On Fri, 2006-04-21 at 04:05 +0200, Andi Kleen wrote:
> Piet Delaney <[email protected]> writes:
> >
> > FreeBSD folks developed a ZERO_COPY_SOCKET facility that uses COW;
> > code looked great.
>
> Linux had patches many years ago (in 2.3.x), but it was never merged
> because it is inherently unscalable on MP. Classical BSD sockets really
> don't work well for zero copy - you need a new interface (like POSIX aio)
> that allows the kernel/user to tell each other when use of data is
> finished and buffers can be reused.
Right, back when I was working on zero copy for 2.4 I noticed that
2.6 seemed to support aio in the socket code, passing the kiocb
pointer as I recall, and support in the socket code for for sendpage
seemed enhanced. I was also wondering about using 2.6 and aio for zero
copy instead of tokens via sendmsg() and recvmsg() cmsghdr structures.
-piet
>
> -Andi
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
--
---
[email protected]
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
[Index of Archives]
[Kernel Newbies]
[Netfilter]
[Bugtraq]
[Photo]
[Stuff]
[Gimp]
[Yosemite News]
[MIPS Linux]
[ARM Linux]
[Linux Security]
[Linux RAID]
[Video 4 Linux]
[Linux for the blind]
[Linux Resources]
