[En-Nut-Discussion] Preview Fix 060626

[Michael Jones]

Here just a little trace of the queue / interrupt behavior...

My conclusion of the below is that somewhere the queue gets "busted".

A = Alloc
E = Empty
R = RxReady
O = RxOverrun
{ = Enter NutEventWait(&ni->ni_rx_rdy, 2000)
} = Leave NutEventWait(&ni->ni_rx_rdy, 2000)
~ = Leave NutEventWait(&ni->ni_rx_rdy, 2000) - Timeout

{.A.E.A.E.R}
{.A.E.A.E.R}
{.A.E.A.E.R}
{.A.E.A.E.R}
{.A.E.A.E.R}
{.A.E.A.E.R}
{.A.E.A.E.R}
{.A.E.A.E.R}
{.A.E.A.E.R}
{.A.E.A.E.R}	//	Flood starts here...
{.A.E.R}		// A	very small time delta from A to B, B to C,
etc..
{.R}			// B
{.R}			// C
{.R			// D	Does not leave the Event Wait
.O			// E	time delta D to E is longer then e.g. C to D
.O			//    which would indicate .O because } never
happened
.O
.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.
.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O
.O.O
.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O
.O.O
.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O
.O.O
.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O
.O.O
.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O
.O.O
.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O
.O.O
.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O
.O.O
.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O
.O.O
.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O
.O.O
.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O.O
.O.O // Flood ends here...
.A.E.A.E.A.E.A.E.A.E.A.E.A.E.A.E.A.E.A.E.A.E.A.E.A.E.A.E.A.E.A.E.A.E.A.E.A.E
.A.E
.A.E.A.E.A.E.A.E.A.E.A.E.A.E.A.E
~}					//	util timeout
{.A.E.A.E.R}	//	back to normal
{.A.E.A.E.R}
{.A.E.A.E.R}
{.A.E.A.E.R}
{.A.E.A.E.R}
{.A.E.A.E.R}
{.A.E.A.E.R}
{.A.E.A.E.R}
{.A.E.A.E.R}
{.A.E.A.E.R}