linux-kernel - Re: [PATCH] VIOC: New Network Device Driver

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-Id: <200609112115.53520.arnd@arndb.de>
Date:	Mon, 11 Sep 2006 21:15:52 +0200
From:	Arnd Bergmann <arnd@...db.de>
To:	"Misha Tomushev" <misha@...ric7.com>
Cc:	jgarzik@...ox.com, netdev@...r.kernel.org,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH] VIOC: New Network Device Driver

On Monday 11 September 2006 19:58, Misha Tomushev wrote:
> The descriptor clean-up does not contribute anything to the performance
> of the driver, it just replenishes the memory pools. It almost does not
> need interrupts. Why would we want to add more cycles to the receive
> logic, when driver is doing useful work for something that can run
> almost at any time?

It can run at almost any time, just not in the interrupt context,
where it needs to schedule the tx softirq first.

Also, the number of tx interrupts you get is a tradeoff between
causing overhead of calling the interrupt handler and stalling
sockets that are waiting for space to become free after transmission.
By using ->poll for it, you can avoid tx interrupts completely
most of the time and free skbs immediately when data comes in.

	Arnd <><
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/