2.5 has the new watchdog.9 component. Here is what I observed:
1. No transition on watchdog.input-0, 0->1 transition on watchdog.enable-in:
a) watchdog.out-ok goes high for watchdog.timeout-0, then goes low since no
transitions.
2. transitions on watchdog.input-0 within watchdog.timeout-0,
0->1 transition on watchdog.enable-in
watchdog.out-ok goes high (OK)
(f) 1->0 transition on watchdog.enable-in:
(c) watchdog.out-ok STAYS high (!!!)
transitions on watchdog.input-0 stop:
(e) watchdog.out-ok STAYS high (!!!)
the watchdog.9 manpage says:
watchdog.enable-in bit in (default: FALSE)
If TRUE, forces out-ok to be false. (b) Additionally, if a
timeout
occurs on any input, this pin must be set FALSE and TRUE again
to re-start the monitoring of input pins.
watchdog.ok-out bit out (default: FALSE)
OK output. This pin is true only if enable-in is TRUE (d) and
no
timeout has been detected (g). This output can be connected to
the
enable input of a charge_pump or stepgen (in v mode), to provide
a heartbeat signal to external monitoring hardware.
(a) contradicts (b). It does go high on enable rising edge, but turns off again
after timout. IMO it should stay low if there are no transitions.
(c) contradicts (d)
(e) contradicts (d) and (g).
---
Since this is potentially dangerous, and I want to avoid shouting 'bug!' when I
might have some setup or conclusion error, I would be grateful if somebody else
could reproduce this.
- Michael
------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure contains a
definitive record of customers, application performance, security
threats, fraudulent activity and more. Splunk takes this data and makes
sense of it. Business sense. IT sense. Common sense.
http://p.sf.net/sfu/splunk-d2dcopy1
_______________________________________________
Emc-developers mailing list
Emc-developers@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-developers
