[Linux-gpib-general] Re: HELP! - HP82341D, linux-gpib 3.1.3, kernel 2.4.5

[Ed S. <ed...@ne...>]

That's excellent, Frank, thanks a lot.

If it's any help, I pulled together all the 9914 and related info onto a
single web page,
  http://www.physics.BrockU.CA/~edik/gpib

In particular, there are hard-to-come-by datasheets for TMS9914 and
SMJ9914A.

I took the old listing of gpib_registers.h (presumably, written as per NI
NAT9914 data sheets) and edited it to reflect the true TI 9914, it is
posted on the above web page as registers.h.  There were only a couple of
small changes.

Unfortunately, all my efforts with HP/Agilent to get the information about
the OTHER 24 i/o locations mapped out have failed so far (the PnP tools
report that the card wants 32 i/o locations, while 9914 only needs 8
registers).  It is also not clear where/how/if DMA mode is to be enabled,
if supported.

The HP82335 cards, for which the support supposedly was there in the
linux-gpib-2.x.x, were MEMORY-mapped (not io-mapped); there the eight
UPPERMOST memory addresses were mapped onto 9914 registers.

I tried to scan the entire isapnp-assigned 32-location space for 9914
"signature" or other info, using raw portio, but the results are puzzling:
multiple reads in a row never yield the same values, pretty much anywhere
in that space, except base+0x09 and base +0x19 which are consistently
zero:

# isapnp isapnp.conf
Board 1 has Identity a9 40 00 18 e8 11 14 f0 22:  HWP1411 Serial No 1073748200 [checksum a9]
HWP1411/1073748200[0]{Hewlett-Packard HP-IB Interface Card}: Port 0x240; IRQ11 --- Enabled OK
# modprobe portio ports='0x240-0x25f'
# grep 240 /proc/ioports
0240-025f : portio
# ./scan-registers.sh
0x240 = 0x50  0x90  0x8e  0x8f  0x8f  0x8e  0x8f  0x90  0x8e  0x8f   
0x242 = 0x8e  0x8d  0x8f  0x8f  0x8e  0x8c  0x8f  0x8d  0x8f  0x8f   
0x243 = 0x8f  0x8e  0x8e  0x8f  0x8f  0x8f  0x8f  0x8e  0x8f  0x8f   
0x244 = 0x91  0x8f  0x8f  0x8e  0x8e  0x8f  0x8f  0x8f  0x8f  0x8f   
0x245 = 0x90  0x8e  0x8f  0x8d  0x8e  0x8e  0x8f  0x8e  0x8f  0x8d   
0x246 = 0x8f  0x8c  0x8c  0x8e  0x8e  0x8f  0x8e  0x8f  0x8e  0x8e   
0x247 = 0x8e  0x8e  0x8d  0x8e  0x8e  0x8f  0x90  0x8e  0x90  0x8f   
0x248 = 0x8f  0x91  0x8f  0x8d  0x8d  0x8e  0x90  0x8e  0x8f  0x8e   
0x249 = 0x00  0x00  0x00  0x00  0x00  0x00  0x00  0x00  0x00  0x00   
0x24A = 0x8e  0x90  0x8e  0x8b  0x8f  0x8e  0x8f  0x8e  0x8f  0x91   
0x24B = 0x90  0x8e  0x8f  0x8c  0x90  0x8e  0x8c  0x90  0x8a  0x8f   
0x24C = 0x90  0x89  0x90  0x8d  0x8f  0x8e  0x8f  0x8f  0x8e  0x8c   
0x24D = 0x8d  0x91  0x8f  0x8f  0x8e  0x90  0x8d  0x8e  0x8d  0x8e   
0x24E = 0x90  0x8e  0x90  0x90  0x8e  0x8f  0x8e  0x8e  0x90  0x8f   
0x24F = 0x8e  0x90  0x8f  0x90  0x8e  0x8e  0x8e  0x8d  0x8f  0x90   
 
0x250 = 0x92  0x90  0x8e  0x8f  0x8e  0x8e  0x8f  0x8e  0x92  0x8e   
0x252 = 0x8f  0x8e  0x8e  0x91  0x8f  0x8e  0x90  0x8f  0x8e  0x8f   
0x253 = 0x8d  0x90  0x8e  0x8f  0x8f  0x8d  0x92  0x8f  0x8e  0x90   
0x254 = 0x8c  0x91  0x8f  0x8b  0x91  0x8b  0x8e  0x8d  0x8f  0x8f   
0x255 = 0x8c  0x92  0x8e  0x8f  0x8e  0x8f  0x8e  0x90  0x8f  0x8f   
0x256 = 0x90  0x8f  0x90  0x8f  0x8a  0x90  0x8e  0x8f  0x8f  0x8e   
0x257 = 0x90  0x8f  0x90  0x8f  0x8f  0x8c  0x8f  0x8f  0x90  0x8e   
0x258 = 0x8f  0x90  0x8e  0x90  0x8c  0x8f  0x8f  0x8f  0x90  0x8f   
0x259 = 0x00  0x00  0x00  0x00  0x00  0x00  0x00  0x00  0x00  0x00   
0x25A = 0x8d  0x8e  0x8e  0x8f  0x8f  0x8d  0x92  0x8e  0x8e  0x90   
0x25B = 0x8e  0x94  0x8e  0x8c  0x90  0x90  0x8f  0x8f  0x8f  0x8e   
0x25C = 0x8d  0x8e  0x8e  0x8d  0x8f  0x8e  0x8f  0x90  0x8f  0x90   
0x25D = 0x91  0x8f  0x8e  0x8f  0x8e  0x8f  0x8d  0x8f  0x8f  0x8e   
0x25E = 0x8e  0x8e  0x8f  0x90  0x8e  0x8f  0x8e  0x8e  0x8e  0x8e   
0x25F = 0x90  0x8f  0x8f  0x8e  0x8f  0x8f  0x8f  0x8d  0x90  0x8f   

Perhaps, some magic write needs to happen before reads return meaningful
values?

So that's where I am stuck.

------------------------------------------------------------------
Ed Sternin, Physics Dept, Brock U, St.Catharines, Ontario, L2S 3A1
       http://www.physics.BrockU.CA/faculty/sternin.html
 voice:(905)688-5550x3414 FAX:(905)682-9020 email: ed...@br...