-
Notifications
You must be signed in to change notification settings - Fork 20
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
No Pluto optimization in matrix multiplication example #119
Comments
Hi there, you may need to add -reg2mem -extract-scop-stmt before
-pluto-opt, sorry for the confusion. Let me know if you need any further
help.
On Mon, 29 Nov 2021 at 05:17, malixian ***@***.***> wrote:
I build the polymer with impact branch, and try to optimize matmul.mlir.
But I find there is no difference between output mlir and input mlir.
The command I used is polymer-opt -pluto-opt ../matmul.mlir.
And the output mlir is:
module {
func @Matmul() {
%0 = alloc() : memref<64x64xf32>
%1 = alloc() : memref<64x64xf32>
%2 = alloc() : memref<64x64xf32>
affine.for %arg0 = 0 to 64 {
affine.for %arg1 = 0 to 64 {
affine.for %arg2 = 0 to 64 {
%3 = affine.load %0[%arg0, %arg2] : memref<64x64xf32>
%4 = affine.load %1[%arg2, %arg1] : memref<64x64xf32>
%5 = mulf %3, %4 : f32
%6 = affine.load %2[%arg0, %arg1] : memref<64x64xf32>
%7 = addf %5, %6 : f32
affine.store %7, %2[%arg0, %arg1] : memref<64x64xf32>
}
}
}
return
}
}
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#119>, or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ACC42R7EU7FVZALV6YUVNWTUOMEFRANCNFSM5I6HTAHQ>
.
--
Ruizhe Zhao
https://kumasento.github.io
|
Thank you for your reply @kumasento . When I run the command
|
Interesting. Thanks for reporting this.
Would you mind trying out the main branch? Or check out any example under
test/ to see if it can run on your end?
Also, you may want to check .github/workflows/buildAndTest.yml to see how
polymer can be built with the main branch, which is different from impact.
On Tue, 30 Nov 2021 at 01:11, malixian ***@***.***> wrote:
Thank you for your reply @kumasento <https://github.com/kumasento> . When
I run the command ./polymer-opt -reg2mem -extract-scop-stmt -pluto-opt
matmul.mlir, there occurs a core dump.
Error log is:
polymer-opt:
/home/mlx/repos/MLIR-repos/polymer/llvm/mlir/lib/Analysis/AffineStructures.cpp:3050:
void getIndependentConstraints(const mlir::FlatAffineConstraints &,
unsigned int, unsigned int, SmallVectorImpl &, SmallVectorImpl &):
Assertion `pos < cst.getNumIds() && "invalid start position"' failed.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#119 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ACC42R6F5UTIL2RNXEMJFELUOQQCPANCNFSM5I6HTAHQ>
.
--
Ruizhe Zhao
https://kumasento.github.io
|
Thanks, I will try it. |
I build the polymer with impact branch, and try to optimize matmul.mlir. But I find there is no difference between output mlir and input mlir.
The command I used is
polymer-opt -pluto-opt ../matmul.mlir
.And the output mlir is:
When I try to build this project with main branch and use the command
./scripts/build-with-polygeist.sh
, I get a compile error:Is there any solutions to solve these problems above? Thanks.
The text was updated successfully, but these errors were encountered: