Allow rw locking of texture caches #2433
Merged
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
To reduce thread contention on the IC/TS caches, change from unique locks to reader-writer locks. * unordered_map_concurrent: Change the shards from spin_lock to spin_rw_lock. Most operations (insert, erase, holding an iterator, and find, which returns an iterator) are unique (write) locks. But the underappreciated retrieve() method is stateless can be a reader lock! * In imagecache.cpp, change a bunch of the find() calls and fiddling with iterators into simpler calls to retrieve(), which should be able to happen concurrently. On a highly threaded machine (52 cores) this resulted in a 6% gain (geometric mean) across a set of production render frames that we like to test with, but that disguises what's really happening. Most of the individual tests in that suite don't change at all (or vary within the +/- ~1% we expect from timing noise and run-to-run variation. But a handful of the tests sped up by 20-40%! So I think you are unlikely to see any change for a few threads, or even many threads for most situations. But for highly threaded renderings where the access pattern really resulting in the cache locks as a major rendering bottleneck, this should help a lot.
|
LGTM! |
|
ok, here goes nothing... |
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Outdated suggestions cannot be applied.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.
To reduce thread contention on the IC/TS caches, change from unique
locks to reader-writer locks.
unordered_map_concurrent: Change the shards from spin_lock to
spin_rw_lock. Most operations (insert, erase, holding an iterator,
and find, which returns an iterator) are unique (write) locks. But
the underappreciated retrieve() method is stateless and can be a reader
lock!
In imagecache.cpp, change a bunch of the find() calls and fiddling
with iterators into simpler calls to retrieve(), which should be
able to happen concurrently.
On a highly threaded machine (52 cores) this resulted in a 6% gain
(geometric mean) across a set of production render frames that we like
to test with, but that disguises what's really happening. Most of the
individual tests in that suite don't change at all (or vary within the
+/- ~1% we expect from timing noise and run-to-run variation. But a
handful of the tests sped up by 20-40%! So I think you are unlikely
to see any change for a few threads, or even many threads for most
situations. But for highly threaded renderings where the access pattern
really results in the cache locks as a major rendering bottleneck,
this should help a lot.