Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 7 Feb 2011 15:52:03 +0300
From:      Lev Serebryakov <lev@freebsd.org>
To:        Mike Tancsa <mike@sentex.net>
Cc:        freebsd-net@freebsd.org, freebsd-stable@freebsd.org, "Vogel, Jack" <jack.vogel@intel.com>
Subject:   Re: em0 hangs without any messages like "Watchdog timeout", only down/up reset it.
Message-ID:  <504473738.20110207155203@serebryakov.spb.ru>
In-Reply-To: <4D4FE3F3.6030604@sentex.net>
References:  <787079038.20110207004023@serebryakov.spb.ru> <4D4FE3F3.6030604@sentex.net>

next in thread | previous in thread | raw e-mail | index | archive | help
Hello, Mike.
You wrote 7 =F4=E5=E2=F0=E0=EB=FF 2011 =E3., 15:22:11:

>>   It hangs under load without any output. When it works with POLLING, it
>> prints "Watchdog timeout" and reset automatically several times a day,
>> but without POLLING it simply hangs with same frequency. It is
>> 8.2-PRERELEASE (from RELENG_8), cvsupped AFTER 22 Jan (last commit to
>> e1000 drivers family).
>>=20
>>   Locally system works, but mbufs are overfilled. "ifconfig em0 down
>> && ifconfig em0 up" solves problem.
> In non polling mode, do you have any special network or driver settings
> in /boot/loader.conf or /etc/sysctl.conf ?

/boot/loader.conf:
hw.em.rxd=3D4096
hw.em.txd=3D4096
net.link.ifqmaxlen=3D16384

/etc/sysctl.conf:
dev.em.0.rx_int_delay=3D200
dev.em.0.tx_int_delay=3D200
dev.em.0.rx_abs_int_delay=3D4000
dev.em.0.tx_abs_int_delay=3D4000
dev.em.0.rx_processing_limit=3D4096


   I'm trying to run with patch from "em driver, 82574L chip, and
 possibly ASPM" thread now under heavy network load.

--=20
// Black Lion AKA Lev Serebryakov <lev@serebryakov.spb.ru>




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?504473738.20110207155203>