Introduce `Rvalue::ShallowInitBox` #89030

nbdd0121 · 2021-09-16T22:14:56Z

Polished version of #88700.

Implements MCP rust-lang/compiler-team#460, and should allow #43596 to go forward.

In short, creating an empty box is split from a nullary-op NullOp::Box into two steps, first a call to exchange_malloc, then a Rvalue::ShallowInitBox which transmutes *mut u8 to a shallow-initialized Box<T>. This allows the exchange_malloc call to unwind. Details can be found in the MCP.

NullOp::Box is not yet removed, purely to make reverting easier in case anything goes wrong as the result of this PR. If revert is needed a reversion of "Use Rvalue::ShallowInitBox for box expression" commit followed by a test bless should be sufficient.

Experiments in #88700 showed a very slight compile-time perf regression due to (supposedly) slightly more time spent in LLVM. We could omit unwind edge generation (in non-oom=panic case) in box expression MIR construction to restore perf; but I don't think it's necessary since runtime perf isn't affected and perf difference is rather small.

rust-highfive · 2021-09-16T22:14:58Z

Some changes occured to rustc_codegen_cranelift

cc @bjorn3

Some changes occurred in src/tools/clippy.

cc @rust-lang/clippy

Some changes occured to the CTFE / Miri engine

cc @rust-lang/miri

rust-highfive · 2021-09-16T22:15:00Z

r? @michaelwoerister

(rust-highfive has picked a reviewer for you, use r? to override)

nbdd0121 · 2021-09-17T03:01:06Z

The error is due to neglected no_global_oom_handling cfg-gating. I filed #89036 to fix it but that means this PR will be blocked until #89036 is merged :(

Fix missing `no_global_oom_handling` cfg-gating Cfg-gate these trait impls that are neglected. These functions compile now because they use `box` syntax which depends on `exchange_malloc` during codegen only; as a result they compiles with cfg `no_global_oom_handling` but shouldn't. Discovered in rust-lang#89030 because that PR makes `box` syntax depend on `exchange_malloc` lang item during MIR construction.

michaelwoerister · 2021-09-22T08:35:32Z

r? @jonas-schievink, who seconded the MCP. I don't know enough about MIR to review this PR.

Fix missing `no_global_oom_handling` cfg-gating Cfg-gate these trait impls that are neglected. These functions compile now because they use `box` syntax which depends on `exchange_malloc` during codegen only; as a result they compiles with cfg `no_global_oom_handling` but shouldn't. Discovered in rust-lang#89030 because that PR makes `box` syntax depend on `exchange_malloc` lang item during MIR construction.

jonas-schievink

Changes look generally good to me! A few notes:

This comment can now be removed
Is NullOp::Box still used anywhere or can/should it be removed?

compiler/rustc_middle/src/mir/mod.rs

jonas-schievink · 2021-09-24T22:19:43Z

compiler/rustc_borrowck/src/type_check/mod.rs

+                    substs: tcx.mk_substs_trait(ty, &[]),
+                };
+
+                self.prove_trait_ref(


I guess this is copied from NullaryOp (can the branches be merged?), but I don't get why this is both doing ensure_place_sized and proving that ty: Sized

Indeed. With a bit of archaeology I found that this check was introduced in #53575, and with #87935 the Sized constraint is verified in rustc_typeck so the error path will never be hit. The unsized path definitely could be removed.

jonas-schievink · 2021-09-24T22:26:46Z

compiler/rustc_middle/src/mir/mod.rs

@@ -2450,6 +2456,10 @@ impl<'tcx> Debug for Rvalue<'tcx> {
                    }),
                }
            }
+
+            ShallowInitBox(ref place, ref ty) => {
+                write!(fmt, "ShallowInitBox({:?}, {:?})", place, ty)


This looks a bit odd in the MIR dumps (with a value and a type being passed the same way), but it does match how NullaryOp is printed, so I guess this can be improved later on.

compiler/rustc_middle/src/mir/tcx.rs

nbdd0121 · 2021-09-24T23:57:43Z

Is NullOp::Box still used anywhere or can/should it be removed?

No, it won't be used anymore. My thought is that if we keep it, it'll be easier to revert if anything goes wrong. Reverting "Use Rvalue::ShallowInitBox for box expression" would be enough to bring back the old code path.

If you feel that it should be removed in this PR, I can do that as well.

jonas-schievink · 2021-09-25T01:15:58Z

No, it won't be used anymore. My thought is that if we keep it, it'll be easier to revert if anything goes wrong. Reverting "Use Rvalue::ShallowInitBox for box expression" would be enough to bring back the old code path.

Seems reasonable!

@bors r+ rollup=never

(for the perf triage team: there's a slight perf impact expected, which was previously measured in #88700 (comment))

bors · 2021-09-25T01:15:59Z

📌 Commit ab64580 has been approved by jonas-schievink

nbdd0121 · 2021-09-25T04:54:22Z

GitHub Actions has encountered an internal error when running your job.

