netdev - Re: [PATCH RFX]: napi

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

Message-Id: <1185340196.1803.482.camel@localhost.localdomain>
Date:	Wed, 25 Jul 2007 15:09:55 +1000
From:	Rusty Russell <rusty@...tcorp.com.au>
To:	David Miller <davem@...emloft.net>
Cc:	netdev@...r.kernel.org, shemminger@...ux-foundation.org,
	jgarzik@...ox.com, hadi@...erus.ca
Subject: Re: [PATCH RFX]: napi_struct V3

On Tue, 2007-07-24 at 21:29 -0700, David Miller wrote:
> From: Rusty Russell <rusty@...tcorp.com.au>
> Date: Wed, 25 Jul 2007 12:33:14 +1000
> 
> > Maybe by adding YA state bit?  Hold on, this might get ugly...
> > 
> > Say netif_rx_schedule_prep() sets the MORE_TODO bit (atomically instead
> > of setting __LINK_STATE_RX_SCHED) if it's going to fail, and
> > netif_rx_complete() returns 0 if it was set, or 1 if it's OK.  Now
> > callers do:
> > 
> > 	reenable_interrupts();
> >         if (rx_pending() || !netif_rx_complete(netdev, napi))
> >                 disable_interrupts();
> 
> This is an interesting idea, and would work, but the extra atomics
> for something that 2 or 3 drivers actually need.... unless I
> misunderstand your suggestion?

Well, netif_rx_schedule_prep() becomes a cmpxchg instead of a
test_and_set_bit:

	/* We either set MORE_TODO or RX_SCHED. */
	unsigned long state;

again:
	state = napi->state;
	if (!(state & __LINK_STATE_RX_SCHED)) {
		if (cmpxchg(&napi->state, state, state|__LINK_STATE_RX_SCHED) == state)
			return 1;
	} else {
		if (cmpxchg(&napi->state, state, state|__LINK_STATE_MORE_TODO) == state)
			return 0;
	}
	goto again;

netif_rx_complete would need something more complex... hmm... 

	unsigned long state;
again:
	state = napi->state;

	if (state & __LINK_STATE_MORE_TODO)
		return 0;

	list_del(&napi->poll_list);
	if (cmpxchg(&napi->state, state, state & ~__LINK_STATE_RX_SCHED) == state)
		return 1;
	list_add_tail(napi->poll_list, &__get_cpu_var(softnet_data).poll_list);
	goto again;

Not pretty 8(

Rusty.


	if (

	




-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html