Folks...

I'm working on some timing bugs in the SCSI driver, and would like some 
assistance in testing them.  Short story:  I replaced the old, horribly 
broken DELAY() macro with a much more accurate function that is 
calibrated during early system initialzation, and doesn't have the 68040 
kludge that the old one did (to work around the cache on the 68040).  
This is the same _delay()/delay()/DELAY() used by the sun3 and mvme68k 
ports.

Now, once I replaced DELAY() with something more accurate, the SCSI 
driver started having problems (see dmesg below).  Basically, the SCSI 
driver was relying on bugs in the old DELAY() macro...

I think I know how to work around the timing bugs for now, and I'm hoping 
that doing so will fix the problems some people (including myself) have 
experienced with Quantum Lightning drives on hp300 systems...(I'm lucky 
enough to have replaced my Quantum drive with an Hitachi drive.)

Don Lewis once reported some timeout problems to me, and my work-arounds 
will be based on his suggestions.  I can test these on my hardware, but 
I'll feel better if I can get others to test them, too.

If you're willing to do so, please send me private e-mail ASAP.

Thanks!

----save the ancient forests - http://www.bayarea.net/~thorpej/forest/----
Jason R. Thorpe                                       thorpej@nas.nasa.gov
NASA Ames Research Center                               Home: 408.866.1912
NAS: M/S 258-6                                          Work: 415.604.0935
Moffett Field, CA 94035                                Pager: 415.428.6939

----- snip -----
syncing disks... done
rebooting...
NetBSD 1.1B (BASALT) #160: Tue May 14 08:46:20 PDT 1996
    thorpej@basalt:/work/netbsd/src/sys/arch/hp300/compile/BASALT
HP9000/380/425 (25MHz MC68040 CPU+MMU+FPU, 4k on-chip physical I/D caches)
delay constant for 25MHz cpu: 81
real mem = 33546240
avail mem = 27181056
using 819 buffers containing 3354624 bytes of memory
hil2: US ASCII keyboard
hil1: security code 10 4 b4 41 46 60 8c 11 41 0 0 0 0 0 0 0
dma0: 98620C, 2 channels, 32 bit
grf0 at addr 0x560000: 1280 x 1024 256 color hi-res catseye display (console)
ite0 at grf0: attached
hpib0 at scode7 ipl 3: Internal HP-IB
rd0 at hpib0 slave 0: 7945A
rd0: 968 cylinders, 7 heads, 108416 blocks, 512 bytes/block
dca0 at scode9 ipl 5: working fifo
dcm0 at scode13 ipl 3 flags 0xe
scsi0 at scode14 ipl 4: 32 bit dma, async, scsi id 7
sd0 at scsi0 slave 0: <HP, 2213A, C023>
sd0: 1457 cylinders, 16 heads, 1296512 blocks, 512 bytes/block
sd1 at scsi0 slave 1: <COMPAQ, M2694ES-512, 952B>
sd1: 1819 cylinders, 15 heads, 2051000 blocks, 512 bytes/block
sd2 at scsi0 slave 2: <TOSHIBA, CD-ROM XM-3301TA, 0272> (SCSI-2)
sd2: CD-ROM, drive empty
st0 at scsi0 slave 3: <ARCHIVE, VIPER 60   21116, -007>
scsi0: abort from icmd: phase=0x49, ssts=0xb6, ints=0x0
-- scsiabort gave up after 1000 tries (startlen = 1024 len = 1023)
st: wrong specs: type 0 qual 0 version 2
le0 at scode21 ipl 5: address 08:00:09:06:48:bb
le0: 8 receive buffers, 2 transmit buffers
unconfigured card id 6 at sc23
Changing root device to sd0a
syncing disks... done
rebooting...
NetBSD 1.1B (BASALT) #158: Sat May 11 09:48:33 PDT 1996
    thorpej@basalt:/work/netbsd/src/sys/arch/hp300/compile/BASALT
HP9000/380/425 (25MHz MC68040 CPU+MMU+FPU, 4k on-chip physical I/D caches)
real mem = 33546240
avail mem = 27181056
using 819 buffers containing 3354624 bytes of memory
hil2: US ASCII keyboard
hil1: security code 10 4 b4 41 46 60 8c 11 41 0 0 0 0 0 0 0
dma0: 98620C, 2 channels, 32 bit
grf0 at addr 0x560000: 1280 x 1024 256 color hi-res catseye display (console)
ite0 at grf0: attached
hpib0 at scode7 ipl 3: Internal HP-IB
rd0 at hpib0 slave 0: 7945A
rd0: 968 cylinders, 7 heads, 108416 blocks, 512 bytes/block
dca0 at scode9 ipl 5: working fifo
dcm0 at scode13 ipl 3 flags 0xe
scsi0 at scode14 ipl 4: 32 bit dma, async, scsi id 7
sd0 at scsi0 slave 0: <HP, 2213A, C023>
sd0: 1457 cylinders, 16 heads, 1296512 blocks, 512 bytes/block
sd1 at scsi0 slave 1: <COMPAQ, M2694ES-512, 952B>
sd1: 1819 cylinders, 15 heads, 2051000 blocks, 512 bytes/block
sd2 at scsi0 slave 2: <TOSHIBA, CD-ROM XM-3301TA, 0272> (SCSI-2)
sd2: CD-ROM, drive empty
st0 at scsi0 slave 3: <ARCHIVE, VIPER 60   21116, -007>
sd3 at scsi0 slave 4: <HITACHI, DK516C, NO34> (SCSI-2)
sd3: 2179 cylinders, 15 heads, 2606400 blocks, 512 bytes/block
le0 at scode21 ipl 5: address 08:00:09:06:48:bb
le0: 8 receive buffers, 2 transmit buffers
unconfigured card id 6 at sc23
Changing root device to sd0a