Subject: Re: bind9 on -current/sparc (pthread problem?)
To: Christian Limpach <chris@pin.lu>
From: Martin Husemann <martin@duskware.de>
List: current-users
Date: 01/11/2004 16:38:02
On Sun, Jan 11, 2004 at 04:15:08PM +0100, Christian Limpach wrote:
> we use: (pthread__lockprim_init in pthread_lock.c)
>  if (ncpu == 1 && rasctl((void *)pthread__lock_ras_start,
>      (size_t)((uintptr_t)pthread__lock_ras_end -
>               (uintptr_t)pthread__lock_ras_start),
>      RAS_INSTALL) == 0) {
> to decide which locking operations to use.

Yes, and rasctl() on sparc always returns EOPNOTSUP, since __HAVE_RAS is
not defined when compiling the kernel.

So either gdb is confused or the stack is smashed?

Martin