On Mon, Feb 19, 2007 at 01:28:17PM -0800, Christopher Layne wrote: > Absolutely. I don't disagree with this. The issue is the magnitude. > The opteron box has 4gigs of ram, scsi 320 disks, and is running water cooled > at 2.8 ghz. Nothing *normal* can explain such a reason why a Celeron-D can > fork the same /bin/echo's 15 times faster than that box, except for the OS > difference. Even then, the difference is insane. > > 70 288 [main] xargs 5956 child_copy: dll data - hp 0x6C8 low > 0x61118000, high 0x6111D040, res 1 > 75657 75945 [main] xargs 5956 child_copy: dll bss - hp 0x6C8 low > 0x61175000, high 0x6117EC30, res 1 > ^ 903 76848 [main] xargs 5956 child_copy: user heap - hp 0x6C8 low > 0xA20000, high 0xA40000, res 1 > | 44 76892 [main] xargs 5956 child_copy: done > | 44 76936 [main] xargs 5956 child_copy: data - hp 0x6C8 low 0x406000, > high 0x406050, res 1 > | 29 76965 [main] xargs 5956 child_copy: bss - hp 0x6C8 low 0x409000, high > 0x4093A0, res 1 > | 21 76986 [main] xargs 5956 child_copy: done > | > |__ not normal
And the plot thickens. http://www.sourceware.org/ml/cygwin/2006-12/msg00494.html http://www.sourceware.org/ml/cygwin/2006-12/msg00711.html -cl -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/