Дилян Палаузов <dilyan.palauzov@xxxxxxxxx> writes:
> diff -u git-1.7.9.orig/config.mak.in git-1.7.9/config.mak.in
> --- git-1.7.9.orig/config.mak.in 2012-01-27 20:51:04.000000000 +0000
> +++ git-1.7.9/config.mak.in 2012-02-12 00:52:41.457968080 +0000
> @@ -74,3 +74,4 @@
> NO_PTHREADS=@NO_PTHREADS@
> PTHREAD_CFLAGS=@PTHREAD_CFLAGS@
> PTHREAD_LIBS=@PTHREAD_LIBS@
> +LINK_CHARSET=@LINK_CHARSET@
> diff -u git-1.7.9.orig/configure.ac git-1.7.9/configure.ac
> --- git-1.7.9.orig/configure.ac 2012-01-27 20:51:04.000000000 +0000
> +++ git-1.7.9/configure.ac 2012-02-12 00:44:29.222967868 +0000
> @@ -836,6 +836,18 @@
> [HAVE_LIBCHARSET_H=YesPlease],
> [HAVE_LIBCHARSET_H=])
> AC_SUBST(HAVE_LIBCHARSET_H)
> +# Define LINK_LIBCHARSET if libiconv does not export the
Because the use of configure is optional in our build infrastructure, I
wouldn't have objected if this comment were missing from configure.ac, but
the new variable *must* be described in Makefile (see the top 250 lines or
so of that file).
I also need to point out that LINK_LIBCHARSET does not sit very well with
the way how existing Makefile variables are named. Perhaps make the new
variable contain the necessary string ("-lcharset" in your case), and name
it CHARSET_LIB or something? By doing so, when we find a platform that
has the necessary locale_charset() not in libcharset.{a,so} but somewhere
else, e.g. libxyzzy.a, we can accomodate it with "CHARSET_LIB = -lxyzzy".
Thanks. Also as Ævar pointed out, please do not forget to sign off your
patch.
--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html