This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: A completely different approach to EH runtime
- To: kenner at vlsi1 dot ultra dot nyu dot edu (Richard Kenner)
- Subject: Re: A completely different approach to EH runtime
- From: Andreas Jaeger <aj at suse dot de>
- Date: 21 Feb 2001 21:35:11 +0100
- Cc: hjl at valinux dot com, gcc at gcc dot gnu dot org
- References: <10102211959.AA00575@vlsi1.ultra.nyu.edu>
kenner@vlsi1.ultra.nyu.edu (Richard Kenner) writes:
> Let me repeat. Any solution for Linux has to be worked out between
> glibc and gcc, not by gcc alone.
>
> How can there *possibly* be any interaction between glibc and gcc? The vast
> majority of systems on which GCC is run don't even *use* glibc.
Glibc is compiled by GCC. The libs are linked with libgcc_s.so.0. As noted before:
$ ldd math/libm.so.6
libc.so.6 => /lib/libc.so.6 (0x4001c000)
libgcc_s.so.0 => /opt/gcc-2.97.test/lib/libgcc_s.so.0 (0x4012f000)
/usr/src/test/glibc-2.2/lib/ld-linux.so.2 => /usr/src/test/glibc-2.2/lib/ld-linux.so.2 (0x40000000)
This causes major problems if libgcc_s is not on the same partition as
libc.so.6.
> The routines in libgcc are meant to be called *only* by calls inserted
> by GCC. Any other usage is incorrect and should be fixed.
Please reread the thread, the whole situation is rather complicated.
Andreas
--
Andreas Jaeger
SuSE Labs aj@suse.de
private aj@arthur.inka.de
http://www.suse.de/~aj