From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756878Ab3FDOTA (ORCPT ); Tue, 4 Jun 2013 10:19:00 -0400 Received: from youngberry.canonical.com ([91.189.89.112]:46828 "EHLO youngberry.canonical.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755660Ab3FDOIS (ORCPT ); Tue, 4 Jun 2013 10:08:18 -0400 From: Luis Henriques To: linux-kernel@vger.kernel.org, stable@vger.kernel.org, kernel-team@lists.ubuntu.com Cc: Mikulas Patocka , Alasdair G Kergon , Luis Henriques Subject: [PATCH 41/70] dm bufio: avoid a possible __vmalloc deadlock Date: Tue, 4 Jun 2013 15:06:57 +0100 Message-Id: <1370354846-14213-42-git-send-email-luis.henriques@canonical.com> X-Mailer: git-send-email 1.8.1.2 In-Reply-To: <1370354846-14213-1-git-send-email-luis.henriques@canonical.com> References: <1370354846-14213-1-git-send-email-luis.henriques@canonical.com> X-Extended-Stable: 3.5 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org 3.5.7.14 -stable review patch. If anyone has any objections, please let me know. ------------------ From: Mikulas Patocka commit 502624bdad3dba45dfaacaf36b7d83e39e74b2d2 upstream. This patch uses memalloc_noio_save to avoid a possible deadlock in dm-bufio. (it could happen only with large block size, at most PAGE_SIZE << MAX_ORDER (typically 8MiB). __vmalloc doesn't fully respect gfp flags. The specified gfp flags are used for allocation of requested pages, structures vmap_area, vmap_block and vm_struct and the radix tree nodes. However, the kernel pagetables are allocated always with GFP_KERNEL. Thus the allocation of pagetables can recurse back to the I/O layer and cause a deadlock. This patch uses the function memalloc_noio_save to set per-process PF_MEMALLOC_NOIO flag and the function memalloc_noio_restore to restore it. When this flag is set, all allocations in the process are done with implied GFP_NOIO flag, thus the deadlock can't happen. This should be backported to stable kernels, but they don't have the PF_MEMALLOC_NOIO flag and memalloc_noio_save/memalloc_noio_restore functions. So, PF_MEMALLOC should be set and restored instead. Signed-off-by: Mikulas Patocka Signed-off-by: Alasdair G Kergon [ luis: backport to 3.5 as suggested: - inlined functions memalloc_noio_save and memalloc_noio_restore - replaced PF_MEMALLOC_NOIO by PF_MEMALLOC ] Signed-off-by: Luis Henriques --- drivers/md/dm-bufio.c | 26 +++++++++++++++++++++++++- 1 file changed, 25 insertions(+), 1 deletion(-) diff --git a/drivers/md/dm-bufio.c b/drivers/md/dm-bufio.c index c0fc827..e9097af 100644 --- a/drivers/md/dm-bufio.c +++ b/drivers/md/dm-bufio.c @@ -321,6 +321,9 @@ static void __cache_size_refresh(void) static void *alloc_buffer_data(struct dm_bufio_client *c, gfp_t gfp_mask, enum data_mode *data_mode) { + unsigned noio_flag; + void *ptr; + if (c->block_size <= DM_BUFIO_BLOCK_SIZE_SLAB_LIMIT) { *data_mode = DATA_MODE_SLAB; return kmem_cache_alloc(DM_BUFIO_CACHE(c), gfp_mask); @@ -334,7 +337,28 @@ static void *alloc_buffer_data(struct dm_bufio_client *c, gfp_t gfp_mask, } *data_mode = DATA_MODE_VMALLOC; - return __vmalloc(c->block_size, gfp_mask, PAGE_KERNEL); + + /* + * __vmalloc allocates the data pages and auxiliary structures with + * gfp_flags that were specified, but pagetables are always allocated + * with GFP_KERNEL, no matter what was specified as gfp_mask. + * + * Consequently, we must set per-process flag PF_MEMALLOC_NOIO so that + * all allocations done by this process (including pagetables) are done + * as if GFP_NOIO was specified. + */ + + if (gfp_mask & __GFP_NORETRY) { + noio_flag = current->flags & PF_MEMALLOC; + current->flags |= PF_MEMALLOC; + } + + ptr = __vmalloc(c->block_size, gfp_mask, PAGE_KERNEL); + + if (gfp_mask & __GFP_NORETRY) + current->flags = (current->flags & ~PF_MEMALLOC) | noio_flag; + + return ptr; } /* -- 1.8.1.2