[pvrusb2] New driver snapshot: pvrusb2-mci-20060329

[Mike Isely]

I've updated the pvrusb2 driver to fix the recently discovered 
instability.  At least I *think* it's fixed.  The root cause was that the 
driver initialization sequence was doing something to disrupt endpoint 
communication (i.e. reseting it) without regard for the fact that other 
threads might be using that same pathway to I2C operations.  Stomping on 
the endpoint at that instance can cause a command to fail, which is how we 
were getting that I2C timeout that precipitated the explosition in 
msp3400.  I did one experiment that positively indicated this interference 
was taking place and since fixing it, nobody who has tried this has seen 
the problem reappear.  This was a single line fix.

You can find the driver & associated info in the usual place:

   http://www.isely.net/pvrusb2.html

Note: Though I haven't tested for this yet, I seriously doubt this fix 
will do anything to help other still lingering problems with the new 
hardware.  The symptoms don't match up.  So no, there's still stability 
work to do involving the new hardware, and hopefully once again the old 
hardware should resume being stable again.

   -Mike

-- 
                         |         Mike Isely          |     PGP fingerprint
      Spammers Die!!     |                             | 03 54 43 4D 75 E5 CC 92
                         |   isely @ pobox (dot) com   | 71 16 01 E2 B5 F5 C1 E8
                         |                             |