:Anyways more work and lots of testing needs to be done. Even though I like
:the new ata stuff and use it on all machines (except for one) this is scary
:since a bunch of 4.0 users could get screwed with older good hardware.
:
:Tonight I will try to define
: ATA_16BIT_ONLY
:in my kernel and try to boot with ata. This may have to be set in GENERIC
:for installs to work ... or the ata driver learn how to figure this out
:automatically on the fly.
:
:BTW this almost hosed me.
:
:Doug A.
I've already tried ATA_16BIT_ONLY - no effect on my problem. I also
tried turning on DMA, which also had no effect. I'm also pretty sure
that my SCSI stuff is not creating any issues since I still get lockups
and a non-working timer interrupt ('ticks' doesn't increment) when I
take the SCSI out.
The weird thing is that what the ata driver appears to be doing is
preventing the timer interrupt from working. The driver itself does
appear to work in terms of issuing commands and getting responses,
at least when there is a drive on the bus for it to probe.
To me this points to cockpit trouble in the driver code somewhere.
The box this trouble is occuring on for me is an old Pentium 90. It's
the oldest working box I still own (but it makes a wonderful backup
machine and CVS respository!).
-Matt
Matthew Dillon
<[EMAIL PROTECTED]>
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message
