github.com/llvm/Polygeist commits | Ecosyste.ms: OpenCollective

Use a different array expansion test case

66ad9bafe031d7e67a70bce8802d5b07051d3c48 authored almost 1 year ago

Added side effect guards when adding def ops (#23)

ba31d6d52d8fb584889298f8a1095633b295e8bd authored almost 1 year ago

Enables pluto tiling for simple cases (#22)

6f4cdef3c504aea3a9dea37dedd443b4bb56019e authored almost 1 year ago

Naive array expansion implementation

bb3c8645e8e4175f0c759ac820e3d356524c0d3b authored almost 1 year ago

Fixed problems in generating loop bounds with outer IVs (#21)

c9e72923f12b67e132be3898e92348d8362522c2 authored almost 1 year ago

Fixed the issue with <scatnames> (#20)

028febddd143983365ead1c4181e7777a45b2d5b authored almost 1 year ago

[ExtractScopStmt] add more tests and fixed bugs (#103)

* [Pass][ExtractScopStmt] internal index_cast

* Added test cases

* Temporarily added aes.m...

183089657ed10dec1d75e0239717bbe4c2af244a authored almost 1 year ago

Reverted the Polygeist version and check-mlir-clang

b52ce95b2b7b6cc9f092491dd30f97529d36e7ce authored almost 1 year ago

Added a nested reduction test case

f5990ea92deebcd9c8ef218009f039b05dedff06 authored almost 1 year ago

Added a transpose test case for nested loops (#18)

a31f7fcf42eb0243512d9c178c10c5da5855908a authored almost 1 year ago

Changed PLUTO compilation & created opt pass (#13)

00ea62a4c482c4df2b7a9d383bf60f77fb8a23c8 authored almost 1 year ago

Scratchpad created in reg2mem now has scop.scratchpad attr

19af6e6a617de6a4f2b2c193dcdffb7137a3b856 authored almost 1 year ago

Added PLUTO to the compilation flow (#10)

400bcb2151ac64ef93a4856c72b2a28aec89fbc2 authored almost 1 year ago

Added missing clast_binary to AffineExpr translation (#9)

d352af9be3cf475fe2d8972c0edc3b1311605fbb authored almost 1 year ago

Removed llvm in the tree

67d91d1290bdf8232d4f1edb9c05a80d0e5d6e30 authored almost 1 year ago

Handle loop reduction in reg2mem

9f99a994854291460fff04842801c4e5080fb21e authored almost 1 year ago

Initial implementation of store-based OpenScop transform (#4)

* Initial implementation of store-based OpenScop transform

* Fixed (partially) unit tests

6bafbd206e9429e7e8b03bf4c6547d20a0dcbd27 authored almost 1 year ago

Added InvariantScopTransform and polymer-opt (#3)

904dd6a076bfcf7b445c55aebdaf5a02f5b7b244 authored almost 1 year ago

Archieved old tests excl. reg2mem

b079e2476694ae17c9bfa8b398812c913be0091d authored almost 1 year ago

Generate MLIR from OpenScop (#2)

* Generate MLIR from OpenScop

* Added simple unittests and fixed the old ones

* Removed re...

358b2b8c75c95f7ea4b49369f128a4144e35feb5 authored almost 1 year ago

Emitting OpenSCoP from MLIR affine (#1)

* Setup the infra for emitting openscop

* Implemented a draft version

* Added LIT regressi...

