git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: PCRE v2 compile error, was Re: What's cooking in git.git (May 2017, #01; Mon, 1)

From
Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Date
May 2, 2017, 21:16 UTC
Message-ID
<CACBZZX5ani4DLj4-63yZFPaJnWXquuFb+pXKvyHL61kja6YghA@mail.gmail.com>
In-Reply-To
<20170502205138.GB17703@alpha.vpn.ikke.info>
On Tue, May 2, 2017 at 10:51 PM, Kevin Daudt <me@ikke.info> wrote:
Show 16 quoted lines
> On Tue, May 02, 2017 at 08:52:21PM +0200, Ęvar Arnfjörš Bjarmason wrote:
>>
>>  * Due to the bizarro existing semantics of the configure script noted
>> upthread if you have a git build script that does --with-libpcre & you
>> have libpcre1 installed, it'll link to it, but now since
>> --with-libpcre defaults to libpcre2 it'll silently skip linking to it
>> if you don't have it installed.
>>
>
> Case in point: The Archlinux git-git aur package[0] (community maintained,
> latest git version) does run ./configure without --with-libpcre, but
>  requests it from make with USE_LIBPCRE=1.
>
> I noticed when trying git grep -P which then failed.
>
> [0]: https://aur.archlinux.org/cgit/aur.git/tree/PKGBUILD?h=git-git

Whatever's going on there is unrelated to the issue I'm talking about, but if that's producing a package where -P doesn't work that looks like a bug in the Makefile.

The way --with-libpcre works now is that it second guesses you, i.e. it'll put USE_LIBPCRE=YesPlease into config.mak.autogen (sourced by the Makefile) only if you supply --with-libpcre *and* it can find a working libpcre on the system, otherwise it silently ignores you.

Whatever you supply to the configure script it'll be overridden by Makefile arguments if present, but even if there was another bug where ./configure arguments took precedence over Makefile arguments I don't see how it would apply here, in this case nothing libpcre related will be written to config.mak.autogen.

So I must be missing something. I don't see how that package build doesn't either fail at compile time because it doesn't have pcre, or work, in which case the -P option will work.

Previous: Kevin DaudtNext: Johannes Schindelin
Message 16 of 36 in “What's cooking in git.git (May 2017, #01; Mon, 1)”
  1. Junio C HamanoMay 1, 2017
  2. Ævar Arnfjörð BjarmasonMay 1, 2017
  3. Brandon WilliamsMay 1, 2017
  4. Ævar Arnfjörð BjarmasonMay 1, 2017
  5. Jeff KingMay 1, 2017
  6. Junio C HamanoMay 1, 2017
  7. Ævar Arnfjörð BjarmasonMay 2, 2017
  8. Junio C HamanoMay 2, 2017
  9. Ævar Arnfjörð BjarmasonMay 2, 2017
  10. Johannes SchindelinMay 2, 2017
  11. PCRE v2 compile error, was Re: What's cooking in git.git (May 2017, #01; Mon, 1)Johannes Schindelin, May 2, 2017
  12. Ævar Arnfjörð BjarmasonMay 2, 2017
  13. Johannes SchindelinMay 2, 2017
  14. Ævar Arnfjörð BjarmasonMay 2, 2017
  15. Kevin DaudtMay 2, 2017
  16. Ævar Arnfjörð BjarmasonMay 2, 2017
  17. Johannes SchindelinMay 3, 2017
  18. Duy NguyenMay 3, 2017
  19. Ævar Arnfjörð BjarmasonMay 3, 2017
  20. Johannes SchindelinMay 4, 2017
  21. Ævar Arnfjörð BjarmasonMay 4, 2017
  22. Johannes SchindelinMay 4, 2017
  23. Ævar Arnfjörð BjarmasonMay 4, 2017
  24. Ævar Arnfjörð BjarmasonMay 8, 2017
  25. Junio C HamanoMay 8, 2017
  26. brian m. carlsonMay 8, 2017
  27. Ævar Arnfjörð BjarmasonMay 9, 2017
  28. brian m. carlsonMay 9, 2017
  29. Johannes SchindelinMay 9, 2017
  30. Ævar Arnfjörð BjarmasonMay 9, 2017
  31. Ævar Arnfjörð BjarmasonMay 9, 2017
  32. demerphqMay 9, 2017
  33. Ævar Arnfjörð BjarmasonMay 9, 2017
  34. Brandon WilliamsMay 2, 2017
  35. Ævar Arnfjörð BjarmasonMay 2, 2017
  36. Ramsay JonesMay 5, 2017

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.