patterns: reject raw pointers that are not just integers #116930

RalfJung · 2023-10-19T06:37:41Z

Matching against 0 as *const i32 is fine, matching against &42 as *const i32 is not.

This extends the existing check against function pointers and wide pointers: we now uniformly reject all these pointer types during valtree construction, and then later lint because of that. See here for some more explanation and context.

Also fixes #116929.

Cc @oli-obk @lcnr

rustbot · 2023-10-19T06:37:49Z

r? @davidtwco

(rustbot has picked a reviewer for you, use r? to override)

RalfJung · 2023-10-19T06:38:08Z

compiler/rustc_lint_defs/src/builtin.rs

    pub POINTER_STRUCTURAL_MATCH,
-    Allow,


I didn't realize this lint is still allow-by-default oO. Seems high time we make it warn-by-default?

RalfJung · 2023-10-19T06:38:50Z

The 2nd commit makes this forbid-by-default so we can crater it.
@bors try

bors · 2023-10-19T06:39:59Z

⌛ Trying commit 0787e62 with merge ebb1189...

patterns: reject raw pointers that are not just integers Matching against `0 as *const i32` is fine, matching against `&42 as *const i32` is not. Cc `@oli-obk` `@lcnr`

RalfJung · 2023-10-19T06:41:19Z

Due to #116929 this lint still has a big gap, it'll only find bad raw pointers "at the root".

I wonder if with this change to valtree construction, we can make it so that some lint always fires when a const is used as a pattern and valtree construction fails? I do think those cases should all become hard errors eventually...

bors · 2023-10-19T08:01:15Z

☀️ Try build successful - checks-actions
Build commit: ebb1189 (ebb118907fa893a392c699ddf120848fa0a8d6c2)

RalfJung · 2023-10-19T08:02:09Z

Let's see how much fallout we're getting from this weak form of the lint.
@craterbot check

craterbot · 2023-10-19T08:02:21Z

👌 Experiment pr-116930 created and queued.
🤖 Automatically detected try build ebb1189
🔍 You can check out the queue and this experiment's details.

ℹ️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

rustbot · 2023-10-19T08:03:27Z

Some changes might have occurred in exhaustiveness checking

cc @Nadrieril

compiler/rustc_mir_build/src/thir/pattern/const_to_pat.rs

RalfJung · 2023-10-20T08:10:53Z

Lol, the doctest for StructuralEq uses a function pointer as a bad example.^^

RalfJung · 2023-10-20T08:31:29Z

library/core/src/marker.rs

@@ -247,6 +247,7 @@ marker_impls! {
 ///
 /// const CFN: Wrap<fn(&())> = Wrap(higher_order);
 ///
+/// #[allow(pointer_structural_match)]


FWIW these entire trait docs are outdated; that's tracked in #115881.

craterbot · 2023-10-20T11:55:53Z

🚧 Experiment pr-116930 is now running

ℹ️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

craterbot · 2023-10-22T12:29:57Z

🎉 Experiment pr-116930 is completed!
📊 876 regressed and 3 fixed (379825 total)
📰 Open the full report.

⚠️ If you notice any spurious failure please add them to the blacklist!
ℹ️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

RalfJung · 2023-10-22T15:55:33Z

@craterbot run mode=check-only crates=https://crater-reports.s3.amazonaws.com/pr-116930/retry-regressed-list.txt

RalfJung · 2023-10-30T07:50:17Z

Those seem to be 3 legit regressions:

this matched on a fn ptr (so already triggered the lint before this PR, but the lint was allow-by-default)
same here
and same here

In particular, none of them is matching on a raw pointer that was not derived from an integer. So IMO we should land this PR to make sure people don't start doing that.

RalfJung · 2023-10-30T07:54:13Z

@rust-lang/lang this PR does three things:

it fixes our old lint that detects match on function pointers and wide pointers to also catch cases where those pointers are hidden inside other types
it makes that lint warn-by-default (but doesn't yet make it show up in dependencies)
it makes the lint also trigger when encountering a raw pointer that was not constructed via int as *const/mut T, but e.g. via &something as *const T -- those allocations don't have guarantees on their ptr equality just like function pointers and vtables, so we should reject them for the same reason.

