Shared fix fanout #7
Closed
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
Signed-off-by: Nadia Derbey <[email protected]>
dycz0fx
approved these changes
Jul 10, 2018
bosilca
reviewed
Jul 10, 2018
| @@ -9,7 +9,7 @@ int mca_coll_shared_reduce_intra(const void *sbuf, void* rbuf, int count, | |||
| ptrdiff_t extent, lower_bound; | |||
| ompi_datatype_get_extent(dtype, &lower_bound, &extent); | |||
| if (count*extent <= 256) { | |||
| mca_coll_shared_reduce_binomial(sbuf, rbuf, count, dtype, op, root, comm, module); | |||
| mca_coll_shared_reduce_linear(sbuf, rbuf, count, dtype, op, root, comm, module); | |||
There was a problem hiding this comment.
Why do we force the linear for small messages ? Logically the binomial seems like a better choice.
|
I forced it because from the slides Xi showed us during a meeting, that was the best algo for small sizes. |
|
Ok, could find the graph.
|
|
We will reevaluate the shared component and will adapt the selection logic. Meanwhile I will close this PR. |
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Outdated suggestions cannot be applied.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.
During last meeting, we saw on Xi's slides that the mca_coll_shared_reduce_linear algo is the best for sizes <= 256 bytes.
So I did 2 changes:
. I fixed the fanout limit to avoid a segfault when calling mca_coll_shared_reduce_linear ()
. I put the call to this algo into mca_coll_shared_reduce_intra()