Date: Wed, 20 May 2009 15:24:48 -0400 From: John Baldwin <jhb@freebsd.org> To: Jeff Roberson <jroberson@jroberson.net> Cc: Dag-Erling =?iso-8859-1?q?Sm=F8rgrav?= <des@des.no>, arch@freebsd.org Subject: Re: lockless file descriptor lookup Message-ID: <200905201524.49090.jhb@freebsd.org> In-Reply-To: <alpine.BSF.2.00.0905200851390.981@desktop> References: <alpine.BSF.2.00.0905111720280.981@desktop> <20090514131613.T1224@besplex.bde.org> <alpine.BSF.2.00.0905200851390.981@desktop>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wednesday 20 May 2009 2:59:52 pm Jeff Roberson wrote: > On Thu, 14 May 2009, Bruce Evans wrote: > > Anyway, you probably need atomics that have suitable memory barriers. > > Memory barriers must affect the compiler and make it perform refreshes > > for them to work, so you shouldn't need any volatile casts. E.g., all > > atomic store operations (including cmpset) have release semantics even > > if they aren't spelled with "_rel" or implemented using inline asm. > > On amd64 and i386, they happen to be implemented using inline asm with > > "memory" clobbers. The "memory" clobbers force refreshes of all > > non-local variables. > > So I think I need an _acq memory barrier on the atomic cmpset of the > refcount to prevent speculative loading of the fd_ofiles array pointer by > the processor and the volatile in the second dereference as I have it > now to prevent caching of the pointer by the compiler. What do you think? > > The references prior to the atomic increment have no real ordering > requirements. Only the ones afterwards need to be strict so that we can > verify the results. I think having the _acq is correct and that the "memory" clobber it contains will force the compiler to reload fd_ofiles without needing the volatile cast (and thus that you can remove the volatile cast altogether and just add the _acq barrier). -- John Baldwin
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200905201524.49090.jhb>