[pvrusb2] HVR-1950 driver loading problem

[JE Geiger]

Well, this has gone from strange to bizarre.

The first machine that I was testing was an AMD 64 x2, dual core.

I tried another machine which was AMD Sempron.  Same problems of repetitive
resets and Oops all the way around.

I finally tried the HVR-1950 on a machine that is Intel Pentium D dual core
and it exhibits none of the problems, driver loads and intializes properly.
I used modprobe 0xfffffff and it is fine.  It is at Fedora 10 and I plan to
move it to Fedora 11 soon, but may wait until I get my HVR-1950 video
capture done.

What is really funny is that the first machine is my main mythtv machine and
had been working properly with the HVR-1950 back in April, prior to
upgrading to the new Fedora.

Am I seeing a gcc bad code generation problem where the code on the Intel is
fine, but the AMD code is not?  Is this even a possibility?  I thought the
compiler generated a "family" of code and each processor just executed it
based on it's instruction set.

Do you think it would make sense to recompile my kernel using an earlier
version of gcc, like the one shipped with Fedora 10?

On Tue, Jun 16, 2009 at 7:57 PM, Mike Isely <isely at isely.net> wrote:

> Wish I could suggest something here, but right now I'm still at a loss.
> if ANYONE else is seeing this that would at least rule out your specific
> environment / hardware combination.
>