We haven't reached a proper conclusion in the recent "match on const" meeting, but the general direction seems to have been towards "it should still work like a pattern" (and not like sugar for ==), and this PR is a crucial step towards ensuring that they do indeed behave like a pattern, with structural properties and all that. Even if the goal is just to future-proof against both options (as has been the strategy so far), we want this warning.

Crater found only 3 cases where the warning triggers, in the entire ecosystem. All of them would already warn on stable if they enabled the lint; the new cases covered by this PR were not detected at all by crater.

davidtwco

Implementation changes look good to me, r=me after decisions from t-lang.

davidtwco · 2023-11-04T01:45:03Z

compiler/rustc_mir_build/src/thir/pattern/const_to_pat.rs

-                        );
-                    }
-                    _ => {}
+                if !have_valtree {


nit: could be part of the condition on the previous line

nikomatsakis · 2023-11-08T16:47:41Z

I'm in favor of this. (I also have thoughts on the semantics of matching on constants -- tl;dr I think I'm ready to accept "match is not an extensible mechanism" and hence will not act the same as == so we can stop talking about it -- but those are best discussed elsewhere, I assume)

WaffleLapkin · 2023-11-08T16:47:43Z

T-lang meeting consensus: let's do this, no FCP needed.

RalfJung · 2023-11-08T19:39:03Z

@bors r=davidtwco

bors · 2023-11-08T19:39:06Z

📌 Commit 3058865 has been approved by davidtwco

It is now in the queue for this repository.

bors · 2023-11-08T20:42:35Z

⌛ Testing commit 3058865 with merge fdaaaf9...

bors · 2023-11-08T22:39:54Z

☀️ Test successful - checks-actions
Approved by: davidtwco
Pushing fdaaaf9 to master...

rust-timer · 2023-11-09T00:20:34Z

Finished benchmarking commit (fdaaaf9): comparison URL.

Overall result: no relevant changes - no action needed

@rustbot label: -perf-regression

Instruction count

This benchmark run did not return any relevant results for this metric.

Max RSS (memory usage)

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	0.6%	[0.4%, 1.4%]	5
Regressions ❌ (secondary)	3.6%	[3.6%, 3.6%]	1
Improvements ✅ (primary)	-0.7%	[-1.1%, -0.5%]	7
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-0.1%	[-1.1%, 1.4%]	12

Cycles

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	0.4%	[0.4%, 0.5%]	2
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.4%	[-0.4%, -0.4%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.1%	[-0.4%, 0.5%]	3

Binary size

This benchmark run did not return any relevant results for this metric.

Bootstrap: 662.394s -> 663.159s (0.12%)
Artifact size: 308.74 MiB -> 308.75 MiB (0.00%)

patterns: reject raw pointers that are not just integers Matching against `0 as *const i32` is fine, matching against `&42 as *const i32` is not. This extends the existing check against function pointers and wide pointers: we now uniformly reject all these pointer types during valtree construction, and then later lint because of that. See [here](rust-lang/rust#116930 (comment)) for some more explanation and context. Also fixes rust-lang/rust#116929. Cc `@oli-obk` `@lcnr`

rustbot assigned davidtwco Oct 19, 2023

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. labels Oct 19, 2023

RalfJung commented Oct 19, 2023

View reviewed changes

This comment has been minimized.

Sign in to view

craterbot added S-waiting-on-crater Status: Waiting on a crater run to be completed. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Oct 19, 2023

RalfJung force-pushed the raw-ptr-match branch from 0787e62 to a91227c Compare October 19, 2023 08:03

RalfJung commented Oct 19, 2023

View reviewed changes

compiler/rustc_mir_build/src/thir/pattern/const_to_pat.rs Outdated Show resolved Hide resolved

RalfJung force-pushed the raw-ptr-match branch from a91227c to 41ab7f7 Compare October 20, 2023 06:35

This comment has been minimized.

Sign in to view

RalfJung force-pushed the raw-ptr-match branch from 2e70736 to 35b0255 Compare October 20, 2023 08:26

RalfJung commented Oct 20, 2023

View reviewed changes

craterbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-crater Status: Waiting on a crater run to be completed. labels Oct 22, 2023

This comment was marked as outdated.

