[Cfp-interest 1454] WG14 IEEE 754-C binding meeting minutes 2020/01/09
Rajan Bhakta
rbhakta at us.ibm.com
Thu Jan 9 09:38:55 PST 2020
Attendees: Rajan, Jim, Mike, Ian, David H.
New agenda items:
None.
Carry over action items:
CFP: Put the tgmath redefinition as a proposal to the standard once we
have a base document with TS Part 3 in it.
Last meeting action items:
Fred: Update erange proposal based on CFP 1437 and CFP teleconference
discussion. - Unknown.
Jim: Submit http://wiki.edg.com/pub/CFP/WebHome/C2x_proposal_-
_NaN_and_infinity_macros_-_20191108.pdf to WG14. - Done.
Rajan: Compile example code in
http://wiki.edg.com/pub/CFP/WebHome/C2x_proposal_-
_why_no_wide_string_strfrom_functions_v2.pdf. - Done.
Jim: Choose how to attach footnote in
http://wiki.edg.com/pub/CFP/WebHome/C2x_proposal_-
_why_no_wide_string_strfrom_functions_v2.pdf; mention the other approach
in the problem description. - Done.
Jim: Draft a paper proposing changing the “cr” prefix for correctly
rounded functions to “cr_”. Include with other suggested changes in
response to Jens’s naming paper. - Done.
David H: Draft words to recommend honoring properties (Ex. even/odd,
periodicity) of math functions that would follow automatically from
correct rounding. - Not done.
Jim: Find a place for words about math function properties.
Coordinate with David H (previous AI). - Not done.
Comments:
Ian: Monoticity as well and making correctly rounded be
recommended practice.
David: We can say if it is correctly rounded, they get the
properties anyways.
Jim: Draft a proposal to change the return words for powr and add the
footnote (presented in the meeting) to justify powr. - Done.
New action items:
Jim: Look at seeing what would be involved in removing Annex J types
that already have the same basic types.
Jim: Get a WG14 document number and submit
C2x_proposal_-_why_no_wide_string_strfrom_functions- 20191230.pdf.
Jim: Get a WG14 document number and submit
C2x_proposal_-_names_and_locations-20200105.pdf.
Jim: Ensure natural log is written as log in the standard (regarding
C2x_proposal_-_powr_justification_wording-20200104.pdf).
Jim: Ensure thin spaces are put where needed in
C2x_proposal_-_powr_justification_wording-20200104.pdf. Ex. Between the e
and y in the "e^y log x" in three places.
Jim: Update links on the C FP wiki to point to the latest drafts of
the TS parts and look into having a section for working drafts.
Rajan: Ensure casting to _Float16 won't cause a problem with the using
INFINITY wherever a constant expression can be used change we proposed.
Next Meeting(s):
Wednesday, February 19th, 2020, 11:00 EST, 8:00 PST, 4PM UTC
Same teleconference number.
Please notify the group if this time slot does not work.
Jim away Jan. 14-28th.
Discussion:
754 revision:
Done!
CFP 1443 has the paper/article David wrote.
Drop this agenda item.
C++ Liaison:
Nothing.
C2X integration (
http://www.open-std.org/jtc1/sc22/wg14/www/docs/n2433.pdf):
Draft with TS 1, 2, and 4a (Latest C2X draft:
http://www.open-std.org/jtc1/sc22/wg14/www/docs/n2454.pdf, also as link on
CFP wiki)
Part 3 - Was in a branch as work in progress prior to Jens stepping
down as editor. As Annex J.
Part 4b - Looking as an updated TS.
Part 5a,b,c,d
C2X target: 2023
Jim: Idea is to propose IEEE 2019 binding in the next C meeting. We
can discuss next CFP meeting.
Jim: Question: Should we remove value redundant types from Annex J?
Ex. _Float32 being the same as float, _Float64 being the same as double,
_Float32x being the same as double or long double. We added this as
consistency, but the initial goal was just to provide nomenclature for the
types.
*AI*: Jim: Look at seeing what would be involved in removing Annex
J types that already have the same basic types.
Action item details:
Fred: Update erange proposal based on CFP 1437 and CFP teleconference
discussion.
