From mboxrd@z Thu Jan 1 00:00:00 1970 From: Trond Myklebust Subject: Re: [PATCH 01/11] svcrdma: Add a type for keeping NFS RPC mapping Date: Tue, 24 Jun 2008 16:38:54 -0400 Message-ID: <1214339935.7156.50.camel@localhost> References: <12120836962076-git-send-email-tom@opengridcomputing.com> <12120836962324-git-send-email-tom@opengridcomputing.com> <20080616194858.GA29446@fieldses.org> <485D2CEF.404@opengridcomputing.com> <20080623182707.GC24373@fieldses.org> <486062CD.6010800@opengridcomputing.com> <20080624195838.GE15786@fieldses.org> Mime-Version: 1.0 Content-Type: text/plain Cc: Tom Tucker , linux-nfs@vger.kernel.org To: "J. Bruce Fields" Return-path: Received: from mail-out1.uio.no ([129.240.10.57]:55814 "EHLO mail-out1.uio.no" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752353AbYFXUjB (ORCPT ); Tue, 24 Jun 2008 16:39:01 -0400 In-Reply-To: <20080624195838.GE15786@fieldses.org> Sender: linux-nfs-owner@vger.kernel.org List-ID: On Tue, 2008-06-24 at 15:58 -0400, J. Bruce Fields wrote: > My understanding is that it's A--e.g. checkpatch.pl has a dumb regex > that just checks for these assignments and whines about them however > they're used. > > Google doesn't find me any more detailed discussion of the policy. We had that discussion about 10 years ago. What followed was a general purge of all zero initialisations of global and static variables in the kernel, and the official policy is that initialising those variables is wrong, and should _not_ be done. The reason is that BSS is automatically zeroed out, and so adding an initialiser causes unnecessary code bloat and slows down the boot process. Trond