P3111R8 Atomic Reduction Operations #8025
Conversation
| \end{note} | ||
| For \tcode{store_max} and \tcode{store_min}, | ||
| the maximum and minimum computation is performed | ||
| as if by \tcode{max} and \tcode{min} algorithms\iref{alg.min.max}, respectively, |
There was a problem hiding this comment.
Might want to say "as if by the max and min algorithms", but I guess, either is technically okay.
There was a problem hiding this comment.
This phrasing is copied from fetch_min/max. I'd suggest to make a separate pull request after the motions merge.
| @@ -4740,6 +5050,35 @@ | |||
| constexpr @\placeholdernc{integral-type}@ fetch_min( @\placeholdernc{integral-type}@, | |||
| memory_order = memory_order::seq_cst) noexcept; | |||
|
|
|||
| void store_add(@\placeholdernc{integral-type}@, | |||
There was a problem hiding this comment.
Am I unfamiliar with something, or why is this using placeholder instead of exposid? According to the comments in macros.tex, placeholder is for placeholders in the middle of identifiers, like N in intN_t, and integral-type really just looks like an exposition-only type to me.
There was a problem hiding this comment.
This is a meta-placeholder, because we stamp out a full specialization for each integer type. There is an implied "foreach" around this subclause. exposid isn't like that.
| \pnum | ||
| \begin{itemize} |
There was a problem hiding this comment.
Missing \remarks here. I actually happened to be in LWG at the time we've looked at this wording and pointed out it's weird to me that we have two separate Remarks specifications in the place, but it was not seen as an issue.
@jwakely having a second Remarks here is intentional, right?
There was a problem hiding this comment.
The automatic checking won't allow duplicate elements.
jensmaurer
force-pushed
the
motions-2025-06-lwg-28
branch
from
722505e to
02fe942
Compare
jensmaurer
force-pushed
the
motions-2025-06-lwg-28
branch
from
02fe942 to
23eada3
Compare
Fixes #7965
Fixes cplusplus/papers#1902