Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932274AbbFIMFq (ORCPT ); Tue, 9 Jun 2015 08:05:46 -0400 Received: from mail-pa0-f54.google.com ([209.85.220.54]:33518 "EHLO mail-pa0-f54.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753008AbbFIMFj (ORCPT ); Tue, 9 Jun 2015 08:05:39 -0400 From: Sergey Senozhatsky To: Andrew Morton Cc: Minchan Kim , Christoph Lameter , Pekka Enberg , Joonsoo Kim , Michal Hocko , David Rientjes , linux-mm@kvack.org, linux-kernel@vger.kernel.org, sergey.senozhatsky.work@gmail.com, Sergey Senozhatsky Subject: [RFC][PATCH 0/5] do not dereference NULL pools in pools' destroy() functions Date: Tue, 9 Jun 2015 21:04:48 +0900 Message-Id: <1433851493-23685-1-git-send-email-sergey.senozhatsky@gmail.com> X-Mailer: git-send-email 2.4.3.368.g7974889 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2645 Lines: 93 Hello, RFC Proposed by Andrew Morton: https://lkml.org/lkml/2015/6/8/583 The existing pools' destroy() functions do not allow NULL pool pointers; instead, every destructor() caller forced to check if pool is not NULL, which: a) requires additional attention from developers/reviewers b) may lead to a NULL pointer dereferences if (a) didn't work First 3 patches tweak - kmem_cache_destroy() - mempool_destroy() - dma_pool_destroy() to handle NULL pointers. Basically, this patch set will: 1) Can prevent us from still undiscovered NULL pointer dereferences. (like the one that was addressed in https://lkml.org/lkml/2015/6/5/262) 2) Make a cleanup possible. Things like: [..] if (xhci->segment_pool) dma_pool_destroy(xhci->segment_pool); .. if (xhci->device_pool) dma_pool_destroy(xhci->device_pool); .. if (xhci->small_streams_pool) dma_pool_destroy(xhci->small_streams_pool); .. if (xhci->medium_streams_pool) dma_pool_destroy(xhci->medium_streams_pool); [..] or [..] fail_dma_pool: if (IS_QLA82XX(ha) || ql2xenabledif) { dma_pool_destroy(ha->fcp_cmnd_dma_pool); ha->fcp_cmnd_dma_pool = NULL; } fail_dl_dma_pool: if (IS_QLA82XX(ha) || ql2xenabledif) { dma_pool_destroy(ha->dl_dma_pool); ha->dl_dma_pool = NULL; } fail_s_dma_pool: dma_pool_destroy(ha->s_dma_pool); ha->s_dma_pool = NULL; [..] may now be simplified. 0004 and 0005 are not so necessary, simply because there are not so many users of these two (added for pool's destroy() functions consistency): -- zpool_destroy_pool() -- zs_destroy_pool() So, 0004 and 0005 can be dropped. - zbud does kfree() in zbud_destroy_pool(), so I didn't touch it. Sergey Senozhatsky (5): mm/slab_common: allow NULL cache pointer in kmem_cache_destroy() mm/mempool: allow NULL `pool' pointer in mempool_destroy() mm/dmapool: allow NULL `pool' pointer in dma_pool_destroy() mm/zpool: allow NULL `zpool' pointer in zpool_destroy_pool() mm/zsmalloc: allow NULL `pool' pointer in zs_destroy_pool() mm/dmapool.c | 3 +++ mm/mempool.c | 3 +++ mm/slab_common.c | 3 +++ mm/zpool.c | 3 +++ mm/zsmalloc.c | 3 +++ 5 files changed, 15 insertions(+) -- 2.4.3.368.g7974889 -- 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/