| Age | Commit message (Collapse) | Author |
|---|
|
Adds the first XeGPU transform op, `xegpu.set_desc_layout`, which attachs a `xegpu.layout` attribute to the descriptor that a `xegpu.create_nd_tdesc` op returns.
|
|
Introduces a Transform-dialect SMT-extension so that we can have an op
to express constrains on Transform-dialect params, in particular when
these params are knobs -- see transform.tune.knob -- and can hence be
seen as symbolic variables. This op allows expressing joint constraints
over multiple params/knobs together.
While the op's semantics are clearly defined, per SMTLIB, the interpreted
semantics -- i.e. the `apply()` method -- for now just defaults to failure. In
the future we should support attaching an implementation so that users
can Bring Your Own Solver and thereby control performance of
interpreting the op. For now the main usage is to walk schedule IR and
collect these constraints so that knobs can be rewritten to constants that
satisfy the constraints.
|
|
This patch adds conversion to LLVM for all existing pointer ops. This is
a stop gap measure to allow users to use the `ptr` dialect now. In the
future some of these conversions will be removed, and added as
translations, for example `ptradd`.
Example:
```mlir
func.func @test_memref_ptradd_indexing(%arg0: memref<10x?x30xf32, #ptr.generic_space>, %arg1: index) -> !ptr.ptr<#ptr.generic_space> {
%0 = ptr.to_ptr %arg0 : memref<10x?x30xf32, #ptr.generic_space> -> <#ptr.generic_space>
%1 = ptr.type_offset f32 : index
%2 = arith.muli %1, %arg1 : index
%3 = ptr.ptr_add %0, %2 : <#ptr.generic_space>, index
return %3 : !ptr.ptr<#ptr.generic_space>
}
// mlir-opt --convert-to-llvm --canonicalize --cse
llvm.func @test_memref_ptradd_indexing(%arg0: !llvm.ptr, %arg1: !llvm.ptr, %arg2: i64, %arg3: i64, %arg4: i64, %arg5: i64, %arg6: i64, %arg7: i64, %arg8: i64, %arg9: i64) -> !llvm.ptr {
%0 = llvm.mlir.zero : !llvm.ptr
%1 = llvm.getelementptr %0[1] : (!llvm.ptr) -> !llvm.ptr, f32
%2 = llvm.ptrtoint %1 : !llvm.ptr to i64
%3 = llvm.mul %2, %arg9 : i64
%4 = llvm.getelementptr %arg1[%3] : (!llvm.ptr, i64) -> !llvm.ptr, i8
llvm.return %4 : !llvm.ptr
}
```
|
|
As part of XeVM dialect upsteaming, covers remaining parts required for XeVM dialect integration and testing.
It has two high level components
- XeVM target and serialization support
- XeVM dialect integration tests using level zero runtime
Co-Authored-by: Artem Kroviakov <artem.kroviakov@intel.com>
|
|
Reland https://github.com/llvm/llvm-project/pull/150805
Shared libs build was broken.
Add `${dialect_libs}` and `${conversion_libs}` to
`MLIRRegisterAllExtensions` because it depends on
`registerConvert***ToLLVMInterface` functions.
|
|
(#151118)
Reverts llvm/llvm-project#150805
Some bots are failing.
|
|
`InitAll***` functions are used by `opt`-style tools to init all MLIR
dialects/passes/extensions. Currently they are implemeted as inline
functions and include essentially the entire MLIR header tree. Each file
which includes this header (~10 currently) takes 10+ sec and multiple GB
of ram to compile (tested with clang-19), which limits amount of
parallel compiler jobs which can be run. Also, flang just includes this
file into one of its headers.
Move the actual registration code to the static library, so it's
compiled only once.
Discourse thread
https://discourse.llvm.org/t/rfc-moving-initall-implementation-into-static-library/87559
|