44a084a1ed08490a746f9bc63f54058e8a61d5a5 authored almost 1 year ago

[PlutoTransform] fix symbol alignment

71132455d89a553c91a0f4229da63dfa03c4e624 authored almost 1 year ago

[Reg2Mem] fixed issues lead to absent scratchpad (#102)

* [Pass][ExtractScopStmt] internal index_cast

* Added test cases

* Removed unwanted tests

6dc67b7acf0c386e87976f11b153d920082641ed authored almost 1 year ago

Updated symlink and doc

ccd49408f42758b0f3ac464553077c25d8601efe authored almost 1 year ago

Increase external Pluto compilation parallelism

f8845e73d7e5f3bdcb6884a3c5e157d3a4c56a97 authored almost 1 year ago

Update README.md

5a4d646780dfe042bf112abb3033656af608fc58 authored almost 1 year ago

Added PLUTO as submodule and its install insts.

fa1c1debef80cfe28fed9331fdd806643606dd1f authored almost 1 year ago

Canonicalize extract-scop-stmt output/skip sitofp

08dc13e4682c677256997d477fed478e0da72882 authored almost 1 year ago

Added LLVM submodule and install instructions

56c1fd597b562dbc94de4ed4ef5cf5988f8c8bea authored almost 1 year ago

Initial commit

bd6cb08ed5ac601b01232ce87e1c985d8dcf29bc authored almost 1 year ago

Improved update-polygeist.sh for continuous use

2b605dcd0bf0e726d01ce616d32398dac4ee873f authored almost 1 year ago

Added docker setup (#78)

243394247a90424659226cc8342ff93b21378c7c authored almost 1 year ago

split sanity check, insert-redundant-load, etc

e3d4302230b19e95cf1cbc6ee93693d04e073ee7 authored almost 1 year ago

Changed again the Polygeist version

35fe7cbfdee571f005bbc3b7276ef65623d15828 authored almost 1 year ago

Added update-polygeist.sh

Signed-off-by: Ruizhe Zhao <[email protected]>

281134ee1283022114e2593df65036e78f6c7baa authored almost 1 year ago

Added -extract-point-loops to gather point loop bands

8422c78cd8d59f4557e6b042b2ce2e4512c90e72 authored almost 1 year ago

Updated Pluto ref

39a298f654da36c707bf21b99dcae1f3c59c5e05 authored almost 1 year ago

Added Polygeist version.

Signed-off-by: Ruizhe Zhao <[email protected]>

131b50290e834a7659737b7b1837bacd819401b0 authored almost 1 year ago

Improved point-loop annotation by handling affine.apply

448bbb57a3d4f4466ed9a1b8f1a1341ca183e9be authored almost 1 year ago

Support naive point loop annotation

a553d78a06b16e560d3c125d8d3242238722fcec authored almost 1 year ago

Update README.md

51282b0dee9c944a65b62b6f2846c693d62580ff authored almost 1 year ago

Support cloogf and cloogl

94005c7fa3447dfe1553789ab5240d2631f38b60 authored almost 1 year ago

Improved CMake configuration

021fcda840d4c5674544510ebd54d67d17a29369 authored almost 1 year ago

Handle void ternary (#381)

589412047e1472f8bacf646681612659448efd5b authored about 1 year ago

Link c++fs only in case it is needed (#378)

2aab07f17ef364ab5b33109c41fd1061d1eae7dd authored about 1 year ago

Fail when clang FE returned an error (#377)

* Fail when clang FE returned an error

- Add check if clang diagnostics engine reported any e...

bc788a0c60bb4169a9ecb7d2e0a03b9deb963663 authored about 1 year ago

Fix for identifying builtin sqrt (#376)

e971ea4c8cf7d0808cd22142170e8fa5d1253688 authored about 1 year ago

Fix memory bug in ConvertParallelToGPU

4f63629a3be8b5e76557890ff05eecd0f3aa537f authored about 1 year ago

Null terminate alternative kernel ID

f0c1f6da3fbce303b21915f77c1965e943b4d9d1 authored about 1 year ago

Fix memory bug in AffineCFG

9b3ebe904391f20a4bd0f8ab7c390d9bbbdb36d7 authored about 1 year ago

Citing (#374)

36118aea1cb7b0dcfb256efd64d3d3f8c2b15810 authored about 1 year ago

Add floatingreal to complex case and relevant test code (#372)

* Update ReadME.md

* 3413

* Update clang-mlir.cc

* append clang-mlir

* update clang-...

d0adb65bfa4eaa055b0016219c5204816322b025 authored about 1 year ago

Add support for num_threads clause in OpenMP parallel pragma (#369)

9bec02e3e71250239d23d9308b52916d915cf198 authored about 1 year ago

Fix bug with Opaque-ifying identified structs with non-identified members (#365)

7abbb3a73980a8505038ebfcc46c47503ef315bd authored about 1 year ago

Change --function default to * (#364)

06bc20e382af2916579a9c20a80c7fa87f55473d authored over 1 year ago

Squashed commit of the following:

* More flexible specification of coarsening factors

* Fix device-side get_global ops

* For PGO...

