From: OGAWA Hirofumi Subject: Re: [PATCH 2/5] fat: add fat_fallocate operation Date: Wed, 30 Oct 2013 10:27:25 +0900 Message-ID: <87zjprwnsy.fsf@devron.myhome.or.jp> References: <1381329640-11871-1-git-send-email-linkinjeon@gmail.com> <87hac2yga3.fsf@devron.myhome.or.jp> Mime-Version: 1.0 Content-Type: text/plain Cc: akpm@linux-foundation.org, linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org, a.sahrawat@samsung.com, Namjae Jeon To: Namjae Jeon Return-path: In-Reply-To: (Namjae Jeon's message of "Wed, 30 Oct 2013 08:24:45 +0900") Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-fsdevel.vger.kernel.org Namjae Jeon writes: >>> + /* Release unwritten fallocated blocks on file release. */ >>> + if (round_up(inode->i_size, sb->s_blocksize) < >>> + MSDOS_I(inode)->i_disksize && inode->i_nlink != 0) >>> + fat_truncate_blocks(inode, inode->i_size); >>> + >>> truncate_inode_pages(&inode->i_data, 0); >>> if (!inode->i_nlink) { >>> inode->i_size = 0; [...] >> And inode->i_size should be ->mmu_private? > Sorry, I don't understand your question yet... > Could you elaborate more ? "if" part is checking "round_up(->i_size) < ->i_disksize", but I guess fallocate() truncation should be only needed if there is uninitialized data. So, I felt that part should check ->mmu_private and ->i_disksize. Thanks. -- OGAWA Hirofumi