-
Notifications
You must be signed in to change notification settings - Fork 5.3k
Use zero-extension inMemory<T>.Pin
#119895
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Open
xtqqczze
wants to merge
1
commit into
dotnet:main
Choose a base branch
from
xtqqczze:nint-memorypin
base: main
Could not load branches
Branch not found: {{ refName }}
Loading
Could not load tags
Nothing to show
Loading
Are you sure you want to change the base?
Some commits from the old base branch may be removed from the timeline,
and old review comments may become outdated.
Open
Changes from all commits
Commits
File filter
Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
There are no files selected for viewing
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 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
Oops, something went wrong.
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Outdated suggestions cannot be applied.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.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
nit: I don't think the comment being moved down is "clearer", it just creates a larger diff.
The same with changing the name from
stringDatatoreference.For the actual
Unsafe.Add, it can just be(uint)_index, you don't also need the cast to(nint). However, it's also worth noting that a lot of modern hardware has zero-cost sign-extension, so I'm not sure the extra churn/complexity is worthwhile.This is unlikely to be a hot path as repeated pinning is itself expensive and the potential 1-cycle savings will get lost in the general noise due to the overhead of GCHandle.Alloc and freeing the handle.
Uh oh!
There was an error while loading. Please reload this page.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We see diffs like this on X64:
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
There's diffs on ARM64 as well. It seems like the primary improvement of the change is actually to addressing modes.
Uh oh!
There was an error while loading. Please reload this page.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The latter isn't "strictly" better. It may actually be more expensive due to using a more expensive
leamode which is restricted to a single port, rather than the simplerleamode which works on multiple ports.The former could also become (almost) the latter if it was better and looks like just a missing JIT opt. Namely, it could be (only differing by the
movsxd):That is, there shouldn't be anything here that blocks the
r14+0x10from being contained since it's just(x+y)+(8*z)and changing tox+(8*z)+yThere was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It seems far more meaningful to investigate what's "blocking" that transform right now and whether it is actually more beneficial to do in typical scenarios.
It would likely have much broader impact to ecosystem and not require manually fixing a bunch of C# code to explicitly cast to
uintThere was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@jakobbotsch Is this an example of the following issue?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Well currently it is not possible because the logic reads
_indexmultiple times, so it cannot be proven to be never negative. We could readvar @this = thisat the beginning of the method, this would certainly help.There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The memory model allows the reads to be hoisted in various scenarios, so there’s really nothing that should be blocking if we aren’t explicitly writing from the same thread