887bbd5b253b97f353b7ea0d1d2136d61e5621a1 authored over 1 year ago

Bump LLVM Version

(merge 65fa61d29f5ea6e85aa731f212c5e3374c634e8f)

50d4f21d6db736a5aa5528fa4c1cf2fef61d6408 authored over 1 year ago

Fix PostDec operator (#361)

* Fix PostDec operator

* Add test: decrement operator

423dd1765ea0bd2a5e6971bd7bb40e7791efc1bb authored over 1 year ago

Correctly emit subindex for pointer offset

9608d7954d50e51358666a0ee785c8df0b580d21 authored over 1 year ago

Add -x <language> option

fd4194b099efbf2b717bd9270541e240072f0b1b authored over 1 year ago

Emit builtin_constant_p

253ac70f275e2a9afbd7ca3c2bb2a43651fb1892 authored over 1 year ago

Implement thread coarsening for GPU kernels (#352)

* Add parallel loop unroll pass

* Canonicalizer for flattening alternatives op

* Add env v...

07ec6caf69b980db2636d92628dc85c459294961 authored over 1 year ago

Add ext (#332)

e288f7b56fe108435b99a6ffcc3f64958dca1ff3 authored over 1 year ago

Fix github action

71a2b1f8bae0ac0a03f46f7a5f8bcaeeb9064c3f authored over 1 year ago

Fix tests

7515cc6da4df033009046c83a8a07d1cf8427c9f authored over 1 year ago

Use the C++ language so this can compile (#353)

f6ae41e5e8f2e9b57366fbb3198379b0b9ec9d33 authored over 1 year ago

Support cudaGetDeviceProperties hip translation (#350)

36f98f594078bfbfcaa50b50e311d040a31940b0 authored over 1 year ago

Support for __constant__ and __device__ vars in CUDA (#349)

b894587bc6cfb54c082c2d861f6c2e08a8d2d2b8 authored over 1 year ago

Add support for compilation to ROCM (#344)

3c48b7e88a6dbc41068c2da003a8d31a476fe33f authored over 1 year ago

Fix incorrect trunc of constants in SelectOfExt

6d10d942938d3f12581bf832f8dd37ffca828c8b authored over 1 year ago

fix shared-library build (#348)

Move a table from being a static variable in the dependee lib into a
function returning a refer...

