[J3] Proposals for new intrinsic procedures

[Gary Klimowicz]

For the record, the only other paper I found was J3/05-185. It just lists the procedure names from HPF version 2.0 paragraph 7.5.6.

05-246r1 is about 20 times larger, by line count. It has the definitions.



> On Jan 12, 2023, at 07:54, Dan Nagle via J3 <j3 at mailman.j3-fortran.org> wrote:
> 
> Hi,
> 
> For a while, the prefix/suffix procedures were on the worklist
> (for f08?).  One reason they were not completed was that the strategy
> subgroup favored was to copy the HPF definitions (we had a ps of the HPF doc),
> but they were found to be too hard to put into ISO language.
> So the project stalled and hasn't restarted until now.
> 
> HTH
> 
> I don't recall implementation effort as the showstopper, some vendors
> already had HPF libraries.  Which is one more advantage of keeping
> the HPF definitions.
> 
> The only paper I found quickly is J3/05-246r1.
> 
> 
> 
> 
>> On Jan 11, 2023, at 18:59, Brad Richardson via J3 <j3 at mailman.j3-fortran.org> wrote:
>> 
>> Hi Malcolm,
>> 
>> Thanks for the suggestions. I did look briefly at HPF. My question
>> would be, do the committee and vendors believe having operator specific
>> versions would be preferable to a single, generic version? My feeling
>> was that the "too few uses to too much work" aspect would be minimized
>> by adding only the generalized version.
>> 
>> Regards,
>> Brad
>> 
>> On Thu, 2023-01-12 at 10:52 +0900, Malcolm Cohen via J3 wrote:
>>> I'd recommend using the HPF_LIBRARY terminology, they had SUM_PREFIX,
>>> SUM_SUFFIX, and SUM_SCATTER. Optional argument EXCLUSIVE controlling
>>> exclusivity. Some vendors already implemented these a long time ago
>>> (possibly only in serial mode).
>>> 
>>> I seem to remember suggesting these before actually, but not many
>>> committee members were in favour at that time. Something alone the
>>> lines of "too few uses + too much work".
>>> 
>>> HPF_LIBRARY also had ALL_PREFIX, ANY_PREFIX, COPY_PREFIX,
>>> COUNT_PREFIX, IALL_PREFIX, IANY_PREFIX, IPARITY_PREFIX,
>>> MAXVAL_PREFIX, MINVAL_PREFIX, PARITY_PREFIX and PRODUCT_PREFIX. (And
>>> the corresponding _SCATTER and _SUFFIX functions.) Perhaps overly
>>> generalised, but may be worth considering.
>>> 
>>> Cheers,
>>> --
>>> ..............Malcolm Cohen, NAG Oxford/Tokyo.
>>> 
>>> -----Original Message-----
>>> From: J3 <j3-bounces at mailman.j3-fortran.org> On Behalf Of Brad
>>> Richardson via J3
>>> Sent: Thursday, January 12, 2023 10:14 AM
>>> To: General J3 interest list <j3 at mailman.j3-fortran.org>
>>> Cc: Brad Richardson <everythingfunctional at protonmail.com>
>>> Subject: Re: [J3] Proposals for new intrinsic procedures
>>> 
>>> To those interested, I have produced drafts of the proposals as PRs
>>> to the j3-fortran/fortran_proposals Github repositories.
>>> 
>>> https://github.com/j3-fortran/fortran_proposals/pull/290
>>> https://github.com/j3-fortran/fortran_proposals/pull/291
>>> 
>>> I welcome any comments or suggestions there. I will upload the papers
>>> to the J3 site after making any suggested changes.
>>> 
>>> Regards,
>>> Brad
>>> 
>>> On Thu, 2023-01-05 at 23:35 +0000, Brad Richardson via J3 wrote:
>>>> Thanks all for the suggestions, direction and background. I'll do
>>>> scan
>>>> (or whatever we end up calling it) as one paper, and the other IEEE
>>>> procedures as one paper. If anybody has suggestions for the
>>>> spelling
>>>> of INCLUSIVE_SCAN and EXCLUSIVE_SCAN let me know. I'll definitely
>>>> have
>>>> it as a straw vote.
>>>> 
>>>> Regards,
>>>> Brad
>>>> 
>>>> On Thu, 2023-01-05 at 19:35 +0000, Brad Richardson via J3 wrote:
>>>>> Hi all,
>>>>> 
>>>>> Would proposals for new intrinsic functions be acceptable to
>>>>> submit
>>>>> for the upcoming meeting? Specifically I've been asked to write
>>>>> up
>>>>> the papers for:
>>>>> 
>>>>> * prefix sum (commonly known as scan, but of course that name's
>>>>> already taken), both inclusive and exclusive
>>>>> * All functions recommended by IEEE-754, most of which Fortran
>>>>> already has, but the whole list is reproduced below
>>>>> 
>>>>> exp
>>>>> expm1
>>>>> exp2
>>>>> exp2m1
>>>>> exp10
>>>>> exp10m1
>>>>> log
>>>>> log2
>>>>> log10
>>>>> logp1
>>>>> log2p1
>>>>> log10p1
>>>>> hypot
>>>>> rSqrt
>>>>> compound
>>>>> rootn
>>>>> pown
>>>>> sin
>>>>> cos
>>>>> tan
>>>>> sinPi
>>>>> cosPi
>>>>> tanPi
>>>>> asin
>>>>> acos
>>>>> atan
>>>>> atan2
>>>>> sinh
>>>>> cosh
>>>>> tanh
>>>>> acosh
>>>>> atanh
>>>>> 
>>>>> Some points of order questions also. Which subgroup is in charge
>>>>> of
>>>>> the intrinsic procedures so that I can coordinate with them?
>>>>> Would
>>>>> it be best to write a separate paper for each new procedure, put
>>>>> them all in a single paper, or some scheme for grouping certain
>>>>> subsets? And whilst we don't have final version of F2023 to draft
>>>>> the edits against, would it be worthwhile to go ahead and detail
>>>>> out
>>>>> what they will look like?
>>>>> 
>>>>> Thanks in advance for any feedback.
>>>>> 
>>>>> Regards,
>>>>> Brad
>>>>> 
>>>> 
>>>> 
>>> 
>>> 
>>> 
>> 
>> 
> 
> 
> --
> 
> Cheers!
> Dan Nagle
> 
>