We have a report (now entered as bug 575) that when open64 is given
the -print-file-name, -print-prog-name, and -print-multi-lib it is
calling the system GCC compiler to handle these instead of using the
GCC that is shipped with open64.
This patch fixes that problem by using the internal GCC instead
of the system one and thus gives a correct answer with something
like:
opencc -print-file-name=libstdc++.a
Before it would return the path to the system libstdc++, now it
will return the path to the one shipped with open64.
Can a Gatekeeper approve this patch for checkin?
Steve Ellcey
s...@cup.hp.com
Index: opt_actions.c
===================================================================
--- opt_actions.c (revision 3290)
+++ opt_actions.c (working copy)
@@ -1735,7 +1735,11 @@ print_file_path (char *fname, int exe)
/* not found, so ask gcc */
int m32 = check_for_saved_option("-m32");
char *argv[4];
- argv[0] = "gcc";
+ phases_t lang = (invoked_lang == L_CC) ? P_gcpp_plus : P_gcpp;
+ if (external_gcc == TRUE)
+ argv[0] = get_phase_name(lang);
+ else
+ argv[0] = get_full_phase_name (lang);
argv[1] = m32 ? "-m32" : "-m64";
asprintf(&argv[2], "-print-%s-name=%s", exe ? "prog" : "file", fname);
argv[3] = NULL;
@@ -1749,7 +1753,11 @@ void
print_multi_lib ()
{
char *argv[3];
- argv[0] = "gcc";
+ phases_t lang = (invoked_lang == L_CC) ? P_gcpp_plus : P_gcpp;
+ if (external_gcc == TRUE)
+ argv[0] = get_phase_name(lang);
+ else
+ argv[0] = get_full_phase_name (lang);
asprintf(&argv[1], "-print-multi-lib");
argv[2] = NULL;
/* MINGW doesn't support execvp, everyone supports execlp */
------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
Open64-devel mailing list
Open64-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/open64-devel
