[Cfp-interest 2625] Preliminary WG14 IEEE 754-C binding meeting minutes - 2023/01/18
Jim Thomas
jaswthomas at sbcglobal.net
Fri Jan 20 21:02:37 PST 2023
WG14 Nnnnn
Meeting notes
C Floating Point Study Group Teleconference
2023-01-18 (continuation of 2023-01-04,05 meeting)
8 AM PST / 10 PM EST / 4 PM UTC
Attendees: Rajan, Jim, Fred, David H, Damian
New agenda items
(https://wiki.edg.com/pub/CFP/WebHome/CFP_continuation_meeting-20230118.pdf):
None.
2023-01-04,05 meeting minutes:
· [Cfp-interest 2594] WG14 IEEE 754-C binding meeting minutes - 2023/01/04-2023/01/05 Update <http://mailman.oakapple.net/pipermail/cfp-interest/2023-January/002608.html> <> Rajan Bhakta
Next Meeting:
February 14th, 2023, 4PM
Please notify the group if this time slot does not work.
Note: The next WG 14 meeting is January 23-27, 2023
Last meeting carry over action items:
David H: Get an example for the scaled reduction functions (perhaps by asking Jason or Jim or looking into the IEEE references).
https://754r.ucbtest.org/background/traps-and-wraps.txt
Deferred
David H: Get an example for the augmented arithmetic functions (perhaps by asking Jason or Jim D or looking into the IEEE references).
Deferred
Last meeting action items:
All: Review CD comments related to CFP.
Done
New action items:
Jim: Update N3081, highlight changes, email to CFP for fast review, and submit to WG 14.
C++ liaison:
Rounding direction specific functions – Did not get to this item.
C23 integration:
CD ballot done (https://wiki.edg.com/pub/CFP/WebHome/ISO-IEC_JTC_1-SC_22_N5777_Text_for_CD_9899.pdf).
· [Cfp-interest 2558] C23 CD <http://mailman.oakapple.net/pipermail/cfp-interest/2022-October/002572.html> <> Jim Thomas
Continued review:
Following https://wiki.edg.com/pub/CFP/WebHome/CFP_continuation_meeting-20230118.pdf ...
· [Cfp-interest 2595] draft CFP response for NB comments and N3071 <http://mailman.oakapple.net/pipermail/cfp-interest/2023-January/002609.html> <> Jim Thomas
Reviewed statement of CFP response to GB-286, GB-287. Agreed.
· [Cfp-interest 2597] Re: draft CFP response for NB comments and N3071 <http://mailman.oakapple.net/pipermail/cfp-interest/2023-January/002611.html> <> Joseph Myers
The CD adds to C18 the requirement for errno setting when overflow occurs in a nondefault rounding mode. Requires diagnosing overflow when a finite result is delivered.
Accommodation for constexpr wasn’t stated correctly.
· [Cfp-interest 2601] Re: draft CFP response for NB comments and N3071 <http://mailman.oakapple.net/pipermail/cfp-interest/2023-January/002615.html> <> Jim Thomas
Reviewed Fred’s comments in CFP 2600 about constexpr examples, and Jim’s fixes. Agreed.
· [Cfp-interest 2603] Re: draft CFP response for NB comments and N3071 <http://mailman.oakapple.net/pipermail/cfp-interest/2023-January/002617.html> <> Jim Thomas
N3081 was submitted to WG14. Reviewed …
GB-164
Prefer no change, but Ok with: Requiring errno setting only in default rounding mode, but
requiring exception regardless of rounding mode. Either no change or both proposed changes.
GB-279
Reviewed N3081 statement of CFP response. Agreed.
GB-287
Reviewed N3081 statement of CFP response. Agreed.
· [Cfp-interest 2607] Re: [SC22WG14.22883] N3071 exactly representable <http://mailman.oakapple.net/pipermail/cfp-interest/2023-January/002621.html> <> Jim Thomas
"compatible" versus "same" -- agreed with N3081 wording.
Change complex SNAN example to be “ok”. Agreed.
Also, add constexpr to static case in 5.2.4.2.2#21 (initialization with SNaN macros). Agreed.
· [Cfp-interest 2605] Fwd: [SC22WG14.22874] CFP views on CA-N3073-006 <http://mailman.oakapple.net/pipermail/cfp-interest/2023-January/002619.html> <> Jim Thomas
Reviewed CA comment.
Prefer no change to 5.2.4.2.2 and normalized and EPSILON.
Noted possible fallback loosening "normalized".
· [Cfp-interest 2615] Re: [SC22WG14.22900] CFP views on CA-N3073-006 <http://mailman.oakapple.net/pipermail/cfp-interest/2023-January/002629.html> <> Jim Thomas
Reviewed remarks about proposal. Agreed.
· [Cfp-interest 2616] Re: [SC22WG14.22918] CFP views on CA-N3073-006 <http://mailman.oakapple.net/pipermail/cfp-interest/2023-January/002630.html> <> Hubert Tong
Reviewed response to [Cfp-interest 2615].
Agreed on recommending no change.
Appreciated concern that existing programs might end up with different results due to these changes from C18.
· [Cfp-interest 2610] Re: [SC22WG14.22886] N3071 exactly representable <http://mailman.oakapple.net/pipermail/cfp-interest/2023-January/002624.html> <> Jₑₙₛ Gustedt
Discussed translation time vs before startup.
For needed change in 5.2.4.2.2 #21 (noted above), liked the suggestion to refer to translation time evaluation instead of listing static, thread, and constexpr as in F.8.4 and F.8.5.
Jim Thomas and Fred Tydeman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.oakapple.net/pipermail/cfp-interest/attachments/20230120/23fd9150/attachment-0001.htm>
More information about the Cfp-interest
mailing list