* [PATCH v2] mm/page_isolation: don't putback unisolated page
@ 2021-09-13 11:50 Miaohe Lin
2021-09-13 12:15 ` David Hildenbrand
0 siblings, 1 reply; 3+ messages in thread
From: Miaohe Lin @ 2021-09-13 11:50 UTC (permalink / raw)
To: akpm
Cc: david, jhubbard, vbabka, iamjoonsoo.kim, linux-mm, linux-kernel,
linmiaohe
If __isolate_free_page() failed, due to zone watermark check, the page is
still on the free list. But this page will be put back to free list again
via __putback_isolated_page() now. This may trigger page->flags checks in
__free_one_page() if PageReported is set. Or we will corrupt the free list
because list_add() will be called for pages already on another list.
Fixes: 3c605096d315 ("mm/page_alloc: restrict max order of merging on isolated pageblock")
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
---
v1->v2:
Add VM_WARN_ON to catch unexpected isolating failure.
---
mm/page_isolation.c | 9 +++++++--
1 file changed, 7 insertions(+), 2 deletions(-)
diff --git a/mm/page_isolation.c b/mm/page_isolation.c
index f93cc63d8fa1..f67c4c70f17f 100644
--- a/mm/page_isolation.c
+++ b/mm/page_isolation.c
@@ -94,8 +94,13 @@ static void unset_migratetype_isolate(struct page *page, unsigned migratetype)
buddy = page + (buddy_pfn - pfn);
if (!is_migrate_isolate_page(buddy)) {
- __isolate_free_page(page, order);
- isolated_page = true;
+ isolated_page = !!__isolate_free_page(page, order);
+ /*
+ * Isolating a free page in an isolated pageblock
+ * is expected to always work as watermarks don't
+ * apply here.
+ */
+ VM_WARN_ON(!isolated_page);
}
}
}
--
2.23.0
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH v2] mm/page_isolation: don't putback unisolated page
2021-09-13 11:50 [PATCH v2] mm/page_isolation: don't putback unisolated page Miaohe Lin
@ 2021-09-13 12:15 ` David Hildenbrand
2021-09-13 12:37 ` Miaohe Lin
0 siblings, 1 reply; 3+ messages in thread
From: David Hildenbrand @ 2021-09-13 12:15 UTC (permalink / raw)
To: Miaohe Lin, akpm; +Cc: jhubbard, vbabka, iamjoonsoo.kim, linux-mm, linux-kernel
On 13.09.21 13:50, Miaohe Lin wrote:
> If __isolate_free_page() failed, due to zone watermark check, the page is
> still on the free list. But this page will be put back to free list again
> via __putback_isolated_page() now. This may trigger page->flags checks in
> __free_one_page() if PageReported is set. Or we will corrupt the free list
> because list_add() will be called for pages already on another list.
>
> Fixes: 3c605096d315 ("mm/page_alloc: restrict max order of merging on isolated pageblock")
> Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
> ---
> v1->v2:
> Add VM_WARN_ON to catch unexpected isolating failure.
> ---
> mm/page_isolation.c | 9 +++++++--
> 1 file changed, 7 insertions(+), 2 deletions(-)
>
> diff --git a/mm/page_isolation.c b/mm/page_isolation.c
> index f93cc63d8fa1..f67c4c70f17f 100644
> --- a/mm/page_isolation.c
> +++ b/mm/page_isolation.c
> @@ -94,8 +94,13 @@ static void unset_migratetype_isolate(struct page *page, unsigned migratetype)
> buddy = page + (buddy_pfn - pfn);
>
> if (!is_migrate_isolate_page(buddy)) {
> - __isolate_free_page(page, order);
> - isolated_page = true;
> + isolated_page = !!__isolate_free_page(page, order);
> + /*
> + * Isolating a free page in an isolated pageblock
> + * is expected to always work as watermarks don't
> + * apply here.
> + */
> + VM_WARN_ON(!isolated_page);
Didn't we discuss that this can never happen?
In that case, the patch description+subject needs updating and the
"Fixes" line would be wrong.
> }
> }
> }
>
--
Thanks,
David / dhildenb
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH v2] mm/page_isolation: don't putback unisolated page
2021-09-13 12:15 ` David Hildenbrand
@ 2021-09-13 12:37 ` Miaohe Lin
0 siblings, 0 replies; 3+ messages in thread
From: Miaohe Lin @ 2021-09-13 12:37 UTC (permalink / raw)
To: David Hildenbrand, akpm
Cc: jhubbard, vbabka, iamjoonsoo.kim, linux-mm, linux-kernel
On 2021/9/13 20:15, David Hildenbrand wrote:
> On 13.09.21 13:50, Miaohe Lin wrote:
>> If __isolate_free_page() failed, due to zone watermark check, the page is
>> still on the free list. But this page will be put back to free list again
>> via __putback_isolated_page() now. This may trigger page->flags checks in
>> __free_one_page() if PageReported is set. Or we will corrupt the free list
>> because list_add() will be called for pages already on another list.
>>
>> Fixes: 3c605096d315 ("mm/page_alloc: restrict max order of merging on isolated pageblock")
>> Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
>> ---
>> v1->v2:
>> Add VM_WARN_ON to catch unexpected isolating failure.
>> ---
>> mm/page_isolation.c | 9 +++++++--
>> 1 file changed, 7 insertions(+), 2 deletions(-)
>>
>> diff --git a/mm/page_isolation.c b/mm/page_isolation.c
>> index f93cc63d8fa1..f67c4c70f17f 100644
>> --- a/mm/page_isolation.c
>> +++ b/mm/page_isolation.c
>> @@ -94,8 +94,13 @@ static void unset_migratetype_isolate(struct page *page, unsigned migratetype)
>> buddy = page + (buddy_pfn - pfn);
>> if (!is_migrate_isolate_page(buddy)) {
>> - __isolate_free_page(page, order);
>> - isolated_page = true;
>> + isolated_page = !!__isolate_free_page(page, order);
>> + /*
>> + * Isolating a free page in an isolated pageblock
>> + * is expected to always work as watermarks don't
>> + * apply here.
>> + */
>> + VM_WARN_ON(!isolated_page);
>
>
> Didn't we discuss that this can never happen?
>
> In that case, the patch description+subject needs updating and the "Fixes" line would be wrong.
>
I thought this might need a Fixes tag... Will change all of these. Many thanks.
>> }
>> }
>> }
>>
>
>
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2021-09-13 12:37 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-09-13 11:50 [PATCH v2] mm/page_isolation: don't putback unisolated page Miaohe Lin
2021-09-13 12:15 ` David Hildenbrand
2021-09-13 12:37 ` Miaohe Lin
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.