From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1422902AbcBZU5X (ORCPT ); Fri, 26 Feb 2016 15:57:23 -0500 Received: from mail3-relais-sop.national.inria.fr ([192.134.164.104]:28998 "EHLO mail3-relais-sop.national.inria.fr" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754972AbcBZU5V (ORCPT ); Fri, 26 Feb 2016 15:57:21 -0500 X-IronPort-AV: E=Sophos;i="5.22,505,1449529200"; d="scan'208";a="166129978" Date: Fri, 26 Feb 2016 21:57:17 +0100 (CET) From: Julia Lawall X-X-Sender: jll@localhost6.localdomain6 To: =?ISO-8859-15?Q?Javier_Gonz=E1lez?= cc: Matias Bjorling , linux-block@vger.kernel.org, linux-kernel@vger.kernel.org, kbuild-all@01.org Subject: [PATCH] pblk: fix call_kern.cocci warnings Message-ID: User-Agent: Alpine 2.02 (DEB 1266 2009-07-14) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Convert GFP_KERNEL to GFP_ATOMIC in a function that may be called with locks held. See the call site on line 751, and the call to spin_lock on line 744. Generated by: scripts/coccinelle/locks/call_kern.cocci Signed-off-by: Fengguang Wu Signed-off-by: Julia Lawall --- Perhaps there is a way to reorganize the code such that pblk_get_blk is not called under the lock. tree: https://github.com/OpenChannelSSD/linux.git pblk head: 01eab75e8310ada5c5d387ee0d77c9051f2ba238 commit: 8e755964c7e8b0266a95fb90557823dacf9da12f [11/21] pblk: do not use mempool for write buffer allocation pblk.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) --- a/drivers/lightnvm/pblk.c +++ b/drivers/lightnvm/pblk.c @@ -282,14 +282,14 @@ static struct pblk_block *pblk_get_blk(s void *data; int nentries = dev->sec_per_blk; - data = kmalloc(dev->sec_per_blk * dev->sec_size, GFP_KERNEL); + data = kmalloc(dev->sec_per_blk * dev->sec_size, GFP_ATOMIC); if (!data) { pr_err("nvm: pblk: cannot allocate write buffer for block\n"); return NULL; } entries = kmalloc(dev->sec_per_blk * sizeof(struct buf_entry), - GFP_KERNEL); + GFP_ATOMIC); if (!entries) { pr_err("nvm: pblk: cannot allocate write buffer for block\n"); kfree(data); @@ -297,7 +297,7 @@ static struct pblk_block *pblk_get_blk(s } sync_bitmap = kmalloc(BITS_TO_LONGS(nentries) * sizeof(unsigned long), - GFP_KERNEL); + GFP_ATOMIC); if (!sync_bitmap) { kfree(data); kfree(entries);