Merged
Conversation
…nd the part that is CUDA-specific (i.e. needs to be linked with CUDA)
* renamed macros
* Missing parameters that are defaulted in CUDA but not HIP for EGP pushing * Specialised genLazyVariableDynamicAllocation for CUDA and HIP because of malloc/alloc naming difference
…`` test parameters based on available backend modules
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #647 +/- ##
==========================================
- Coverage 88.94% 88.62% -0.33%
==========================================
Files 106 111 +5
Lines 14786 15062 +276
==========================================
+ Hits 13152 13348 +196
- Misses 1634 1714 +80 ☔ View full report in Codecov by Sentry. |
tnowotny
approved these changes
Jan 28, 2025
Member
tnowotny
left a comment
tnowotny
left a comment
There was a problem hiding this comment.
Yes, this makes sense & let's hope that HIP has a future and is not replaced by yet another API ...
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
So, HIP is a clone of CUDA with all the functions and types prefixed with
hipinstead ofcuda. This means it's much closer to CUDA than e.g. OpenCL so I have added another layer of abstraction betweenBackendSIMTand the CUDA backend -BackendCUDAHIP- to reflect this. The CUDA and HIP backends are then derived off this and provide virtuals which provide the prefixes for the various libraries and do anything that actually involves interfacing with the device (gettingcudaDeviceProp). Other than that heavy-lifting, this PR:This is the type of change that is going to be absolutely horrible to merge other stuff onto so, while this isn't really ready for anyone to use until I get to test on an actual AMD card, I'd like to get it merged so I can do other things. It definitely doesn't break CUDA and the HIP backend works fine with the NVIDIA backend which basically just renamed everything back.