Minor vfork patch

Nick Burrett nick at dsvr.net
Fri Dec 5 01:22:22 PST 2003



David Marston wrote:
> It doesn't seem to be related to the current issue, but there's a patch
> for vfork attached to fix the position of the child process pid stored
> in the jmp_buf.
> 
> On a different subject I was looking at the other set/longjmp code. Can
> anyone tell me if ___arm_alloca_longjmp and ___arm_alloca_setjmp in
> gcc/libgcc/lib1aof.s are still needed. They don't appear to be called
> anywhere anymore (since version 1.5 of gcc/gcc/config/arm/arm.h a year
> ago) unless I'm missing something.

Although they appear not to be used at the moment, they may come back 
into action in the future.

They relate to exception handling and the freeing of alloca chunks when 
we longjmp to higher level functions.  They're not currently used 
because I was experimenting with using GCC's own 
__builtin_{setjmp,longjmp} implementations and DWARF2 frame unwinding 
for exception handling.

-- 
Nick Burrett
Network Engineer, Designer Servers Ltd.   http://www.dsvr.co.uk





More information about the gcc mailing list