SpeedTouch and ISO mode, stats, ADSL Max.

[David Woodhouse]

On Tue, 2006-03-21 at 14:58 +0100, Duncan Sands wrote:
> Another possibility is that using a tasklet to process received urbs
> was a bad idea: too slow.  Sounds unlikely to me, maybe you have some
> insight? 

I agree. It'll introduce latency but not throughput -- as long as the
tasklet is correctly handling more than one packet at a time (which it
is), it really shouldn't be a problem.

>  In order to determine this, what I really need is a way in the
> kernel to measure the wall-time between being at point A in the code,
> and being at point B, where the time difference is small (this makes
> using jiffies problematic).  That way I can measure eg: the time
> between scheduling the tasklet and when it is actually run.  Any
> suggestions?

#include <asm/timex.h> and use get_cycles(). It uses the TSC on i386,
timebase on PPC, etc. 

It's interesting to note what Alistair said about the red light coming
on when bulk transfers weren't sufficient, but never when it maxes out
in ISO mode -- it really does sound like the bottleneck could be in the
modem itself, and we have to prod it differently to make it go faster.

-- 
dwmw2