In strict mode, we should return 0 if there is any hole in pageblock. If
we successfully isolated pages at beginning at pageblock and then have a
bogus compound_order outside pageblock in next page. We will abort search
loop with blockpfn > end_pfn. Although we will limit blockpfn to end_pfn,
we will treat it as a successful isolation in strict mode as blockpfn is
not < end_pfn and return partial isolated pages. Then
isolate_freepages_range may success unexpectly with hole in isolated
range.

Fixes: 9fcd6d2e052ee ("mm, compaction: skip compound pages by order in free scanner")
Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
---
 mm/compaction.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)

diff --git a/mm/compaction.c b/mm/compaction.c
index a40550a33aee..b4d03c9ffe7c 100644
--- a/mm/compaction.c
+++ b/mm/compaction.c
@@ -631,6 +631,14 @@ static unsigned long isolate_freepages_block(struct compact_control *cc,
 				page += (1UL << order) - 1;
 				nr_scanned += (1UL << order) - 1;
 			}
+			/*
+			 * There is a tiny chance that we have read bogus
+			 * compound_order(), so be careful to not go outside
+			 * of the pageblock.
+			 */
+			if (unlikely(blockpfn >= end_pfn))
+				blockpfn = end_pfn - 1;
+
 			goto isolate_fail;
 		}
 
@@ -678,8 +686,7 @@ static unsigned long isolate_freepages_block(struct compact_control *cc,
 		spin_unlock_irqrestore(&cc->zone->lock, flags);
 
 	/*
-	 * There is a tiny chance that we have read bogus compound_order(),
-	 * so be careful to not go outside of the pageblock.
+	 * Be careful to not go outside of the pageblock.
 	 */
 	if (unlikely(blockpfn > end_pfn))
 		blockpfn = end_pfn;
-- 
2.30.0

On Sat, Aug 26, 2023 at 11:36:13PM +0800, Kemeng Shi wrote:
> In strict mode, we should return 0 if there is any hole in pageblock. If
> we successfully isolated pages at beginning at pageblock and then have a
> bogus compound_order outside pageblock in next page. We will abort search
> loop with blockpfn > end_pfn. Although we will limit blockpfn to end_pfn,
> we will treat it as a successful isolation in strict mode as blockpfn is
> not < end_pfn and return partial isolated pages. Then
> isolate_freepages_range may success unexpectly with hole in isolated
> range.
> 
> Fixes: 9fcd6d2e052ee ("mm, compaction: skip compound pages by order in free scanner")
> Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>

Modify the (likely(order <= MAX_ORDER)) block to avoid ever updating
blockpfn past the end of the pageblock. Then remove the second redundant
check.

-- 
Mel Gorman
SUSE Labs

on 8/29/2023 5:39 PM, Mel Gorman wrote:
> On Sat, Aug 26, 2023 at 11:36:13PM +0800, Kemeng Shi wrote:
>> In strict mode, we should return 0 if there is any hole in pageblock. If
>> we successfully isolated pages at beginning at pageblock and then have a
>> bogus compound_order outside pageblock in next page. We will abort search
>> loop with blockpfn > end_pfn. Although we will limit blockpfn to end_pfn,
>> we will treat it as a successful isolation in strict mode as blockpfn is
>> not < end_pfn and return partial isolated pages. Then
>> isolate_freepages_range may success unexpectly with hole in isolated
>> range.
>>
>> Fixes: 9fcd6d2e052ee ("mm, compaction: skip compound pages by order in free scanner")
>> Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
> 
> Modify the (likely(order <= MAX_ORDER)) block to avoid ever updating
> blockpfn past the end of the pageblock. Then remove the second redundant
> check.
> 
Sure, I will improve this in next version.

On 8/26/2023 11:36 PM, Kemeng Shi wrote:
> In strict mode, we should return 0 if there is any hole in pageblock. If
> we successfully isolated pages at beginning at pageblock and then have a
> bogus compound_order outside pageblock in next page. We will abort search
> loop with blockpfn > end_pfn. Although we will limit blockpfn to end_pfn,
> we will treat it as a successful isolation in strict mode as blockpfn is
> not < end_pfn and return partial isolated pages. Then
> isolate_freepages_range may success unexpectly with hole in isolated
> range.
> 
> Fixes: 9fcd6d2e052ee ("mm, compaction: skip compound pages by order in free scanner")
> Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>

LGTM.
Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>

> ---
>   mm/compaction.c | 11 +++++++++--
>   1 file changed, 9 insertions(+), 2 deletions(-)
> 
> diff --git a/mm/compaction.c b/mm/compaction.c
> index a40550a33aee..b4d03c9ffe7c 100644
> --- a/mm/compaction.c
> +++ b/mm/compaction.c
> @@ -631,6 +631,14 @@ static unsigned long isolate_freepages_block(struct compact_control *cc,
>   				page += (1UL << order) - 1;
>   				nr_scanned += (1UL << order) - 1;
>   			}
> +			/*
> +			 * There is a tiny chance that we have read bogus
> +			 * compound_order(), so be careful to not go outside
> +			 * of the pageblock.
> +			 */
> +			if (unlikely(blockpfn >= end_pfn))
> +				blockpfn = end_pfn - 1;
> +
>   			goto isolate_fail;
>   		}
>   
> @@ -678,8 +686,7 @@ static unsigned long isolate_freepages_block(struct compact_control *cc,
>   		spin_unlock_irqrestore(&cc->zone->lock, flags);
>   
>   	/*
> -	 * There is a tiny chance that we have read bogus compound_order(),
> -	 * so be careful to not go outside of the pageblock.
> +	 * Be careful to not go outside of the pageblock.
>   	 */
>   	if (unlikely(blockpfn > end_pfn))
>   		blockpfn = end_pfn;