Jim: Submit http://wiki.edg.com/pub/CFP/WebHome/C2x_proposal_-
_NaN_and_infinity_macros_-_20191108.pdf to WG14. (
http://wiki.edg.com/pub/CFP/WebHome/n2469.pdf)
Looks good.
Rajan: Compile example code in
http://wiki.edg.com/pub/CFP/WebHome/C2x_proposal_-
_why_no_wide_string_strfrom_functions_v2.pdf. (See Rajan’s CFP 1441)
Looks good.
Jim: Choose how to attach footnote in
http://wiki.edg.com/pub/CFP/WebHome/C2x_proposal_-
_why_no_wide_string_strfrom_functions_v2.pdf; mention the other approach
in the problem description.
http://wiki.edg.com/pub/CFP/WebHome/C2x_proposal_-_why_no_wide_string_strfrom_functions-
20191230.pdf
Looks good.
*AI*: Jim: Get a WG14 document number and submit
C2x_proposal_-_why_no_wide_string_strfrom_functions- 20191230.pdf.
Jim: Draft a paper proposing changing the “cr” prefix for correctly
rounded functions to “cr_”. Include with other suggested changes in
response to Jens’s naming paper.
http://wiki.edg.com/pub/CFP/WebHome/C2x_proposal_-_names_and_locations-20200105.pdf
cr_ was never proposed by any paper.
Moving the SNAN macros does follow the suggestion of renaming, but
also moves it which was not previously proposed.
*AI*: Jim: Get a WG14 document number and submit
C2x_proposal_-_names_and_locations-20200105.pdf.
David H: Draft words to recommend honoring properties of math
functions that would follow automatically from correct rounding.
Still to do.
Jim: Find a place for words about math function properties. Coordinate
with David H (previous AI).
Suggest adding a title to the table in F.3#23.
David: Functions that IEC 60559 recommends be correctly rounded
David: The first table title should also be changed to distinguish
this table. Ex. Binding of operations required by IEC 60559; Binding of
functions or operations recommended by IEC 60559.
Jim: Or "Bindings of operations required by IEC 60559"; "Bindings
of operations recommended by IEC 60559" See C FP 1453
Suggest adding a second recommended practice to F.10.
Jim: Draft a proposal to change the return words for powr and add the
footnote (presented in the meeting) to justify powr.
http://wiki.edg.com/pub/CFP/WebHome/C2x_proposal_-_powr_justification_wording-20200104.pdf
*AI*: Jim: Ensure natural log is written as log in the standard
(regarding C2x_proposal_-_powr_justification_wording-20200104.pdf).
*AI*: Jim: Ensure thin spaces are put where needed in
C2x_proposal_-_powr_justification_wording-20200104.pdf. Ex. Between the e
and y in the "e^y log x" in three places.
Other issues:
WG14 editor change
Jim and Rajan meeting with David Keaton regarding finishing off the
work.
Is “Floating-point extensions for C” still an appropriate description
of CFP work? See Mike’s email CFP 1445.
Jim: "IEC 60559 Floating point support in C"?
Sounds good.
Jim: Also remove reference to 2008 on the website.
Jim: Update links to the published TS's on the website.
Rajan: Link published for all, and have working drafts for any
ones with changes or that are not in C2X.
*AI*: Jim: Update links on the C FP wiki to point to the latest
drafts of the TS parts and look into having a section for working drafts.
Issue for _Float16: INFINITY has type float so widens operations with
_Float16 operands.
Jim: Ideally NAN and INFINITY macros would be _Float16 if it was
supported. We could add a comment suggesting putting a cast to _Float16 in
front of INFINITY if you really wanted operations to stay in _Float16
type.
Ian: Doesn't work for SNAN's.
Jim: Not an issue since we have type'd versions of those.
May be an issue for constant expressions.
*AI*: Rajan: Ensure casting to _Float16 won't cause a problem with
the using INFINITY wherever a constant expression can be used change we
proposed.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.oakapple.net/pipermail/cfp-interest/attachments/20200109/4ad1d399/attachment-0001.html
More information about the Cfp-interest
mailing list