Maybe SCO had a point
Posted Aug 22, 2003 1:26 UTC (Fri) by
danw6144 (guest, #14336)
Parent article:
Maybe SCO had a point
There exists a much simpler and innocent explanation for the code similarities
in SysVr4 and linux. Simple necessity -- the mother of all invention.
The ASSERTs and other code similarities didn't exist in older unixes because
they weren't SMP. When implementing new SMP code with locking functions it is
completely natural to have the debugging ASSERTs at points where the locking
is being implimented. It simply wouldn't make sense to sprinkle ASSERTs among
older fragments of the 32v code. Raymond plainly states there were obvious
differences in the two implimentations.
ESR states "The first ASSERT actually differs in a way that isn't trivial (the
Linux version excludes a size argument of zero). And there is a simple
functional reason for the locking calls; 32V didn't do SMP (Symmetric
Multi-Processing), but both SysVr4 and Linux do."
If you were implementing new SMP code where would your debug points be placed ?
Another hint is Raymond's steak dinner remark. It implies SysVr4 suffered from
lack of maintenance -- old debugging junk was left in the code.
(
Log in to post comments)