On 2/5/25 3:49 PM, Bruno Haible wrote:
> Eli Schwartz wrote:
>> Bisected, and landed on:
>>
>>
>> 043355371a76de8ea7d06f79a69fde905af7cc45 is the first bad commit
>> commit 043355371a76de8ea7d06f79a69fde905af7cc45
>> Author:     Paul Eggert <[email protected]>
>> AuthorDate: Fri Sep 6 22:32:54 2024
>> Commit:     Paul Eggert <[email protected]>
>> CommitDate: Mon Sep 16 01:50:38 2024
>>
>>     X == -1 → X < 0
> 
> Oh, I see. On Haiku, system-defined errno values are not positive (as
> demanded by POSIX), but negative [1]. Since there's no chance that this
> will change (because changing errno values would mean that all programs
> need to be recompiled), we need to bite the sour apple and revert to
> == -1 for errno values.


Thanks. I was *wondering* whether the errno changes were the cause --
they did seem the most logical source for a platform-specific
inconsistency after all.


> With the attached patch, the tests concat-git-diff and remember-backup-files
> now pass. The only remaining test failure is preserve-mode-and-timestamp,
> for which we have an explanation.


Hmm, so this also fixes an outstanding issue from the previous version?
Interesting.


-- 
Eli Schwartz

Attachment: OpenPGP_signature.asc
Description: OpenPGP digital signature

  • Re: patch-2.7.... Eli Schwartz
    • Re: patch... Bug reports, suggestions, general discussion for GNU patch.
      • Re: p... Eli Schwartz
        • R... Bug reports, suggestions, general discussion for GNU patch.
          • ... Eli Schwartz
            • ... Bug reports, suggestions, general discussion for GNU patch.
          • ... Bug reports, suggestions, general discussion for GNU patch.
          • ... Paul Eggert
            • ... Andreas Grünbacher
            • ... Bug reports, suggestions, general discussion for GNU patch.

Reply via email to