On Tue, Jul 28, 1998 at 10:57:52AM +0900, Scott Stone wrote: > it returns EINVAL on NetBSD - Linux's behavior could well be different. I wasn't very clear. Yes, EINVAL is returned by NetBSD - but this is a bug itself. The API uses unsigned lengths, so -1 isn't a bogus value (well, perhaps it is actually, if you consider the fact that buffer_start + 0xFFFFFFFFul) >= process space end address, then EINVAL is legitimate). -cw
This archive was generated by hypermail 2b30 : Fri Apr 13 2001 - 14:08:43 PDT