[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20160118230324.GF14531@node.shutemov.name>
Date: Tue, 19 Jan 2016 01:03:24 +0200
From: "Kirill A. Shutemov" <kirill@...temov.name>
To: Arnd Bergmann <arnd@...db.de>,
Andrew Morton <akpm@...ux-foundation.org>
Cc: Ebru Akagunduz <ebru.akagunduz@...il.com>,
dan.carpenter@...cle.com, linux-mm@...ck.org,
Linus Torvalds <torvalds@...ux-foundation.org>,
"Kirill A. Shutemov" <kirill.shutemov@...ux.intel.com>,
Rik van Riel <riel@...hat.com>, linux-kernel@...r.kernel.org,
linux-arm-kernel@...ts.infradead.org
Subject: Re: [PATCH] mm: avoid uninitialized variable in tracepoint
On Mon, Jan 18, 2016 at 09:50:26PM +0100, Arnd Bergmann wrote:
> A newly added tracepoint in the hugepage code uses a variable in the
> error handling that is not initialized at that point:
>
> include/trace/events/huge_memory.h:81:230: error: 'isolated' may be used uninitialized in this function [-Werror=maybe-uninitialized]
>
> The result is relatively harmless, as the trace data will in rare
> cases contain incorrect data.
>
> This works around the problem by adding an explicit initialization.
>
> Signed-off-by: Arnd Bergmann <arnd@...db.de>
> Fixes: 7d2eba0557c1 ("mm: add tracepoint for scanning pages")
There's the same patch in mm tree, but it got lost on the way to Linus'
tree:
https://ozlabs.org/~akpm/mmots/broken-out/mm-make-optimistic-check-for-swapin-readahead-fix.patch
Andrew?
>
> diff --git a/mm/huge_memory.c b/mm/huge_memory.c
> index b2db98136af9..bb3b763b1829 100644
> --- a/mm/huge_memory.c
> +++ b/mm/huge_memory.c
> @@ -2320,7 +2320,7 @@ static void collapse_huge_page(struct mm_struct *mm,
> pgtable_t pgtable;
> struct page *new_page;
> spinlock_t *pmd_ptl, *pte_ptl;
> - int isolated, result = 0;
> + int isolated = 0, result = 0;
> unsigned long hstart, hend;
> struct mem_cgroup *memcg;
> unsigned long mmun_start; /* For mmu_notifiers */
>
--
Kirill A. Shutemov
Powered by blists - more mailing lists