Using SCSI driver sym53c8xx.c - no nvram - no symbios BIOS

Using SCSI driver sym53c8xx.c - no nvram - no symbios BIOS

[Ron Bianco]

Is anyone using a sym53c8xx family SCSI chip on their main board?  i.e. not on a plugin
PCI card with extra stuff.

I'm using the 53c895.
If there is no NVRAM or SCSI biosROM, then does one have to tune the driver by modifying
some #define's ?

I know there is a few things that the driver will use from the pci setup code, and that
this driver supports kernel command line options, for some stuff.  But I'd like to have a
proper default setup.

Right now I'm having trouble with ncr_getclock()  which is calling ncrgetfreq(),  always
reporting 0kHz.  I'm tempted to just force it to 40000khz which is the freq. of the osc
connected to the chip's SCLK input, but I'm not sure if this is the clock referred to in
the code. (NCR SCSI clock freq. )

This problem could also be due to a hardware problem, but the code detects the chip and
reads the config data OK.

Another problem I'm having is:  CACHE TEST FAILED: timeout.  I'm still tracing this.
I'm pretty sure I set a legal cache line size and latency in the sandpoint_pci.c config
code.

I shall forward this msg to the author of the driver tomorrow if no joy.

regards, Ron

Ron Bianco
Computer Systems Technologist
Level Control Systems
email: ronb@junction.net
phone: 250-549-2558 Ext 8
web:  www.lcsaudio.com


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

RE: Using SCSI driver sym53c8xx.c - no nvram - no symbios BIOS

[Ron Bianco]

Referring to my previous email:

Oops, I found the #define options in:  sym53c8xx_defs.h !
I also found a few config pins that were not set properly and probably explain the
problems I was having.

Still would like to know if anyone else is using the driver tho'.

Also, should I set the 53c895 to little endian mode?  It is set this way at the moment and
the chip is detected and config data read OK.

So, with the PPC in big endian and the SCSI chip in little endian, and the kernel compiled
accordingly, I should be able to access files on a disk that were written by an intel box,
properly, right?

Ron


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/