SCSI Card Complaining LFS 4.0

Michael Havens bmike1 at vei.net
Mon May 24 19:14:24 PDT 2004


I am getting an error that is very annoying. Whenever I open an alternate 
terminal it begins to spit an error message which is:
   scsi1: PCI error Interrupt at seqaddr = 0x8 
   scsi1: Data Parity Error Detected durring address or write data phase 
   scsi1: WARNING WARNING WARNING 
   scsi1: Too many PCI parity errors  observed as a target. 
   scsi1: Some device on this bus is generating bad parity. 
   scsi1: This is an error *observed by*, not *generated by* this controller. 
   scsi1: PCI parity checking has been disabled. 
   scsi1: WARNING WARNING WARNING 
In an attempt to find the offending PCI card I unplugged all non-essential PCI 
cards one at a time but the error kept reappearing.
 
In my research I found that some motherboards don't like too many pci cards 
plugged in and that if this error appears to disable parity checking. Well, I 
have been searching for the means to disable parity checking for quite some 
time now but can't find the means. I would be most grateful if you could tell 
me.
-- 
<:-)Mike(-:>
Thanks.
bmike1 at vei.net

On Mon, 23 Jun 2003, Ken Moffat wrote that Matt Darcy wrote:
> Hi all,
>
> This is not stricly an LFS problem, but as the input from this group is
> excellent and it on an LFS system I'll ask, as I get a better respone than
> alt.linux
>

[ snip details of system ]
>
> Now for the problem /interesting part
>
> At the console I keep getting alerts which look like
>
> scsi0: PCI error Interrupt at seqaddr = 0x9
> scsi0: Data Parity Error Detected during address or write data phase
> scsi0: PCI error Interrupt at seqaddr = 0x7
> scsi0: Data Parity Error Detected during address or write data phase
>
>
> I am not sure what this means however it looks like an interupt problem.
> On investigation I see in /proc/interrupts
>
>            CPU0
[snip]
>  10:     344236          XT-PIC  eth1
>  11:     238104          XT-PIC  aic7xxx, eth0
[snip]
> You'll notice that interupt "11" has both SCSI card and eth0 network card
> sharing and interupt.On research this is not unusual but the output from
> dmesg leads me to believe more
>
> eth0: incorrect signature 0xffff
> PP_addr=0xffff

>
> I find it suspect that the 2 devices that share interupt 11 are both
> working, but complaining too.
>
> My question is what does the error mean, and does anyone have any
> suggestions / comments.
>

 Definitely looks like an interrupt problem.  Can you try moving the
cards around until /proc/interrupts shows the network and scsi
interrupts are not shared ?

 My reading of the messages is that the two cards are conflicting with
each other, causing the busses to float high.  Maybe not all the time,
or you wouldn't be able to use either device, so it might just be
interference between them when they are both trying to use the PCI bus.

 On the other hand, if the performance is "adequate", you could always
leave it...   In my personal opinion, the eth0 errors can probably be
ignored (they're dumped in the log) unless you are seeing poor network
throughput, but scsi errors tend to make me nervous.  Of course, in some
situations we all run out of suitable alternative PCI slots.

Ken
-- 
You'll always find me out to lunch
-- 
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe lfs-support' in the subject header of the message


 
 






This message has been scanned for viruses by the VEI Internet
Automatic Email Spam and Virus Scanner, and is believed to be free of spam or viruses.
Please report spam to spamtrap at vei.net. If you would like 98.9 % spam blocked from your
E-mail then go to VEI Internet for details. Anti-spam/Anti-virus is FREE with every account. 


http://www.vei.net/
mailtospamtrap at vei.net



More information about the lfs-support mailing list