On Feb 19, 2016, at 11:08 AM, Noel Chiappa <jnc at
mercury.lcs.mit.edu> wrote:
So I recently discovered that the LTC feature on the BDV11 bootstrap board
has a minor issue: the latch that stores the 'LTC enabled' bit is _not_
cleared by INIT (unlike every other PDP-11 device I've ever heard of), but
only by a direct store into the LTC CSR, _or_ power cycling (BDCOK, to be
exact).
This means that once you turn the LTC on using the BDV11, neither an INIT
instruction, nor a 'Start' command to ODT, will disable it! Needless to say,
it tends to scramble the booting process when an LTC interrupt shows up
before the software is ready for one...
Yuck.
I suppose this doesn't show up as an issue in practice because normal boot sequencing
starts with the processor at PR7, and it stays there until device setup has been done.
Presumably OSs that support this misbegotten device know to clear the CSR before lowering
the processor priority.
paul