ICH7 SATA and em interrupt sharing

Patrick M. Hausen hausen at punkt.de
Mon Aug 21 15:07:48 UTC 2006


Again ...

On Mon, Aug 21, 2006 at 04:53:28PM +0200, Patrick M. Hausen wrote:

> Seems finally I am able to reproduce part of the problems
> some people see with their hardware.
> 
> I used device.hints to force em0 to share its irq
> with atapci1 - the SATA300 controller in my machine.
> 
> atapci1: <Intel ICH7 SATA300 controller> port 0x30e8-0x30ef,0x30dc-0x30df,0x30e0-0x30e7,0x30d8-0x30db,0x30b0-0x30bf mem 0xdc500400-0xdc5007ff irq 19 at device 31.2 on pci0
> em0: <Intel(R) PRO/1000 Network Connection Version - 6.0.5> port 0x4000-0x401f mem 0xdc080000-0xdc09ffff,0xdc000000-0xdc07ffff irq 19 at device 0.0 on pci4
> 
> And pronto:
> 
> em0: Missing Tx completion interrupt!
> em0: Missing Tx completion interrupt!
> em0: Missing Tx completion interrupt!
> em0: Missing Tx completion interrupt!
> em0: Missing Tx completion interrupt!
> em0: Missing Tx completion interrupt!
> 
> And the system is really sluggish when used via SSH over that NIC.
> I've yet to provoke the "detached disk" problem, though.

make -j4 buildworld and ... *boom*

ad6: TIMEOUT - READ_DMA retrying ...
ad4: TIMEOUT - READ_DMA retrying ...

HTH,

Patrick M. Hausen
Leiter Netzwerke und Sicherheit
-- 
punkt.de GmbH         Internet - Dienstleistungen - Beratung
Vorholzstr. 25        Tel. 0721 9109 -0 Fax: -100
76137 Karlsruhe       http://punkt.de


More information about the freebsd-stable mailing list