[mythtv-users] Mythfrontend idle cpu consumption help

[Michael T. Dean]

On 11/16/2009 10:17 AM, Mark Garland wrote:
> I don't know anything about eglibc.
>
> http://www.eglibc.org/home/ says
> "Embedded GLIBC (EGLIBC) is a variant of the GNU C Library (GLIBC)  that is
> designed to work well on embedded systems. EGLIBC strives to be source and
> binary compatible with GLIBC. EGLIBC's goals include reduced footprint,
> configurable components, better support for cross-compilation and
> cross-testing."
> I'm guessing we're using it for drawing the frontend in some way, although I
> thought that was QT.
>
> If we are really confident that our issue is not with the myth code and is
> caused by this library in particular, how would you (or someone else) feel
> about raising an issue on http://www.eglibc.org/issues/ ?
>
> I would offer to do this myself although as I have no idea what this library
> really does, I'm not sure I would be able to push things along too well,
> although would be happy to try if people think this is a good idea and
> no-one else is willing/able.

We'd need a lot more testing, and--ideally--identification of the actual 
problem.

At a minimum, we'd need to find out if it's only the people using eglibc 
(Ubuntu/Debian/???) having issues.  And, we need to define what "having 
issues" is.  Based on my recollection of the thread, we seem to have 
both users with around 15% CPU usage (of a "real" processor) and users 
with 100% CPU usage (possibly only because they're underpowered) processors.

And, even if it is eglibc users, it's quite possible (based on the Qt 
code) that the problem isn't eglibc, but is Qt's configure-time checking 
of the system's support for CLOCK_MONOTONIC--or some other Qt 
pre-compile test--that "does the wrong thing".

So, for right now, let's focus on finding out just who/what distros are 
affected and how much so.

Mike