netdev
[Top] [All Lists]

Re: [PATCH] skb_{over,under}_panic optimizations

To: Ben LaHaise <bcrl@xxxxxxxxxx>
Subject: Re: [PATCH] skb_{over,under}_panic optimizations
From: "David S. Miller" <davem@xxxxxxxxxx>
Date: Mon, 16 Jul 2001 15:36:46 -0700 (PDT)
Cc: <netdev@xxxxxxxxxxx>
In-reply-to: <Pine.LNX.4.33.0107152228490.16774-100000@xxxxxxxxxxxxxxxxxxxxxxxxxx>
References: <Pine.LNX.4.33.0107152228490.16774-100000@xxxxxxxxxxxxxxxxxxxxxxxxxx>
Sender: owner-netdev@xxxxxxxxxxx
Ben LaHaise writes:
 > Below is a patch that moves the skb_{over,under}_panic function calls in
 > skb_put and skb_push out of line on i386 and hopefully improves register
 > allocation.  If this turns out to be useful, there are bunch of other
 > places we can move these kinds useful debugging assertions out of line.
 > Comments?

I can't see how this can make all that much of a difference, as Andi
has stated already.  Especially on x86.

I mean, GCC is going to generate a branch around the "call
skb_foo_panic" code block, and that code block will likely
sit in (at least mostly) it's own cache line.

x86 code is so dense anyways, so this change can hardly make a
difference.

I'm willing to, and would happily be, proven wrong.  Show me
some numbers.

Later,
David S. Miller
davem@xxxxxxxxxx

<Prev in Thread] Current Thread [Next in Thread>