Skip to content

[fft][1 of 3] build system and helpers to support cuFFT and MKL #5855

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.

Sign up for GitHub

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

Jump to bottom
Merged
ezyang merged 3 commits into from
Mar 19, 2018
Merged

[fft][1 of 3] build system and helpers to support cuFFT and MKL #5855

ezyang merged 3 commits into from
Mar 19, 2018

Conversation

@ssnl
Copy link
Collaborator

@ssnl ssnl commented Mar 17, 2018 •
edited
Loading

This is the first of three PRs that #5537 will be split into.

This PR adds mkl headers to included files, and provides helper functions for MKL fft and cuFFT.
In particular, on POSIX, headers are using mkl-include from conda, and on Windows, it is from a new file @yf225 and I made and uploaded to s3.

cc @soumith @apaszke @ezyang

@o8ht88z00f o8ht88z00f mentioned this pull request Mar 17, 2018
Closed
@ssnl ssnl mentioned this pull request Mar 17, 2018
Merged
ezyang
ezyang reviewed Mar 19, 2018
View reviewed changes
.jenkins/win-build.sh Outdated

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

ezyang
ezyang reviewed Mar 19, 2018
View reviewed changes
aten/src/TH/cmake/FindMKL.cmake Outdated

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

ezyang
ezyang reviewed Mar 19, 2018
View reviewed changes
aten/src/ATen/native/cuda/CuFFTUtils.h Outdated

This comment was marked as off-topic.

Sign in to view

ezyang
ezyang reviewed Mar 19, 2018
View reviewed changes
aten/src/ATen/native/cuda/CuFFTUtils.h Outdated

This comment was marked as off-topic.

Sign in to view

This comment was marked as off-topic.

Sign in to view

@ezyang
Copy link
Contributor

ezyang commented Mar 19, 2018

Are the mkl implementations always superior or do you want to turn them off sometimes, in the same way as cudnn.enabled? Maybe this is moot because we'll come up with a new dispatching strategy in C10.

@ssnl
Copy link
Collaborator Author

ssnl commented Mar 19, 2018

@ezyang Currently there is only mkl implementation of fft. I looked into writing a naive inefficient fft in case there is no mkl, but it was a bit tricky and very slow. Since we ship with mkl anyways, I think this should be fine for now.

@ezyang ezyang merged commit 22ef8e5 into pytorch:master Mar 19, 2018
@ejoebstl
Copy link
Contributor

ejoebstl commented Mar 19, 2018 •
edited
Loading

I'm afraid this broke the build from source. It seems like my mkl lib was found, but still there is a variable that's not set.

-- Checking for [mkl_gf_lp64 - mkl_gnu_thread - mkl_core - gomp - pthread - m - dl]
--   Library mkl_gf_lp64: /home/emi/anaconda2/lib/libmkl_gf_lp64.so
--   Library mkl_gnu_thread: /home/emi/anaconda2/lib/libmkl_gnu_thread.so
--   Library mkl_core: /home/emi/anaconda2/lib/libmkl_core.so
--   Library gomp: -fopenmp
--   Library pthread: /usr/lib/x86_64-linux-gnu/libpthread.so
--   Library m: /usr/lib/x86_64-linux-gnu/libm.so
--   Library dl: /usr/lib/x86_64-linux-gnu/libdl.so
-- MKL library found
-- Found a library with BLAS API (mkl).
-- Found a library with LAPACK API (mkl).
-- Found cuDNN: v6.0.21  (include: /usr/local/cudnn/include, library: /usr/local/cudnn/lib64/libcudnn.so.6)
-- Could NOT find NNPACK (missing:  NNPACK_INCLUDE_DIR NNPACK_LIBRARY CPUINFO_LIBRARY PTHREADPOOL_LIBRARY) 
-- NNPACK not found. Compiling without NNPACK support
-- Using python found in /home/emi/anaconda2/bin/python
disable contrib because ATEN_NO_CONTRIB is set
CMake Error: The following variables are used in this project, but they are set to NOTFOUND.
Please set them or make sure they are set and tested correctly in the CMake files:
MKL_INCLUDE_DIR (ADVANCED)

update

Okay, the problem was the new mkl-include dependency, which can only be installed with a newer conda version than I was using. A more descriptive error-message for this case might be useful.

