Skip to contentSkip to conversation

mm: support free hugepage pre zero out

RFC PATCH In review Liang Li 3 participants 4 messages Dec 22, 2020

Subsystems (2)
lore
Share & tools
lore

Liang Li

author
#0 lore raw

RFC PATCH 3/3 mm: support free hugepage pre zero out

This patch add support of pre zero out free hugepage, we can use this feature to speed up page population and page fault handing.

Cc Alexander Duyck <[email protected]>
Cc Mel Gorman <[email protected]>
Cc Andrea Arcangeli <[email protected]>
Cc Dan Williams <[email protected]>
Cc Dave Hansen <[email protected]>
Cc David Hildenbrand <[email protected]>
Cc Michal Hocko <[email protected]>
Cc Andrew Morton <[email protected]>
Cc Alex Williamson <[email protected]>
Cc Michael S. Tsirkin <[email protected]>
Cc Jason Wang <[email protected]>
Cc Mike Kravetz <[email protected]>
Cc Liang Li <[email protected]>
---
 mm/page_prezero.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)
mm/page_prezero.c +17 -0
index c8ce720bfc54..dff4e0adf402 100644
@@ -26,6 +26,7 @@ static unsigned long delay_millisecs = 1000;
 static unsigned long zeropage_enable __read_mostly;
 static DEFINE_MUTEX(kzeropaged_mutex);
 static struct page_reporting_dev_info zero_page_dev_info;
+static struct page_reporting_dev_info zero_hugepage_dev_info;
 
 inline void clear_zero_page_flag(struct page *page, int order)
 {
@@ -69,9 +70,17 @@ static int start_kzeropaged(void)
 		zero_page_dev_info.delay_jiffies = msecs_to_jiffies(delay_millisecs);
 
 		err = page_reporting_register(&zero_page_dev_info);
+
+		zero_hugepage_dev_info.report = zero_free_pages;
+		zero_hugepage_dev_info.mini_order = mini_page_order;
+		zero_hugepage_dev_info.batch_size = batch_size;
+		zero_hugepage_dev_info.delay_jiffies = msecs_to_jiffies(delay_millisecs);
+
+		err |= hugepage_reporting_register(&zero_hugepage_dev_info);
 		pr_info("Zero page enabled\n");
 	} else {
 		page_reporting_unregister(&zero_page_dev_info);
+		hugepage_reporting_unregister(&zero_hugepage_dev_info);
 		pr_info("Zero page disabled\n");
 	}
 
@@ -90,7 +99,15 @@ static int restart_kzeropaged(void)
 		zero_page_dev_info.batch_size = batch_size;
 		zero_page_dev_info.delay_jiffies = msecs_to_jiffies(delay_millisecs);
 
+		hugepage_reporting_unregister(&zero_hugepage_dev_info);
+
+		zero_hugepage_dev_info.report = zero_free_pages;
+		zero_hugepage_dev_info.mini_order = mini_page_order;
+		zero_hugepage_dev_info.batch_size = batch_size;
+		zero_hugepage_dev_info.delay_jiffies = msecs_to_jiffies(delay_millisecs);
+
 		err = page_reporting_register(&zero_page_dev_info);
+		err |= hugepage_reporting_register(&zero_hugepage_dev_info);
 		pr_info("Zero page enabled\n");
 	}
 
Signature · 1 line
2.18.2
#1 lore raw
Quoted 58 lines · Liang Li

This patch add support of pre zero out free hugepage, we can use this feature to speed up page population and page fault handing.

Cc Alexander Duyck <[email protected]>
Cc Mel Gorman <[email protected]>
Cc Andrea Arcangeli <[email protected]>
Cc Dan Williams <[email protected]>
Cc Dave Hansen <[email protected]>
Cc David Hildenbrand <[email protected]>
Cc Michal Hocko <[email protected]>
Cc Andrew Morton <[email protected]>
Cc Alex Williamson <[email protected]>
Cc Michael S. Tsirkin <[email protected]>
Cc Jason Wang <[email protected]>
Cc Mike Kravetz <[email protected]>
Cc Liang Li <[email protected]>
---
 mm/page_prezero.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)
mm/page_prezero.c +11 -0
index c8ce720bfc54..dff4e0adf402 100644
@@ -26,6 +26,7 @@ static unsigned long delay_millisecs = 1000;
 static unsigned long zeropage_enable __read_mostly;
 static DEFINE_MUTEX(kzeropaged_mutex);
 static struct page_reporting_dev_info zero_page_dev_info;
+static struct page_reporting_dev_info zero_hugepage_dev_info;
 
