Date: Fri, 06 Sep 1996 13:03:47 -0700 From: Wayne Scott <wscott@ichips.intel.com> To: "David E. Tweten" <tweten@frihet.com> Cc: Bruce Evans <bde@zeta.org.au>, current@freebsd.org Subject: Re: fixing accesses to volatile variable `time' Message-ID: <199609062003.NAA29277@pdxlx008.intel.com> In-Reply-To: Your message of "Thu, 05 Sep 1996 19:22:16 PDT." <199609060222.TAA05756@ns.frihet.com>
next in thread | previous in thread | raw e-mail | index | archive | help
> bde@zeta.org.au said:
> >1. time.tv_sec is long, and accesses to longs are not guaranteed to
> > be atomic. They happen to be atomic on i386's.
>
> To satisfy my curiosity, just who makes this "guarantee?" Obviously,
> access to a bit is inherently atomic, but I don't recall reading any C
> language specification indicating that chars, shorts, or int accesses are
> atomic whereas longs aren't.
Intel guarantees that 32-bit accesses on a 32-bit boundary are atomic on
all 32-bit Intel Architecture processors.
(Pentium Pro manual volume 3 page 7-2)
So as long as you align the time data structure correctly, there is no problem.
-----------
Wayne Scott MD6 Architecture
wscott@ichips.intel.com Work #: (503) 264-4165
Disclaimer: All views expressed are my own opinions, and not necessarily
those of Intel Corporation.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199609062003.NAA29277>