Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753140AbZA3Pw7 (ORCPT ); Fri, 30 Jan 2009 10:52:59 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751751AbZA3Pwv (ORCPT ); Fri, 30 Jan 2009 10:52:51 -0500 Received: from mba.ocn.ne.jp ([122.1.235.107]:65099 "EHLO smtp.mba.ocn.ne.jp" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751651AbZA3Pwv (ORCPT ); Fri, 30 Jan 2009 10:52:51 -0500 Date: Sat, 31 Jan 2009 00:52:50 +0900 (JST) Message-Id: <20090131.005250.95064457.anemo@mba.ocn.ne.jp> To: matthias@kaehlcke.net Cc: akpm@linux-foundation.org, David.Woodhouse@intel.com, vapier.adi@gmail.com, linux-mtd@lists.infradead.org, linux-kernel@vger.kernel.org Subject: Re: [BUG] physmap_flash_probe() frees memory still in use From: Atsushi Nemoto In-Reply-To: <20090130134317.GV9390@traven> References: <20090130134317.GV9390@traven> X-Fingerprint: 6ACA 1623 39BD 9A94 9B1A B746 CA77 FE94 2874 D52F X-Pgp-Public-Key: http://wwwkeys.pgp.net/pks/lookup?op=get&search=0x2874D52F X-Mailer: Mew version 5.2 on Emacs 21.4 / Mule 5.0 (SAKAKI) Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3602 Lines: 115 On Fri, 30 Jan 2009 14:43:18 +0100, Matthias Kaehlcke wrote: > as i reported earlier in this month > (http://lkml.org/lkml/2009/1/12/141) MTD partition names are screwed > up in 2.6.29-rcX. > > commit 176bf2e0f10ecf1d20a97db3bd5bb2e6ba0b5668 deals with a memory > leak and frees the pointer array of mtd_partition after the call to > add_mtd_partitions(). the problem is that mtd_table[x]->name still > points to the freed memory. Oh Yes, I had missed that point. And looking at physmap driver again, I found that del_mtd_partition() and del_mtd_device() are abused. Though these APIs are robust enough, deleting right things in right order would be better. Could you test this patch? I cannot test it until Monday. ------------------------------------------------------ Subject: [MTD] physmap: Fix wrong free and del_mtd_{partition,device} From: Atsushi Nemoto commit 176bf2e0f10ecf1d20a97db3bd5bb2e6ba0b5668 deals with a memory leak and frees the pointer array of mtd_partition after the call to add_mtd_partitions(). the problem is that mtd_table[x]->name still points to the freed memory. Aldo physmap_flash_remove() should call del_mtd_partitions() or del_mtd_device() only once. Reported-by: Matthias Kaehlcke Signed-off-by: Atsushi Nemoto --- diff --git a/drivers/mtd/maps/physmap.c b/drivers/mtd/maps/physmap.c index 8774366..4b122e7 100644 --- a/drivers/mtd/maps/physmap.c +++ b/drivers/mtd/maps/physmap.c @@ -29,6 +29,7 @@ struct physmap_flash_info { struct map_info map[MAX_RESOURCES]; #ifdef CONFIG_MTD_PARTITIONS int nr_parts; + struct mtd_partition *parts; #endif }; @@ -45,25 +46,26 @@ static int physmap_flash_remove(struct platform_device *dev) physmap_data = dev->dev.platform_data; -#ifdef CONFIG_MTD_CONCAT - if (info->cmtd != info->mtd[0]) { +#ifdef CONFIG_MTD_PARTITIONS + if (info->nr_parts) { + del_mtd_partitions(info->cmtd); + kfree(info->parts); + } else if (physmap_data->nr_parts) + del_mtd_partitions(info->cmtd); + else del_mtd_device(info->cmtd); +#else + del_mtd_device(info->cmtd); +#endif + +#ifdef CONFIG_MTD_CONCAT + if (info->cmtd != info->mtd[0]) mtd_concat_destroy(info->cmtd); - } #endif for (i = 0; i < MAX_RESOURCES; i++) { - if (info->mtd[i] != NULL) { -#ifdef CONFIG_MTD_PARTITIONS - if (info->nr_parts || physmap_data->nr_parts) - del_mtd_partitions(info->mtd[i]); - else - del_mtd_device(info->mtd[i]); -#else - del_mtd_device(info->mtd[i]); -#endif + if (info->mtd[i] != NULL) map_destroy(info->mtd[i]); - } } return 0; } @@ -86,9 +88,6 @@ static int physmap_flash_probe(struct platform_device *dev) int err = 0; int i; int devices_found = 0; -#ifdef CONFIG_MTD_PARTITIONS - struct mtd_partition *parts; -#endif physmap_data = dev->dev.platform_data; if (physmap_data == NULL) @@ -167,10 +166,11 @@ static int physmap_flash_probe(struct platform_device *dev) goto err_out; #ifdef CONFIG_MTD_PARTITIONS - err = parse_mtd_partitions(info->cmtd, part_probe_types, &parts, 0); + err = parse_mtd_partitions(info->cmtd, part_probe_types, + &info->parts, 0); if (err > 0) { - add_mtd_partitions(info->cmtd, parts, err); - kfree(parts); + add_mtd_partitions(info->cmtd, info->parts, err); + info->nr_parts = err; return 0; } -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/