> Andrew Cagney wrote:
> 
> 
>> Using this patch, current GCC can at least configure and build 
>> (.../configure ; gmake all-gdb).  It barfs while trying to cross compile 
>> some of the libraries vis:
>> 
>> [ libiberty/strerror.c ]
> 
> 
> Note that we have the following change in the NetBSD tree (this is
> against binutils 2.11 but should be pretty much the same for gcc's
> libiberty):



FYI,

It was a red herring.

The compile barf was caused by a misconfigured multi-lib tree - it had 
incorrectly guessed that the host didn't have sys_errlist[].

That in turn was caused by the just compiled GCC not having -D__ELF__ in 
its list of definitions :-/

> Note that we have the following change in the NetBSD tree (this is
> against binutils 2.11 but should be pretty much the same for gcc's
> libiberty):


They should be identical.  GCC's libiberty is the master.  Any changes 
are double committed to the binutils/gdb libiberty.

> Index: strerror.c
> ===================================================================
> RCS file: /cvsroot/gnusrc/gnu/dist/toolchain/libiberty/strerror.c,v
> retrieving revision 1.1.1.2
> retrieving revision 1.2
> diff -d -p -u -r1.1.1.2 -r1.2
> --- strerror.c	2001/04/23 12:28:30	1.1.1.2
> +++ strerror.c	2001/05/30 09:51:10	1.2
> @@ -463,7 +463,7 @@ static int num_error_names = 0;
>  static int sys_nerr;
>  static const char **sys_errlist;
>  
> -#else
> +#elif !defined(__NetBSD__)
>  
>  extern int sys_nerr;
>  extern char *sys_errlist[];
> 
> 
> I suspect this patch would break very old versions of NetBSD though.


It wouldn't be accepted either :-)  To be honest, looking at strerror() 
I'm thinking the code has a bit of an attitude problem.  It insists on 
using the name sys_errlist, even doing this results in the diging of an 
ever deeper hole .... :-)

	Andrew




	Andrew