On Wed, 5 Nov 2003, David Xu wrote:
Jun Su wrote:
I noticed that Jeff Roberson implement this already. Is whi will be commit?
http://kerneltrap.org/node/view/1531
I google this because I found this feature is listed in the list of Kernel
Improvement of WindowsXP. :-)
Thanks,
Jun Su
I have almost done this experiment about 10 months ago.
http://people.freebsd.org/~davidxu/fastsyscall/
The patch is out of date and still not complete.
Also it can give you some performance improve, but I think too many
things need to be changed,
and this really makes user ret code very dirty, some syscalls, for
example, pipe() can not use
this fast syscall, becaues pipe() seems using two registers to return
file handle, the performance gain
is immediately lost when the assemble code becomes more complex. I don't
think this hack is worth
to do on IA32, I heard AMD has different way to support fast syscall,
that may already in FreeBSD
AMD 64 branch.
This works with every syscall. I have a patch in perforce that doesn't
require any changes to userret(). The performance gain is not so
substantial for most things but I feel that it is worth it. Mini is
probably going to finish this up over the next week or so.
Cheers,
Jeff
David Xu
___
[EMAIL PROTECTED] mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to [EMAIL PROTECTED]
___
[EMAIL PROTECTED] mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to [EMAIL PROTECTED]