Looks like a bogus failure.

jonas-schievink · 2021-09-25T10:20:48Z

@bors retry

bors · 2021-09-25T11:01:16Z

⌛ Testing commit ab64580 with merge e9f29a8...

bors · 2021-09-25T13:43:47Z

☀️ Test successful - checks-actions
Approved by: jonas-schievink
Pushing e9f29a8 to master...

rust-timer · 2021-09-25T16:51:06Z

Finished benchmarking commit (e9f29a8): comparison url.

Summary: This change led to large relevant regressions 😿 in compiler performance.

Large regression in instruction counts (up to 1.9% on incr-patched: println builds of syn)

If you disagree with this performance assessment, please file an issue in rust-lang/rustc-perf.

Next Steps: If you can justify the regressions found in this perf run, please indicate this with @rustbot label: +perf-regression-triaged along with sufficient written justification. If you cannot justify the regressions please open an issue or create a new PR that fixes the regressions, add a comment linking to the newly created issue or PR, and then add the perf-regression-triaged label to this PR.

@rustbot label: +perf-regression

rylev · 2021-09-28T13:01:40Z

@nbdd0121 @jonas-schievink I see there was a bit of a justification for the perf regression given when opening this PR:

Experiments in #88700 showed a very slight compile-time perf regression due to (supposedly) slightly more time spent in LLVM. We could omit unwind edge generation (in non-oom=panic case) in box expression MIR construction to restore perf; but I don't think it's necessary since runtime perf isn't affected and perf difference is rather small.

The perf regressions aren't huge but I wouldn't classify themselves as small either. We have some non-noisy, real world crate test cases that are showing regressions that are 9 times larger than the calculated significance threshold. This isn't nearly big enough to merit a revert, but if there's a clear path forward for trying to gain the performance back it's probably worth trying to do that sooner rather than later (considering this is for a feature flagged feature and doesn't really benefit any users at this point). Thoughts?

nbdd0121 · 2021-09-28T16:40:12Z

The wall-clock time of some of these "regressions" are improved though; instruction count isn't the full picture.

Introduce `Rvalue::ShallowInitBox` Polished version of rust-lang#88700. Implements MCP rust-lang/compiler-team#460, and should allow rust-lang#43596 to go forward. In short, creating an empty box is split from a nullary-op `NullOp::Box` into two steps, first a call to `exchange_malloc`, then a `Rvalue::ShallowInitBox` which transmutes `*mut u8` to a shallow-initialized `Box<T>`. This allows the `exchange_malloc` call to unwind. Details can be found in the MCP. `NullOp::Box` is not yet removed, purely to make reverting easier in case anything goes wrong as the result of this PR. If revert is needed a reversion of "Use Rvalue::ShallowInitBox for box expression" commit followed by a test bless should be sufficient. Experiments in rust-lang#88700 showed a very slight compile-time perf regression due to (supposedly) slightly more time spent in LLVM. We could omit unwind edge generation (in non-`oom=panic` case) in box expression MIR construction to restore perf; but I don't think it's necessary since runtime perf isn't affected and perf difference is rather small.

rylev · 2021-09-28T18:13:17Z

@nbdd0121 wall-clock time is an extremely noisy metric which we don't usually look at unless the gains are consistent and large which is not the case here. Of course, instruction count is only part of the picture, but it's generally the best we have for determining runtime performance changes, so I still think discussing the targeted fixed is warranted.

Remove `NullOp::Box` Follow up of rust-lang#89030 and MCP rust-lang/compiler-team#460. ~1 month later nothing seems to be broken, apart from a small regression that rust-lang#89332 (1aac85bb716c09304b313d69d30d74fe7e8e1a8e) shows could be regained by remvoing the diverging path, so it shall be safe to continue and remove `NullOp::Box` completely. r? `@jonas-schievink` `@rustbot` label T-compiler

rust-highfive assigned michaelwoerister Sep 16, 2021

rust-highfive added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Sep 16, 2021

nbdd0121 mentioned this pull request Sep 16, 2021

[experimental] Rvalue::InitBox prototype #88700

Closed

This comment has been minimized.

Sign in to view

nbdd0121 mentioned this pull request Sep 17, 2021

Fix missing no_global_oom_handling cfg-gating #89036

Merged

est31 mentioned this pull request Sep 19, 2021

Remove box syntax from compiler and tools #87781

Merged

rust-highfive assigned jonas-schievink and unassigned michaelwoerister Sep 22, 2021

nbdd0121 force-pushed the box2 branch from 434c277 to 0a5093c Compare September 22, 2021 22:42

This comment has been minimized.

