Subject: Re: aic7880 & MP
To: Peter Eisch <peter@boku.net>
From: Greg A. Woods <woods@weird.com>
List: netbsd-users
Date: 03/03/2005 18:03:21
[ On Thursday, February 17, 2005 at 15:40:11 (-0600), Peter Eisch wrote: ]
> Subject: aic7880 & MP
>
> 
> Anyone have a patch for the ahc(4) driver that will work with an MP (perhaps
> i386) kernel?  The driver spews after "waiting 2 seconds for devices to
> settle" when it tries to scan these onboard controllers:

Might be an i386-only problem.

I have the recent ahc(4) driver backported to my netbsd-1.6.x kernels
and I'm using what I believe is a multiple channel chip on a 2-processor
AlphaPC 264DP with no problem:

	NetBSD 1.6.2_STABLE (PC264DP.MP) #0: Fri Feb 28 02:44:26 EST 2025

	cpu0 at mainbus0: ID 0 (primary), 21264-4
	cpu0: Architecture extensions: 303<PAT,MVI,FIX,BWX>
	cpu1 at mainbus0: ID 1, 21264-4
	cpu1: Architecture extensions: 303<PAT,MVI,FIX,BWX>

	ahc0 at pci0 dev 6 function 0
	ahc0: interrupting at dec 6600 irq 19
	ahc0: aic7895C: Ultra Wide Channel A, SCSI Id=7, 32/253 SCBs
	scsibus0 at ahc0: 16 targets, 8 luns per target
	ahc1 at pci0 dev 6 function 1
	ahc1: interrupting at dec 6600 irq 18
	ahc1: aic7895C: Ultra Wide Channel B, SCSI Id=7, 32/253 SCBs
	scsibus1 at ahc1: 16 targets, 8 luns per target

	scsibus0: waiting 2 seconds for devices to settle...
	sd0 at scsibus0 target 0 lun 0: <DEC, RZ29B    (C) DEC, 0016> SCSI2 0/direct fixed
	sd0: 4091 MB, 3708 cyl, 20 head, 113 sec, 512 bytes/sect x 8380080 sectors
	sd0: sync (172.0ns offset 8), 16-bit (11.626MB/s) transfers, tagged queueing
	scsibus1: waiting 2 seconds for devices to settle...
	sd1 at scsibus1 target 1 lun 0: <DEC, RZ29B    (C) DEC, 0016> SCSI2 0/direct fixed
	sd1: 4091 MB, 3708 cyl, 20 head, 113 sec, 512 bytes/sect x 8380080 sectors
	sd1: sync (172.0ns offset 8), 16-bit (11.626MB/s) transfers, tagged queueing
	sd2 at scsibus1 target 2 lun 0: <DEC, RZ29B    (C) DEC, 0016> SCSI2 0/direct fixed
	sd2: 4091 MB, 3708 cyl, 20 head, 113 sec, 512 bytes/sect x 8380080 sectors
	sd2: sync (172.0ns offset 8), 16-bit (11.626MB/s) transfers, tagged queueing

17:56 [5] $ df
Filesystem    512-blocks     Used    Avail %Cap Mounted on
/dev/sd0a        2383134  1214110  1049866  53% /
/dev/sd1d        6169056    91020  5769580   1% /var
mfs:118           496222        2   471408   0% /tmp
/dev/sd0d        3748544        4  3561112   0% /scratch
/dev/sd2d        6169056  3079300  2781300  52% /var/packages
/dev/ld0a       66056376 13281184 49472372  21% /build
/dev/ld1a       38975900 10949548 26077556  29% /usr/pkg


It's built NetBSD from scratch to ISO (incl. xsrc), and quite a few
packages too as you can see, with no problems whatsoever.


Note the onboard controller on my system is two PCI functions on the
same PCI device (one "slot") whereas your onboard controller is two
separate PCI devices (effectively two "slots").  That shouldn't make any
difference though -- indeed it should help, not hurt.

-- 
						Greg A. Woods

H:+1 416 218-0098  W:+1 416 489-5852 x122  VE3TCP  RoboHack <woods@robohack.ca>
Planix, Inc. <woods@planix.com>          Secrets of the Weird <woods@weird.com>