mythtranscode segfault [patch]

[adiemus <mythtv@…>]

I recently upgraded my 0.21-fixed myth environment to 0.25-fixes. I'm chasing down several problems, once of which is that my newly transcoded content looks terrible (it's very blocky and seems to frequently corrupt a bit on scene changes) and that mythtranscode refuses to respect the sampling rate I set for audio transcode.

Anyway, in trying to troubleshoot these issues by running mythtranscode manually, I ran into an immediate and easily reproducible segfault. (Command, output and backtrace attached)

The source of the crash is ultimately at dtvrecorder.cpp:35 in the LOC preprocessor macro which dereferences tvrec without verifying that it's not NULL. (In the case of a transcode, it's explicitly NULL, leading to the crash in DTVRecorder::ResetForNewFile?())

I'm far too unfamiliar with myth's codebase to understand the proper way to fix the problem, but a simple workaround is to validate tvrec first, which seems to at least allow me to run mythtranscode manually so I can get about finding the source of my other issues.

Very simple patch attached. (Hopefully in a useful format)

[adiemus <mythtv@…>]

manual command and output

[adiemus <mythtv@…>]

gdb backtrace

[adiemus <mythtv@…>]

simple patch

[Daniel Thor Kristjansson <danielk@…>]

In ac36239755c1eb4e241e801c4c3b30f2f41b123d/mythtv:

Fixes #10668. Fixes segfault in mythtranscode use of DTVRecorder.

This was discovered and diagnosed by Chris Tracy <mythtv@…>.

The problem is a debugging macro that assumes the recorder is being controlled by TVRec.

[Daniel Thor Kristjansson <danielk@…>]

In d06433833bd0411cd3ff82e3810f006f4529da81/mythtv:

Fixes #10668. Fixes segfault in mythtranscode use of DTVRecorder.

This was discovered and diagnosed by Chris Tracy <mythtv@…>.

The problem is a debugging macro that assumes the recorder is being controlled by TVRec.