5.4-STABLE problem on 164SX?

[Bernd Walter]

On Sun, Jun 05, 2005 at 10:28:38PM +0200, Wilko Bulte wrote:
> On Sun, Jun 05, 2005 at 09:54:05PM +0200, Bernd Walter wrote..
> > On Sun, Jun 05, 2005 at 08:44:56PM +0200, Wilko Bulte wrote:
> > > Folks,
> > > 
> > > After just having built the June 2005 6.0 and 5.4-STABLE
> > > snapshots I tested them on my 164SX.
> > > 
> > > 6.0 works OK, but 5.4-STABLE after a test install gives me a hang
> > > when rebooting from the freshly installed disk:
> > > 
> > > cd0 at ahc0 bus 0 target 0 lun 0
> > > cd0: <TOSHIBA CD-ROM XM-6201TA 1206> Removable CD-ROM SCSI-2 device 
> > > cd0: 10.000MB/s transfers (10.000MHz, offset 15)
> > > cd0: Attempt to query device size failed: NOT READY, Medium not present
> > > da0 at ahc0 bus 0 target 1 lun 0
> > > da0: <IBM DDRS-39130W S92A> Fixed Direct Access SCSI-2 device 
> > > da0: Serial Number REF38392        
> > > da0: 40.000MB/s transfers (20.000MHz, offset 8, 16bit), Tagged Queueing
> > > Enabled
> > > da0: 8715MB (17850000 512 byte sectors: 255H 63S/T 1111C)
> > > (ahc0:A:0:0): Sending SDTR period 19, offset f
> > > (ahc0:A:0:0): Received SDTR period 19, offset f
> > >         Filtered to period 19, offset f
> > > (cd0:ahc0:0:0:0): error 6
> > > (cd0:ahc0:0:0:0): Unretryable Error
> > > (ahc0:A:0:0): Sending SDTR period 19, offset f
> > > (ahc0:A:0:0): Received SDTR period 19, offset f
> > >         Filtered to period 19, offset f
> > > (cd0:ahc0:0:0:0): error 6
> > > (cd0:ahc0:0:0:0): Unretryable Error
> > > GEOM: Configure da0a, start 0 length 268435456 end 268435455
> > > GEOM: Configure da0b, start 268435456 length 2143109120 end 2411544575
> > > GEOM: Configure da0c, start 0 length 9139200000 end 9139199999
> > > GEOM: Configure da0d, start 2411544576 length 268435456 end 2679980031
> > > GEOM: Configure da0e, start 2679980032 length 268435456 end 2948415487
> > > GEOM: Configure da0f, start 2948415488 length 6190784512 end 9139199999
> > > Mounting root from ufs:/dev/da0a
> > > start_init: trying /sbin/init
> > > 
> > > <hang>
> > > 
> > > On the DS10 the same thing works just fine.
> > > 
> > > I would appreciate if people with other Alpha models could try the SNAP
> > > to see if it is just my box.  
> > 
> > I personally never had success with any ahc controller in any alpha.
> 
> In what sense?  Booting an ahc?  Only a few SRM versions / machine models
> can do that.  But the 164SX allows me to install and boot 6.0-current
> without problems.

Using it under FreeBSD.
I tried 3985, 2940, 2940UW and 2944W in NoName, PC164 and AS4100 using
memory and io mode, all with similar results - randomly ahc won't
attach, because transport tests with io and memory space failed and
if it attaches you see various kind of data corruption or possibly
timeouts while probing.
All of the cards were verified to run under i386 based systems.
This happened to me since early 4.x times.
Especially the ahc self test, that sometimes failed, was very promissing
to debug things, but I got no reply when asking for details about this
test last year, when I did lots of testing with AS4100 and bridged cards.
IIRC this test is only run if you are using memory range, which requires
setting AHC_ALLOW_MEMIO in kernel, which is a requirement for bridged
cards on alpha, but I asume io space suffers from excactly the same
problem since there shouldn't be any technically differences between
those modes on alpha.

-- 
B.Walter                   BWCT                http://www.bwct.de
bernd at bwct.de                                  info at bwct.de