Re: A PPC patch for glibc 2.1

[Geoff Keating <geoffk@ozemail.com.au>]

> From: hjl@lucon.org (H.J. Lu)
> Date: Wed, 2 Dec 1998 22:24:44 -0800 (PST)
> Cc: libc-hacker@cygnus.com (GNU C Library)
> 
> Hi,
> 
> This patch is needed by my upcoming gdb 4.17.0.6.

What's wrong with the existing files?  The following works fine for
me in 2.0.104:

[geoffk@geoffk geoffk]$ cat > x.c
#include <sys/types.h>
#include <sys/user.h>
#include <sys/procfs.h>
[geoffk@geoffk geoffk]$ gcc x.c -c
[geoffk@geoffk geoffk]$ 

Your sys/procfs.h seems to be nearly identical to the existing one, it
just uses the sys/ header files not the asm/ ones.  If this is how
it's supposed to work, it would be better to change the generic
version.  Your sys/user.h seems to be effectively identical to the
existing one.

> --- /dev/null	Tue May  5 13:32:27 1998
> +++ sysdeps/unix/sysv/linux/powerpc/sys/elf.h	Wed Dec  2 21:34:45 1998

> +#define ELF_EXEC_PAGESIZE	4096

Shouldn't this be 64k?  Or is it supposed to vary with CPU type?

> +/* This is the location that an ET_DYN program is loaded if exec'ed.  Typical
> +   use of this is to invoke "./ld.so someprog" to test out a new version of
> +   the loader.  We need to make sure that it is out of the way of the program
> +   that it will "exec", and that there is sufficient room for the brk.  */
> +
> +#define ELF_ET_DYN_BASE         (0x08000000)

This varies by kernel version.  You must test it at runtime if you
need it.  On my machine it is 0x2aaab000.  On some previous linux-ppc
versions it was 0x30000000.  It may change again.  Really, ld.so
should relocate itself if it detects a conflict.

Perhaps it would be better if you only defined the things you really
needed.  But I can't see anything at all in this file that you should
need!  They are all either stored in the core file, or don't seem to
be particularly relevant to a debugger.  What have I missed?

-- 
Geoffrey Keating <geoffk@ozemail.com.au>