KellenSunderland opened a new pull request #10569: Use default nvcc when 
configured nvcc not present.
   ## Description ##
   This fixes an issue with the nvcc path resolution logic in the Makefile 
(#10553).  In prior releases the makefile would use $(USE_CUDA_PATH)/bin/nvcc 
for nvcc no matter what was specified in config. For some users this was 
actually a good thing if 'nvcc' was not on the path, as it would specify the 
binary with an absolute path.  However the behaviour in the prior release 
caused issues when cross compiling. You may have an arm and x86 cuda path 
specified, and you need to select the correct nvcc for your platform.
   This PR checks the configured nvcc to ensure it is actually on the path.  If 
it isn't it will set the binary to the absolute path.  This should match the 
behaviour of the prior release for users without nvcc on the path, and also 
allow for cross-compilation.
   ## Checklist ##
   ### Essentials ###
   Please feel free to remove inapplicable items for your PR.
   - [x] The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to 
the relevant [JIRA issue]( 
created (except PRs with tiny changes)
   - [x] Changes are complete (i.e. I finished coding on this PR)
   - [x] All changes have test coverage:
   - Unit tests are added for small changes to verify correctness (e.g. adding 
a new operator)
   - Nightly tests are added for complicated/long-running ones (e.g. changing 
distributed kvstore)
   - Build tests will be added for build configuration changes (e.g. adding a 
new build option with NCCL)
   - [x] Code is well-documented: 
   - For user-facing API changes, API doc string has been updated. 
   - For new C++ functions in header files, their functionalities and arguments 
are documented. 
   - For new examples, is added to explain the what the example does, 
the source of the dataset, expected performance on test set and reference to 
the original paper if applicable
   - Check the API doc at$PR_ID/$BUILD_ID/index.html
   - [x] To the my best knowledge, examples are either not affected by this 
change, or have been fixed to be compatible with this change
   ## Comments ##
   - Note on why this was missed in CI: we do not have any CUDA CI 
configurations that do not have nvcc on the path.

This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:

With regards,
Apache Git Services

Reply via email to