtomic* instructions, also when a pointer argument is coming from an `inttoptr .. to` instruction that caused problems earlier. Existing test cases are updated accordingly. This fixes https://github.com/llvm/llvm-project/issues/127491
https://github.com/KhronosGroup/SPIRV-Tools/pull/5407 introduces a check for WorkgroupSize variable to be a 3-component 32-bit int vector, and indeed, we see this requirement in https://registry.khronos.org/vulkan/specs/latest/man/html/WorkgroupSize.html#VUID-WorkgroupSize-WorkgroupSize-04427 However, OpenCL imposes different requirements, documented here: https://registry.khronos.org/OpenCL/specs/3.0-unified/html/OpenCL_Env.html#_built_in_variables OpenCL environment requires WorkgroupSize variable to have components of size_t size that will be 32 or 64 depending on a target. This is the way how the SPIR-V Backend implements this, by querying pointer size of the current platform/target. To allow spirv-val to account target environments difference, this PR adds `--target-env <env>` to test cases referring to the BuiltIn WorkgroupSize variable.
This PR is to address legacy issues with module analysis that currently uses a complicated and not so efficient approach to trace dependencies between SPIR-V id's via a duplicate tracker data structures and an explicitly built dependency graph. Even a quick performance check without any specialized benchmarks points to this part of the implementation as a biggest bottleneck. This PR specifically: * eliminates a need to build a dependency graph as a data structure, * updates the test suite (mainly, by fixing incorrect CHECK's referring to a hardcoded order of definitions, contradicting the spec requirement to allow certain definitions to go "in any order", see https://registry.khronos.org/SPIR-V/specs/unified1/SPIRV.html#_logical_layout_of_a_module), * improves function pointers implementation so that it now passes EXPENSIVE_CHECKS (thus removing 3 XFAIL's in the test suite). As a quick sanity check of whether goals of the PR are achieved, we can measure time of translation for any big LLVM IR. While testing the PR in the local development environment, improvements of the x5 order have been observed. For example, the SYCL test case "group barrier" that is a ~1Mb binary IR input shows the following values of the naive performance metric that we can nevertheless apply here to roughly estimate effects of the PR. before the PR: ``` $ time llc -O0 -mtriple=spirv64v1.6-unknown-unknown _group_barrier_phi.bc -o 1 --filetype=obj real 3m33.241s user 3m14.688s sys 0m18.530s ``` after the PR ``` $ time llc -O0 -mtriple=spirv64v1.6-unknown-unknown _group_barrier_phi.bc -o 1 --filetype=obj real 0m42.031s user 0m38.834s sys 0m3.193s ``` Next work should probably address Duplicate Tracker further, as it needs analysis now from the perspective of what parts of it are not necessary now, after changing the approach to implementation of the module analysis step.
The test cases marked with XFAIL by this commit are not yet supported by the SPIR-V backend with LLVM_ENABLE_EXPENSIVE_CHECKS enabled.
This PR improves general validity of emitted code between passes due to generation of `TargetOpcode::PHI` instead of `SPIRV::OpPhi` after Instruction Selection, fixing generation of OpTypePointer instructions and using of proper virtual register classes. Using `TargetOpcode::PHI` instead of `SPIRV::OpPhi` after Instruction Selection has a benefit to support existing optimization passes immediately, as an alternative path to disable those passes that use `MI.isPHI()`. This PR makes it possible thus to revert https://github.com/llvm/llvm-project/pull/116060 actions and get back to use the `MachineSink` pass. This PR is a solution of the problem discussed in details in https://github.com/llvm/llvm-project/pull/110507. It accepts an advice from code reviewers of the PR #110507 to postpone generation of OpPhi rather than to patch CodeGen. This solution allows to unblock improvements wrt. expensive checks and makes it unrelated to the general points of the discussion about OpPhi vs. G_PHI/PHI. This PR contains numerous small patches of emitted code validity that allows to substantially pass rate with expensive checks. Namely, the test suite with expensive checks set ON now has only 12 fails out of 569 total test cases. FYI @bogner
This PR is to fix the following issues: * the SPIR-V Backend didn't generate Alignment decoration for alloca instructions, * we need to use types from demangled function declarations to specify types for opaque pointers.