Date: Wed, 11 Nov 2015 16:18:39 +0000 From: Steven Hartland <steven@multiplay.co.uk> To: "freebsd-net@freebsd.org" <freebsd-net@freebsd.org> Cc: Jack F Vogel <jfv@FreeBSD.org> Subject: Re: Intel XL710 broken link down detection? Message-ID: <56436A5F.4020102@multiplay.co.uk> In-Reply-To: <564357E0.1050002@freebsd.org> References: <564357E0.1050002@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Comparing this to the Linux driver which does detect the link down I've discovered it actually polls the link status by default in its watchdog. Disabling this with "ethtool --set-priv-flags eth1 LinkPolling off" and the Linux driver also fails to detect link down. So this seems like a firmware or even hardware bug where it should be reporting down events and the Linux driver has been updated to workaround the problem? On 11/11/2015 14:59, Steven Hartland wrote: > Seems the Intel XL710 (ixl) has broken link detection, it triggers on > link up but not on link down. > > I'm currently running HEAD (r290610) on the box which has 1.4.3 and > have tried the latest driver from downloads.intel.com which is listed > as 1.4.8 but reports 1.4.5. > > We've updated to the latest NVM image today to see if that would help > but no change. > > Tracing with dtrace shows that ixl_do_adminq process a link event and > fires off ixl_link_event for link up but not for link down for some > reason. > > Running ifconfig against the interface updates the status as runs > ixl_update_link_status in ixl_media_status, so the device can > determine the link is down its just not noticing the event. > > From dmesg: > ixl0: <Intel(R) Ethernet Connection XL710 Driver, Version - 1.4.3> mem > 0xc5000000-0xc57fffff,0xc5808000-0xc580ffff irq 34 at device 0.0 on pci5 > ixl0: Using MSIX interrupts with 33 vectors > ixl0: f4.40 a1.4 n04.53 e80001ccc > ixl0: Using defaults for TSO: 65518/35/2048 > ixl0: Ethernet address: 0c:c4:7a:3a:2d:60 > ixl0: PCI Express Bus: Speed 8.0GT/s Width x8 > queues is 0xfffffe00015bd000 > ixl0: netmap queues/slots: TX 32/1024, RX 32/1024 > ixl1: <Intel(R) Ethernet Connection XL710 Driver, Version - 1.4.3> mem > 0xc4800000-0xc4ffffff,0xc5800000-0xc5807fff irq 34 at device 0.1 on pci5 > ixl1: Using MSIX interrupts with 33 vectors > ixl1: f4.40 a1.4 n04.53 e80001ccc > ixl1: Using defaults for TSO: 65518/35/2048 > ixl1: Ethernet address: 0c:c4:7a:3a:2d:61 > ixl1: PCI Express Bus: Speed 8.0GT/s Width x8 > queues is 0xfffffe0001ad1000 > ixl1: netmap queues/slots: TX 32/1024, RX 32/1024 > > When its up it reports 10GBase-KR > > Any ideas? > > Regards > Steve
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?56436A5F.4020102>