[MIR trans] Optimize trans for biased switches #33566
Merged
+26
−10
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
|
r? @arielb1 (rust_highfive has picked a reviewer for you, use r? to override) |
| // If a single target basic blocks is predominant, promote that to be the | ||
| // default case for the switch instruction to reduce the size of the generated | ||
| // code. This is especially helpful in cases like an if-let on a huge enum. | ||
| Some((&&bb, &c)) if c as f32 / targets.len() as f32 >= 0.05 => { |
| // instruction so LLVM knows that | ||
| let unreachable_blk = self.unreachable_block(); | ||
| let switch = bcx.switch(discr, unreachable_blk.llbb, targets.len()); | ||
| let mut bb_hist = FnvHashMap(); |
There was a problem hiding this comment.
You should add a FIXME here to check for the validity of such transformation. While currently the otherwise branch is always unreachable (and contains a panic), this might not be the case after a bunch of optimisation passes are added to the MIR thus making this transformation invalid.
There was a problem hiding this comment.
TerminatorKind::Switch can't have an otherwise branch in any case.
There was a problem hiding this comment.
There's also the assert_eq!(adt_def.variants.len(), targets.len()); which should in theory make sure we've covered all the cases.
Currently, all switches in MIR are exhausitive, meaning that we can have a lot of arms that all go to the same basic block, the extreme case being an if-let expression which results in just 2 possible cases, be might end up with hundreds of arms for large enums. To improve this situation and give LLVM less code to chew on, we can detect whether there's a pre-dominant target basic block in a switch and then promote this to be the default target, not translating the corresponding arms at all. In combination with rust-lang#33544 this makes unoptimized MIR trans of nickel.rs as fast as using old trans and greatly improves the times for optimized builds, which are only 30-40% slower instead of ~300%. cc rust-lang#33111
|
I guess this is fine to land. I’d still prefer it to become a proper MIR transformation down the line but this can be done later. @bors r+ |
|
📌 Commit 49b2cdf has been approved by |
Manishearth
added a commit
to Manishearth/rust
that referenced
this pull request
May 12, 2016
[MIR trans] Optimize trans for biased switches Currently, all switches in MIR are exhausitive, meaning that we can have a lot of arms that all go to the same basic block, the extreme case being an if-let expression which results in just 2 possible cases, be might end up with hundreds of arms for large enums. To improve this situation and give LLVM less code to chew on, we can detect whether there's a pre-dominant target basic block in a switch and then promote this to be the default target, not translating the corresponding arms at all. In combination with rust-lang#33544 this makes unoptimized MIR trans of nickel.rs as fast as using old trans and greatly improves the times for optimized builds, which are only 30-40% slower instead of ~300%. cc rust-lang#33111
eddyb
added a commit
to eddyb/rust
that referenced
this pull request
May 12, 2016
[MIR trans] Optimize trans for biased switches Currently, all switches in MIR are exhausitive, meaning that we can have a lot of arms that all go to the same basic block, the extreme case being an if-let expression which results in just 2 possible cases, be might end up with hundreds of arms for large enums. To improve this situation and give LLVM less code to chew on, we can detect whether there's a pre-dominant target basic block in a switch and then promote this to be the default target, not translating the corresponding arms at all. In combination with rust-lang#33544 this makes unoptimized MIR trans of nickel.rs as fast as using old trans and greatly improves the times for optimized builds, which are only 30-40% slower instead of ~300%. cc rust-lang#33111
eddyb
added a commit
to eddyb/rust
that referenced
this pull request
May 13, 2016
[MIR trans] Optimize trans for biased switches Currently, all switches in MIR are exhausitive, meaning that we can have a lot of arms that all go to the same basic block, the extreme case being an if-let expression which results in just 2 possible cases, be might end up with hundreds of arms for large enums. To improve this situation and give LLVM less code to chew on, we can detect whether there's a pre-dominant target basic block in a switch and then promote this to be the default target, not translating the corresponding arms at all. In combination with rust-lang#33544 this makes unoptimized MIR trans of nickel.rs as fast as using old trans and greatly improves the times for optimized builds, which are only 30-40% slower instead of ~300%. cc rust-lang#33111
Manishearth
added a commit
to Manishearth/rust
that referenced
this pull request
May 14, 2016
[MIR trans] Optimize trans for biased switches Currently, all switches in MIR are exhausitive, meaning that we can have a lot of arms that all go to the same basic block, the extreme case being an if-let expression which results in just 2 possible cases, be might end up with hundreds of arms for large enums. To improve this situation and give LLVM less code to chew on, we can detect whether there's a pre-dominant target basic block in a switch and then promote this to be the default target, not translating the corresponding arms at all. In combination with rust-lang#33544 this makes unoptimized MIR trans of nickel.rs as fast as using old trans and greatly improves the times for optimized builds, which are only 30-40% slower instead of ~300%. cc rust-lang#33111
bors
added a commit
that referenced
this pull request
May 14, 2016
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.
Currently, all switches in MIR are exhausitive, meaning that we can have
a lot of arms that all go to the same basic block, the extreme case
being an if-let expression which results in just 2 possible cases, be
might end up with hundreds of arms for large enums.
To improve this situation and give LLVM less code to chew on, we can
detect whether there's a pre-dominant target basic block in a switch
and then promote this to be the default target, not translating the
corresponding arms at all.
In combination with #33544 this makes unoptimized MIR trans of
nickel.rs as fast as using old trans and greatly improves the times for
optimized builds, which are only 30-40% slower instead of ~300%.
cc #33111