-
Notifications
You must be signed in to change notification settings - Fork 13.8k
StateTransform: Only load pin field once. #147493
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
base: master
Are you sure you want to change the base?
Conversation
Some changes occurred to MIR optimizations cc @rust-lang/wg-mir-opt |
@bors try @rust-timer queue |
This comment has been minimized.
This comment has been minimized.
StateTransform: Only load pin field once.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
Finished benchmarking commit (8a45031): comparison URL. Overall result: ❌✅ regressions and improvements - please read the text belowBenchmarking this pull request means it may be perf-sensitive – we'll automatically label it not fit for rolling up. You can override this, but we strongly advise not to, due to possible changes in compiler perf. Next Steps: If you can justify the regressions found in this try perf run, please do so in sufficient writing along with @bors rollup=never Instruction countOur most reliable metric. Used to determine the overall result above. However, even this metric can be noisy.
Max RSS (memory usage)Results (secondary -5.2%)A less reliable metric. May be of interest, but not used to determine the overall result above.
CyclesResults (secondary 0.3%)A less reliable metric. May be of interest, but not used to determine the overall result above.
Binary sizeResults (primary -0.0%, secondary -1.8%)A less reliable metric. May be of interest, but not used to determine the overall result above.
Bootstrap: 472.232s -> 472.472s (0.05%) |
This comment has been minimized.
This comment has been minimized.
This PR was rebased onto a different master commit. Here's a range-diff highlighting what actually changed. Rebasing is a normal part of keeping PRs up to date, so no action is needed—this note is just to help reviewers. |
The current implementation starts by transforming all instances of
_1
into(*_1)
, and then traverses the body again to transform(*_1)
into(*(_1.0))
, and again forDerefer
.This PR changes the implementation to only traverse the body once. As
_1.0
cannot be not modified inside the body (we just changed its type!), we have no risk of loading from the wrong pointer.