ded532c6931fe5c20d0d2ef7c69b18db526d05df authored over 1 year ago

Implement __builtin_operator_{new,delete}

eda0c6cbf5ae960940e9f57bc5ffe845e6ac7657 authored over 1 year ago

Add option to enable/disable buffer elimination

a02ad762e7baeafc0f5b179468968d4e9724f759 authored over 1 year ago

PGO fixes

004eb023d6097fea2aa7b9d0ee8a831df2fc80c3 authored over 1 year ago

Fix nvvm.maxntidx attr name

33da4958cf28f40bc80d050378f3cf6b4b5720d4 authored over 1 year ago

Add support for using PGO for polygeist alternatives op

9597ff86ca2876972e2884be94e9cc056fdc4eda authored over 1 year ago

Add option to control emitting gpu launch bounds

79f5456dd41055b8b7b80d7bd853b590f87205e9 authored over 1 year ago

Fix parallel to gpu tests

abb9545e8a5889ee69065f06441b0152ea4a3f3d authored over 1 year ago

Switch to Cuda runtime API

c3ea4e84f08cc3fc24a8de1f34b1ff80bf2f5f88 authored over 1 year ago

Add triple and data layout before translating gpu mlir module to llvmir

cb635eabbd184d0a908299586e4ca90b6253eb1d authored over 1 year ago

Add flag to print ir between passes

d62661c38b1201c54f4b6d5c86588cfdeda2a71e authored over 1 year ago

Add polygeist NoopOp

Use a noop to preserve structure of gpu kernels through optimizations

58827e0d5b83d347dbf4ab42fa10f4945c008ba7 authored over 1 year ago

LICM CSE Passes ordering fixes

eabb71923ac0c3c2a53921e0426afeea5907c443 authored over 1 year ago

Add flag to preserve GPU kernel structure while still enabling some parallel optimisations (#331)

* Add wrapper operations for gpu blocks and grids that prevent certain optimizations

* Handle...

0126f75dd142b8586612e38505a385813890ef67 authored almost 2 years ago

Fix use after free bug (#329)

a27def31a2b0742308e7c8db4c0f40a37727269a authored almost 2 years ago

Warn but enable unregistered dialects for pragma (#324)

d36626d45a8d8b77baf24446e3b134a88990a047 authored almost 2 years ago

Heuristic for deciding the block size of GPU kernels (#327)

* Take occupancy into account when emitting GPU kernels

* Add launch bounds to kernels

* A...

2aeaad0a46fb1e230e661e613ef9a99ee851ec1a authored almost 2 years ago

Not all block args dominate while loop to be converted to for loop (#326)

1774948a915c374d65a911c691ffb5700581a944 authored almost 2 years ago

Support sizeof... expressions (#325)

* Support sizeof... expressions

Generate an arith.constant with the length of the input pack ...

b8db7f8837475cff4e62cd07f87bf1610e5bdba3 authored almost 2 years ago

Enable compilation to CUDA (#273)

* Remove device functions from host module

* Integrate gpu to nvvm lowering in PolygeistToL...

e5d8b2bd9e5b70acfae50318adc89da034134d4a authored about 2 years ago

Fix parallel lower preinline (#313)

f7b222bd11e009917667bb2cfaac4aee35823ead authored about 2 years ago

Fix atomic load (#312)

* Fix atomic load

* Fix addrspace conversion

* Parallel lower fixup

d8b6e4f5c83bd25a48bb681fc3180ba3f04611f6 authored about 2 years ago

Misc LULESH fixes (#310)

* Misc LULESH fixes

* Fix glob

* fixes

c235e45170edfa5455acec788be53bab7a6fe68e authored about 2 years ago

Fix builtin addressof handling and reference type initexpr (#304)

* Fix builtin addressof handling and reference type initexpr

* clang-format

* Fix warning ...

911b6a5c621e5aa65b4fcf95814d9b2d1e6c9cdb authored about 2 years ago

Keep InitLLVM around until we exit so that we can use -debug-only= (#303)

f42428658c7a9eeb4156410c1f600015b836f0c9 authored about 2 years ago

Fix async lowering for exec body that lowers to multiple blocks (#302)

* Fix async lowering for exec body that lowers to multiple blocks

* format

* Remove prints...

bcd804bf5b4d376f23f9b9ea205fae4aee6fadca authored about 2 years ago

Fix handling of complex floats (#299)

* Init expr

* Support struct-abi=0 for complex nums

65bc29b6fd44d54e4b229ea8317ac28850d30cee authored about 2 years ago

Fix and add test for async lowering (#301)

* Fix and add test for async lowering

* generalize test

Co-authored-by: Ivan Radanov Ivano...

1621eda8a1a4829fd336d2f3ab056cad671000e8 authored about 2 years ago

Creating an affine if with results needs an else (#300)

4370bbae940ffccd1edf058a82fd12453bfe4b53 authored about 2 years ago

Add cmake option for linker (#295)

d2e4c05782cba96a08ce4efc4318e878200a62e0 authored about 2 years ago