On Thu, Feb 22, 2018 at 11:47:57AM -0800, Daniel Lustig wrote: > On 2/22/2018 10:27 AM, Peter Zijlstra wrote: > > On Thu, Feb 22, 2018 at 10:13:17AM -0800, Paul E. McKenney wrote: > >> So we have something that is not all that rare in the Linux kernel > >> community, namely two conflicting more-or-less concurrent changes. > >> This clearly needs to be resolved, either by us not strengthening the > >> Linux-kernel memory model in the way we were planning to or by you > >> strengthening RISC-V to be no weaker than PowerPC for these sorts of > >> externally viewed release-acquire situations. > >> > >> Other thoughts? > > > > Like said in the other email, I would _much_ prefer to not go weaker > > than PPC, I find that PPC is already painfully weak at times. > > Sure, and RISC-V could make this work too by using RCsc instructions > and/or by using lightweight fences instead. It just wasn't clear at > first whether smp_load_acquire() and smp_store_release() were RCpc, > RCsc, or something else, and hence whether RISC-V would actually need > to use something stronger than pure RCpc there. Likewise for > spin_unlock()/spin_lock() and everywhere else this comes up. > > As Paul's email in the other thread observed, RCpc seems to be > OK for smp_load_acquire()/smp_store_release() at least according > to the current LKMM herd spec. Unlock/lock are stronger already > I guess. But if there's an active proposal to strengthen them all > to something stricter than pure RCpc, then that's good to know. > > My understanding from earlier discussions is that ARM has no plans > to use their own RCpc instruction for smp_load_acquire() instead > of their RCsc instructions. Is that still true? If they were to > use the RCpc load there, that would cause them to have the same > problem we're discussing here, right? Just checking.
Agreed. No plans to use the LDAPR instruction in Linux. Will