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.
Goals of the PR are:
* to ensure that correct types are applied to virtual registers which
were used as return values in call lowering. A reproducer is attached as
a new test case, before the PR it fails because spirv-val considers
output invalid due to wrong result/operand types in OpPhi's;
* improve type inference by speeding up postprocessing of types: by
limiting iterations by checking what remains to process, and processing
each instruction just once for any number of operands with uncomplete
types;
* improve type inference by more accurate work with uncomplete types
(pass uncomplete property to dependent operands, ensure consistency of
uncomplete-types data structure);
* change processing order and add traversing of PHI nodes when type
inference apply instructions results to specify/update/cast operands
type (fixes an issue with OpPhi's result type mismatch with operand
types).
This PR fixes generation of OpConstantFunctionPointerINTEL instruction
for the SPIR-V extension SPV_INTEL_function_pointers. Result type of
OpConstantFunctionPointerINTEL must be OpTypePointer with Storage Class
operand equal to CodeSectionINTEL.
See also https://github.com/llvm/llvm-project/pull/116636
CC: @MrSidims
This PR reworks implementation of OpSpecConstantOp with ptr-cast
operation (PtrCastToGeneric, GenericCastToPtr). Previous implementation
didn't take into account a lot of use cases, including multiple
inclusion of pointers, reference to a pointer from OpName, etc. A
reproducer is attached as a new test case.
This PR also fixes wrong type inference for IR patterns which generate
new virtual registers without SPIRV type. Previous implementation
assumed always that result has the same address space as a source that
is not the fact, and, for example, led to impossibility to emit a
ptr-cast operation in the reproducer, because wrong type inference
rendered source and destination with the same address space, eliminating
translation of G_ADDRSPACE_CAST.
This PR continues https://github.com/llvm/llvm-project/pull/101732
changes in virtual register processing aimed to improve correctness of
emitted MIR between passes from the perspective of MachineVerifier.
Namely, the following changes are introduced:
* register classes (lib/Target/SPIRV/SPIRVRegisterInfo.td) and
instruction patterns (lib/Target/SPIRV/SPIRVInstrInfo.td) are corrected
and simplified (by removing unnecessary sophisticated options) -- e.g.,
this PR gets rid of duplicating 32/64 bits patterns, removes ANYID
register class and simplifies definition of the rest of register
classes,
* hardcoded LLT scalar types in passes before instruction selection are
corrected -- the goal is to have correct bit width before instruction
selection, and use 64 bits registers for pattern matching in the
instruction selection pass; 32-bit registers remain where they are
described in such terms by SPIR-V specification (like, for example,
creation of virtual registers for scope/mem semantics operands),
* rework virtual register type/class assignment for calls/builtins
lowering,
* a series of minor changes to fix validity of emitted code between
passes:
- ensure that that bitcast changes the type,
- fix the pattern for instruction selection for OpExtInst,
- simplify inline asm operands usage,
- account for arbitrary integer sizes / update legalizer rules;
* add '-verify-machineinstrs' to existed test cases.
See also https://github.com/llvm/llvm-project/issues/88129 that this PR
may resolve.
This PR fixes a great number of issues reported by MachineVerifier and,
as a result, reduces a number of failed test cases for the mode with
expensive checks set on from ~200 to ~57.
