dist: Change default library to libnccl-net-ofi.so #808
+19
−4
Conversation
This file contains 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
The NGC containers have added some logic to try and support multiple plugins by renaming the OFI plugin to libnccl-net-aws.so. Ignoring the naming for a second, it's clearly useful for us to conform to Sylvain's original expectation that everyone named plugins as libnccl-net-<network>.so and providing a symlink to the generic name where appropriate. This patch causes us to generate the plugin as libnccl-net-ofi.so and create a symlink from libnccl-net-ofi.so to libnccl-net.so for backwards compatibility. Signed-off-by: Brian Barrett <[email protected]>
rauteric
approved these changes
Mar 14, 2025
rauteric
reviewed
Mar 14, 2025
| @@ -248,6 +248,12 @@ AS_IF([test -d "${srcdir}/.git" -a -z "${enable_werror}"], | |||
| CXXFLAGS="${werror_flags} ${CXXFLAGS} " | |||
| CFLAGS="${werror_flags} ${CFLAGS} "]) | |||
|
|
|||
| AC_ARG_ENABLE([nccl-net-symlink], | |||
| [AS_HELP_STRING([--disable-nccl-net-symlink], | |||
| [Create symlink libnccl-net.so to point at libnccl-net-ofi.so. Useful in all cases where | |||
There was a problem hiding this comment.
This help string is worded incorrectly since it is describing the --disable-nccl-net-symlink option. It should say "Do not create ..." and so on.
|
mind adding a 'fixes: #508' to the commit message? |
rajachan
approved these changes
Mar 14, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
The NGC containers have added some logic to try and support multiple plugins by renaming the OFI plugin to libnccl-net-aws.so. Ignoring the naming for a second, it's clearly useful for us to conform to Sylvain's original expectation that everyone named plugins as libnccl-net-.so and providing a symlink to the generic name where appropriate. This patch causes us to generate the plugin as libnccl-net-ofi.so and create a symlink from libnccl-net-ofi.so to libnccl-net.so for backwards compatibility.
By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.