miri: make NaN generation non-deterministic by RalfJung · Pull Request #116551 · rust-lang/rust

RalfJung · 2023-10-09T05:49:10Z

This implements the LLVM semantics for NaN generation. I will soon submit an RFC to make this also officially the Rust semantics, but it has been our de-facto semantics for a long time so there's no reason Miri has to wait for that RFC. This PR just better aligns Miri with codegen.

This PR does that just for the operations that have MIR primitives; a future PR will adjust the intrinsics.

rustbot · 2023-10-09T05:49:17Z

r? @davidtwco

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

rustbot · 2023-10-09T05:49:21Z

Some changes occurred to the CTFE / Miri engine

cc @rust-lang/miri

The Miri subtree was changed

cc @rust-lang/miri

RalfJung · 2023-10-09T05:50:28Z

+        _inputs: &[F1],
+    ) -> F2 {
+        // By default we always return the preferred NaN.
+        F2::NAN


This is a change in behavior for some cases: apfloat does sometimes propagate the payload of an input, e.g. for casts; with this PR const-eval is changed to never propagate and always generate a positive preferred NaN.

rustbot · 2023-10-09T05:51:45Z

The Miri subtree was changed

cc @rust-lang/miri

Some changes occurred to the CTFE / Miri engine

cc @rust-lang/miri

RalfJung · 2023-10-09T05:51:50Z

r? @oli-obk

oli-obk · 2023-10-10T09:58:16Z

I didn't check the tests, just the logic, but that lgtm

@bors r+

bors · 2023-10-10T09:58:19Z

📌 Commit 08deb0d has been approved by oli-obk

It is now in the queue for this repository.

bors · 2023-10-10T11:42:29Z

⌛ Testing commit 08deb0d with merge 061c330...

bors · 2023-10-10T13:28:02Z

☀️ Test successful - checks-actions
Approved by: oli-obk
Pushing 061c330 to master...

rust-timer · 2023-10-10T15:17:27Z

Finished benchmarking commit (061c330): 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
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-4.6%	[-4.6%, -4.6%]	1
All ❌✅ (primary)	-	-	0

Cycles

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

Binary size

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

Bootstrap: 627.375s -> 625.33s (-0.33%)
Artifact size: 270.86 MiB -> 270.81 MiB (-0.02%)

RalfJung added 2 commits October 8, 2023 20:35

miri: make NaN generation non-deterministic

6796c57

ensure unary minus propagates NaN payloads exactly

615d738

rustbot assigned davidtwco Oct 9, 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 9, 2023

RalfJung commented Oct 9, 2023

View reviewed changes

RalfJung force-pushed the nondet-nan branch from a61dc2a to 4a06b79 Compare October 9, 2023 05:51

rustbot assigned oli-obk and unassigned davidtwco Oct 9, 2023

float-to-float casts also have non-deterministic NaN results

08deb0d

RalfJung force-pushed the nondet-nan branch from 4a06b79 to 08deb0d Compare October 9, 2023 06:31

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 Oct 10, 2023

bors added the merged-by-bors This PR was explicitly merged by bors. label Oct 10, 2023

bors merged commit 061c330 into rust-lang:master Oct 10, 2023

rustbot added this to the 1.75.0 milestone Oct 10, 2023

RalfJung deleted the nondet-nan branch October 11, 2023 17:09

Dajamante mentioned this pull request Oct 12, 2023

Pull upstream master 2023 10 12 ferrocene/ferrocene#48

Merged

RalfJung mentioned this pull request Oct 25, 2023

Value of f32.is_sign_positive() differs from native code rust-lang/miri#3139

Closed

dtolnay mentioned this pull request Oct 26, 2023

Fix handling of NaN sign throughout parsing and serialization toml-rs/toml#637

Merged

This was referenced Oct 26, 2023

Lint for `-f64::NAN rust-lang/rust-clippy#11717

Open

Lint for floating64 as f32 losing the sign of NAN rust-lang/rust-clippy#11718

Open

Lint for sign-ambiguous NAN constants rust-lang/rust-clippy#11720

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

miri: make NaN generation non-deterministic#116551

miri: make NaN generation non-deterministic#116551
bors merged 3 commits into
rust-lang:masterfrom
RalfJung:nondet-nan

RalfJung commented Oct 9, 2023

Uh oh!

rustbot commented Oct 9, 2023

Uh oh!

rustbot commented Oct 9, 2023

Uh oh!

RalfJung Oct 9, 2023

Uh oh!

rustbot commented Oct 9, 2023

Uh oh!

RalfJung commented Oct 9, 2023

Uh oh!

oli-obk commented Oct 10, 2023

Uh oh!

bors commented Oct 10, 2023

Uh oh!

bors commented Oct 10, 2023

Uh oh!

bors commented Oct 10, 2023

Uh oh!

rust-timer commented Oct 10, 2023

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

Uh oh!

Conversation

RalfJung commented Oct 9, 2023

Uh oh!

rustbot commented Oct 9, 2023

Uh oh!

rustbot commented Oct 9, 2023

Uh oh!

RalfJung Oct 9, 2023

Choose a reason for hiding this comment

Uh oh!

rustbot commented Oct 9, 2023

Uh oh!

RalfJung commented Oct 9, 2023

Uh oh!

oli-obk commented Oct 10, 2023

Uh oh!

bors commented Oct 10, 2023

Uh oh!

bors commented Oct 10, 2023

Uh oh!

bors commented Oct 10, 2023

Uh oh!

rust-timer commented Oct 10, 2023

Overall result: no relevant changes - no action needed

Instruction count

Max RSS (memory usage)

Cycles

Binary size

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants