Hi, all.

I'm wondering how many people out there are using PMAZ-A TURBOchannel
option card on 3MIN or MIXINE.

I have an unproven hypothesis that it might be unnecessary to
'slow-down' NCR 53C94 clock rate when PMAZ-A is attached with
slowly-clocked (12.5MHz) TURBOchannel option slots found in 3MIN or
MAXINE.  A crude try with my MAXINE shows PMAZ-A still works at 25MHz
clock rate, but I'm not sure it's globally true.

So, someone who are brave enough to trash disk contents, please try to

o make a change in file sys/dev/tc/asc_tc.c;

*** asc_tc.c    Fri Oct 31 22:32:36 1997
--- /tmp/asc_tc.c       Thu Jun  4 10:01:41 1998
***************
*** 127,132 ****
--- 127,133 ----
        asc->dma_start = tc_dma_start;
        asc->dma_end = tc_dma_end;
  
+ #if 0
        /*
         * Now for timing. The 3max has a 25Mhz tb whereas the 3min and
         * maxine are 12.5Mhz.
***************
*** 145,153 ****
                speed = ASC_SPEED_12_5_MHZ;
                break;
        };
! 
        ascattach(asc, speed);
- 
        /* tie pseudo-slot to device */
        tc_intr_establish(parent, t->ta_cookie, TC_IPL_BIO,
                          asc_intr, asc);
--- 146,155 ----
                speed = ASC_SPEED_12_5_MHZ;
                break;
        };
! #else
!       speed = ASC_SPEED_25_MHZ;
! #endif
        ascattach(asc, speed);
        /* tie pseudo-slot to device */
        tc_intr_establish(parent, t->ta_cookie, TC_IPL_BIO,
                          asc_intr, asc);
    
o then, with a newly compiled kernel, try to I/O intensitive jobs on
  disks attached on the PMAZ-A, and watch closely how things go.  Do
  disk get probed correctly?  Are I/O errors happening on data transfer?
  I'm willing to have reports.

Tohru Nishimura
Nara Institute of Science and Technology