Sign in to view

craterbot removed the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Oct 22, 2023

RalfJung force-pushed the raw-ptr-match branch from 5ab33a9 to 70a8e15 Compare October 30, 2023 07:49

RalfJung added the I-lang-nominated The issue / PR has been nominated for discussion during a lang team meeting. label Oct 30, 2023

RalfJung added the S-waiting-on-team Status: Awaiting decision from the relevant subteam (see the T-<team> label). label Oct 30, 2023

remove some dead code

03b24f2

davidtwco approved these changes Nov 4, 2023

View reviewed changes

avoid unnecessary nested conditionals

3058865

bors added the merged-by-bors This PR was explicitly merged by bors. label Nov 8, 2023

bors merged commit fdaaaf9 into rust-lang:master Nov 8, 2023
12 checks passed

rustbot added this to the 1.75.0 milestone Nov 8, 2023

This was referenced Nov 8, 2023

make matching on NaN a hard error, and remove the rest of illegal_floating_point_literal_pattern #116284

Merged

mark ty::Const::Error when meet unsupport ty for const generic params #117176

Merged

matthiaskrgr mentioned this pull request Nov 11, 2023

ICE: Got a scalar pair where a scalar was expected #117806

Closed

RalfJung deleted the raw-ptr-match branch November 11, 2023 09:58

matthiaskrgr mentioned this pull request Dec 24, 2023

ICE: shouldn't have created a ValTree for *const .. #119270

Closed

RalfJung mentioned this pull request Jan 28, 2024

update indirect_structural_match and pointer_structural_match lints to match RFC and to show up for dependencies #120423

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

patterns: reject raw pointers that are not just integers #116930

patterns: reject raw pointers that are not just integers #116930

RalfJung commented Oct 19, 2023 •

edited

rustbot commented Oct 19, 2023

RalfJung Oct 19, 2023

RalfJung commented Oct 19, 2023

bors commented Oct 19, 2023

RalfJung commented Oct 19, 2023

This comment has been minimized.

bors commented Oct 19, 2023

RalfJung commented Oct 19, 2023

craterbot commented Oct 19, 2023

rustbot commented Oct 19, 2023

This comment has been minimized.

RalfJung commented Oct 20, 2023

RalfJung Oct 20, 2023

craterbot commented Oct 20, 2023

craterbot commented Oct 22, 2023

RalfJung commented Oct 22, 2023

This comment was marked as outdated.

RalfJung commented Oct 30, 2023

RalfJung commented Oct 30, 2023 •

edited

davidtwco left a comment

davidtwco Nov 4, 2023

nikomatsakis commented Nov 8, 2023 •

edited

WaffleLapkin commented Nov 8, 2023

RalfJung commented Nov 8, 2023

bors commented Nov 8, 2023

bors commented Nov 8, 2023

bors commented Nov 8, 2023

rust-timer commented Nov 9, 2023

patterns: reject raw pointers that are not just integers #116930

patterns: reject raw pointers that are not just integers #116930

Conversation

RalfJung commented Oct 19, 2023 • edited

rustbot commented Oct 19, 2023

RalfJung Oct 19, 2023

Choose a reason for hiding this comment

RalfJung commented Oct 19, 2023

bors commented Oct 19, 2023

RalfJung commented Oct 19, 2023

This comment has been minimized.

bors commented Oct 19, 2023

RalfJung commented Oct 19, 2023

craterbot commented Oct 19, 2023

rustbot commented Oct 19, 2023

This comment has been minimized.

RalfJung commented Oct 20, 2023

RalfJung Oct 20, 2023

Choose a reason for hiding this comment

craterbot commented Oct 20, 2023

craterbot commented Oct 22, 2023

RalfJung commented Oct 22, 2023

This comment was marked as outdated.

RalfJung commented Oct 30, 2023

RalfJung commented Oct 30, 2023 • edited

davidtwco left a comment

Choose a reason for hiding this comment

davidtwco Nov 4, 2023

Choose a reason for hiding this comment

nikomatsakis commented Nov 8, 2023 • edited

WaffleLapkin commented Nov 8, 2023

RalfJung commented Nov 8, 2023

bors commented Nov 8, 2023

bors commented Nov 8, 2023

bors commented Nov 8, 2023

rust-timer commented Nov 9, 2023

Overall result: no relevant changes - no action needed

Instruction count

Max RSS (memory usage)

Cycles

Binary size

RalfJung commented Oct 19, 2023 •

edited

RalfJung commented Oct 30, 2023 •

edited

nikomatsakis commented Nov 8, 2023 •

edited