From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Wed, 31 Jan 2007 11:11:46 -0800 From: Andrew Morton Subject: Re: [patch] simplify shmem_aops.set_page_dirty method Message-Id: <20070131111146.2b29d851.akpm@osdl.org> In-Reply-To: References: Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org Return-Path: To: Hugh Dickins Cc: Ken Chen , linux-mm@kvack.org List-ID: On Wed, 31 Jan 2007 17:17:10 +0000 (GMT) Hugh Dickins wrote: > 2. Please remind me what good __mark_inode_dirty will do for shmem: None that I can think of - tmpfs inodes don't get written back to swap (do they?) > in my patch the equivalent function did nothing beyond SetPageDirty > (your TestSetPageDirty looks better, less redirtying the cacheline). Will test_and_set_bit() avoid dirtying the cacheline? I guess it _could_ do this, and perhaps this depends upon the architecture. Perhaps if (!PageDirty(page)) SetPageDirty(page); would be better here. > The world may have moved on and __mark_inode_dirty now be important, > but I suspect still not - I think it just puts the inode on some > hashlist which serves no good purpose for nowriteback mappings. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org