On Thu, Sep 6, 2012 at 7:08 PM, warmcat <email address hidden> wrote:
> In terms of decoding V8 I don't know how smooth we should expect it. It
> will be software-only, looking at top might give a clue if it's cpu
> bound as might be expected.
Yeah, it can be optimized but I believe the best solution here would
be to use gstreamer with hwdecode support, something that it might be
possible with the gstreamer specific changes from TI.
> About the crash, since we found that Androidization patch nastiness the
> default suspicion about low probability mm trouble is that. So retry
> with tilt-3.4 basis b71a842 or better will be interesting.
This is the most interesting point of this bug, will make sure the
latest kernel is part of tomorrow's image.
On Thu, Sep 6, 2012 at 7:08 PM, warmcat <email address hidden> wrote:
> In terms of decoding V8 I don't know how smooth we should expect it. It
> will be software-only, looking at top might give a clue if it's cpu
> bound as might be expected.
Yeah, it can be optimized but I believe the best solution here would
be to use gstreamer with hwdecode support, something that it might be
possible with the gstreamer specific changes from TI.
> About the crash, since we found that Androidization patch nastiness the
> default suspicion about low probability mm trouble is that. So retry
> with tilt-3.4 basis b71a842 or better will be interesting.
This is the most interesting point of this bug, will make sure the
latest kernel is part of tomorrow's image.