Re: hardware question: scsi disk and SS10

[Michael <macallan%netbsd.org@localhost>]

Hello,

On Tue, 11 Nov 2014 00:01:56 -0800
Erik Fair <fair%netbsd.org@localhost> wrote:

> On Nov 10, 2014, at 20:56, Michael <macallan%netbsd.org@localhost> wrote:
> 
> > Yeah, those hme/wide scsi boards are really nice. Might want to look
> > for QLogic ( 'Antares' ) boards too - that would give you UW-SCSI, I’m using one of those in my U1
> > ( isp* at sbus? )
> 
> Are you getting 40MB/s observed performance from it, or a reasonable percentage thereof?

I verified, and the answer is yes, I do:

cpu0 at mainbus0: SUNW,UltraSPARC @ 167.002 MHz, UPA id 0
cpu0: 16K instruction (32 b/l), 16K data (32 b/l), 512K external (64 b/l)
...
isp0 at sbus0 slot 1 offset 0x10000 vector 3 ipl 3 for QLGC,isp
...
sd0 at scsibus0 target 0 lun 0: <SEAGATE, SX173404LC, BD13> disk fixed
sd0: 70007 MB, 14100 cyl, 24 head, 423 sec, 512 bytes/sect x 143374738 sectors
sd0: sync (100.00ns offset 15), 8-bit (10.000MB/s) transfers, tagged queueing
...
sd1 at scsibus1 target 1 lun 0: <NETAPP, X243_FAL7E073S10, NA00> disk fixed
sd1: 70092 MB, 29902 cyl, 8 head, 600 sec, 512 bytes/sect x 143550456 sectors
sd1: sync (50.00ns offset 8), 16-bit (40.000MB/s) transfers, tagged queueing

~# dd if=/dev/rsd0c of=/dev/null bs=1m count=64
64+0 records in
64+0 records out
67108864 bytes transferred in 8.711 secs (7703921 bytes/sec)
~# dd if=/dev/rsd1c of=/dev/null bs=1m count=256
256+0 records in
256+0 records out
268435456 bytes transferred in 7.769 secs (34552124 bytes/sec)

sd0 is the 8bit happy Seagate disk I mentioned earlier, connected to
the onboard esp, while sd1 is one of those that I couldn't get to work
reliably on an 8bit bus. They're fine at anything 16bit though.

have fun
Michael