[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20170424104539.qa2jva2jytwboeam@hirez.programming.kicks-ass.net>
Date: Mon, 24 Apr 2017 12:45:39 +0200
From: Peter Zijlstra <peterz@...radead.org>
To: Kees Cook <keescook@...omium.org>
Cc: linux-kernel@...r.kernel.org, Eric Biggers <ebiggers3@...il.com>,
Christoph Hellwig <hch@...radead.org>,
"axboe@...nel.dk" <axboe@...nel.dk>,
James Bottomley <James.Bottomley@...senpartnership.com>,
Elena Reshetova <elena.reshetova@...el.com>,
Hans Liljestrand <ishkamiel@...il.com>,
David Windsor <dwindsor@...il.com>, x86@...nel.org,
Ingo Molnar <mingo@...nel.org>, Arnd Bergmann <arnd@...db.de>,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
Jann Horn <jann@...jh.net>, davem@...emloft.net,
linux-arch@...r.kernel.org, kernel-hardening@...ts.openwall.com,
PaX Team <pageexec@...email.hu>
Subject: Re: [PATCH] x86/refcount: Implement fast refcount_t handling
On Fri, Apr 21, 2017 at 03:09:39PM -0700, Kees Cook wrote:
> +static __always_inline __must_check bool refcount_inc_not_zero(refcount_t *r)
> +{
> + const int a = 1;
> + const int u = 0;
> + int c, old;
> +
> + c = atomic_read(&(r->refs));
> + for (;;) {
> + if (unlikely(c == (u)))
> + break;
> + old = atomic_cmpxchg(&(r->refs), c, c + (a));
Please use atomic_try_cmpxchg(), that generates saner code.
> + if (likely(old == c))
> + break;
> + c = old;
> + }
> + return c != u;
> +}
Powered by blists - more mailing lists