sPHENIX MAPS tracker discussion

[Martin Purschke <purschke AT bnl.gov>]

All (this merits a wider audience than just Jo and me)

Jo had upgraded the mvtxgtm firmware to v49 and found that the gtm 
doesn't start any longer.

I had first thought that this is due to a known effect that sometimes, 
once in 5 or so reboots, there's a clock glitch and the FFO board is 
dead in the water. That is NOT the cause (and come to think of it, it 
only happens with the external clock that we aren't using here).

The support software was a really old version that predates the new 
features with the local modes etc. Bottom line, v49 needs the latest 
software that I just put on. That in turn needs some support libraries, 
such as the postgres libs and other stuff. I upgraded all of that.

Also, on the main GL1/GTMs I had split the setup into a part where we 
start the rpc server, and the part where we initialize registers at 
boot. The former is done with systemd, and the latter is done in 
/etc/rc.local. So the two can now be done independently.

We need to init the regs only once after a boot (hence /etc/rc.local), 
while there may be reasons (new versions, etc) to restart the server. 
That in itself will not touch any regs; the server will map itself to 
the register space and sees what is already there.

So your gtm is now like the others, and it works reliably. I made it so 
that only vGTM4 is enabled, and here it is:

> root@mvtxgtm:~# gtm_status
> 0x10
> root@mvtxgtm:~# gtm_start
> root@mvtxgtm:~# gtm_status
> 0x100010

Enjoy!

Best,
        Martin

--
Martin L. Purschke, Ph.D.        ;   purschke AT bnl.gov
                                 ;   http://www.phenix.bnl.gov/~purschke
                                 ;
Brookhaven National Laboratory   ;   phone: +1-631-344-5244
Physics Department Bldg 510 C    ;   fax:   +1-631-344-3253
Upton, NY 11973-5000             ;   skype: mpurschke
-----------------------------------------------------------------------