Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754755Ab1CLAkg (ORCPT ); Fri, 11 Mar 2011 19:40:36 -0500 Received: from arroyo.ext.ti.com ([192.94.94.40]:48387 "EHLO arroyo.ext.ti.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752830Ab1CLAke (ORCPT ); Fri, 11 Mar 2011 19:40:34 -0500 From: Omar Ramirez Luna To: Greg KH , l-m Cc: Felipe Contreras , l-o , Ohad Ben-Cohen , Fernando Guzman Lugo , Tuomas Kulve , Omar Ramirez Luna Subject: [PATCH] staging: tidspbridge: protect dmm_map properly Date: Fri, 11 Mar 2011 18:29:06 -0600 Message-Id: <1299889746-16502-1-git-send-email-omar.ramirez@ti.com> X-Mailer: git-send-email 1.7.1 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3553 Lines: 117 Hi Greg, Please consider to apply this patch in the staging tree, as the description says it fixes a crash in tidspbridge driver, this bug was already present but it seems to have surfaced by recent tests made by Felipe and Tuomas. It is an urgent fix for 2.6.38. More on this discussion: http://www.gossamer-threads.com/lists/linux/kernel/1351446 Thanks, - omar From: Felipe Contreras We need to protect not only the dmm_map list, but the individual map_obj's, otherwise, we might be building the scatter-gather list with garbage. So, use the existing proc_lock for that. I observed race conditions which caused kernel panics while running stress tests, also, Tuomas Kulve found it happening quite often in Gumstix Over. This patch fixes those. Cc: Tuomas Kulve Signed-off-by: Felipe Contreras Signed-off-by: Omar Ramirez Luna --- drivers/staging/tidspbridge/rmgr/proc.c | 19 ++++++++++++++----- 1 files changed, 14 insertions(+), 5 deletions(-) diff --git a/drivers/staging/tidspbridge/rmgr/proc.c b/drivers/staging/tidspbridge/rmgr/proc.c index b47d7aa..e2fe165 100644 --- a/drivers/staging/tidspbridge/rmgr/proc.c +++ b/drivers/staging/tidspbridge/rmgr/proc.c @@ -781,12 +781,14 @@ int proc_begin_dma(void *hprocessor, void *pmpu_addr, u32 ul_size, (u32)pmpu_addr, ul_size, dir); + mutex_lock(&proc_lock); + /* find requested memory are in cached mapping information */ map_obj = find_containing_mapping(pr_ctxt, (u32) pmpu_addr, ul_size); if (!map_obj) { pr_err("%s: find_containing_mapping failed\n", __func__); status = -EFAULT; - goto err_out; + goto no_map; } if (memory_give_ownership(map_obj, (u32) pmpu_addr, ul_size, dir)) { @@ -795,6 +797,8 @@ int proc_begin_dma(void *hprocessor, void *pmpu_addr, u32 ul_size, status = -EFAULT; } +no_map: + mutex_unlock(&proc_lock); err_out: return status; @@ -819,21 +823,24 @@ int proc_end_dma(void *hprocessor, void *pmpu_addr, u32 ul_size, (u32)pmpu_addr, ul_size, dir); + mutex_lock(&proc_lock); + /* find requested memory are in cached mapping information */ map_obj = find_containing_mapping(pr_ctxt, (u32) pmpu_addr, ul_size); if (!map_obj) { pr_err("%s: find_containing_mapping failed\n", __func__); status = -EFAULT; - goto err_out; + goto no_map; } if (memory_regain_ownership(map_obj, (u32) pmpu_addr, ul_size, dir)) { pr_err("%s: InValid address parameters %p %x\n", __func__, pmpu_addr, ul_size); status = -EFAULT; - goto err_out; } +no_map: + mutex_unlock(&proc_lock); err_out: return status; } @@ -1726,9 +1733,8 @@ int proc_un_map(void *hprocessor, void *map_addr, (p_proc_object->hbridge_context, va_align, size_align); } - mutex_unlock(&proc_lock); if (status) - goto func_end; + goto unmap_failed; /* * A successful unmap should be followed by removal of map_obj @@ -1737,6 +1743,9 @@ int proc_un_map(void *hprocessor, void *map_addr, */ remove_mapping_information(pr_ctxt, (u32) map_addr, size_align); +unmap_failed: + mutex_unlock(&proc_lock); + func_end: dev_dbg(bridge, "%s: hprocessor: 0x%p map_addr: 0x%p status: 0x%x\n", __func__, hprocessor, map_addr, status); -- 1.7.1 -- 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/