Hasten macro parsing by nnethercote · Pull Request #68848 · rust-lang/rust

nnethercote · 2020-02-05T05:15:06Z

r? @eddyb

nnethercote · 2020-02-05T05:15:44Z

This is a hacky but easy change that gets some of the wins that #68836 would get.

nnethercote · 2020-02-05T05:59:51Z

@bors try @rust-timer queue

rust-timer · 2020-02-05T05:59:53Z

Awaiting bors try build completion

bors · 2020-02-05T06:00:04Z

⌛ Trying commit e30f495ecec1a30ac97215bee2107b4fdaa1a884 with merge 0c7ea93de45e21ba54a804998b0b7563386ba795...

Centril · 2020-02-05T06:08:40Z

cc @petrochenkov

bors · 2020-02-05T08:43:32Z

☀️ Try build successful - checks-azure
Build commit: 0c7ea93de45e21ba54a804998b0b7563386ba795 (0c7ea93de45e21ba54a804998b0b7563386ba795)

rust-timer · 2020-02-05T08:43:35Z

Queued 0c7ea93de45e21ba54a804998b0b7563386ba795 with parent 002287d, future comparison URL.

nnethercote · 2020-02-05T19:41:23Z

As expected, the perf results are excellent for html5ever and are negligible for all other benchmarks.

petrochenkov · 2020-02-05T21:03:27Z

I like this, the idea is pretty simple and I wouldn't even call it a hack.
I've left one comment.

petrochenkov · 2020-02-08T12:42:33Z

The Directory change is not obvious to me.
Even if we now clone parser, the directory cloning is

still needed much rarer than the parser cloning,
parser cloning is more expensive if there's an owned directory inside it.

The change is justified if the directory cloning is so relatively cheap that it is lost in the noise from the whole parser cloning.
Have you measured the effect?
(Otherwise, I'd make this change because it makes code simpler.)

nnethercote · 2020-02-09T01:08:47Z

Parser cloning is very rare on this code path, so much so that there is no measurable benefit to making it cheaper.

I previously made Directory a Cow to avoid allocations on this hot path. This new Parser-is-Cow optimization subsumes the old Directory-is-Cow optimization.

Does that help? I'm not sure how else to explain it.

nnethercote · 2020-02-10T04:45:43Z

@petrochenkov are you happy to r+ this instead of @eddyb?

petrochenkov · 2020-02-10T06:13:55Z

Yes, r? @petrochenkov

I left a couple of comments, r=me after addressing them.

Currently, every iteration of the main loop in `generic_extension` instantiates a `Parser`, which is expensive because `Parser` is a large type. Many of those instantiations are only used immutably, particularly for simple-but-repetitive macros of the sort seen in `html5ever` and PR 68836. This commit initializes a single parser outside the loop, and then uses `Cow` to avoid cloning it except for the mutating iterations. This speeds up `html5ever` runs by up to 15%.

The previous commit wrapped `Parser` within a `Cow` for the hot macro parsing path. As a result, there's no need for the `Cow` within `Directory`, because it lies within `Parser`.

This is a small win, because `Failure` is much more common than `Success`.

nnethercote · 2020-02-10T22:42:22Z

I addressed the comments, and added a trivial commit fixing a typo in a variable name.

@bors r=petrochenkov

bors · 2020-02-10T22:42:51Z

📌 Commit 2a13b24 has been approved by petrochenkov

@eddyb

…=petrochenkov Hasten macro parsing r? @eddyb

@eddyb

…=petrochenkov Hasten macro parsing r? @eddyb

@ghost

Rollup of 9 pull requests Successful merges: - #67642 (Relax bounds on HashMap/HashSet) - #68848 (Hasten macro parsing) - #69008 (Properly use parent generics for opaque types) - #69048 (Suggestion when encountering assoc types from hrtb) - #69049 (Optimize image sizes) - #69050 (Micro-optimize the heck out of LEB128 reading and writing.) - #69068 (Make the SGX arg cleanup implementation a NOP) - #69082 (When expecting `BoxFuture` and using `async {}`, suggest `Box::pin`) - #69104 (bootstrap: Configure cmake when building sanitizer runtimes) Failed merges: r? @ghost

bors · 2020-02-13T08:14:50Z

☔ The latest upstream changes (presumably #69118) made this pull request unmergeable. Please resolve the merge conflicts.

nnethercote · 2020-02-13T23:36:46Z

Something weird happened here. Somehow an old version of this branch got merged. I will have to put the orphaned changes in a separate PR.

nnethercote · 2020-02-14T02:45:18Z

Something weird happened here. Somehow an old version of this branch got merged. I will have to put the orphaned changes in a separate PR.

#69150 is the PR.

