maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/mlir/lib/Dialect
History
Christian Sigg f03826f896 Pass GPU events instead of streams across async regions. 
Lower !gpu.async.tokens returned from async.execute regions to events instead of streams.

Make !gpu.async.token returned from !async.execute single-use.
This allows creating one event per use and destroying them without leaking or ref-counting.
Technically we only need this for stream/event-based lowering. I kept the code separate
from the rest of the gpu-async-region pass so that we can make this optional or move
to a separate pass as needed.

Reviewed By: herhut

Differential Revision: https://reviews.llvm.org/D96965
 		2021-02-25 13:18:18 +01:00
..
AVX512 [mlir][vector][avx512] move avx512 lowering pass into general vector lowering 2020-12-03 17:23:46 -08:00
Affine Revert "[MLIR] Create memref dialect and move several dialect-specific ops from std." 2021-02-18 12:49:52 +01:00
ArmNeon [mlir][ArmNeon][RFC] Add a Neon dialect 2020-12-11 13:49:40 +00:00
ArmSVE [mlir][NFC] Remove StandardTypes.h now that all usages point to BuiltinTypes.h 2020-12-16 20:41:42 -08:00
Async [mlir] Async: add a separate pass to lower from async to async.coro and async.runtime 2021-01-26 03:33:20 -08:00
Complex [mlir] Add Complex dialect. 2021-01-15 19:58:10 +01:00
GPU Pass GPU events instead of streams across async regions. 2021-02-25 13:18:18 +01:00
LLVMIR [mlir][IR] Refactor the `getChecked` and `verifyConstructionInvariants` methods on Attributes/Types 2021-02-22 17:37:49 -08:00
Linalg [mlir] Add a builder to `linalg.tiled_loop`. 2021-02-24 14:47:27 +01:00
Math [mlir] Add polynomial approximation for math::LogOp (using builders API) 2021-02-24 07:50:25 -08:00
OpenACC [mlir][Types][NFC] Move all of the builtin Type classes to BuiltinTypes.h 2020-12-03 18:02:10 -08:00
OpenMP [MLIR][OpenMP] Attribute to include WsLoop upperbound 2021-01-08 14:42:18 +00:00
PDL [mlir][PDL] Use ODS for defining PDL types 2021-01-08 12:32:28 -08:00
PDLInterp [mlir][Types][NFC] Move all of the builtin Type classes to BuiltinTypes.h 2020-12-03 18:02:10 -08:00
Quant [mlir][IR] Refactor the `getChecked` and `verifyConstructionInvariants` methods on Attributes/Types 2021-02-22 17:37:49 -08:00
SCF [mlir] NFC - Use declarative assembly for scf::YieldOp 2021-02-23 11:17:30 +00:00
SDBM Separate the Registration from Loading dialects in the Context 2020-08-19 01:19:03 +00:00
SPIRV [mlir][spirv] Define spv.GLSL.Ldexp 2021-02-24 13:07:46 -05:00
Shape [mlir][Shape] Fix a crash when folding nary broadcast ops 2021-02-22 20:06:37 +01:00
StandardOps [mlir] Add constBuilderCall to TypeAttr to simplify builders 2021-02-24 13:04:03 -05:00
Tensor [mlir] Add folding of tensor.cast -> subtensor_insert 2021-02-19 17:24:16 +00:00
Tosa [mlir][nfc] Fix typo in documentation comment 2021-02-25 08:32:14 +01:00
Utils Reland "[MLIR] Make structured op tests permutation invariant" 2021-02-18 18:09:49 -08:00
Vector Reland "[MLIR] Make structured op tests permutation invariant" 2021-02-18 18:09:49 -08:00
CMakeLists.txt Reland "[MLIR] Make structured op tests permutation invariant" 2021-02-18 18:09:49 -08:00
Traits.cpp [mlir][Shape] Generalize cstr_broadcastable folding for n-ary broadcasts 2021-02-17 11:44:52 +01:00