Re: fix in --help output
От | Zdenek Kotala |
---|---|
Тема | Re: fix in --help output |
Дата | |
Msg-id | 47BF3868.5070405@sun.com обсуждение исходный текст |
Ответ на | Re: fix in --help output (Zdenek Kotala <Zdenek.Kotala@Sun.COM>) |
Ответы |
Re: fix in --help output
|
Список | pgsql-patches |
Ups. Patch attached Zdenek Zdenek Kotala napsal(a): > Tom Lane napsal(a): >> Zdenek Kotala <Zdenek.Kotala@Sun.COM> writes: >>> I attached patch which replaces any "--..." occurrence with "-c..." >>> on command line. >> >> Please see whether forcibly using src/port/getopt.c fixes this, >> instead. A saner patch would probably add something like this >> to configure.in: >> >> if test "$PORTNAME" = "solaris"; then >> AC_LIBOBJ(getopt) >> AC_LIBOBJ(getopt_long) >> fi > > Yeah, this is the best solution. I attach a patch. Is possible to > backport it back to 8.3, 8.2? > > Just for completeness Solaris getopt function has a extension which > processes long option argument. This extension is called CLIP. This > implementation collides with unusual getopt usage for long option > processing in PostgreSQL. > > More info are there: > http://docs.sun.com/app/docs/doc/819-2243/getopt-3c?l=en&a=view > http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/lib/libc/port/gen/getopt.c > > http://docs.sun.com/app/docs/doc/819-2239/6n4hsf6e5?l=en&a=view > > By the way old solaris getopt version is similary with postgres > implementation. :-) > > http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/lib/libbc/libc/gen/common/getopt.c > > > > Thanks for help > Zdenek > > ---------------------------(end of broadcast)--------------------------- > TIP 2: Don't 'kill -9' the postmaster Index: configure.in =================================================================== RCS file: /zfs_data/cvs_pgsql/cvsroot/pgsql/configure.in,v retrieving revision 1.551 diff -c -r1.551 configure.in *** configure.in 19 Feb 2008 01:05:28 -0000 1.551 --- configure.in 22 Feb 2008 20:16:05 -0000 *************** *** 1102,1107 **** --- 1102,1115 ---- AC_LIBOBJ(getopt_long) fi + # Solaris has getopt implementation which is not compatible with unusual + # parsing for long options used in PostgreSQL. Replace it by own implementation + # is best solution how to avoid complication. + if test "$PORTNAME" = "solaris"; then + AC_LIBOBJ(getopt) + AC_LIBOBJ(getopt_long) + fi + # Win32 support if test "$PORTNAME" = "win32"; then AC_REPLACE_FUNCS(gettimeofday)
В списке pgsql-patches по дате отправления: