pstarkcdpr

Thanks @MaheshRavishankar. The actual FFT implementation is one large change (`RewriteFft.cpp`) and a few minor ones. I don't think that can't really be separated out easily. Fixing the tests that...

Thanks @MaheshRavishankar. I'm very new to making changes in this project, so I wouldn't be surprised if there are things that could be done better. The changes in `RewriteFft.cpp` still...

There was a CI failure due to a mistake I made with an excluded file in a `BUILD.bazel`. Should be fixed now. Not sure if there is a way to...

Fixed the pre-commit formatting changes. It's my first PR in this repo so I had no idea that existed. I had been trying to stay within 80 character line limit...

Sorry to revive an old bug (let me know if I should log a new issue instead), but I ran into the same thing. I don't use the VMVX backend...

Thanks @MaheshRavishankar. I wanted to break this out as a separate issue since we are actively running into it. I added `util.optimization_barrier %9#1` to our local compiler build to work...

@MaheshRavishankar Sorry for the delay. I finally got around to checking if `--iree-llvmcpu-fail-on-out-of-bounds-stack-allocation=false` works around the issue. On Linux, it seems to. On Windows, however, the runtime segfaults. ``` iree-run-module...

I wondered about that, but I'm not very familiar with MLIR files. Do you know what good linalg code would look like here (sorry if that's in the linked PR...

Oh. I wondered why the input tensors to `linalg_ext.fft` were in `outs`. I guess it's an in-place FFT. I modified the StableHLO lowering code to use a new tensor initialized...

Hi @hanhanW. I've been looking into this more. I worked around the problem with inputs, now I just have a problem with outputs. I mentioned this above, but the trouble...