On Fri, Nov 14, 2014 at 02:44:15PM +0300, kbuild test robot wrote:
> [ You would have to enable transparent huge page tables on a 32 bit
> system to trigger this bug and I don't think that's possible.
It is. We have THP on 32-bit x86.
> I don't think Smatch will complain about this if you have the cross
> function database turned on because it knows the value of size in that
> case. But most people don't build the database so it might be worth
> silencing this bug? Should I even bother sending these email for
> non-bugs? Let me know. -dan ]
>
> tree: git://git.cmpxchg.org/linux-mmotm.git master
> head: e668fb4c5c5e6de5b9432bd36d83b3a0b4ce78e8
> commit: be7c8db9daa43935912bc8c898ecea99b32d805b [120/306] mm: fix huge zero page accounting in smaps report
>
> fs/proc/task_mmu.c:474 smaps_account() warn: should 'size << 12' be a 64 bit type?
This should fix the issue.
diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c
index 8fd00743bd4d..de80a887d98e 100644
--- a/fs/proc/task_mmu.c
+++ b/fs/proc/task_mmu.c
@@ -464,17 +464,16 @@ static void smaps_account(struct mem_size_stats *mss, struct page *page,
mss->shared_dirty += size;
else
mss->shared_clean += size;
- mss->pss += (size << PSS_SHIFT) / mapcount;
+ mss->pss += ((u64)size << PSS_SHIFT) / mapcount;
} else {
if (dirty || PageDirty(page))
mss->private_dirty += size;
else
mss->private_clean += size;
- mss->pss += (size << PSS_SHIFT);
+ mss->pss += (u64)size << PSS_SHIFT;
}
}
-
static void smaps_pte_entry(pte_t *pte, unsigned long addr,
struct mm_walk *walk)
{
--
Kirill A. Shutemov
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>