em WOL_MCAST repowers machine after poweroff [Was: Re: em JumboFrame improovement and PCIe addon-card regression]

Jack Vogel jfvogel at gmail.com
Wed May 5 21:53:26 UTC 2010


Oppps, forgot about the lem case, sorry, and here I made lem so i wouldnt
have to touch that code, lol :)

As for WOL not working at all, that's something I'll have to check on.

Jack


On Wed, May 5, 2010 at 2:19 PM, Harald Schmalzbauer <
h.schmalzbauer at omnilan.de> wrote:

> Jack Vogel schrieb am 27.04.2010 23:58 (localtime):
>
>  Thanks Harald,
>>
>> Have already been made aware of this, its due to the broadcast WOL being
>> enabled, I will be
>> fixing the problem shortly. Sorry for the inconvenience.
>>
>
> Hello Jack,
>
> I saw your RELENG_8 change and recompiled one kernel today. It seems that
> you only disabled WOL_*CAST for 7.0.5 adapters. Here's one S3200SHLX board,
> which has two onboard nics (em0 with WOL_*CAST disabled and em1 with WOL_*
> enabled):
>
> em0: <Intel(R) PRO/1000 Network Connection 7.0.5> port 0x30e0-0x30ff mem
> 0xe1b00000-0xe1b1ffff,0xe1b20000-0xe1b20fff irq 20 at device 25.0 on pci0
> em0: Using MSI interrupt
> em0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu 1500
>
> options=219b<RXCSUM,TXCSUM,VLAN_MTU,VLAN_HWTAGGING,VLAN_HWCSUM,TSO4,WOL_MAGIC>
>
> em1: <Intel(R) PRO/1000 Legacy Network Connection 1.0.1> port 0x1000-0x103f
> mem 0xe1920000-0xe193ffff,0xe1900000-0xe191ffff irq 18 at device 2.0 on pci4
> em1: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu 1500
> options=389b<RXCSUM,TXCSUM,VLAN_MTU,VLAN_HWTAGGING,VLAN_HWCSUM,WOL_UCAST,WOL_MCAST,WOL_MAGIC>
>
> I set this options manually on another box and with these em1 settings the
> machine powers up immediately.
> But I have a much more severe problem:
> I disabled WOL_MCAST with ifconfig for one of my cold-standby machines. But
> the machine doesn't wake up any more, even if I can see the WOL packets
> arriving at the NIC (LED blinks).
> Since the machine also doesn't power up on WOL when shutted down via ESXi I
> thought that was an issue with BIOS49, which I recently upgraded to. I filed
> an intel reseller case -> status still in progress.
> Is it possible that the new code also broke intentionally WOL events?
>
> Thanks,
>
> -Harry
>
>


More information about the freebsd-stable mailing list