@petrochenkov

…ochenkov Follow-up to rust-lang#68848 This PR contains some late changes to rust-lang#68848 that somehow didn't get included when that PR was merged in a roll-up. r? @petrochenkov

@ghost

Rollup of 7 pull requests Successful merges: - #68129 (Correct inference of primitive operand type behind binary operation) - #68475 (Use a `ParamEnvAnd<Predicate>` for caching in `ObligationForest`) - #68856 (typeck: clarify def_bm adjustments & add tests for or-patterns) - #69051 (simplify_try: address some of eddyb's comments) - #69128 (Fix extra subslice lowering) - #69150 (Follow-up to #68848) - #69164 (Update pulldown-cmark dependency) Failed merges: r? @ghost

rust-highfive assigned eddyb Feb 5, 2020

rust-highfive added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Feb 5, 2020

petrochenkov self-assigned this Feb 5, 2020

petrochenkov reviewed Feb 5, 2020
View reviewed changes

Comment threadsrc/librustc_expand/mbe/macro_rules.rs Outdated

petrochenkov added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Feb 5, 2020

petrochenkov reviewed Feb 5, 2020
View reviewed changes

Comment threadsrc/librustc_expand/mbe/macro_rules.rs Outdated

nnethercote force-pushed the hasten-macro-parsing branch from e30f495 to 2a13b24Compare February 6, 2020 05:56

petrochenkov added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Feb 6, 2020

petrochenkov reviewed Feb 8, 2020
View reviewed changes

Comment threadsrc/librustc_expand/lib.rs Outdated
Comment threadsrc/librustc_expand/mbe/macro_rules.rs Outdated

petrochenkov added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Feb 8, 2020

rust-highfive unassigned eddyb Feb 10, 2020

nnethercote added 2 commits February 11, 2020 08:43

Remove the Cow from Directory.
63bc4b3
The previous commit wrapped `Parser` within a `Cow` for the hot macro parsing path. As a result, there's no need for the `Cow` within `Directory`, because it lies within `Parser`.

nnethercote added 2 commits February 11, 2020 08:43

Change condition ordering in parse_tt.
e6e6bb5
This is a small win, because `Failure` is much more common than `Success`.

Fix a typo in a variable name.
6851f9a

nnethercote force-pushed the hasten-macro-parsing branch from 2a13b24 to 6851f9aCompare February 10, 2020 22:41

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Feb 10, 2020

jonas-schievink mentioned this pull request Feb 11, 2020
Rollup of 11 pull requests #69056
Closed

jonas-schievink mentioned this pull request Feb 11, 2020
Rollup of 11 pull requests #69075
Closed

Dylan-DPC-zz pushed a commit to Dylan-DPC-zz/rust that referenced this pull request Feb 12, 2020
Rollup merge ofrust-lang#68848- nnethercote:hasten-macro-parsing, r…
8f813b7
…=petrochenkov Hasten macro parsing r? @eddyb

Dylan-DPC-zz mentioned this pull request Feb 12, 2020
Rollup of 10 pull requests #69085
Closed

Dylan-DPC-zz pushed a commit to Dylan-DPC-zz/rust that referenced this pull request Feb 13, 2020
Rollup merge ofrust-lang#68848- nnethercote:hasten-macro-parsing, r…
87ba8f2
…=petrochenkov Hasten macro parsing r? @eddyb

Dylan-DPC-zz mentioned this pull request Feb 13, 2020
Rollup of 9 pull requests #69118
Merged

bors added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Feb 13, 2020

nnethercote closed this Feb 13, 2020

nnethercote deleted the hasten-macro-parsing branch February 13, 2020 23:52

nnethercote mentioned this pull request Feb 14, 2020
Follow-up to #68848 #69150
Merged

JohnTitor mentioned this pull request Feb 14, 2020
Rollup of 7 pull requests #69172
Merged

Uh oh!

Uh oh!

Conversation

nnethercote commented Feb 5, 2020

Uh oh!

nnethercote commented Feb 5, 2020

Uh oh!

nnethercote commented Feb 5, 2020

Uh oh!

rust-timer commented Feb 5, 2020

Uh oh!

bors commented Feb 5, 2020

Uh oh!

Centril commented Feb 5, 2020

Uh oh!

bors commented Feb 5, 2020

Uh oh!

rust-timer commented Feb 5, 2020

Uh oh!

nnethercote commented Feb 5, 2020

Uh oh!

Uh oh!

petrochenkov commented Feb 5, 2020

Uh oh!

Uh oh!

Uh oh!

Uh oh!

petrochenkov commented Feb 8, 2020

Uh oh!

nnethercote commented Feb 9, 2020

Uh oh!

nnethercote commented Feb 10, 2020

Uh oh!

petrochenkov commented Feb 10, 2020

Uh oh!

nnethercote commented Feb 10, 2020• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

bors commented Feb 10, 2020

Uh oh!

bors commented Feb 13, 2020

Uh oh!

nnethercote commented Feb 13, 2020

Uh oh!

nnethercote commented Feb 14, 2020

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

nnethercote commented Feb 10, 2020•
edited
Loading