From mboxrd@z Thu Jan  1 00:00:00 1970
From: vapier.adi@gmail.com (Mike Frysinger)
Date: Sat, 14 Feb 2009 12:37:01 -0500
Subject: [ltt-dev] [RFC git tree] Userspace RCU (urcu) for Linux (repost)
In-Reply-To: <20090214150622.GF7194@linux.vnet.ibm.com>
References: <20090213155507.GA2838@Krystal> <20090213173352.GB4684@Krystal>
	<alpine.LFD.2.00.0902130945350.3179@localhost.localdomain>
	<alpine.LFD.2.00.0902130956400.3179@localhost.localdomain>
	<20090213185411.GB7124@Krystal>
	<20090213193619.GH6854@linux.vnet.ibm.com>
	<8bd0f97a0902132107r3651ce4rc5bc4e027268e2aa@mail.gmail.com>
	<20090214052030.GB7194@linux.vnet.ibm.com>
	<8bd0f97a0902132146t2badab79ue76e0a43d74c95ca@mail.gmail.com>
	<20090214150622.GF7194@linux.vnet.ibm.com>
Message-ID: <8bd0f97a0902140937v22917306m44e062d5b0dfacf6@mail.gmail.com>

On Sat, Feb 14, 2009 at 10:06, Paul E. McKenney wrote:
> Other potential issues include unrelated processes that share memory via
> shmget() or mmap() -- presumably groups of such processes would need to
> be bound to a single CPU?

using PROT_WRITE with MAP_SHARED is a mess with no-mmu already ... in
other words, it isnt in use in practice as most of the time, the
kernel wont even grant it

i dont believe we've looked at shm/ipc ... i'll open a tracker item on
our side to review it
-mike