Bert Freudenberg wrote:
> 
> On Thu, 13 Jul 2000, Raab, Andreas wrote:
> 
> > > On my machine I get a 25% speedup from this:
> > >
> > > 0 tinyBenchmark went up from
> > >     '45325779 bytecodes/sec; 1455887 sends/sec'
> > > to
> > >     '55555555 bytecodes/sec; 1636262 sends/sec'
> > > Get it from CVS at squeak.sourceforge.net :-)
> > >
> > > Now, is this still slower than the Windows VM? This is a 2xPIII/600.
> >
> > PIII/750 (Dell Inspiron 5000):
> >       '72562368 bytecodes/sec; 2080785 sends/sec'
> >
> > Seems about right to me considering the 1/3 speed advantage.
> 
> I found a Windows machine in our lab that seems very similar to my linux
> box (2xPIII/600) and I get
>         '63054187 bytecodes/sec; 1740489 sends/sec'
> there, which still leaves a significant difference.
> 
> On my linux machine, even when Squeak is idle, it takes up more than 50%
> cpu time. Anybody else experiencing this?

Do you use the last version from SF (I think so)? There was an
improvement, which has changed the CPU load from 100% to about 10% for
me:

I wrote:
> 
> New Unix_2.8 alpha_02 in SourceForge: added fairly portable sleep if
> there isn't a socketPollFunction (see below).
> 
> Avoids eating of processor cycles...
> 
> Stephan
> 
> src/sqXWindow.c:

Please check this file!

> 
> int ioRelinquishProcessorForMicroseconds(int microSeconds)
> {
>   /* sleep in select() for immediate response to socket i/o */
> #ifdef HEADLESS
>   if ( socketPollFunction ) { socketPollFunction(microSeconds,0);}
>   /* aioPollForIO(microSeconds, 0); */
> #else
>   if ( socketPollFunction ) { socketPollFunction(microSeconds,stXfd);}
>   /* aioPollForIO(microSeconds, stXfd); */
> #endif
>   else { /* fairly portable sleep if there isn't a socketPollFunction */
>     struct timeval tv;
>     tv.tv_sec = microSeconds / 1000000;
>     tv.tv_usec = microSeconds % 1000000;
>     select(0, NULL, NULL, NULL, &tv);
>   }
>   return microSeconds;
> }
> 

Greetings,

Stephan

> 
> -- Bert

-- 
Stephan Rudlof (sr at evolgo.de)
   "Genius doesn't work on an assembly line basis.
    You can't simply say, 'Today I will be brilliant.'"
    -- Kirk, "The Ultimate Computer", stardate 4731.3