Uh oh!
There was an error while loading. Please reload this page.
clarify float min/max behavios for NaNs and signed zeros - #149239
Merged
Conversation
rustbot
commented
Nov 23, 2025
Collaborator
rustbot
commented
Nov 23, 2025
Collaborator
RalfJungforce-pushed
the
float-intrinsics
branch
from
November 23, 2025 14:01
fd218fe to
500033dComparerustbot
commented
Nov 23, 2025
Collaborator
This comment has been minimized.
This comment has been minimized.
RalfJungforce-pushed
the
float-intrinsics
branch
from
November 23, 2025 14:27
500033d to
6ddca27CompareRalfJung
commented
Nov 24, 2025
| /// | ||
| /// If one of the arguments is NaN, then the other argument is returned. If the inputs compare equal | ||
| /// (such as for the case of `+0.0` and `-0.0`), either input may be returned non-deterministically. | ||
| /// The last point makes this not quite the same as IEEE 754-2008 minNum. |
MemberAuthor
There was a problem hiding this comment.
Actually I think I misunderstood here -- looking at the 2008 version of the standard, it says
minNum(x, y) is the canonicalized number x if x < y, y if y < x, the canonicalized number if one
operand is a number and the other a quiet NaN. Otherwise it is either x or y, canonicalized (this
means results might differ among implementations). When either x or y is a signalingNaN, then the
result is according to 6.2.
So what we do actually does exactly match IEEE 754-2008 minNum, except for the usual differences in NaN treatment.
Member
There was a problem hiding this comment.
Yeah, this was the difference between 2008 minNum and 2019 minimumNumber. I perpetually forget these details so wrote them down in a compiler-builtins issue that I refer to rust-lang/compiler-builtins#838.
MemberAuthor
There was a problem hiding this comment.
Yeah, IEEE apparently has three versions of minimum/maximum now, if you also consider historic versions (that are still wide-spread). And their names are all super easy to mix up. What a mess. I am strongly inclined to come up with more distinct names for our intrinsics, but I'm not sure it's worth it given that everything around us would still use the confusing names...
RalfJungforce-pushed
the
float-intrinsics
branch
from
November 24, 2025 07:21
6ddca27 to
8b87b25Comparetgross35
approved these changes
Nov 26, 2025
tgross35
commented
Nov 26, 2025
Member
@bors r+ rollup |
bors
commented
Nov 26, 2025
Collaborator
Zalathar added a commit
to Zalathar/rust
that referenced
this pull request
Nov 26, 2025
clarify float min/max behavios for NaNs and signed zeros The first comment is internal, it only documents the intrinsics to more clearly say what they do. This makes the currently implemented semantics more explicit, so one does not have to go look for the publicly exposed version of the operation to figure out what exactly should happen. The second commit adds a NaN test to the doc comment for `min`/`max`, which matches what we already have for `minimum`/`maximum`.
Zalathar added a commit
to Zalathar/rust
that referenced
this pull request
Nov 26, 2025
clarify float min/max behavios for NaNs and signed zeros The first comment is internal, it only documents the intrinsics to more clearly say what they do. This makes the currently implemented semantics more explicit, so one does not have to go look for the publicly exposed version of the operation to figure out what exactly should happen. The second commit adds a NaN test to the doc comment for `min`/`max`, which matches what we already have for `minimum`/`maximum`.
Zalathar added a commit
to Zalathar/rust
that referenced
this pull request
Nov 26, 2025
clarify float min/max behavios for NaNs and signed zeros The first comment is internal, it only documents the intrinsics to more clearly say what they do. This makes the currently implemented semantics more explicit, so one does not have to go look for the publicly exposed version of the operation to figure out what exactly should happen. The second commit adds a NaN test to the doc comment for `min`/`max`, which matches what we already have for `minimum`/`maximum`.
bors added a commit
that referenced
this pull request
Nov 26, 2025
Rollup of 19 pull requests Successful merges: - #148048 (Stabilize `maybe_uninit_write_slice`) - #148641 (Add a diagnostic attribute for special casing const bound errors for non-const impls) - #148765 (std: split up the `thread` module) - #149074 (Add Command::get_env_clear) - #149097 (num: Implement `uint_gather_scatter_bits` feature for unsigned integers) - #149131 (optimize `slice::Iter::next_chunk`) - #149190 (Forbid `CHECK: br` and `CHECK-NOT: br` in codegen tests (suggest `br {{.*}}` instead)) - #149239 (clarify float min/max behavios for NaNs and signed zeros) - #149243 (Fix typo and clarify bootstrap change tracker entry) - #149270 (implement `Iterator::{exactly_one, collect_array}`) - #149295 (Suggest _bytes versions of endian-converting methods) - #149301 (Motor OS: make decode_error_kind more comprehensive) - #149306 (bootstrap: Miri now handles jemalloc like everything else) - #149325 (rustdoc: add regression test for #140968) - #149332 (fix rustdoc search says “Consider searching for "null" instead.” #149324) - #149349 (Fix typo in comment.) - #149353 (Tidying up UI tests [3/N]) - #149355 (Document that `build.description` affects symbol mangling and crate IDs) - #149360 (Enable CI download for windows-gnullvm) r? `@ghost` `@rustbot` modify labels: rollup
bors added a commit
that referenced
this pull request
Nov 27, 2025
Rollup of 12 pull requests Successful merges: - #147115 (More robust stack protector testing) - #148048 (Stabilize `maybe_uninit_write_slice`) - #148641 (Add a diagnostic attribute for special casing const bound errors for non-const impls) - #149074 (Add Command::get_env_clear) - #149097 (num: Implement `uint_gather_scatter_bits` feature for unsigned integers) - #149131 (optimize `slice::Iter::next_chunk`) - #149190 (Forbid `CHECK: br` and `CHECK-NOT: br` in codegen tests (suggest `br {{.*}}` instead)) - #149239 (clarify float min/max behavios for NaNs and signed zeros) - #149243 (Fix typo and clarify bootstrap change tracker entry) - #149301 (Motor OS: make decode_error_kind more comprehensive) - #149306 (bootstrap: Miri now handles jemalloc like everything else) - #149325 (rustdoc: add regression test for #140968) r? `@ghost` `@rustbot` modify labels: rollup
Uh oh!
There was an error while loading. Please reload this page.
rust-timer added a commit
that referenced
this pull request
Nov 27, 2025
Rollup merge of #149239 - RalfJung:float-intrinsics, r=tgross35 clarify float min/max behavios for NaNs and signed zeros The first comment is internal, it only documents the intrinsics to more clearly say what they do. This makes the currently implemented semantics more explicit, so one does not have to go look for the publicly exposed version of the operation to figure out what exactly should happen. The second commit adds a NaN test to the doc comment for `min`/`max`, which matches what we already have for `minimum`/`maximum`.
github-actionsBot
pushed a commit
to model-checking/verify-rust-std
that referenced
this pull request
Jan 20, 2026
clarify float min/max behavios for NaNs and signed zeros The first comment is internal, it only documents the intrinsics to more clearly say what they do. This makes the currently implemented semantics more explicit, so one does not have to go look for the publicly exposed version of the operation to figure out what exactly should happen. The second commit adds a NaN test to the doc comment for `min`/`max`, which matches what we already have for `minimum`/`maximum`.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The first comment is internal, it only documents the intrinsics to more clearly say what they do.
This makes the currently implemented semantics more explicit, so one does not have to go look for the publicly exposed version of the operation to figure out what exactly should happen.
The second commit adds a NaN test to the doc comment for
min/max, which matches what we already have forminimum/maximum.