[WIP] Test performance of running MIR inliner on inline(always) function calls #82280

wesleywiser · 2021-02-19T01:41:27Z

r? @ghost

wesleywiser · 2021-02-19T01:42:08Z

@bors try @rust-timer queue

rust-timer · 2021-02-19T01:42:10Z

Awaiting bors try build completion.

@rustbot label: +S-waiting-on-perf

bors · 2021-02-19T01:42:21Z

⌛ Trying commit c2230893d572fed6a1809edc67098f5bf08be1c9 with merge 386d02a615f9f2266eeae4d38d62aba0d4b994de...

bors · 2021-02-19T02:42:43Z

☀️ Try build successful - checks-actions
Build commit: 386d02a615f9f2266eeae4d38d62aba0d4b994de (386d02a615f9f2266eeae4d38d62aba0d4b994de)

rust-timer · 2021-02-19T02:42:44Z

Queued 386d02a615f9f2266eeae4d38d62aba0d4b994de with parent 0148b97, future comparison URL.

rust-timer · 2021-02-19T04:22:32Z

Finished benchmarking try commit (386d02a615f9f2266eeae4d38d62aba0d4b994de): comparison url.

Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. Please note that if the perf results are neutral, you should likely undo the rollup=never given below by specifying rollup- to bors.

Importantly, though, if the results of this run are non-neutral do not roll this PR up -- it will mask other regressions or improvements in the roll up.

@bors rollup=never
@rustbot label: +S-waiting-on-review -S-waiting-on-perf

wesleywiser · 2021-02-19T15:20:48Z

A few small wins, lots of significant losses. The main culprit seems to be increased incremental cache load times.

This is probably proof that addressing #80536 will be critical before we see compile time wins from this optimization.

compiler/rustc_mir/src/transform/inline.rs

wesleywiser · 2021-03-02T01:14:54Z

@bors try @rust-timer queue

rust-timer · 2021-03-02T01:14:56Z

Awaiting bors try build completion.

@rustbot label: +S-waiting-on-perf

bors · 2021-03-02T01:15:10Z

⌛ Trying commit 71afdfc4d02f8d9b030954134b2242bde5106c7c with merge bd198f8ac3c38a1e932bacccc5a2cdfb569c1972...

bors · 2021-03-02T02:15:37Z

☀️ Try build successful - checks-actions
Build commit: bd198f8ac3c38a1e932bacccc5a2cdfb569c1972 (bd198f8ac3c38a1e932bacccc5a2cdfb569c1972)

rust-timer · 2021-03-02T02:15:39Z

Queued bd198f8ac3c38a1e932bacccc5a2cdfb569c1972 with parent 4f20caa, future comparison URL.

rust-timer · 2021-03-02T05:21:20Z

Finished benchmarking try commit (bd198f8ac3c38a1e932bacccc5a2cdfb569c1972): comparison url.

Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. Please note that if the perf results are neutral, you should likely undo the rollup=never given below by specifying rollup- to bors.

Importantly, though, if the results of this run are non-neutral do not roll this PR up -- it will mask other regressions or improvements in the roll up.

@bors rollup=never
@rustbot label: +S-waiting-on-review -S-waiting-on-perf

oli-obk · 2021-03-02T09:32:33Z

Ok, from the results it looks like we need to turn inlining off during incremental compilation, too.

compiler/rustc_mir/src/transform/mod.rs

Const eval no longer runs MIR optimizations so unless this is getting run as part of a MIR optimization like const-prop, there can be unused type parameters even if polymorphization is enabled.

rust-log-analyzer · 2021-10-01T02:15:45Z

The job x86_64-gnu-llvm-10 failed! Check out the build log: (web) (plain)

Click to see the possible cause of the failure (guessed by this bot)

   Compiling rustc_parse v0.0.0 (/checkout/compiler/rustc_parse)
   Compiling rustc_ast_lowering v0.0.0 (/checkout/compiler/rustc_ast_lowering)
   Compiling rustc_middle v0.0.0 (/checkout/compiler/rustc_middle)
   Compiling chalk-engine v0.55.0
error: internal compiler error: Encountered errors `[FulfillmentError(Obligation(predicate=Binder(TraitPredicate(<<Q as query::config::QueryConfig>::Value as std::fmt::Debug>), []), depth=0),Unimplemented)]` resolving bounds after type-checking
  = note: delayed at compiler/rustc_trait_selection/src/traits/codegen.rs:124:24


thread 'rustc' panicked at 'no errors encountered even though `delay_span_bug` issued', compiler/rustc_errors/src/lib.rs:1165:13

error: internal compiler error: unexpected panic

note: the compiler unexpectedly panicked. this is a bug.
note: the compiler unexpectedly panicked. this is a bug.

note: we would appreciate a bug report: https://github.com/rust-lang/rust/issues/new?labels=C-bug%2C+I-ICE%2C+T-compiler&template=ice.md

note: rustc 1.57.0-nightly (68d4429ee 2021-10-01) running on x86_64-unknown-linux-gnu

note: compiler flags: -Z macro-backtrace -Z tls-model=initial-exec -Z unstable-options -Z binary-dep-depinfo -Z force-unstable-if-unmarked -C opt-level=3 -C embed-bitcode=no -C debuginfo=0 -C debug-assertions=on -C overflow-checks=off -C link-args=-Wl,-rpath,$ORIGIN/../lib -C prefer-dynamic -C llvm-args=-import-instr-limit=10 --crate-type lib
note: some of the compiler flags provided by cargo are hidden

query stack during panic:
end of query stack

…_message, r=michaelwoerister Correct caller/callsite confusion in inliner message `callee_body` is the MIR `Body` for the `callsite.callee` so this message basically says `"Inline {bar span} into bar"` when it should say `"Inline bar into foo"`. Extracted out of rust-lang#82280

mati865 · 2021-11-24T18:37:15Z

AFAIK we were using old LLVM pass manager back when this wad tested. Do you think it'd be worth to retest with the new pass manager?

wesleywiser · 2021-12-04T02:29:52Z

I think it definitely would be worth retesting with NewPM but a naïve rebase ICEs the compiler during bootstrap and I haven't had time to investigate further.

bors · 2021-12-05T06:49:56Z

☔ The latest upstream changes (presumably #91475) made this pull request unmergeable. Please resolve the merge conflicts.

…ll, r=oli-obk Enable MIR inlining Continuation of rust-lang#82280 by `@wesleywiser.` rust-lang#82280 has shown nice compile time wins could be obtained by enabling MIR inlining. Most of the issues in rust-lang#81567 are now fixed, except the interaction with polymorphization which is worked around specifically. I believe we can proceed with enabling MIR inlining in the near future (preferably just after beta branching, in case we discover new issues). Steps before merging: - [x] figure out the interaction with polymorphization; - [x] figure out how miri should deal with extern types; - [x] silence the extra arithmetic overflow warnings; - [x] remove the codegen fulfilment ICE; - [x] remove the type normalization ICEs while compiling nalgebra; - [ ] tweak the inlining threshold.

rustbot added the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Feb 19, 2021

jyn514 added A-mir-opt-inlining Area: MIR inlining T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. labels Feb 19, 2021

This comment has been minimized.

# to view

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-perf Status: Waiting on a perf run to be completed. labels Feb 19, 2021

wesleywiser closed this Feb 19, 2021

oli-obk reviewed Mar 1, 2021

View reviewed changes

compiler/rustc_mir/src/transform/inline.rs Outdated Show resolved Hide resolved

wesleywiser reopened this Mar 2, 2021

wesleywiser force-pushed the enable_mir_inlining_inline_all branch from c223089 to 71afdfc Compare March 2, 2021 01:01

This comment has been minimized.

# to view

rustbot added the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Mar 2, 2021

rustbot removed the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Mar 2, 2021

oli-obk reviewed Mar 2, 2021

View reviewed changes

compiler/rustc_mir/src/transform/mod.rs Outdated Show resolved Hide resolved

This comment has been minimized.

# to view

wesleywiser added 2 commits September 30, 2021 22:03

Test enabling MIR inliner

448815d

Don't assert polymorphization has taken effect in const eval

d28bda0

Const eval no longer runs MIR optimizations so unless this is getting run as part of a MIR optimization like const-prop, there can be unused type parameters even if polymorphization is enabled.

wesleywiser force-pushed the enable_mir_inlining_inline_all branch from 7bf524b to d28bda0 Compare October 1, 2021 02:04

bors added the S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. label Dec 5, 2021

This was referenced Dec 10, 2021

Introduce MIR summary to avoid loading large bodies without inlining them #89708

Closed

Use DefPathHash instead of HirId to break inlining cycles. #85321

Merged

Enable MIR inlining #91743

Merged

wesleywiser closed this Dec 27, 2021

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[WIP] Test performance of running MIR inliner on inline(always) function calls #82280

[WIP] Test performance of running MIR inliner on inline(always) function calls #82280

wesleywiser commented Feb 19, 2021

wesleywiser commented Feb 19, 2021

rust-timer commented Feb 19, 2021

bors commented Feb 19, 2021

This comment has been minimized.

bors commented Feb 19, 2021

rust-timer commented Feb 19, 2021

rust-timer commented Feb 19, 2021

wesleywiser commented Feb 19, 2021

wesleywiser commented Mar 2, 2021

This comment has been minimized.

rust-timer commented Mar 2, 2021

bors commented Mar 2, 2021

bors commented Mar 2, 2021

rust-timer commented Mar 2, 2021

rust-timer commented Mar 2, 2021

oli-obk commented Mar 2, 2021

This comment has been minimized.

rust-log-analyzer commented Oct 1, 2021

mati865 commented Nov 24, 2021

wesleywiser commented Dec 4, 2021

bors commented Dec 5, 2021

[WIP] Test performance of running MIR inliner on inline(always) function calls #82280

[WIP] Test performance of running MIR inliner on inline(always) function calls #82280

Conversation

wesleywiser commented Feb 19, 2021

wesleywiser commented Feb 19, 2021

rust-timer commented Feb 19, 2021

bors commented Feb 19, 2021

This comment has been minimized.

bors commented Feb 19, 2021

rust-timer commented Feb 19, 2021

rust-timer commented Feb 19, 2021

wesleywiser commented Feb 19, 2021

wesleywiser commented Mar 2, 2021

This comment has been minimized.

rust-timer commented Mar 2, 2021

bors commented Mar 2, 2021

bors commented Mar 2, 2021

rust-timer commented Mar 2, 2021

rust-timer commented Mar 2, 2021

oli-obk commented Mar 2, 2021

This comment has been minimized.

rust-log-analyzer commented Oct 1, 2021

mati865 commented Nov 24, 2021

wesleywiser commented Dec 4, 2021

bors commented Dec 5, 2021