Date: Fri, 17 Jul 1998 13:30:35 -0500 From: Doug Ledford <dledford@dialnet.net> To: "Patrick W. Bryant" <daspwb@Queequeg.Gsu.EDU> Cc: aic7xxx Mailing List <AIC7xxx@FreeBSD.ORG> Subject: Re: 5.1.0-pre2-to-pre3 Message-ID: <35AF984B.66D71D82@dialnet.net> References: <Pine.LNX.3.96.980717121605.4985A-100000@Queequeg.Gsu.EDU>
next in thread | previous in thread | raw e-mail | index | archive | help
Patrick W. Bryant wrote:
>
> Hi all. My aic7890 Ultra2 is still hanging on "Downloading sequencer
> code," though it reports "407 instructions downloaded," which is one step
> beyond where I got with the pre2 patch. The only device on the adapter is
> a Quantum Viking II 9.1 GB (Ultra2), and the SCSI bios has all default
> values. Oh yea, kernel is 2.0.34.
>
> Any suggestions?
>
> Here's the output with aic7xxx="verbose:0x1ffff,panic_on_abort":
>
> (scsi0) Adaptec AIC-7890/1 Ultra2 SCSI host adapter
> (scsi0) Wide Channel, SCSI ID=7, 32/255 SCBs
> (scsi0) Downloading sequencer code... 407 instructions downloaded
> aic7xxx driver version 5.1.0pre3
> Controller type:
> Adaptec AIC-7890/1 Ultra2 SCSI host adapter
> p->flags=0x10804501, p->chip=0x406, p->features=0xf6, sequencer is paused
> SIMODE0=0X0, SIMODE1=0Xa4, SSTAT0=0x0, SSTAT1=0x0, INTSTAT=0x0
> SCSISIGI=0x0, SXFRCTL0=0x88, SXFRCTL1=0x27, SCSISEQ=0x0, SEQCTL=0x90
> SCSIRATE=0x0, SCSIOFFSET=0, SEQADDR=0x1, LASTPHASE=P_MESGIN
> QINFIFO: (TAG)
> Current SCB: (SCBPTR/TAG/CONTROL) 0/255/0x0
> WAITING_SCBS: (SCBPTR/TAG/CONTROL)
> DISCONNECTED_SCBS: (SCBPTR/TAG/CONTROL)
OK...this actually looks like a different problem. Turn off the
panic_on_abort and verbose options and see if it will boot any further
normally. The thing I'm seeing here is an immediate interrupt storm as soon
as we enable the card is the likely cause of all this, and that may not
actually be coming from the card, it may be coming from elsewhere and we
just get hit by it once that card is enabled. However, there is something
else to consider here as well, there was no bus reset in the above info.
Please make sure that the no_reset option is not in use, and that the
Adaptec SCSI BIOS is configured to reset the scsi bus at IC init time. I
left the no_reset option for people that *truly* need it (for example, they
have a SCSI bus shared between two machines and a reset would cause the
other machine to interrupt whatever it is doing, such as a possible tape
backup or CD burn, ruining whatever was interrupted). The semantics of
negotiation under the 5.0.x and 5.1.x drivers have changed enough that the
no_reset code path is not guaranteed to be reliable, and I'm not even going
to guarantee that your machine will boot using that code. The no_reset
option forces the driver to do things that I know have proven to lock up
certain devices.
--
Doug Ledford <dledford@dialnet.net>
Opinions expressed are my own, but
they should be everybody's.
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe aic7xxx" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?35AF984B.66D71D82>