[LEDE-DEV] revisited: Nanostation m5 XW ethernet patch gone

[Daniel Golle]

Hi!

I just witnessed the Ethernet issue returning on OpenWrt snapshot on
Ubiquiti NanoStation M5 XW (NSM5-XW):
ag71xx.0: connected to PHY at ag71xx-mdio.0:00 [uid=004dd043, driver=Atheros AR8216/AR8236/AR8316]
After about three days uptime eth0 links were lost and could be
recovered by resetting the device. The device is connected to another
ubnt XW device (NB NBM5) which also runs OpenWrt.
The logs didn't show anything suspicious. How should I debug the
situation next time it occurrs?

Cheers

Daniel



On Thu, Oct 19, 2017 at 02:03:35PM -0700, Joe Ayers wrote:
> Was the fix included in the Generic PHY?  If not, possible some of the
> following are still using the generic and therefore would still have
> the bug?:
> 
> NBE/PBE-M5-300 XW: ag71xx ag71xx.0: connected to PHY at
> ag71xx-mdio.0:01 [uid=004dd023, driver=Generic PHY]
> NSM5-loco-XW:  ag71xx ag71xx.0: connected to PHY at ag71xx-mdio.0:01
> [uid=004dd023, driver=Generic PHY]
> Airgrid M5 XW:  ag71xx ag71xx.0: connected to PHY at ag71xx-mdio.0:01
> [uid=004dd023, driver=Generic PHY]
> NBM5/16 XW:  ag71xx ag71xx.0: connected to PHY at ag71xx-mdio.0:01
> [uid=004dd023, driver=Generic PHY]
> NBM5/19 XW: ag71xx ag71xx.0: connected to PHY at ag71xx-mdio.0:01
> [uid=004dd023, driver=Generic PHY]
> 
> There was a bounty source on this defect -- how to get the original
> OpenWRT defect closed?
> 
> Regards,
> Joe AE6XE
> 
> 
> 
> 
> On Thu, Oct 19, 2017 at 12:58 PM, Felix Fietkau <nbd at nbd.name> wrote:
> >
> > Hi Daniel,
> >
> > The patch that you mentioned is not related to your issue at all, since
> > it only deals with the AT8032 PHY, which the NanoStation M does not
> > have. Maybe you can provide a more detailed description of what symptoms
> > you're seeing.
> >
> > I did indeed clean up the AT8032 mess and solved it in the PHY driver
> > (controlled by platform data) instead of adding GPIO toggle hackery to
> > the Ethernet driver.
> >
> > - Felix
> >
> > On 2017-10-19 20:20, Daniel Golle wrote:
> > > Hi Felix!
> > > Hi everyone!
> > >
> > > It looks like Tiziano is right and this patch did get lost somehow.
> > > At least I can't find any work-arounds regarding PHY hangs committed in
> > > neither target/linux/generic/files/* nor target/linux/ar71xx/files/*.
> > >
> > > The thread on lede-dev also seems to have stagnated after
> > > http://lists.infradead.org/pipermail/lede-dev/2016-December/004406.html
> > >
> > > Felix, you said you had a more clean patch for this in your staging
> > > tree?
> > >
> > > I'm asking because I might have just hit that bug once again on
> > > UBNT-NM-XW, ie. uid=004dd043, driver=Atheros AR8216/AR8236/AR8316...
> > >
> > > Anyone? I've also had a lot of problems with that bug on XW hardware
> > > and have since simply tried to avoid it (the hardware). Now that also
> > > non-loco NanoStations are only available in their XW-version I'd be
> > > happy to finally have a reliable work-around for that hardware problem.
> > >
> > >
> > > Cheers
> > >
> > >
> > > Daniel
> > >
> > >
> > > ----- Forwarded message from Tiziano Bacocco <tizbac2 at gmail.com> -----
> > >
> > > Date: Mon, 6 Feb 2017 16:34:24 +0100
> > > From: Tiziano Bacocco <tizbac2 at gmail.com>
> > > To: openwrt-devel at lists.openwrt.org
> > > Subject: [OpenWrt-Devel] Nanostation m5 XW ethernet patch gone
> > >
> > > Hello everyone
> > > Could this patch http://gerrit.aredn.org/#/c/57 be merged in trunk?
> > > I've just tested it with trunk and it works properly on my nanostation loco
> > > m5 xw , the PHY chip is correctly reset every like 8 hours or so and no
> > > connectivity loss since then
> > >
> > > Bug report:
> > > https://dev.openwrt.org/ticket/19085
> > >
> > > System log after applying patch
> > > Sun Feb  5 23:40:02 2017 kern.info kernel: [716206.656361] br-lan: port
> > > 1(eth0) entered forwarding state
> > > Sun Feb  5 23:40:02 2017 daemon.notice netifd: Network device 'eth0' link
> > > is up
> > > Sun Feb  5 23:40:04 2017 kern.info kernel: [716208.654021] br-lan: port
> > > 1(eth0) entered forwarding state
> > > Mon Feb  6 05:05:21 2017 kern.info kernel: [735725.556613]
> > > ag71xx_check_reset: expected: 004d, got: 0000
> > > Mon Feb  6 05:05:21 2017 kern.info kernel: [735725.562190]
> > > ag71xx_gpio_reset triggered
> > > Mon Feb  6 05:05:21 2017 kern.info kernel: [735725.571328] eth0: link down
> > > Mon Feb  6 05:05:21 2017 daemon.notice netifd: Network device 'eth0' link
> > > is down
> > > Mon Feb  6 05:05:21 2017 kern.info kernel: [735725.574702] br-lan: port
> > > 1(eth0) entered disabled state
> > > Mon Feb  6 05:05:23 2017 kern.info kernel: [735727.567752] eth0: link up
> > > (100Mbps/Full duplex)
> > > Mon Feb  6 05:05:23 2017 kern.info kernel: [735727.572533] br-lan: port
> > > 1(eth0) entered forwarding state
> > > Mon Feb  6 05:05:23 2017 kern.info kernel: [735727.578224] br-lan: port
> > > 1(eth0) entered forwarding state
> > > Mon Feb  6 05:05:23 2017 daemon.notice netifd: Network device 'eth0' link
> > > is up
> > >
> > >
> > > I'm not commenting to the bug report because i' both unable to register and
> > > to comment because trac complains that i'm spamming, my ip address range
> > > is 89.202.181.128 - 89.202.181.143
> > >
> > > _______________________________________________
> > > openwrt-devel mailing list
> > > openwrt-devel at lists.openwrt.org
> > > https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
> > >
> > >
> > > ----- End forwarded message -----
> > >
> >