This is the mail archive of the
gdb@sources.redhat.com
mailing list for the GDB project.
Re: GDB 5.0.92 available
- To: gdb at sources dot redhat dot com, gdb at thewrittenword dot com
- Subject: Re: GDB 5.0.92 available
- From: Andrew Cagney <ac131313 at cygnus dot com>
- Date: Thu, 01 Nov 2001 06:35:05 -0500
- References: <3BDF7F79.6050205@cygnus.com> <20011101085219.A22826@oolong.il.thewrittenword.com>
Just FYI,
Just FYI, HP/UX is documented as not working on the 5.1 branch. See the
file README. I'll apply tweeks where applicable but won't delay 5.1
because HP/UX doesn't work.
> On Tue, Oct 30, 2001 at 11:35:05PM -0500, Andrew Cagney wrote:
>
>> GDB 9.0.92 is available for download. See:
>>
>> ftp://sources.redhat.com/pub/gdb/snapshots/
>
>
> config/mh-papic has:
> PICFLAG=-fPIC
>
> This is fine if you're building with GCC. On HP-UX though, the -fPIC
> equivalent is +Z. config/mh-sparcpic has:
> PICFLAG=`case '${LIBCFLAGS} ${LIBCXXFLAGS}' in *-fpic* ) \
> echo -fpic ;; * ) echo -fPIC ;; esac`
>
> Ditto for config/mt-papic.
>
> Should we do something like:
> PICFLAG=`case '${CC}' in cc) echo +Z ;; *) echo -fPIC ;; esac`
Hmm, I don't know who maintains that. src/MAINTAINERS says ``Ask''.
> Also, libiberty/regex.c has:
> /* Integer type for pointers. */
> # if !defined _LIBC
> typedef unsigned long int uintptr_t;
> # endif
>
> This conflicts with <sys/_inttypes.h> on HP-UX 10.20 and 11.00 where:
> typedef unsigned long uintptr_t;
>
> Should we just rename the typedef?
Liberty is maintained by gcc (gcc@gcc.gnu.org). Any change to liberty
needs to first be approved by that list. Only once something has been
approved for the trunk will I import it into the branch.
Andrew