Repository navigation
Refactor representation of thread local state. - #751
Merged
Matthew Parkinson (mjp41) merged 8 commits intoMar 21, 2025
Merged
Conversation
Matthew Parkinson (mjp41)
force-pushed
the
thread_local_refactor
branch
2 times, most recently
from
February 20, 2025 13:52
318a880 to
0614a01
Compare
Matthew Parkinson (mjp41)
changed the title
Matthew Parkinson (mjp41)
force-pushed
the
thread_local_refactor
branch
3 times, most recently
from
February 21, 2025 14:22
80d24e0 to
0fd6768
Compare
Contributor
|
Only found time to make it into "Restrict ThreadAlloc usage to globalalloc", but so far so great. Thanks for nwfifying the commit stack. :D |
Matthew Parkinson (mjp41)
force-pushed
the
thread_local_refactor
branch
5 times, most recently
from
February 21, 2025 20:46
41392d1 to
654bcfd
Compare
Nathaniel Wesley Filardo (nwf)
left a comment
Nathaniel Wesley Filardo (nwf)
left a comment
Contributor
There was a problem hiding this comment.
This continues to look really nice. (Read up through "WIP", exclusive.)
| return self->alloc_not_small<zero_mem, CheckInitDefault>(size); | ||
| }, | ||
| size, | ||
| this); |
Contributor
There was a problem hiding this comment.
It's a bit of a pity that something so formulaic can't, AFAIK, be templated away.
Matthew Parkinson (mjp41)
force-pushed
the
thread_local_refactor
branch
from
February 22, 2025 20:17
654bcfd to
fdab65d
Compare
Member
Author
|
Nathaniel Wesley Filardo (@nwf) thanks for the comments. I've addressed them. And I landed the first set of commits in #750. |
Matthew Parkinson (mjp41)
force-pushed
the
thread_local_refactor
branch
2 times, most recently
from
February 27, 2025 16:25
997eed4 to
ab58e2f
Compare
Matthew Parkinson (mjp41)
force-pushed
the
thread_local_refactor
branch
from
March 7, 2025 16:02
295eaa4 to
ed81ba7
Compare
Matthew Parkinson (mjp41)
changed the title
Matthew Parkinson (mjp41)
requested a review
from Nathaniel Wesley Filardo (nwf)
March 7, 2025 20:59
Matthew Parkinson (mjp41)
requested a review
from Schrodinger ZHU Yifan (SchrodingerZhu)
March 14, 2025 16:14
The check init code was tightly integrated into LocalAllocator. This commit pull that code out into ThreadAlloc, and passes a template parameter into the remaining LocalAllocator to perform the relevant TLS manipulations. This removes some of the awkward layering around register_clean_up.
Fully disable lotsofthreads test Need to investigate if the test is unreliable, or we have actually regressed perf. A quick mimalloc-bench didn't show any regressions.
This introduces one additional branch on when processing a batch of messages, but it is likely to only be hit when a lot of messages are processed.
Matthew Parkinson (mjp41)
force-pushed
the
thread_local_refactor
branch
from
March 14, 2025 16:49
ed81ba7 to
48fa1c4
Compare
This combines the notion of CoreAlloc, LocalAlloc and LocalCache into a single class. Previously, these were separated so that a more complex structure would be stored directly in the TLS. This however, proved to be bad for compatibility if the allocator is part of the libc implementation. This commit collapses all the stages of the allocator into a single class. This simplifies the sequencing and overall is a nice reduction in complexity.
Matthew Parkinson (mjp41)
force-pushed
the
thread_local_refactor
branch
from
March 14, 2025 17:05
48fa1c4 to
6f7c8ec
Compare
Matthew Parkinson (mjp41)
force-pushed
the
thread_local_refactor
branch
from
March 21, 2025 14:20
55ed776 to
35d5f46
Compare
Matthew Parkinson (mjp41)
force-pushed
the
thread_local_refactor
branch
from
March 21, 2025 14:42
35d5f46 to
bbc41b6
Compare
Member
Author
|
I have been through this carefully today, and am going to merge. There is one outstanding issue that I am not sure how we should modify the domestication test. There is complexity with how BatchIt interacts. I will raise an issue to fix this down the line, but this shouldn't block other work. |
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
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.
Change thread local state to be a single pointer to a backend allocated structure. Previously, the TLS contained a complex structure, but that didn't interact well with loading numerous copies at once.