From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail.linuxfoundation.org ([140.211.169.12]:47516 "EHLO mail.linuxfoundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1732033AbeHDB0m (ORCPT ); Fri, 3 Aug 2018 21:26:42 -0400 Date: Fri, 3 Aug 2018 16:28:18 -0700 From: Andrew Morton To: Minchan Kim Cc: LKML , Sergey Senozhatsky , Tino Lehnig , stable@vger.kernel.org, Jens Axboe Subject: Re: [PATCH 1/2] zram: remove BD_CAP_SYNCHRONOUS_IO with writeback feature Message-Id: <20180803162818.8c4719232d6dfece21ebea30@linux-foundation.org> In-Reply-To: <20180803025143.GA86818@rodete-desktop-imager.corp.google.com> References: <20180802051112.86174-1-minchan@kernel.org> <20180802141304.d0589ddc5f8213429ab3b565@linux-foundation.org> <20180803025143.GA86818@rodete-desktop-imager.corp.google.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: stable-owner@vger.kernel.org List-ID: On Fri, 3 Aug 2018 11:51:43 +0900 Minchan Kim wrote: > > Is it legitimate to be altering the bdi capabilities at this level? Or > > is this hacky? > > Most of device's bdi capability seems to be static but there are few drivers > which can change capability. For example, BDI_CAP_STABLE_WRITES > > drivers/scsi/iscsi_tcp.c > drivers/md/raid5.c > > I believe it's driver itself advertisement stuff so I hope it's not hack. The bdi is per-disk (per-queue). So if zram changes the bdi for a particular partition then it is accidentally mucking with the other partitions as well, and it shouldn't do that. At least, I think that's how it works - it's been a while. Jens, wake up ;)