netdev - Re: [STRAW MAN PATCH] sch_teql doesn't load-balance ppp(oatm) slaves

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: <1332755149.2379.52.camel@shinybook.infradead.org>
Date:	Mon, 26 Mar 2012 10:45:50 +0100
From:	David Woodhouse <dwmw2@...radead.org>
To:	David Miller <davem@...emloft.net>
Cc:	netdev@...r.kernel.org
Subject: Re: [STRAW MAN PATCH] sch_teql doesn't load-balance ppp(oatm)
 slaves

On Mon, 2012-03-26 at 09:32 +0100, David Woodhouse wrote:
> It does indeed stop the queue. I think it then wakes it right back up
> again in ppp_xmit_process(), *before* returning NETDEV_TX_OK. So the
> offending calls to skb_dequeue() which are putting it back to the front
> of the list are going to be from the softirq trying to feed the device.

A WARN_ON() in teql_dequeue() confirms that.

> I'll confirm that, then try fixing the PPP code so it doesn't stop and
> immediately restart the queue. If it only stops the queue
> *conditionally*, that may well fix the problem. 

Yes, this gives me my full upload bandwidth by spreading packets over
both lines correctly, without touching sch_teql.c.

It returns a value from ppp_xmit_process() which indicates whether the
queue should be stopped. It doesn't actually *remove* the call to
netif_wake_queue(), because other code paths (especially from
ppp_output_wakeup()) still need it to be there. But if the queue wasn't
stopped anyway, it's a no-op. So we just make ppp_start_xmit() stop the
queue based on the return value of ppp_xmit_process(), and it all works
nicely without any of the gratuitous softirq invocations which are a
waste of time in the normal case, and actively harmful with teql.

--- drivers/net/ppp/ppp_generic.c~	2012-01-26 00:39:32.000000000 +0000
+++ drivers/net/ppp/ppp_generic.c	2012-03-26 10:32:31.286744147 +0100
@@ -235,7 +235,7 @@ struct ppp_net {
 /* Prototypes. */
 static int ppp_unattached_ioctl(struct net *net, struct ppp_file *pf,
 			struct file *file, unsigned int cmd, unsigned long arg);
-static void ppp_xmit_process(struct ppp *ppp);
+static int ppp_xmit_process(struct ppp *ppp);
 static void ppp_send_frame(struct ppp *ppp, struct sk_buff *skb);
 static void ppp_push(struct ppp *ppp);
 static void ppp_channel_push(struct channel *pch);
@@ -968,9 +968,9 @@ ppp_start_xmit(struct sk_buff *skb, stru
 	proto = npindex_to_proto[npi];
 	put_unaligned_be16(proto, pp);
 
-	netif_stop_queue(dev);
 	skb_queue_tail(&ppp->file.xq, skb);
-	ppp_xmit_process(ppp);
+	if (!ppp_xmit_process(ppp))
+		netif_stop_queue(dev);
 	return NETDEV_TX_OK;
 
  outf:
@@ -1048,10 +1048,11 @@ static void ppp_setup(struct net_device
  * Called to do any work queued up on the transmit side
  * that can now be done.
  */
-static void
+static int
 ppp_xmit_process(struct ppp *ppp)
 {
 	struct sk_buff *skb;
+	int ret = 0;
 
 	ppp_xmit_lock(ppp);
 	if (!ppp->closing) {
@@ -1061,10 +1062,13 @@ ppp_xmit_process(struct ppp *ppp)
 			ppp_send_frame(ppp, skb);
 		/* If there's no work left to do, tell the core net
 		   code that we can accept some more. */
-		if (!ppp->xmit_pending && !skb_peek(&ppp->file.xq))
+		if (!ppp->xmit_pending && !skb_peek(&ppp->file.xq)) {
 			netif_wake_queue(ppp->dev);
+			ret = 1;
+		}
 	}
 	ppp_xmit_unlock(ppp);
+	return ret;
 }
 
 static inline struct sk_buff *


-- 
dwmw2

Download attachment "smime.p7s" of type "application/x-pkcs7-signature" (5818 bytes)