MIME-Version: 1.0
In-Reply-To: <55AC8CC4.1020802@monom.org>
References: <20150622190553.GZ3644@twins.programming.kicks-ass.net>
	<5589285C.2010100@bmw-carit.de>
	<20150623143411.GA25159@twins.programming.kicks-ass.net>
	<558973A7.6010407@bmw-carit.de>
	<20150623175012.GD3644@twins.programming.kicks-ass.net>
	<20150623193624.GH18673@twins.programming.kicks-ass.net>
	<20150624084648.GB27873@gmail.com>
	<558A75EA.40905@bmw-carit.de>
	<55938163.7020606@monom.org>
	<CA+55aFy0nKSbr6YRUV62Ns9E53yJ9Wf1KHd2+mh+G12Co-tsiA@mail.gmail.com>
	<20150702094155.GW19282@twins.programming.kicks-ass.net>
	<55AC8CC4.1020802@monom.org>
Date: Mon, 20 Jul 2015 11:44:04 -0700
Message-ID: <CA+55aFw+pO=qxnust4mwSTDMyZPZrqyJMfwJkNopUh_4xncsLg@mail.gmail.com>
Subject: Re: [RFC][PATCH 00/13] percpu rwsem -v2
From: Linus Torvalds <torvalds@linux-foundation.org>
To: Daniel Wagner <wagi@monom.org>
Cc: Peter Zijlstra <peterz@infradead.org>, Ingo Molnar <mingo@kernel.org>,
        Oleg Nesterov <oleg@redhat.com>,
        Paul McKenney <paulmck@linux.vnet.ibm.com>, Tejun Heo <tj@kernel.org>,
        Ingo Molnar <mingo@redhat.com>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        der.herr@hofr.at, Davidlohr Bueso <dave@stgolabs.net>,
        Rik van Riel <riel@redhat.com>, Al Viro <viro@zeniv.linux.org.uk>,
        Jeff Layton <jlayton@poochiereds.net>
Content-Type: text/plain; charset=UTF-8
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1422
Lines: 36

On Sun, Jul 19, 2015 at 10:53 PM, Daniel Wagner <wagi@monom.org> wrote:
>
> Turning them back into arch_spinlock_t gives almost the same numbers as
> with spinlock_t.
>
> Then Peter suggested to change the code to
>
>         preempt_disable();
>         spin_unlock();
>         preempt_enable_no_resched();
>
> to verify if arch_spin_lock() is buggy and does not disable preemption
> and we see a lock holder preemption on non virt setups.

Hmm. "arch_spin_lock()" isn't _supposed_ to disable preemption. The
caller should do that (possibly by disabling interrupts). See
include/linux/spinlock_api_smp.h for details.

But yes, that's a *very* subtle difference between "arch_spin_lock()"
and "spin_lock()". The former doesn't do lockdep or other debugging
and it doesn't disable preemption. So they are not interchangeable.

The current lglocks uses arch_spin_lock exactly because it does not
*want* lockdep tracking (it does its own) and because it does its own
preemption handling.

So saying "verify if arch_spin_lock() is buggy and does not disable
preemption" is complete BS. If arch_spin_lock() were to disable
preemption, _that_ would be a bug.

                Linus
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/