incidence of errno in scientific programming

David Hough sun!Eng!David.Hough
Thu Jun 27 20:09:22 PDT 1991


Just to see how often errno was relied upon in
scientific programming, I examined the source codes for some applications that
we have accumulated at Sun for compiler testing:

IMSL C library and validation suite, 5MB source:
	no uses

UCB ERL spice3b1, 2 MB source:	
	one use after getc (not ANSI standard)

lamd, mm1sim, walker, 100K source total, unrelated customer applications:
	no uses



More information about the Numeric-interest mailing list