[mythtv-users] everyday i have to make sure myth worked the day before

[Tom Lichti]

On Thu, Apr 21, 2011 at 8:44 AM, Bruce Taber <b.taber at comcast.net> wrote:
> On 04/21/2011 07:43 AM, Bruce Taber wrote:
> Sorry for the multiple posts on this. Just trying to provide detail on
> what is being found in this process.
>> On 04/21/2011 07:13 AM, Bruce Taber wrote:
>>> 1. git bisect start
>>> 2. git bisect bad
>>> 3. git bisect good <your commit version from above>
> This step is only needed for the other systems. I am not doing git
> bisect on the three (MBE, SBE, FE) systems. The FE system is used for
> the bisect. This step synchronizes the backend systems.
>>> 4. git checkout <your commit version from above>
>>> 5. compile and run tests - most likely needs make distclean & configure
>>> 6. git bisect good or bad depending on test results
>>> 6a. git bisect next - git chooses the version for next test
> Do steps 4-6a until the bad commit is found substituting the version in
> step 4 for the next version git chooses. That can be determined with a
> git describe on the system used for the bisect.
>
>>> 7. git bisect reset - once the bad commit is found

Well I feel a bit stupid. I managed to get the git bisect thing to
work, and did the 7 required compiles, and did not come across the
problem at all during testing. The only problem is I forgot to test
the slave backend during all of it. As soon as I started it up, it
killed the master, the same as before. So at least now I know the
trigger, I will go through the git bisect process again, but this time
testing the SBE in the mix...at least my system is stable without it
though, so the WAF is better.

<sheepish grin>

Tom