Re: GDB: Does "catch vfork" work?
Re: GDB: Does "catch vfork" work?
- Subject: Re: GDB: Does "catch vfork" work?
- From: "Justin C. Walker" <email@hidden>
- Date: Sun, 9 Oct 2005 08:07:47 -0700
My 0.016 Euros:
On Oct 9, 2005, at 06:51 , Jim Correia wrote:
On Oct 8, 2005, at 6:55 AM, Damien Bobillot wrote:
If you use fork, the whole parent memory is copied.
There is no reason the kernel could't (or doesn't) use copy on
write semantics for fork, AFAIK. Whether or not it actually does it
is another question.
It does use copy-on-write semantics. In fact, for reasons that
escape me, from NextStep on, COW was "architectural", rather than an
implementation detail.
If you use vfork, the memory is lazily copied : the system wait
until it really need to make the copy, ie if both processes don't
write anything in a given memory page, the system won't make the
copy, but if one do write something, the system make the copy just
before effectively writing things. This use the copy-on-write flag
of the paged memory manager.
The only difference between fork and vfork is that vfork is more
efficient : for the developer, even when debugging, both functions
seem to work the same way.
Again, I'm not a complete expert on the topic, but my understanding
is that vfork does not use copy on write semantics, but rather
shares the parent's address space until the exec. If you do
anything after the fork and before the exec you are making changes
in the parent's address space (and they are probably unexpected
changes from the parent's POV.) My experience is consistent with
this. (In other words, I believe the man page to be correct.)
This is the difference between fork() and vfork(): in the former, the
child runs in a copy of the parent's address space (whether the copy
is ever completed is a detail); in the latter, the child uses the
parent's address space.
This was a "hack" made at a time when copy-on-write had either not
been invented, or was not possible on much of the hardware (i.e.,
pdp-11's) in use. The intent was to speed up the typical fork()/exec
() sequence by avoiding the copy of the process space.
Regards,
Justin
--
Justin C. Walker, Curmudgeon-At-Large
Institute for General Semantics
--------
Men are from Earth.
Women are from Earth.
Deal with it.
--------
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Xcode-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden