[OpenMP] Add skewed iteration distribution on hybrid systems #69946
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
This commit adds skewed distribution of iterations in nonmonotonic:dynamic schedule (static steal) for hybrid systems when thread affinity is assigned. Currently, it distributes the iterations at 60:40 ratio. Consider this loop with dynamic schedule type, for (int i = 0; i < 100; ++i) In a hybrid system with 20 hardware threads (16 CORE and 4 ATOM core), 88 iterations will be assigned to performance cores and 12 iterations will be assigned to efficient cores. Each thread with CORE core will process 5 iterations + extras and with ATOM core will process 3 iterations.
jpeyton52
requested review from
TerryLWilmarth,
hansangbae,
nawrinsu and
jdoerfert
jdoerfert
reviewed
Oct 24, 2023
openmp/runtime/src/kmp.h
Outdated
| unsigned unused : 28; | ||
| #endif | ||
| #elif KMP_AFFINITY_SUPPORTED && KMP_STATIC_STEAL_ENABLED | ||
| unsigned use_hybrid : 1; |
There was a problem hiding this comment.
Is it really worth adding all these cases for 1 bit?
There was a problem hiding this comment.
Removed the macros here and put comments for relevant macros
openmp/runtime/src/kmp_dispatch.cpp
Outdated
| @@ -90,6 +90,69 @@ static inline int __kmp_get_monotonicity(ident_t *loc, enum sched_type schedule, | |||
| return monotonicity; | |||
| } | |||
|
|
|||
| #if KMP_WEIGHTED_ITERATIONS_SUPPORTED | |||
| static inline float __kmp_get_float_val(float num) { | |||
There was a problem hiding this comment.
Add a comment, get_float_val float->float is really not explaining much
There was a problem hiding this comment.
Added comment, and changed function name for clarity.
|
✅ With the latest revision this PR passed the C/C++ code formatter. |
|
@jdoerfert , friendly ping on this |
|
@jdoerfert , friendly ping on this. |
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.
This commit adds skewed distribution of iterations in
nonmonotonic:dynamicschedule (static steal) for hybrid systems when thread affinity is assigned. Currently, it distributes the iterations at 60:40 ratio. Consider this loop with dynamic schedule type,for (int i = 0; i < 100; ++i). In a hybrid system with 20 hardware threads (16 CORE and 4 ATOM core), 88 iterations will be assigned to performance cores and 12 iterations will be assigned to efficient cores. Each thread with CORE core will process 5 iterations + extras and with ATOM core will process 3 iterations.Original Phabricator Patch: https://reviews.llvm.org/D152955