This is the mail archive of the
binutils@sources.redhat.com
mailing list for the binutils project.
Re: The ia64 linker problems.
- To: "H . J . Lu" <hjl at valinux dot com>
- Subject: Re: The ia64 linker problems.
- From: Jim Wilson <wilson at cygnus dot com>
- Date: Tue, 06 Feb 2001 16:51:27 -0800
- cc: Richard Henderson <rth at redhat dot com>, binutils at sourceware dot cygnus dot com
I noticed that you are adding your patches to binutils development sources,
but not to the binutils 2.11 sources. Is this intentional?
> mov.sptk b0 = r2, shlib_overriddencall2#
The address is just a hint for the branch predictor. It is OK for the address
to be wrong. The program will still work correctly, it will just work a little
slower if the hint is wrong.
Currently, though, gcc doesn't make use of these hints as we have no easy
way to specify them. If we did use them, I doubt that we would be using
global symbol addresses for the hints, unless something very unusual was
going on.
> * * config/tc-ia64.h (TC_RELOC_RTSYM_LOC_FIXUP): Do fixup if
> there is no relocation.
This looks OK to me.
Jim