[mythtv] PVR-250 support

[Dennis Cartier]

> It indicates that there's some data corruption somewhere, but,
> I'm not really
> sure how it could happen/where it could be -- could be in the
> frontend, in
> the wire transfer, or in the backend.

Hmm the Seg Fault is still happening. Anywhere from once every few hours, to
within a few minutes of starting. Unfortunately I am running a CVS version
built from a hacked up copy of the Gentoo ebuild and therefore have no
debugging symbols, but here is what I do have.

First Run - Ran for ~2+ hours:

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 131081 (LWP 30124)]
0x0805fb91 in
QValueListPrivate<QString>::QValueListPrivate(QValueListPrivate<QString>
const&) ()
(gdb) bt
#0  0x0805fb91 in
QValueListPrivate<QString>::QValueListPrivate(QValueListPrivate<QString>
const&) ()
#1  0x42107380 in ?? ()
Cannot access memory at address 0x8
(gdb)

Next Run - Ran for >10 minutes:

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 131081 (LWP 30143)]
0x40678aed in QString::QString(QString const&) ()
   from /usr/qt/3/lib/libqt-mt.so.3
(gdb) bt
#0  0x40678aed in QString::QString(QString const&) ()
   from /usr/qt/3/lib/libqt-mt.so.3

I no longer think this is related to the previous message 'unknown command:
??ERY_RECORDER' as I have seen it go by a few times with no Segfault
resulting.

I am going to back my memory timings off a bit as I might have made them too
aggressive yesterday when I was trying to wring every last bit of
performance out of this machine to eliminate the stutering.

Oops just checked, wrong machine, it was the frontend machine I was
tweaking, this is the backend dual Celeron and it's ram settings are using
the SPD setting.

Any ideas?