Is it supposed to work that way? via-rhine + setiathome

Hello,

Every now and then, when system load is high, I get the following in dmesg.

[995091.005311] ------------[ cut here ]------------
[995091.005335] WARNING: at
/playzone/kernel/debian/build/source_i386_none/net/sched/sch_generic.c:255
dev_watchdog+0xaf/0x100()
[995091.005341] Hardware name: K7VT4APro
[995091.005344] NETDEV WATCHDOG: eth0 (via-rhine): transmit queue 0
timed out
[995091.005348] Modules linked in: isofs snd_seq_dummy binfmt_misc
nouveau ttm drm_kms_helper drm i2c_algo_bit mxm_wmi wmi video nfsd
nls_utf8 nls_cp437 vfat fat fuse eeprom ppdev snd_via82xx gameport
snd_ac97_codec ac97_bus snd_pcm_oss snd_mixer_oss snd_pcm snd_page_alloc
snd_mpu401_uart snd_seq_midi snd_rawmidi snd_seq_midi_event snd_seq
snd_timer snd_seq_device snd parport_pc processor shpchp thermal_sys
pcspkr button parport evdev pci_hotplug i2c_viapro i2c_core soundcore
nfs lockd fscache auth_rpcgss nfs_acl sunrpc ext3 jbd mbcache usbhid hid
ide_cd_mod cdrom sd_mod crc_t10dif ata_generic uhci_hcd pata_via
ehci_hcd sata_via libata floppy scsi_mod usbcore via82cxxx ide_core
via_rhine mii [last unloaded: scsi_wait_scan]
[995091.005437] Pid: 9328, comm: setiathome_6.03 Not tainted
3.0.0-1jk-486 #1
[995091.005441] Call Trace:
[995091.005452] [<c10228c0>] ? warn_slowpath_common+0x6a/0x7d
[995091.005457] [<c11e41e6>] ? dev_watchdog+0xaf/0x100
[995091.005462] [<c1022939>] ? warn_slowpath_fmt+0x28/0x2c
[995091.005467] [<c11e41e6>] ? dev_watchdog+0xaf/0x100
[995091.005481] [<c1026cb3>] ? local_bh_enable+0x2/0x2
[995091.005490] [<c102bd72>] ? run_timer_softirq+0x158/0x1ee
[995091.005496] [<c11e4137>] ? netif_tx_unlock+0x31/0x31
[995091.005501] [<c1026cb3>] ? local_bh_enable+0x2/0x2
[995091.005506] [<c1026d36>] ? __do_softirq+0x83/0x118
[995091.005512] [<c1026cb3>] ? local_bh_enable+0x2/0x2
[995091.005515] <IRQ> [<c1026ece>] ? irq_exit+0x31/0x7b
[995091.005528] [<c1012b3b>] ? smp_apic_timer_interrupt+0x54/0x5c
[995091.005538] [<c125f351>] ? apic_timer_interrupt+0x31/0x40
[995091.005543] ---[ end trace df24697873a41f2d ]---
[995091.005693] via-rhine 0000:00:12.0: eth0: Transmit timed out, status
0000, PHY status 786d, resetting...
[995091.006566] via-rhine 0000:00:12.0: eth0: link up, 100Mbps,
full-duplex, lpa 0x45E1
....

Googling a bit reveals that similar stuff has been happening for quite
some time.
https://lkml.org/lkml/2011/11/20/22
https://lkml.org/lkml/2010/11/30/576
https://lkml.org/lkml/2010/10/30/182
https://lkml.org/lkml/2002/7/20/46

Looks like the more recent posts in that mailing list did not get any
replies.
It is probably less frequent to occur than at the beginning,
but I am starting to wonder if this is the intended mode of operation
for via-rhine. Any ideas for a work around?

bye,
JK