Cuda and OpenCL fixes #8
Open
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
jbedorf
reviewed
Nov 5, 2023
| fprintf(stderr, "ERROR: is not implemented in OpenCL, only in CUDA. Please"); | ||
| fprintf(stderr, "ERROR: file an issue on GitHub if you need this combination."); | ||
| exit(1); | ||
| #else |
There was a problem hiding this comment.
I guess there's no need for the else? Given that there's the exit above? In that case maybe change the #else into #endif
There was a problem hiding this comment.
The problem here actually was that float3 was undefined when compiling with OpenCL, causing a compiler error. So we could move the perThreadSM = ... line outside of the #ifdef, but then it wouldn't compile for OpenCL.
I think I later saw a header somewhere that aliases some OpenCL types to CUDA-like names, so maybe float3 can be added there to fix it instead, I'll have a look.
| ocl_free(); | ||
| cmalloc(src.n, DeviceMemFlags); | ||
| ocl_free(); | ||
| allocate(src.n, DeviceMemFlags); |
There was a problem hiding this comment.
Yes, this looks to be the right thing todo.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Here are some minimal updates to make sapporo2 compile with newer versions of CUDA and OpenCL.
The
cmallocfunction seems to be called by the OpenCL code, but is missing. Possibly it's present in the codes that use sapporo, and since everything is linked statically it gets taken from there when that code is linked? It seems like theallocatemember function is what should have been called there, but I'm not sure.I've tested this with CUDA, and can run the test codes, but with nVidia OpenCL they crash with
Since OpenCL is obsolete anyway, maybe we don't bother going into this?
(I have more and bigger changes in the pipeline, but I'm going to try to split them into small, easy to review PRs to hopefully not take too much of your time. Thanks!)