Sign in to view

nbdd0121 force-pushed the box2 branch from 0a5093c to 18502e9 Compare September 22, 2021 23:26

jonas-schievink reviewed Sep 24, 2021

View reviewed changes

nbdd0121 force-pushed the box2 branch from 18502e9 to 1c32b07 Compare September 24, 2021 23:53

This comment has been minimized.

Sign in to view

nbdd0121 added 4 commits September 25, 2021 01:08

Introduce Rvalue::ShallowInitBox

c38da2e

Use Rvalue::ShallowInitBox for box expression

511333f

Report heap allocation instead of non-const fn for exchange_malloc call

19eaee2

Bless tests

ab64580

nbdd0121 force-pushed the box2 branch from 1c32b07 to ab64580 Compare September 25, 2021 00:08

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-review Status: Awaiting review from the assignee but also interested parties. labels Sep 25, 2021

bors added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Sep 25, 2021

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-review Status: Awaiting review from the assignee but also interested parties. labels Sep 25, 2021

bors added the merged-by-bors This PR was explicitly merged by bors. label Sep 25, 2021

bors merged commit e9f29a8 into rust-lang:master Sep 25, 2021

rustbot added this to the 1.57.0 milestone Sep 25, 2021

bors mentioned this pull request Sep 25, 2021

Remove most box syntax uses from the testsuite except for src/test/ui/issues #88316

Merged

rustbot added the perf-regression Performance regression. label Sep 25, 2021

nbdd0121 deleted the box2 branch September 26, 2021 00:27

nbdd0121 mentioned this pull request Sep 28, 2021

[experiment] remove diverge_from in box expr building #89332

Closed

nbdd0121 mentioned this pull request Oct 6, 2021

Implement -Z oom=panic #88098

Merged

nbdd0121 mentioned this pull request Oct 20, 2021

Remove NullOp::Box #90102

Merged

bjorn3 mentioned this pull request Dec 2, 2021

Rust 1.57 Rust-for-Linux/linux#586

Merged

est31 mentioned this pull request Apr 23, 2023

Implement builtin # syntax and use it for offset_of!(...) #110694

Merged

gurry mentioned this pull request Dec 3, 2023

ice: ProjectionKinds Deref and Field were mismatched #118144

Closed

RalfJung mentioned this pull request Aug 18, 2024

box syntax creates Box<T> where T is not yet initialized #97270

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Introduce `Rvalue::ShallowInitBox` #89030

Introduce `Rvalue::ShallowInitBox` #89030

nbdd0121 commented Sep 16, 2021 •

edited

Loading

rust-highfive commented Sep 16, 2021

rust-highfive commented Sep 16, 2021

This comment has been minimized.

nbdd0121 commented Sep 17, 2021

michaelwoerister commented Sep 22, 2021

This comment has been minimized.

jonas-schievink left a comment

jonas-schievink Sep 24, 2021

nbdd0121 Sep 24, 2021

jonas-schievink Sep 24, 2021

nbdd0121 commented Sep 24, 2021

This comment has been minimized.

jonas-schievink commented Sep 25, 2021

bors commented Sep 25, 2021

nbdd0121 commented Sep 25, 2021

jonas-schievink commented Sep 25, 2021

bors commented Sep 25, 2021

bors commented Sep 25, 2021

rust-timer commented Sep 25, 2021

rylev commented Sep 28, 2021

nbdd0121 commented Sep 28, 2021

rylev commented Sep 28, 2021

Introduce Rvalue::ShallowInitBox #89030

Introduce Rvalue::ShallowInitBox #89030

Conversation

nbdd0121 commented Sep 16, 2021 • edited Loading

rust-highfive commented Sep 16, 2021

rust-highfive commented Sep 16, 2021

This comment has been minimized.

nbdd0121 commented Sep 17, 2021

michaelwoerister commented Sep 22, 2021

This comment has been minimized.

jonas-schievink left a comment

Choose a reason for hiding this comment

jonas-schievink Sep 24, 2021

Choose a reason for hiding this comment

nbdd0121 Sep 24, 2021

Choose a reason for hiding this comment

jonas-schievink Sep 24, 2021

Choose a reason for hiding this comment

nbdd0121 commented Sep 24, 2021

This comment has been minimized.

jonas-schievink commented Sep 25, 2021

bors commented Sep 25, 2021

nbdd0121 commented Sep 25, 2021

jonas-schievink commented Sep 25, 2021

bors commented Sep 25, 2021

bors commented Sep 25, 2021

rust-timer commented Sep 25, 2021

rylev commented Sep 28, 2021

nbdd0121 commented Sep 28, 2021

rylev commented Sep 28, 2021

Introduce `Rvalue::ShallowInitBox` #89030

Introduce `Rvalue::ShallowInitBox` #89030

nbdd0121 commented Sep 16, 2021 •

edited

Loading