Date: Fri, 21 Jul 1995 04:32:13 +1000 From: Bruce Evans <bde@zeta.org.au> To: dillon@blob.best.net, terry@cs.weber.edu Cc: bugs@FreeBSD.org, davidg@Root.COM, mckusick@McKusick.COM Subject: Re: possible ffs_vget() race condition Message-ID: <199507201832.EAA04386@godzilla.zeta.org.au>
next in thread | raw e-mail | index | archive | help
>A lot of the system configuration stuff is bogus/unportable/both. The >use of manifest constants to do the descriptor structs in sysctl.h >is patently bogus. >The CTLTYPE_QUAD definition and the size variant CTLTYPE_INT are both >rather annoying; these should be sized values, and QUAD should probably >not be allowed at all, since 64 bit values are (as yet) intolerably >unportable. Enumerated or defined types are more flexible. You can support an arbitrary number of scalar types or type/representation pairs by converting the values using a routine indexed by the type. >Not to mention the use of a shift of an embeded constant '63' and a -1 >to get the RLIM_INFINITY value in resource.h. >Instead of (~(u_quad_t)0). The shift method has the advantage of giving the correct result :-). You meant ((~(u_quad_t)0) >> 1). This depends on u_quad_t being larger than int. Bruce
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199507201832.EAA04386>