Polygeist
Polygeist copied to clipboard
llvm
C/C++ frontend for MLIR. Also features polyhedral optimizations, parallel optimizations, and more!
Using the latest commit. Here is a minimal reproducible code: ```C #include void foo(uint16_t *m, uint16_t len) { for(uint16_t i = 0; i < len; i++){ m[i] = 1; }...
Here is a minimal reproducible code: ```C #include void foo(uint16_t *a, uint16_t len) { uint16_t c = 1; for (uint16_t i = 0; i < len; i++) { c *=...
Hi! I am using this commit: https://github.com/llvm/Polygeist/commit/a27def31a2b0742308e7c8db4c0f40a37727269a to transform C++ code to MLIR. With this simple C++ code (**test.cpp**): ``` #include std::string test(int x) { return std::to_string(x); } ``` I...
I'm trying to use polygeist to convert C to MLIR with the following command: ``` clang -E aes.c -o aes-pre.c cgeist aes-pre.c -S --function=* --raise-scf-to-affine -o aes.affine.mlir ``` And the...
Hi, developers, I found that there is a branch gpu_comp in polygeist, but it seems only lower the IR generated by cuda-lower, do you have plan to support IR generated...
Hi, I tried to convert polybench c to mlir through cgeist and lower to gpu dialect through mlir-opt, however, seems mlir generated by cgeist is invalid to mlir-opt, do you...
for the following case, if it raise to affine through cgeist, matmul can be parallized by polygeist-opt(affine-parallize), but DSPF_sp_mat_trans_cplx can't be parallize, if it raise to affine through polygeist, then...
**Background**: I was cross-compiling(64bit arch-> 32bit arch) polybench using polygeist while meeting an error. After debugging, I found the datalayout information is lost in frontend. I add the compiled options:...
```cpp void Cast1() { void *mem = malloc(100); char *arr = (char *)mem; arr[0] = '%'; arr[1] = '#'; arr[2] = '\0'; printf("%s\n" , arr); free(mem); } /* func.func @Cast1()...
