From mboxrd@z Thu Jan 1 00:00:00 1970 From: Linus Torvalds Subject: Re: [git patches] xfs and block fixes for virtually indexed arches Date: Thu, 17 Dec 2009 08:16:12 -0800 (PST) Message-ID: References: <20091216043618.GB9104@hera.kernel.org> <20091217132256.GO28962@bombadil.infradead.org> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Return-path: Received: from smtp1.linux-foundation.org ([140.211.169.13]:59180 "EHLO smtp1.linux-foundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1761484AbZLQQRB (ORCPT ); Thu, 17 Dec 2009 11:17:01 -0500 In-Reply-To: <20091217132256.GO28962@bombadil.infradead.org> Sender: linux-arch-owner@vger.kernel.org List-ID: To: Kyle McMartin Cc: linux-parisc@vger.kernel.org, Linux Kernel Mailing List , James.Bottomley@suse.de, hch@infradead.org, linux-arch@vger.kernel.org, Jens Axboe On Thu, 17 Dec 2009, Kyle McMartin wrote: > > Linus, any word on these? I'd really love to get jejb and hch off > my back. :) I hate them. I don't see what the point of allowing kernel virtual addresses in bio's is. It's wrong. The fact that XFS does that sh*t is an XFS issue. Handle it there. Fix XFS. Or convince me with some really good arguments, and make sure that Jens signs off on the cr*p too. In other words, the thing I object to isn't even the new flushing thing, it's this idiocy: - save off virtual address: .. bio->bi_private = data; .. - do vmalloc_to_page: + if (is_vmalloc_addr(data)) { + flush_kernel_dcache_addr(data); + page = vmalloc_to_page(data); + } else + page = virt_to_page(data); WTF? Why the hell should the block layer support this kind of absolute crap? When has "use random kernel virtual address" ever been a valid thing to do for IO? Why aren't you doing this before you submit the vmalloc range for IO? So no. Not a way in hell do I pull this for 33. And preferably never. Linus