This is the mail archive of the
libc-alpha@sources.redhat.com
mailing list for the glibc project.
Re: __libc_ia64_register_backing_store_base
- To: Ulrich Drepper <drepper at cygnus dot com>
- Subject: Re: __libc_ia64_register_backing_store_base
- From: Jeff Sturm <jsturm at one-point dot com>
- Date: Wed, 29 Aug 2001 18:51:53 -0400 (EDT)
- cc: rth at cygnus dot com, java-patches at gcc dot gnu dot org, libc-alpha at sourceware dot cygnus dot com
On 29 Aug 2001, Ulrich Drepper wrote:
> rth@cygnus.com writes:
> > A temporary workaround for bohem-gc is attached, but I think the
> > proper solution is to move the init of this variable into libc.
> > Specifically, into _dl_start_user for ld.so, and __libc_init_first
> > for the static libc. Uli, this seem right to you?
>
> Does this mean the GC is starting to do some work before the
> constructors are run? This seems just wrong. If not all constructors
> have finished their work the system is not yet in a consistent state
> and things like this can happen. Note that this variable is set
> before the applications constructors are run so the latter are fine.
There has been at least one proposal to kill static initializers in
libjava:
http://gcc.gnu.org/ml/java/2001-05/msg00046.html
The patch defers GC startup until after all static constructors, but it
was never installed. IIRC the concern was thread safety in class
registration.
Jeff