[J3] [SC22WG5.6420] [ukfortran] Fwd: N2204-1 is available

John Reid John.Reid at stfc.ac.uk
Sun Aug 14 09:46:15 UTC 2022


I have looked at the changes from the recent meeting and found all well. 
However, I do have some comments on the changes imported from corr 2 and 
on your draft Editor’s Report mjc001, see attached file.

Thanks for your hard and careful work.


Steve Lionel wrote:
> See following note from Malcolm. I'm waiting to submit the DIS until 
> Bill Ash returns next week and submits Corrigendum 2. Feel free to 
> look over Malcolm's Editor's Report and the draft in the Meeting 
> Files/Edit section of the J3 website (login required.)
> Steve
> -------- Forwarded Message --------
> Subject: 	N2204-1 is available
> Date: 	Mon, 8 Aug 2022 15:04:17 +0900
> From: 	Malcolm Cohen <malcolm at nag-j.co.jp>
> To: 	Steve Lionel <steve at stevelionel.com>
> Hi Steve,
> I have completed the draft of the DIS, it is N2204-1.pdf in the Edit 
> folder of Meeting Files.
> I made a number of extra edits in the process. Attached is my draft 
> Editor’s Report, which I will make a J3 paper in due course.
> Do you plan on having anyone review it before submitting it to SC22?
> Cheers,
> -- 
> ..............Malcolm Cohen, NAG Oxford/Tokyo.
> _______________________________________________
> ukfortran mailing list
> https://lists.accu.org/mailman/listinfo/ukfortran

-------------- next part --------------

1. With regard to changes imported from Corr 2:
  (a) There are five places where I think what was "ultimate component" 
      in coarray contexts in F2018 needs to be "potential subobject component"
      in F2023. Here are the places. Restrictions on entities associated with dummy arguments,
      para 1, (3) (e)
      para 1, (4) (e)
      final note (NOTE 5), first sentence.
   16.9.54 CO_BROADCAST, argument A,  final sentence (twice).

   (b) In 17.10 Summary of the procedures, para 3, item ES, "pure elemental" 
      has been changed to "simple elemental". The corresponding change has been 
	  made in 
      17.11.5 IEEE_GET_FLAG (FLAG, FLAG_VALUE) and in 
      I agree with these changes, but they are not documented in mjc001.
   (c) In interp F18/040 (in N2205) these words
      "and if in a recursive procedure at the same depth of recursion"
      appear at the end of the edit but are not in the edit for Subclause in N2203 (TC2) and are not in the DIS. Is this a mistake? 
2. With regard to mjc001:

   (a) 16.9.107 IMAGE_INDEX (COARRAY, SUB) ..., para 3, argument COARRAY. 
      Malcolm asks what happens if TEAM_NUMBER appears and the current team 
	  is the initial team. We have failed to cover this case. There is a 
	  separate sentence for it in 9.6 Image selectors, para 3. We intended 
	  the rules for IMAGE_INDEX to be consistent with the rules in an image 
	  selector, so I think we need a sentence here very like that in 9.6. 
	  I suggest this edit:
      In 16.9.107 IMAGE_INDEX (COARRAY, SUB) ..., para 3, argument COARRAY, 
	  before "If TEAM appears" add
      "If TEAM_NUMBER appears and the current team is the initial team, it 
	  shall be established in the initial team and the value of TEAM_NUMBER
      shall be the team number for the initial team."
      However, is this acceptable as an editorial change? 

     (b) I agree with both these editorial suggestions 
      In A.2 Processor dependencies, bullet "the ordering between records 
	  written by different iterations of a DO CONCURRENT construct ...", 
      change (11.1.7) to (
      In 3.131.1 component specification expression, after "10.1.11" add 
      "paragraph 9".

More information about the J3 mailing list