Re: FreeBSD /bin/sh 'set' bug

[Ralf Wildenhues]

Hi Paul,

* Paul Eggert wrote on Fri, Apr 28, 2006 at 12:05:12AM CEST:
> Thanks for catching that.  The change looks fine, except:
> 
> >      esac
> > -)dnl
> > +) | sort dnl
> 
> Please put the "| sort" inside the parentheses, instead of outside,

Indeed.  Thanks for the review!  Applied like this:

Cheers,
Ralf

        * doc/autoconf.texi (Limitations of Builtins): Document FreeBSD
        /bin/sh set unsorted output.
        * lib/autoconf/general.m4 (_AC_CACHE_DUMP): Adjust.
        * tests/local.at

Index: doc/autoconf.texi
===================================================================
RCS file: /cvsroot/autoconf/autoconf/doc/autoconf.texi,v
retrieving revision 1.1001
diff -u -r1.1001 autoconf.texi
--- doc/autoconf.texi   26 Apr 2006 20:30:55 -0000      1.1001
+++ doc/autoconf.texi   28 Apr 2006 04:16:32 -0000
@@ -12165,6 +12165,9 @@
 set -ex
 @end example
 
+The @command{set} of the address@hidden 6.0 shell does not sort its
+output.
+
 
 @item @command{shift}
 @c ------------------
Index: lib/autoconf/general.m4
===================================================================
RCS file: /cvsroot/autoconf/autoconf/lib/autoconf/general.m4,v
retrieving revision 1.912
diff -u -r1.912 general.m4
--- lib/autoconf/general.m4     24 Apr 2006 19:36:02 -0000      1.912
+++ lib/autoconf/general.m4     28 Apr 2006 04:16:32 -0000
@@ -1852,7 +1852,8 @@
       # `set' quotes correctly as required by POSIX, so do not add quotes.
       sed -n ["/^[_$as_cr_alnum]*_cv_[_$as_cr_alnum]*=/p"]
       ;;
-    esac
+    esac |
+    sort
 )dnl
 ])# _AC_CACHE_DUMP
 
Index: tests/local.at
===================================================================
RCS file: /cvsroot/autoconf/autoconf/tests/local.at,v
retrieving revision 1.29
diff -u -r1.29 local.at
--- tests/local.at      19 Apr 2006 04:13:48 -0000      1.29
+++ tests/local.at      28 Apr 2006 04:16:32 -0000
@@ -181,7 +181,7 @@
 # leaves temporary files.  AT_CONFIG_CMP uses the variable dumps to confirm 
that
 # tests have the same side effects regardless of caching.
 m4_defun([AC_STATE_SAVE],
-[(set) >state-env.$][1 2>&1
+[(set) 2>&1 | sort >state-env.$][1
 ls -1 | sed '/^at-/d;/^state-/d;/^config\./d' | sort >state-ls.$][1
 ])# AC_STATE_SAVE
 ]])