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

Re: [PATCH v2 2/4] compat/regex: include alloca.h before undef it

From
Ramsay Jones <ramsay@ramsayjones.plus.com>
Date
Apr 27, 2020, 16:28 UTC
Message-ID
<f3af49c7-6c35-d2f6-c87c-e7342b1c2b6f@ramsayjones.plus.com>
In-Reply-To
<20200427010821.GD14800@danh.dev>

On 27/04/2020 02:08, Danh Doan wrote: [snip]

Show 8 quoted lines
>>> musl's alloca.h is available here:
>>>
>>> https://git.musl-libc.org/cgit/musl/tree/include/alloca.h
>>
>> Hmm, OK, so that partly explains the problem. I wonder if the
>> musl guys would accept a bug report?
> 
> I don't think they have a policy of no `#undef` whatsoever.
That's fair enough.
> But, I think they're picky when come to C-correctly and
> POSIX-correctly.
> Does C or POSIX define alloca(3) at all?

No alloca() is not in either the POSIX or C standard(s). This was an extension from the early days of BSD Unix.

For some reason, I thought you had to explicitly '#include <alloca.h>' to use it, but it appears that (by default) you get a bonus include from the <stdlib.h> header, unless you restrict the headers using the various macros and/or compiler command-line options.

As it happens, even on glibc systems, the <alloca.h> header is included by the <stdlib.h> header, unless you take steps to suppress it. So, we would have had the same issue, if it wasn't for the aforementioned '#undef alloca' the the glibc header file.

When I need to look at pp output, while debugging things like this, I cherry-pick a patch to the Makefile:

  $ git diff
  diff --git a/Makefile b/Makefile
  index 6d5cee270c..cd8753bf54 100644
  --- a/Makefile
  +++ b/Makefile
  @@ -2423,6 +2423,9 @@ $(ASM_OBJ): %.o: %.S GIT-CFLAGS $(missing_dep_dirs)
   %.s: %.c GIT-CFLAGS FORCE
          $(QUIET_CC)$(CC) -o $@ -S $(ALL_CFLAGS) $(EXTRA_CPPFLAGS) $<
   
  +%.i: %.c GIT-CFLAGS FORCE
  +       $(QUIET_CC)$(CC) $(ALL_CFLAGS) $(EXTRA_CPPFLAGS) -E $< >$*.i
  +
   ifdef USE_COMPUTED_HEADER_DEPENDENCIES
   # Take advantage of gcc's on-the-fly dependency generation
   # See <http://gcc.gnu.org/gcc-3.0/features.html>.
  @@ -2474,7 +2477,7 @@ http-walker.sp http-walker.s http-walker.o: EXTRA_CPPFLAGS = -DNO_EXPAT
   endif
   
   ifdef NO_REGEX
  -compat/regex/regex.sp compat/regex/regex.o: EXTRA_CPPFLAGS = \
  +compat/regex/regex.i compat/regex/regex.sp compat/regex/regex.o: EXTRA_CPPFLAGS = \
          -DGAWK -DNO_MBSUPPORT
   endif
   
  $ 

[The second hunk above is not actually part of the cherry-picked patch, but I needed it in this instance to get GAWK and NO_MBSUPPORT passed to the compiler!]

  $ make NO_REGEX=1 compat/regex/regex.i
      CC compat/regex/regex.i
  $ vim compat/regex/regex.i
  $ 

... which shows <alloca.h> is indeed being #included from <stdlib.h>. [it is protected by a __USE_MISC pp variable, but I didn't bother to track it down! ;-)]

ATB, Ramsay Jones

  
Previous: Danh DoanNext: Danh Doan
Message 34 of 51 in “fix make sparse warning”
  1. 0/4 fix make sparse warningĐoàn Trần Công Danh, Apr 23, 2020
  2. 1/4 C: s/0/NULL/ for pointer typeĐoàn Trần Công Danh, Apr 23, 2020
  3. Ramsay JonesApr 24, 2020
  4. Junio C HamanoApr 24, 2020
  5. Danh DoanApr 24, 2020
  6. Junio C HamanoApr 24, 2020
  7. Luc Van OostenryckMay 14, 2020
  8. 2/4 compat/regex: silence `make sparse` warningĐoàn Trần Công Danh, Apr 23, 2020
  9. Ramsay JonesApr 24, 2020
  10. Danh DoanApr 24, 2020
  11. 3/4 graph.c: limit linkage of internal variableĐoàn Trần Công Danh, Apr 23, 2020
  12. Ramsay JonesApr 24, 2020
  13. 4/4 progress.c: silence cgcc suggestion about internal linkageĐoàn Trần Công Danh, Apr 23, 2020
  14. Ramsay JonesApr 24, 2020
  15. Jeff KingApr 24, 2020
  16. 5/4 fmt-merge-msg.c: fix `make sparse` on nextĐoàn Trần Công Danh, Apr 23, 2020
  17. Ramsay JonesApr 23, 2020
  18. Danh DoanApr 23, 2020
  19. Ramsay JonesApr 24, 2020
  20. 0/4 Fix Sparse WarningĐoàn Trần Công Danh, Apr 24, 2020
  21. 1/4 test-parse-pathspec-file.c: s/0/NULL/ for pointer typeĐoàn Trần Công Danh, Apr 24, 2020
  22. 2/4 compat/regex: include alloca.h before undef itĐoàn Trần Công Danh, Apr 24, 2020
  23. Ramsay JonesApr 24, 2020
  24. Danh DoanApr 24, 2020
  25. Ramsay JonesApr 24, 2020
  26. Danh DoanApr 24, 2020
  27. Ramsay JonesApr 25, 2020
  28. Danh DoanApr 26, 2020
  29. Danh DoanApr 26, 2020
  30. Ramsay JonesApr 26, 2020
  31. Ramsay JonesApr 26, 2020
  32. Junio C HamanoApr 26, 2020
  33. Danh DoanApr 27, 2020
  34. Ramsay JonesApr 27, 2020
  35. Danh DoanApr 27, 2020
  36. Ramsay JonesApr 27, 2020
  37. 3/4 graph.c: limit linkage of internal variableĐoàn Trần Công Danh, Apr 24, 2020
  38. 4/4 progress.c: silence cgcc suggestion about internal linkageĐoàn Trần Công Danh, Apr 24, 2020
  39. Ramsay JonesApr 24, 2020
  40. Johannes SchindelinApr 25, 2020
  41. Danh DoanApr 26, 2020
  42. Ramsay JonesApr 26, 2020
  43. Johannes SchindelinMay 1, 2020
  44. 0/4 Partial fix `make sparse`Đoàn Trần Công Danh, Apr 27, 2020
  45. 1/4 test-parse-pathspec-file.c: s/0/NULL/ for pointer typeĐoàn Trần Công Danh, Apr 27, 2020
  46. 2/4 compat/regex: move stdlib.h up in inclusion chainĐoàn Trần Công Danh, Apr 27, 2020
  47. Ramsay JonesApr 27, 2020
  48. 3/4 graph.c: limit linkage of internal variableĐoàn Trần Công Danh, Apr 27, 2020
  49. 4/4 progress.c: silence cgcc suggestion about internal linkageĐoàn Trần Công Danh, Apr 27, 2020
  50. Johannes SchindelinMay 1, 2020
  51. Ramsay JonesApr 27, 2020

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.