2015-08-04 12:20:06

by Wang Xiaoqiang

Subject: [PATCH] memory-failure/hwpoison_user_mappings: move the comment about swap cache pages' check to proper location

Hi Naoya,

This patch just move the comment about swap cache pages' check to the
proper location in 'hwpoison_user_mappings' function.

Signed-off-by: Wang Xiaoqiang <[email protected]>
---
mm/memory-failure.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/mm/memory-failure.c b/mm/memory-failure.c
index 1cf7f29..3253abb 100644
--- a/mm/memory-failure.c
+++ b/mm/memory-failure.c
@@ -945,10 +945,6 @@ static int hwpoison_user_mappings(struct page *p, unsigned long pfn,
if (!(PageLRU(hpage) || PageHuge(p)))
return SWAP_SUCCESS;

- /*
- * This check implies we don't kill processes if their pages
- * are in the swap cache early. Those are always late kills.
- */
if (!page_mapped(hpage))
return SWAP_SUCCESS;

@@ -957,6 +953,10 @@ static int hwpoison_user_mappings(struct page *p, unsigned long pfn,
return SWAP_FAIL;
}

+ /*
+ * This check implies we don't kill processes if their pages
+ * are in the swap cache early. Those are always late kills.
+ */
if (PageSwapCache(p)) {
printk(KERN_ERR
"MCE %#lx: keeping poisoned page in swap cache\n", pfn);
--
1.7.10.4



--
thx!
Wang Xiaoqiang

2015-08-04 23:06:43

by Naoya Horiguchi

Subject: Re: [PATCH] memory-failure/hwpoison_user_mappings: move the comment about swap cache pages' check to proper location

On Tue, Aug 04, 2015 at 08:20:38PM +0800, Wang Xiaoqiang wrote:
> Hi Naoya,
>
> This patch just move the comment about swap cache pages' check to the
> proper location in 'hwpoison_user_mappings' function.
>
> Signed-off-by: Wang Xiaoqiang <[email protected]>

Thank you for finding out this.

Acked-by: Naoya Horiguchi <[email protected]>

> ---
> mm/memory-failure.c | 8 ++++----
> 1 file changed, 4 insertions(+), 4 deletions(-)
>
> diff --git a/mm/memory-failure.c b/mm/memory-failure.c
> index 1cf7f29..3253abb 100644
> --- a/mm/memory-failure.c
> +++ b/mm/memory-failure.c
> @@ -945,10 +945,6 @@ static int hwpoison_user_mappings(struct page *p, unsigned long pfn,
> if (!(PageLRU(hpage) || PageHuge(p)))
> return SWAP_SUCCESS;
>
> - /*
> - * This check implies we don't kill processes if their pages
> - * are in the swap cache early. Those are always late kills.
> - */
> if (!page_mapped(hpage))
> return SWAP_SUCCESS;
>
> @@ -957,6 +953,10 @@ static int hwpoison_user_mappings(struct page *p, unsigned long pfn,
> return SWAP_FAIL;
> }
>
> + /*
> + * This check implies we don't kill processes if their pages
> + * are in the swap cache early. Those are always late kills.
> + */
> if (PageSwapCache(p)) {
> printk(KERN_ERR
> "MCE %#lx: keeping poisoned page in swap cache\n", pfn);
> --
> 1.7.10.4
>
>
>
> --
> thx!
> Wang Xiaoqiang
>
> -