Re: STREAMS: interface versus implementation

Alan Cox (alan@lxorguk.ukuu.org.uk)
Wed, 16 Sep 1998 20:11:58 +0100 (BST)


> > 3. The "severe overhead argument" is complete shit. You are
> > adding
> >
> > if(*(u32 *)pathname='/dev')
> >
> > to most opens. 2 clocks at most with branch prediction in
>
> I see an alignment trap too, but anyway...

Not on X86. Also statistically speaking most strings are naturally aligned.
Analyse the calls to open() and you'll see that pretty fast.

> The kernel version is free. We already have to handle device
> operations and ioctls in the kernel. Libc would need an extra check,
> while the kernel would not.

The kernel version is not free. It uses kernel memory it needs maintaining

> performance reasons. The device file and ioctls that were requested
> would help STREAMS apps without messing up the network stack.

And they dont need to be in kernel space, the dont need to be in glibc
proper, they can be completely isolated, and nobody will measure the
performance difference. It'll work with kernels prior to the idea even
coming up.

Alan

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/