Date: Fri, 23 May 2025 18:59:30 +0300
Hi all,
Here is another PR against the ub-ifndr branch: https://github.com/cplusplus/draft/pull/7888
This PR breaks up some `\ubdef` entries in the UB annex into multiple, more granular entries, for those cases of UB where there is value in distinguishing them.
This PR represents the second half of the work Josh and I did on enumerating UB (the first part of which was merged via PR #7864 <https://github.com/cplusplus/draft/pull/7864>) and makes the list of UB in the `ub-ifndr` branch identical to the list of UB that we published in Appendix A of P3100R2 <https://wg21.link/p3100r2>.
Please review and merge!
Thanks,
Timur
> On 23 Apr 2025, at 15:49, Timur Doumler via SG12 <sg12_at_[hidden]> wrote:
>
> Hi Herb,
>
> Thanks for your email!
>
>> On 23 Apr 2025, at 06:15, Herb Sutter via SG12 <sg12_at_[hidden]> wrote:
>> The first step is to document the current UB/IFNDR we already have: tag all instances and create Annexes. Shafik (thanks!) with help from some others (especially Timur and Josh, thanks!) has a pretty complete first pass as a draft PR here for explicit cases of IFNDR (25 cases) and UB (63 cases):
>>
>
> While Shafik has done great work, as stated in my series of emails from last week, this PR is missing ~20 cases of UB, including some common ones like reading an indeterminate value, so I would not yet call that "pretty complete". I am currently on vacation with very shaky internet so I can't do much until I'm back home on the 2nd of May, but in my absence Josh has promised to prepare PRs with the missing instances of UB that we found. Not sure where those are at right now (Josh, any update?) but I suggest we merge those in first (and also address the other comments from my series of emails addressing potential defects in Shafik's PR that Jens said he would prepare PRs for) before calling it ready for CWG.
>
> Does that make sense?
>
> Thanks,
> Timur
> _______________________________________________
> SG12 mailing list
> SG12_at_[hidden]
> Subscription: https://lists.isocpp.org/mailman/listinfo.cgi/sg12
> Link to this post: http://lists.isocpp.org/sg12/2025/04/1047.php
Here is another PR against the ub-ifndr branch: https://github.com/cplusplus/draft/pull/7888
This PR breaks up some `\ubdef` entries in the UB annex into multiple, more granular entries, for those cases of UB where there is value in distinguishing them.
This PR represents the second half of the work Josh and I did on enumerating UB (the first part of which was merged via PR #7864 <https://github.com/cplusplus/draft/pull/7864>) and makes the list of UB in the `ub-ifndr` branch identical to the list of UB that we published in Appendix A of P3100R2 <https://wg21.link/p3100r2>.
Please review and merge!
Thanks,
Timur
> On 23 Apr 2025, at 15:49, Timur Doumler via SG12 <sg12_at_[hidden]> wrote:
>
> Hi Herb,
>
> Thanks for your email!
>
>> On 23 Apr 2025, at 06:15, Herb Sutter via SG12 <sg12_at_[hidden]> wrote:
>> The first step is to document the current UB/IFNDR we already have: tag all instances and create Annexes. Shafik (thanks!) with help from some others (especially Timur and Josh, thanks!) has a pretty complete first pass as a draft PR here for explicit cases of IFNDR (25 cases) and UB (63 cases):
>>
>
> While Shafik has done great work, as stated in my series of emails from last week, this PR is missing ~20 cases of UB, including some common ones like reading an indeterminate value, so I would not yet call that "pretty complete". I am currently on vacation with very shaky internet so I can't do much until I'm back home on the 2nd of May, but in my absence Josh has promised to prepare PRs with the missing instances of UB that we found. Not sure where those are at right now (Josh, any update?) but I suggest we merge those in first (and also address the other comments from my series of emails addressing potential defects in Shafik's PR that Jens said he would prepare PRs for) before calling it ready for CWG.
>
> Does that make sense?
>
> Thanks,
> Timur
> _______________________________________________
> SG12 mailing list
> SG12_at_[hidden]
> Subscription: https://lists.isocpp.org/mailman/listinfo.cgi/sg12
> Link to this post: http://lists.isocpp.org/sg12/2025/04/1047.php
Received on 2025-05-23 15:59:41