[mythtv-users] A post about no problem, but still wondering ...

[Gary Buhrmaster]

On Thu, Apr 10, 2014 at 7:32 PM, Saul A. Peebsen <jaglover at gmail.com> wrote:
....
> Is there an explanation to this?

<WAG>

Possibly.  Some hardware (drivers) require
dedicated real memory, and/or additional kernel
memory, and some devices/drivers do not deal
well with OoM conditions in the kernel ("unable
to handle kernel NULL pointer dereference" is
a common result of interesting coding).  The
kernel crash and backtrace would likely tell you
something useful.  Even headless, you can
compile/build/run the kernel with serial debugging
support, and attach a serial cable for the crash
results (or sometimes use your network attached
remote KVM for the debug console).  Or ignore
it all now that it works.

</WAG>