This is the mail archive of the
libc-hacker@cygnus.com
mailing list for the glibc project.
Re: last call
- To: Philip Blundell <pb@nexus.co.uk>
- Subject: Re: last call
- From: Andreas Schwab <schwab@issan.informatik.uni-dortmund.de>
- Date: 29 Jan 1999 10:21:34 +0100
- Cc: drepper@cygnus.com (Ulrich Drepper), libc-hacker@cygnus.com
- References: <E105WS7-0008Kp-00@fountain.nexus.co.uk>
Philip Blundell <pb@nexus.co.uk> writes:
|> But I wonder if the idea of implementing vfork using clone() from user space
|> (which is what we do for ARM) is actually a good one. If the kernel is too
|> old the effect is presumably to give just clone(CLONE_VM) which would result
|> in both the parent and child running simultaneously on the same stack. Ouch.
|> Am I missing something here?
No, that's a fundamental problem, and that's the only reason why
linux/m68k will continue to use the vfork syscall, even though the m68k
does not have the stupid register problems as the ix86.
Andreas.