Add FMV feature names priority[ABCDE]

[alicecarlotti]

These feature names allow developers to specify a priority order for their function versions that differs from the default priority order for the specified architectural features.

This builds upon #370.

---

name: Pull request about: Technical issues, document format problems, bugs in scripts or feature proposal.

---

Thank you for submitting a pull request!

If this PR is about a bugfix:

Please use the bugfix label and make sure to go through the checklist below.

If this PR is about a proposal:

We are looking forward to evaluate your proposal, and if possible to make it part of the Arm C Language Extension (ACLE) specifications.

We would like to encourage you reading through the contribution guidelines, in particular the section on submitting a proposal.

Please use the proposal label.

As for any pull request, please make sure to go through the below checklist.

Checklist: (mark with X those which apply)

• [ ] If an issue reporting the bug exists, I have mentioned it in the PR (do not bother creating the issue if all you want to do is fixing the bug yourself).
• [ ] I have added/updated the SPDX-FileCopyrightText lines on top of any file I have edited. Format is SPDX-FileCopyrightText: Copyright {year} {entity or name} <{contact informations}> (Please update existing copyright lines if applicable. You can specify year ranges with hyphen , as in 2017-2019, and use commas to separate gaps, as in 2018-2020, 2022).
• [ ] I have updated the Copyright section of the sources of the specification I have edited (this will show up in the text rendered in the PDF and other output format supported). The format is the same described in the previous item.
• [ ] I have run the CI scripts (if applicable, as they might be tricky to set up on non-*nix machines). The sequence can be found in the contribution guidelines. Don't worry if you cannot run these scripts on your machine, your patch will be automatically checked in the Actions of the pull request.
• [ ] I have added an item that describes the changes I have introduced in this PR in the section Changes for next release of the section Change Control/Document history of the document. Create Changes for next release if it does not exist. Notice that changes that are not modifying the content and rendering of the specifications (both HTML and PDF) do not need to be listed.
• [ ] When modifying content and/or its rendering, I have checked the correctness of the result in the PDF output (please refer to the instructions on how to build the PDFs locally).
• [ ] The variable draftversion is set to true in the YAML header of the sources of the specifications I have modified.
• [ ] Please DO NOT add my GitHub profile to the list of contributors in the README page of the project.

[jroelofs]

LGTM. Needs rebase and separation from #370. Perhaps it's not the most intuitive syntax but does the job in my opinion. @jroelofs @DanielKristofKiss what are your thoughts?

bikeshed: maybe priority1/priority2/etc. would be more clear naming-wise? seems fine either way, to me.

[labrinea]

bikeshed: maybe priority1/priority2/etc. would be more clear naming-wise? seems fine either way, to me.

I had the exact same thought. I don't mind either way.

[alicecarlotti]

Re. the bikeshedding - this may well be a cultural difference, but to me it's unclear whether priority1 would mean the first (highest) priority, or the lowest one, whereas it feels unambiguous to me lettered priorities.

Some mostly non-technical examples: Smaller numbers are ranked higher - e.g. 1st place in a competition. Larger numbers are ranked higher - e.g. English GCSE exam grades, or the priority numbers in the table of features in this document.

Earlier letters are ranked higher - e.g. Many UK exam grade systems, UK road classification (A, B, C roads, ignoring motorways) Later letters are ranked higher - I can't think of any examples for this.

I recognise that my vague associations with letter/number orderings might be specific to my own life experience, and other people may see this differently, so I'm happy to change this if you disagree with my thinking.

[jroelofs]

All ambiguity would be removed if they were named e.g. priority9998/priority9999/etc. That said, I don't have a preference here. Letters seem fine, since if you're having to adjust priorities manually, you're probably going to have to look at the table anyway.

[DanielKristofKiss]

my 2 cents,

can we specify priorityA+priorityB+crc' ? or even priorityA+priorityA+crc' ?

If we go with priorityN then I'd prefer to specify it with a formula instead of typing in all numbers. This could be a pro besides priority\d+ format.

Original intention was developers rarely need to know about the actual priorities as allocated values should mostly fine for the usual cases. But if they need to deal with the ordering of the selection then maybe exact values are better.

Bit different idea: more complex case could be specify order\d+ and the selection algorithm goes over the ordered item(0toN)first then priority (M to default).

[labrinea]

@DanielKristofKiss can you please explain a little more as I am not sure I understood. Are you suggesting we use an explicit priority syntax like the one for example in RISC-V: https://github.com/riscv-non-isa/riscv-c-api-doc/pull/85/files ? That would be harder to implement, for the current proposal the machinery is in place, only these pseudo fmv features are needed to be implemented.

can we specify priorityA+priorityB+crc ? or even priorityA+priorityA+crc ?

Yes, like with every other feature, these special features can be used in conjuction with others. Repeating the same feature in a version does not get diagnosed, but has no effect (at least in LLVM).

[DanielKristofKiss]

@DanielKristofKiss can you please explain a little more as I am not sure I understood. Are you suggesting we use an explicit priority syntax like the one for example in RISC-V: https://github.com/riscv-non-isa/riscv-c-api-doc/pull/85/files ? That would be harder to implement, for the current proposal the machinery is in place, only these pseudo fmv features are needed to be implemented.

Something similar, so for me the feature and the priority are two different things that could expressed by the developers. (; or recognising the priority/order meta features are similar enough) I see from implementation point of view the current proposal is simpler but that should be the goal or the usability/expressivity?

FYI this is even more complex proposal: https://youtu.be/LTM1was1dTU?t=1720

[AlfieRichardsArm]

Proposal can get a +1 from me. Seems like a good idea

I lean towards priority\d+ personally (I think priority and a number clearly indicates 1 would be highest). But would approve this with either syntax.

[AlfieRichardsArm]

Added a proposal document for this: https://github.com/ARM-software/acle/issues/403

[alicecarlotti]

I just rebased this; the branch no longer adds the commit for #370 since that is now merged to main.

[chrisbazley]

Sorry, but "priorityA+priorityB+crc" seems nonsensical to me. I'd prefer syntax that makes such errors impossible. I don't think that such strings should be used to expand the range of priorities that can be encoded, either.

Give the fact that this is a bit of a niche feature, I don't think that using a number would be problematic (so long as the meaning is documented somewhere). FWIW, I'd expect the 'highest' priority to be the highest number.

[AlfieRichardsArm]

@chrisbazley The priority=N;arch proposal was brought up in the proposal doc for this: https://github.com/ARM-software/acle/issues/403 I think we'll likely close this out in favor of that as we seem to be in agreement its preferred

[AlfieRichardsArm]

@andrewcarlotti Are you okay to close this, as we opted for the other PR