Date: Tue, 15 Jun 2010 17:05:00 -0700 From: Pyun YongHyeon <pyunyh@gmail.com> To: Artem Kim <artem_kim@inbox.ru> Cc: freebsd-stable@freebsd.org, yongari@freebsd.org Subject: Re: Problems with bge (possibly related to r208993) Message-ID: <20100616000500.GN4257@michelle.cdnetworks.com> In-Reply-To: <201006160350.39502.artem_kim@inbox.ru> References: <201006131934.11389.artem_kim@inbox.ru> <201006160257.01213.artem_kim@inbox.ru> <20100615232108.GM4257@michelle.cdnetworks.com> <201006160350.39502.artem_kim@inbox.ru>
next in thread | previous in thread | raw e-mail | index | archive | help
--DBIVS5p969aUjpLe
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
On Wed, Jun 16, 2010 at 03:50:39AM +0400, Artem Kim wrote:
> On Wednesday 16 June 2010 03:21:09 Pyun YongHyeon wrote:
>
> > Hmm, why you need link0 flag? The link0 flag is used to force the
> > interface MASTER. Normally this configuration is automatically
> > done during auto-negotiation such that one is configured as MASTER
> > and the other is configured as SLAVE. If you manually configure
> > this setting you should be very careful not to use the same
> > configuration of MASTER/SLAVE of link partner. If you have to use
> > link0 option, the link partner should be configured to use SLAVE.
> > Normally you should always use auto-negotiation on 1000baseT unless
> > link partner is severely broken to support NWAY.
> >
> > It seems link partner does not agree on resolved speed/duplex
> > configuration of bge1. Check link partner's resolved link
> > configuration.
> >
>
> In any case, now I do not use the flag link0.
> Now the master/slave is assigned through auto-negotiation.
>
> link0 not been set before the problem occurred. I reset link0 flag on NAS2
> when I got a problem the first time.
>
> Now x900 port1.0.12 and bge0 configured automatically.
>
> >bge1: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu 1500
> > media: Ethernet autoselect (1000baseT <full-duplex>)
> > status: active
>
> >awplus>show int port1.0.12 status
> >Port Name Status Vlan Duplex Speed Type
> >port1.0.12 connected 55 a-full a-1000 1000BASE-T
Would you try attached patch and let me know what you can see on
your console? The patch will display error code on console. Note,
it may spam your console a lot if you see many RX errors so please
don't apply it on your production server.
--DBIVS5p969aUjpLe
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment; filename="bge.show_err.diff"
Index: sys/dev/bge/if_bge.c
===================================================================
--- sys/dev/bge/if_bge.c (revision 209211)
+++ sys/dev/bge/if_bge.c (working copy)
@@ -3382,6 +3382,9 @@
stdcnt++;
m = sc->bge_cdata.bge_rx_std_chain[rxidx];
if (cur_rx->bge_flags & BGE_RXBDFLAG_ERROR) {
+#if 1
+ printf("%04x ", cur_rx->bge_error_flag);
+#endif
bge_rxreuse_std(sc, rxidx);
continue;
}
--DBIVS5p969aUjpLe--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20100616000500.GN4257>