ezyang added a commit that referenced this pull request Mar 19, 2018
@ezyang
Revert "[fft][1 of 3] build system and helpers to support cuFFT and M…
34a8d29 
…KL (#5855)"

This reverts commit 22ef8e5.
soumith added a commit that referenced this pull request Mar 19, 2018
@soumith
Revert "[fft][1 of 3] build system and helpers to support cuFFT and M…
4d34b63 
…KL (#5855)"

This reverts commit 22ef8e5.
soumith added a commit that referenced this pull request Mar 19, 2018
@soumith
Revert "Enable resetting of batchnorm running stats and cumulative ("…
7e13138 
…simple") moving average" (#5892)

* Revert "Port ATen and JIT C++ tests to Catch2 (#5788)"

This reverts commit 6f80023.

* Revert "Fix error message for cat-ing zero-dim tensors (#5819)"

This reverts commit cf2e176.

* Revert "Softmax symbolic should account for negative dim (#5846)"

This reverts commit ba64724.

* Revert "[fft][1 of 3] build system and helpers to support cuFFT and MKL (#5855)"

This reverts commit 22ef8e5.

* Revert "Don't modify requires_grad when running DataParallel in no_grad mode (#5880)"

This reverts commit d11b7fb.

* Revert "fix some methods not showing up in doc (#5882)"

This reverts commit 24fca0e.

* Revert "ReduceOps cleanup and set_num_threads (#5723)"

This reverts commit 84400d5.

* Revert "introduce shape_as_tensor and reshape_from_variable_shape (#5824)"

This reverts commit f446b82.

* Revert "Enable resetting of batchnorm running moments and cumulative ("simple") moving average (#5766)"

This reverts commit 99b1f6c.
@ssnl ssnl deleted the fft_build branch March 19, 2018 22:19
jekbradbury pushed a commit to jekbradbury/pytorch that referenced this pull request Mar 21, 2018
@ssnl @jekbradbury
[fft][1 of 3] build system and helpers to support cuFFT and MKL (pyto…
78eddbe 
…rch#5855)

This is the first of three PRs that pytorch#5537 will be split into.

This PR adds mkl headers to included files, and provides helper functions for MKL fft and cuFFT.
In particular, on POSIX, headers are using mkl-include from conda, and on Windows, it is from a new file @yf225 and I made and uploaded to s3.

* add mkl-include to required packages

* include MKL headers; add AT_MKL_ENABLED flag; add a method to query MKL availability

* Add MKL and CUFFT helpers
jekbradbury pushed a commit to jekbradbury/pytorch that referenced this pull request Mar 21, 2018
@soumith @jekbradbury
Revert "Enable resetting of batchnorm running stats and cumulative ("…
52c8d76 
…simple") moving average" (pytorch#5892)

* Revert "Port ATen and JIT C++ tests to Catch2 (pytorch#5788)"

This reverts commit 6f80023.

* Revert "Fix error message for cat-ing zero-dim tensors (pytorch#5819)"

This reverts commit cf2e176.

* Revert "Softmax symbolic should account for negative dim (pytorch#5846)"

This reverts commit ba64724.

* Revert "[fft][1 of 3] build system and helpers to support cuFFT and MKL (pytorch#5855)"

This reverts commit 22ef8e5.

* Revert "Don't modify requires_grad when running DataParallel in no_grad mode (pytorch#5880)"

This reverts commit d11b7fb.

* Revert "fix some methods not showing up in doc (pytorch#5882)"

This reverts commit 24fca0e.

* Revert "ReduceOps cleanup and set_num_threads (pytorch#5723)"

This reverts commit 84400d5.

* Revert "introduce shape_as_tensor and reshape_from_variable_shape (pytorch#5824)"

This reverts commit f446b82.

* Revert "Enable resetting of batchnorm running moments and cumulative ("simple") moving average (pytorch#5766)"

This reverts commit 99b1f6c.
wuhuikx pushed a commit to wuhuikx/pytorch that referenced this pull request Jan 30, 2020
@ezyang
Revert "[fft][1 of 3] build system and helpers to support cuFFT and M…
b868867 
…KL (pytorch#5855)"

This reverts commit 22ef8e5.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@ezyang ezyang ezyang left review comments

@peterjc123 peterjc123 peterjc123 left review comments

Assignees

No one assigned

Labels

open source

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

4 participants

@ssnl @ezyang @ejoebstl @peterjc123