IGNORE: 2910C problems

[Simon Whiting]

G'day All,

Sorry for the excess noise.  My problem turned out to be my MB Bios having a partial
implementation of SCSI (which obviously is broken).  I swapped MB, and everything is
running smoothly.

Cheers,
Simon.


Simon Whiting wrote:

> G'day Doug and aic7xxx people,
>
> I am sorry if this has been asked before, but I couldn't find an archive of the
> list, so here goes:
>
> I am now the proud owner of a 2910C, however things aint well.  My machine is:
>     P100 / 23M / 2x510M IDE Drives
>     Adaptec 2910 (rev3) BIOS 1.11
>     Redhat 5.2 (Bastardised to handle 2.2 kernels)
>     Kernel 2.2.9
>     aic7xxx version 5.1.17/3.2.4 (compiled into the kernel, ie not a module)
>     IRQ 10, SCSI Host 7, Parity OFF, on board term DISABLED (though I have tried
> EVERY combination)
>     the drives/tape unit I am playing with having been working fine on my other
> system with a 2940 for months
>
> Problems:
>
> a). On boot if I put the 2910C with nothing connected, then the system hangs
> while it is probing for SCSI devices (This may be normal, I just don't know).
>
> b). With a terminator (SUN SCSI-2 TERM) on either chain, with no devices, the
> system starts booting into linux of the IDEs, then "scsi0: aborting command due
> to timeouts" start occuring when it hits the Adaptec driver section.
>
> c). Connecting a drive/tape unit(int/ext) doesn't help, though the drives
> sometimes while splash up it has found it.
>
> d). "stpwlev" and "override_term" tweaks didn't seem to make a difference,
> though not to sure if I was aggrevating the problem.
>
> e). "dump_card" gave some interesting (not sure what it means, but intersting
> non the less) :
>     PCI_COMMAND    0x7
>     DEVCONFIG          0x100
>     Loading Serial EEPROM...done
>     BIOS disabled, IO Port 0x6000, IRQ 10
>     IO MEM   0xe1000000, MMAP MEM  0xc280200
>     EEPROM is not PRESENT
>     SE Low Byte term enabled
> <=== I could change this with "override_term"
>     1 host
>     Downloading <SNIP> 406 instructions <SNIP>
>     scsi: aborting command due to timeout
>     scsi: <SNIP> <lots of stuff referencing 0's only (eg LUN, ...etc...)>
>
> f). "dump_sequencer" gave some even MORE interesting stuff:
>     (scsi0) BRKADRINT error (0x20)
>             Scratch Ram/SCB Array Ram Parity Error
>     (scsi0) SEQADDR=0x9
> <=== this was sometimes "0xa" I think
>
> What I have tried:
>     Searhing News/Web servers exhaustively
>     Reading the "README.aic7xxx" from the Linux kernel source
>     Patching the Kernel
>     Every @#$%^& combination of:
>         2910C card settings (ie parity ON/OFF, ...etc...)
>         every combination I could think of for the "stpwlev" and "override_term"
>
>         every combination of drives/tape unit (ie changing SCSI ids & order, but
> not the other jumpers on the
>                 units)
>         different types of terminators (eg SUN SCSI-2, SUN SCSI-3, IBM SCSI-1,
> misc SCSI-1 & SCSI-2)
>         hitting my head against the wall; kicking the cat; watching the cricket
>
> What am I doing wrong, or is the CARD stuffed (the "SCB Array Ram Parity Error"
> sounds nasty).
>
> All help/suggestions greatly appreciated.
>
> Cheers,
> Simon Whiting.
>
> To Unsubscribe: send mail to majordomo at FreeBSD.org
> with "unsubscribe aic7xxx" in the body of the message
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Simon.Whiting.vcf
Type: text/x-vcard
Size: 430 bytes
Desc: Card for Simon Whiting
Url : http://lists.freebsd.org/pipermail/aic7xxx/attachments/19990621/69f08e4a/Simon.Whiting.vcf