From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-pg0-f72.google.com (mail-pg0-f72.google.com [74.125.83.72]) by kanga.kvack.org (Postfix) with ESMTP id 584F0831F4 for ; Tue, 9 May 2017 06:02:17 -0400 (EDT) Received: by mail-pg0-f72.google.com with SMTP id m13so91006522pgd.12 for ; Tue, 09 May 2017 03:02:17 -0700 (PDT) Received: from szxga02-in.huawei.com (szxga02-in.huawei.com. [45.249.212.188]) by mx.google.com with ESMTPS id i123si15691047pfc.167.2017.05.09.03.02.15 for (version=TLS1 cipher=AES128-SHA bits=128/128); Tue, 09 May 2017 03:02:16 -0700 (PDT) Message-ID: <5911927A.6080708@huawei.com> Date: Tue, 9 May 2017 17:57:14 +0800 From: zhong jiang MIME-Version: 1.0 Subject: Re: [PATCH] mm: fix the memory leak after collapsing the huge page fails References: <1494317557-49680-1-git-send-email-zhongjiang@huawei.com> <7d5fd103-f997-e445-2ce6-2e44deed33d8@suse.cz> In-Reply-To: <7d5fd103-f997-e445-2ce6-2e44deed33d8@suse.cz> Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org List-ID: To: Vlastimil Babka Cc: akpm@linux-foundation.org, mgorman@techsingularity.net, hannes@cmpxchg.org, kirill.shutemov@linux.intel.com, linux-mm@kvack.org On 2017/5/9 16:42, Vlastimil Babka wrote: > On 05/09/2017 10:12 AM, zhongjiang wrote: >> From: zhong jiang >> >> Current, when we prepare a huge page to collapse, due to some >> reasons, it can fail to collapse. At the moment, we should >> release the preallocate huge page. > Yeah, looks like the leak is there... > >> Signed-off-by: zhong jiang >> --- >> mm/khugepaged.c | 1 + >> 1 file changed, 1 insertion(+) >> >> diff --git a/mm/khugepaged.c b/mm/khugepaged.c >> index 7cb9c88..3f5749e 100644 >> --- a/mm/khugepaged.c >> +++ b/mm/khugepaged.c >> @@ -1080,6 +1080,7 @@ static void collapse_huge_page(struct mm_struct *mm, >> result = SCAN_SUCCEED; >> out_up_write: >> up_write(&mm->mmap_sem); >> + put_page(new_page); > This doesn't seem correct. > - the put_page() will be called also on success, so a premature free? > - the out_nolock: case should be also handled > - collapse_shmem() seems to have the same problem > >> out_nolock: >> trace_mm_collapse_huge_page(mm, isolated, result); >> return; >> > > . > >Subject: [PATCH v2] mm: fix the memory leak after collapsing the huge page fails Current, when we prepare a huge page to collapse, due to some reasons, it can fail to collapse. At the moment, we should release the preallocate huge page. Signed-off-by: zhong jiang --- mm/khugepaged.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/mm/khugepaged.c b/mm/khugepaged.c index 7cb9c88..586b1f1 100644 --- a/mm/khugepaged.c +++ b/mm/khugepaged.c @@ -1082,6 +1082,8 @@ static void collapse_huge_page(struct mm_struct *mm, up_write(&mm->mmap_sem); out_nolock: trace_mm_collapse_huge_page(mm, isolated, result); + if (page != NULL && result != SCAN_SUCCEED) + put_page(new_page); return; out: mem_cgroup_cancel_charge(new_page, memcg, true); @@ -1555,6 +1557,8 @@ static void collapse_shmem(struct mm_struct *mm, } out: VM_BUG_ON(!list_empty(&pagelist)); + if (page != NULL && result != SCAN_SUCCEED) + put_page(new_page); /* TODO: tracepoints */ } -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org