[Vm-dev] Re: nanoseconds problem in cocoa cog

[Esteban Lorenzano]

mmm... no, it looks like now it is hanging in aioPoll :(
I'm testing with a squeak 4.2 and the problem is evident and automatic... you can see the problem, if you want :)

cheers,
Esteban

El 11/04/2011, a las 6:00p.m., Esteban Lorenzano escribió:

> Hi,
> I'm experimenting some "random hangs" with my latest builds of cocoa cog (any jit version, I don't know jet if stack too)
> The application was hanging (but not crashing) each X random time and with no information, even running with gdb. After a while testing, I found that the problem was gere: 
> 
> /* sleep for microSeconds or until i/o becomes possible, avoiding
>    sleeping in select() if timeout too small */
> 
> int aioSleepForUsecs(int microSeconds)
> {
> #if defined(HAVE_NANOSLEEP)
>   if (microSeconds < (1000000/60))	/* < 1 timeslice? */
>     {
>       if (!aioPoll(0))
> 	{
> 	  struct timespec rqtp= { 0, microSeconds * 1000 };
> 	  struct timespec rmtp;
> 	  nanosleep(&rqtp, &rmtp);
> 	  microSeconds= 0;			/* poll but don't block */
> 	}
>     }
> #endif
>   return aioPoll(microSeconds);
> }
> 
> and more precisely when calling nanosleep
> for now, I disabled HAVE_NANOSLEEP definition, and I'm testing (right now, no hangs), but I don't know how this should work. 
> 
> Someone knows?
> 
> Cheers,
> Esteban

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.squeakfoundation.org/pipermail/vm-dev/attachments/20110411/c1764858/attachment.htm