Unaccounted memory usage...

[Bjorn Gustavsson]

There is a correction for a memory problem (heap fragments being
larger than needed) in R9B-1.

/Bjorn

Shawn Pearce <spearce@REDACTED> writes:

> Just an interesting note, R9B-0 on Windows XP with the same test
> case uses 24 MB, peak memory.
> 
> Odd how the Linux code uses so much, but the Windows code doesn't.
> 
> Of course, Windows didn't have a large enough backlog on the socket
> with {backlog, 5} to handle more than 64 simultaneous incoming
> connections (Linux did).  :( 
> 
> 
> Shawn Pearce writes:
> > Can someone help me explain this a little?  My beam process is 
> > currently at a little over 200 MB resident heap size.  
> 
> -- 
> Shawn.
> 
>   You'll feel much better once you've given up hope.
> 

-- 
Björn Gustavsson            Ericsson Utvecklings AB
bjorn@REDACTED      ÄT2/UAB/F/P
			    BOX 1505
+46 8 727 56 87 	    125 25 Älvsjö