netdev - Re: [3.5 regression / mcs7830 / bisected] bridge constantly toggeling between disabled and forwarding

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: <20121011112649.7d50d605@xenia.leun.net>
Date:	Thu, 11 Oct 2012 11:26:49 +0200
From:	Michael Leun <lkml20120218@...ton.leun.net>
To:	Ondrej Zary <linux@...nbow-software.org>
Cc:	Greg KH <gregkh@...uxfoundation.org>, davem@...emloft.net,
	netdev@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [3.5 regression / mcs7830 / bisected] bridge constantly
 toggeling between disabled and forwarding

On Thu, 11 Oct 2012 10:14:14 +0200
Ondrej Zary <linux@...nbow-software.org> wrote:

> On Tuesday 09 October 2012, Michael Leun wrote:
> > On Tue, 9 Oct 2012 09:21:31 +0200
> >
> > Ondrej Zary <linux@...nbow-software.org> wrote:
> > > On Tuesday 09 October 2012, Michael Leun wrote:
> > > > On Thu, 27 Sep 2012 10:39:05 -0700
> > > >
> > > > Greg KH <gregkh@...uxfoundation.org> wrote:
> > > > > On Tue, Jul 24, 2012 at 01:36:34AM +0200, Michael Leun wrote:
> > > > > > On Mon, 23 Jul 2012 09:15:04 +0200
> > > > > > Michael Leun <lkml20120218@...ton.leun.net> wrote:
> > > > > >
> > > > > > [see issue description below]
> > > > > >
> > > > > > Bisecting yielded
> > > > > >
> > > > > > b1ff4f96fd1c63890d78d8939c6e0f2b44ce3113 is the first bad
> > > > > > commit commit b1ff4f96fd1c63890d78d8939c6e0f2b44ce3113
> > > > > > Author: Ondrej Zary <linux@...nbow-software.org>
> > > > > > Date:   Fri Jun 1 10:29:08 2012 +0000
> > > > > >
> > > > > >     mcs7830: Implement link state detection
> > > > > >
> > > > > >     Add .status callback that detects link state changes.
> > > > > >     Tested with MCS7832CV-AA chip (9710:7830, identified as
> > > > > > rev.C by the driver). Fixes
> > > > > > https://bugzilla.kernel.org/show_bug.cgi?id=28532
> > > > > >
> > > > > >     Signed-off-by: Ondrej Zary <linux@...nbow-software.org>
> > > > > >     Signed-off-by: David S. Miller <davem@...emloft.net>
> > > > > >
> > > > > > :040000 040000 5480780cb5e75c57122a621fc3bab0108c16be27
> > > > > >
> > > > > > d97efd9cc0a465dff76bcd3a3c547f718f2a5345 M    drivers
> > > > > >
> > > > > >
> > > > > > Reverting that from 3.5 makes the issue go away.
> > > > >
> > > > > Did this ever get resolved in 3.6-rc7 or any older kernel?  I
> > > > > can't revert the patch from 3.5.y unless it's also fixed in
> > > > > Linus's tree.
> > > >
> > > > Please excuse me for answering a bit late.
> > > >
> > > > No, that never got resolved, I still have the problem with 3.6
> > > > but I'm not shure about the correct solution.
> > > >
> > > > Maybe link state detection just does not work with some of that
> > > > devices and we should have an possibility to enable/disable it
> > > > per device, maybe it can be handeled with an blacklist of not
> > > > working devices, maybe it could be fixed - I do not know and
> > > > also do not know how to find out.
> > > >
> > > > But I'm willing to test.
> > >
> > > Does the problem appear only in a bridge? Or the link state
> > > detection
> > >is completely wrong?
> >
> > As testing with patch below shows, it also happens without bridge,
> > but I did not notice so far, because that log message I was
> > complainig about originally of course comes from the bridge code.
> >
> > > Can you please apply this debug patch and provide the output?
> > >
> > > --- a/drivers/net/usb/mcs7830.c
> > > +++ b/drivers/net/usb/mcs7830.c
> > > @@ -638,6 +638,7 @@ static void mcs7830_status(struct usbnet *dev,
> > > struct urb *urb) return;
> > >
> > >  	link = !(buf[1] & 0x20);
> > > +	printk("netif_carrier_ok=%d, link=%d, buf[0]=0x%02x,
> > > buf[1]=0x%02x\n", netif_carrier_ok(dev->net), link, buf[0],
> > > buf[1]); if (netif_carrier_ok(dev->net) != link) { if (link) {
> > >  			netif_carrier_on(dev->net);
> >
> > As far as I've seen:
> >
> > Bring device up - link state toggles a few times and stabilizes to
> > up
> >
> > Start ping - link state toggles for every packet
> >
> > Test below done without bridge involved.
> >
> > Note: I renamed the interface to ue7 (I happen to have several of
> > that devices).
> 
> Please try this patch, it should fix the problem.
> (Remove the previous debug patch first.)
> 
> 
> mcs7830: Fix link state detection

For me this works. I've tested with the original bridged setup and do
not longer see that bridge port state changes.

I would pretty much appreciate to see this in mainline and then 3.5 and
3.6 stable ASAP.

Thank you!

-- 
MfG,

Michael Leun

--
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