Red Hat Bugzilla – Bug 32429
rpm 4.0.2 compile problem on Solaris 8 with vendor C compiler
Last modified: 2008-03-13 15:18:51 EDT
From Bugzilla Helper:
User-Agent: Mozilla/4.76 [en] (X11; U; OSF1 V5.1 alpha)
while compiling lib/install.c, the Sun Forte 6.1 C compiler flags the
"../rpmio/rpmio_internal.h", line 8: identifier redeclared: fdFileno
current : static function(pointer to void) returning int
previous: function(pointer to void) returning int : "../rpmio/rpmio.h",
Steps to Reproduce:
1. ./configure; make
on sparc-sun-solaris2.8 with the full development environment.
Since it's declared static in the _internal header but then declared
static qualifier in the public `rpmio.h' header, I just applied the
--- rpm-4.0.2.orig/rpmio/rpmio_internal.h Wed Jan 10 16:26:34 2001
+++ rpm-4.0.2/rpmio/rpmio_internal.h Mon Mar 19 17:50:39 2001
@@ -5,7 +5,7 @@
* \file rpmio/rpmio_internal.h
-static inline int fdFileno(void * cookie);
+/* static inline int fdFileno(void * cookie); */
Created attachment 13163 [details]
patch to comment out static declaration
this is not a wolverine / Red Hat Linux bug.
The patch "works" but goes the wrong way. I wish to get fdFoo out of the API
using instead the fdio vectors. I've had to retrofit several routines in the
order to control for API creep, but, as soon as I can manage a soname change,
fdFileno() et al will be accessible only through fdio vectors.
Deferred until then.
> "this is not a wolverine / Red Hat Linux bug."
Is this a message for Red Hat internal use or does this mean that the inital
shouldn't have been posted here? If so on the RPM home page I read:
Bugzilla, bug tracker for Red Hat, including RPM
so you should ask the RPM maintainers to fix their home page.
Jeff Johnson has asked that all rpm-related problems be reported via Bugzilla.
The fact is that RPM has gone beyond being a RedHat-only or Linux-only thing to
an OpenSource package management system that's used on a wide variety of UNIX
Since there are no options for "other" in the OS/platform area, I just chose an
option that seemed
to be the "least wrong".
You're of course right that this isn't a RedHat or even Linux issue, but it is
an RPM quality issue.