Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Sender: Minchan Kim <minchan.kim@gmail.com>
Date:   Wed, 11 May 2022 14:46:12 -0700
From:   Minchan Kim <minchan@kernel.org>
To:     John Hubbard <jhubbard@nvidia.com>
Cc:     Andrew Morton <akpm@linux-foundation.org>,
        linux-mm <linux-mm@kvack.org>,
        LKML <linux-kernel@vger.kernel.org>,
        "Paul E . McKenney" <paulmck@kernel.org>,
        John Dias <joaodias@google.com>,
        David Hildenbrand <david@redhat.com>
Subject: Re: [PATCH v4] mm: fix is_pinnable_page against on cma page
Message-ID: <YnwupNzDNv7IbjRQ@google.com>
References: <20220510211743.95831-1-minchan@kernel.org>
 <857d21da-5de2-fa3e-b1ce-41cc1cfb0191@nvidia.com>
 <Ynr10y6pkH6WwbQo@google.com>
 <bb6e76f1-2cbc-893d-c8ab-3ecd3fcae2a5@nvidia.com>
 <Ynr+wTCQpyh8+vOD@google.com>
 <2ffa7670-04ea-bb28-28f8-93a9b9eea7e8@nvidia.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <2ffa7670-04ea-bb28-28f8-93a9b9eea7e8@nvidia.com>
Precedence: bulk

On Tue, May 10, 2022 at 09:32:05PM -0700, John Hubbard wrote:
> On 5/10/22 17:09, Minchan Kim wrote:
> > On Tue, May 10, 2022 at 04:58:13PM -0700, John Hubbard wrote:
> > > On 5/10/22 4:31 PM, Minchan Kim wrote:
> > > > > > +	int __mt = get_pageblock_migratetype(page);
> > > > > > +	int mt = __READ_ONCE(__mt);
> > > > > 
> > > > > Although I saw the email discussion about this in v2, that discussion
> > > > > didn't go far enough. It started with "don't use volatile", and went
> > > > > on to "try __READ_ONCE() instead", but it should have continued on
> > > > > to "you don't need this at all".
> > > > 
> > > > That's really what I want to hear from experts so wanted to learn
> > > > "Why". How could we prevent refetching of the mt if we don't use
> > > > __READ_ONCE or volatile there?
> > > > 
> > > > > 
> > > > > Because you don't. There is nothing you are racing with, and adding
> > > > > __READ_ONCE() in order to avoid a completely not-going-to-happen
> > > > > compiler re-invocation of a significant code block is just very wrong.
> > > > > 
> > > > > So let's just let it go entirely. :)
> > > > 
> > > > Yeah, once it's clear for everyone, I am happy to remove the
> > > > unnecessary lines.
> > > > 
> > > > > 
> > > > > > +
> > > > > > +	if (mt == MIGRATE_CMA || mt == MIGRATE_ISOLATE)
> > > > > 
> > > 
> > > With or without __READ_ONCE() or volatile or anything else,
> > > this code will do what you want. Which is: loosely check
> > > for either of the above.
> > > 
> > > What functional problem do you think you are preventing
> > > with __READ_ONCE()? Because I don't see one.
> > 
> > I discussed the issue at v1 so please take a look.
> > 
> > https://lore.kernel.org/all/YnFvmc+eMoXvLCWf@google.com/
> 
> I read that, but there was never any real justification there for needing
> to prevent a re-read of mt, just a preference: "I'd like to keep use the local
> variable mt's value in folloing conditions checks instead of refetching
> the value from get_pageblock_migratetype."
> 
> But I don't believe that there is any combination of values of mt that
> will cause a problem here.
> 
> I also think that once we pull in experts, they will tell us that the
> compiler is not going to re-run a non-trivial function to re-fetch a
> value, but I'm not one of those experts, so that's still arguable. But
> imagine what the kernel code would look like if every time we call
> a large function, we have to consider if it actually gets called some
> arbitrary number of times, due to (anti-) optimizations by the compiler.
> This seems like something that is not really happening.

Maybe, I might be paranoid since I have heard too subtle things
about how compiler could changes high level language code so wanted
be careful especially when we do lockless-stuff.

Who cares when we change the large(?) function to small(?) function
later on? I'd like to hear from experts to decide it.