From mboxrd@z Thu Jan 1 00:00:00 1970 From: Al Viro Subject: Re: [PATCH 12/23] initrd: switch initrd loading to struct file based APIs Date: Mon, 27 Jul 2020 02:50:38 +0100 Message-ID: <20200727015038.GA795125@ZenIV.linux.org.uk> References: <20200714190427.4332-1-hch@lst.de> <20200714190427.4332-13-hch@lst.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <20200714190427.4332-13-hch@lst.de> Sender: linux-kernel-owner@vger.kernel.org To: Christoph Hellwig Cc: linux-kernel@vger.kernel.org, "H. Peter Anvin" , Song Liu , Linus Torvalds , linux-raid@vger.kernel.org, linux-fsdevel@vger.kernel.org List-Id: linux-raid.ids On Tue, Jul 14, 2020 at 09:04:16PM +0200, Christoph Hellwig wrote: > static int __init > -identify_ramdisk_image(int fd, int start_block, decompress_fn *decompressor) > +identify_ramdisk_image(struct file *file, int start_block, > + decompress_fn *decompressor) > { .... > - ksys_lseek(fd, start_block * BLOCK_SIZE, 0); > kfree(buf); > return nblocks; > } You do realize that you've changed behaviour of that thing if start_block != 0? Old one used to leave the things for subsequent reads to start at start_block * 512; new one will ignore that. So after > - nblocks = identify_ramdisk_image(in_fd, rd_image_start, &decompressor); > + nblocks = identify_ramdisk_image(in_file, rd_image_start, &decompressor); you'll have in_file->f_pos left at 0 instead of rd_image_start * 512. ... affecting this > - if (crd_load(in_fd, out_fd, decompressor) == 0) > + if (crd_load(in_file, out_file, decompressor) == 0) ... and this > - ksys_read(in_fd, buf, BLOCK_SIZE); > - ksys_write(out_fd, buf, BLOCK_SIZE); > + kernel_read(in_file, buf, BLOCK_SIZE, &in_file->f_pos); > + kernel_write(out_file, buf, BLOCK_SIZE, &out_file->f_pos); FWIW, I would suggest *not* bothering with ->f_pos and using two global (well, file-static, obviously) variables instead. And kill 'pos' in identify_ramdisk_image() as well - use the in_pos instead.