From mboxrd@z Thu Jan 1 00:00:00 1970 From: Sergey Senozhatsky Subject: Re: [rfc patch] drivers/block/zram: Replace bit spinlocks with rtmutex for -rt Date: Mon, 4 Apr 2016 09:27:57 +0900 Message-ID: <20160404002756.GA6164@swordfish> References: <1458641979.15742.6.camel@gmail.com> <20160330085622.GA21849@linutronix.de> <1459329836.4175.16.camel@gmail.com> <1459390108.4030.5.camel@gmail.com> <20160401104221.GB29603@linutronix.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Mike Galbraith , Minchan Kim , Nitin Gupta , Sergey Senozhatsky , RT To: Sebastian Andrzej Siewior Return-path: Received: from mail-pa0-f46.google.com ([209.85.220.46]:33466 "EHLO mail-pa0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752276AbcDDA1h (ORCPT ); Sun, 3 Apr 2016 20:27:37 -0400 Received: by mail-pa0-f46.google.com with SMTP id zm5so131888069pac.0 for ; Sun, 03 Apr 2016 17:27:37 -0700 (PDT) Content-Disposition: inline In-Reply-To: <20160401104221.GB29603@linutronix.de> Sender: linux-rt-users-owner@vger.kernel.org List-ID: On (04/01/16 12:42), Sebastian Andrzej Siewior wrote: > >They're nondeterministic, and lead to ___might_sleep() splats in -rt. > >OTOH, they're a lot less wasteful than an rtmutex per page. > Applied. > Although I'm disapointed about Sergey's argument against the spinlock in > general due to a few bytes used by lockdep which can be disabled. Not to > mention the usefull debug facility lockdep provides if one decides to > use. I'll take a look. -ss