[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Problems fork code



Wolfgang Lux writes:
> Juergen Lock wrote:
> >  [...]
> > 
> >  but what is definitely broken now is vfork().  at least for me it is...
> > 
> 
> Hmmm, maybe I've missed something.  Could you perhaps tell me what's
> wrong with vfork() now?

Hmm so vfork() still works for you?  i have a cron and (older) ksh that
spawn childs using vfork/exec and they both now get bus/address errors
somewhere in/after the vfork() (address apparently always < 0x50 so i'd
guess a bad sp or overwritten stack...)  cron i think gets this every time
(it fork()s before so the grandparent keeps running), ksh only if it
fork()d a subshell and actually did run the (fork) parent and child in
parallel i.e. when doing something like  (ls -l; ls -l) |cat  with more
data than what fits in a pipe.

 (this is partly from memory and without having attached a gdb or anything
so take that description with a grain of salt...  i can do `real' tests
if your patient, or just send you the cron source if you want to look
for yourself.)

 oh and both are -mbaserel, could that make a difference?

 hope this helps
	Juergen