 inline void clear_zero_page_flag(struct page *page, int order)
 {
@@ -69,9 +70,17 @@ static int start_kzeropaged(void)
 		zero_page_dev_info.delay_jiffies = msecs_to_jiffies(delay_millisecs);
 
 		err = page_reporting_register(&zero_page_dev_info);
+
+		zero_hugepage_dev_info.report = zero_free_pages;
+		zero_hugepage_dev_info.mini_order = mini_page_order;
+		zero_hugepage_dev_info.batch_size = batch_size;
+		zero_hugepage_dev_info.delay_jiffies = msecs_to_jiffies(delay_millisecs);
+
+		err |= hugepage_reporting_register(&zero_hugepage_dev_info);
 		pr_info("Zero page enabled\n");
 	} else {
 		page_reporting_unregister(&zero_page_dev_info);
+		hugepage_reporting_unregister(&zero_hugepage_dev_info);
 		pr_info("Zero page disabled\n");
 	}
 
@@ -90,7 +99,15 @@ static int restart_kzeropaged(void)
 		zero_page_dev_info.batch_size = batch_size;
 		zero_page_dev_info.delay_jiffies = msecs_to_jiffies(delay_millisecs);
 
+		hugepage_reporting_unregister(&zero_hugepage_dev_info);
+

mm/page_prezero.c static int restart_kzeropaged(void) · line 113

Quoted 10 lines
mm/page_prezero.c +6 -0
+		zero_hugepage_dev_info.report = zero_free_pages;
+		zero_hugepage_dev_info.mini_order = mini_page_order;
+		zero_hugepage_dev_info.batch_size = batch_size;
+		zero_hugepage_dev_info.delay_jiffies = msecs_to_jiffies(delay_millisecs);
+
 		err = page_reporting_register(&zero_page_dev_info);
+		err |= hugepage_reporting_register(&zero_hugepage_dev_info);
 		pr_info("Zero page enabled\n");
 	}
 

Free page reporting in virtio-balloon doesn't give you any guarantees regarding zeroing of pages. Take a look at the QEMU implementation - e.g., with vfio all reports are simply ignored.

Also, I am not sure if mangling such details ("zeroing of pages") into the page reporting infrastructure is a good idea.

Signature · 3 lines
Thanks,

David / dhildenb
#2 lore raw
Quoted 74 lines · David Hildenbrand
Quoted 58 lines · Liang Li

This patch add support of pre zero out free hugepage, we can use this feature to speed up page population and page fault handing.

Cc Alexander Duyck <[email protected]>
Cc Mel Gorman <[email protected]>
Cc Andrea Arcangeli <[email protected]>
Cc Dan Williams <[email protected]>
Cc Dave Hansen <[email protected]>
Cc David Hildenbrand <[email protected]>
Cc Michal Hocko <[email protected]>
Cc Andrew Morton <[email protected]>
Cc Alex Williamson <[email protected]>
Cc Michael S. Tsirkin <[email protected]>
Cc Jason Wang <[email protected]>
Cc Mike Kravetz <[email protected]>
Cc Liang Li <[email protected]>
---
 mm/page_prezero.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)
mm/page_prezero.c +11 -0
index c8ce720bfc54..dff4e0adf402 100644
@@ -26,6 +26,7 @@ static unsigned long delay_millisecs = 1000;
 static unsigned long zeropage_enable __read_mostly;
 static DEFINE_MUTEX(kzeropaged_mutex);
 static struct page_reporting_dev_info zero_page_dev_info;
+static struct page_reporting_dev_info zero_hugepage_dev_info;
 
 inline void clear_zero_page_flag(struct page *page, int order)
 {
@@ -69,9 +70,17 @@ static int start_kzeropaged(void)
 		zero_page_dev_info.delay_jiffies = msecs_to_jiffies(delay_millisecs);
 
 		err = page_reporting_register(&zero_page_dev_info);
+
+		zero_hugepage_dev_info.report = zero_free_pages;
+		zero_hugepage_dev_info.mini_order = mini_page_order;
+		zero_hugepage_dev_info.batch_size = batch_size;
+		zero_hugepage_dev_info.delay_jiffies = msecs_to_jiffies(delay_millisecs);
+
+		err |= hugepage_reporting_register(&zero_hugepage_dev_info);
 		pr_info("Zero page enabled\n");
 	} else {
 		page_reporting_unregister(&zero_page_dev_info);
+		hugepage_reporting_unregister(&zero_hugepage_dev_info);
 		pr_info("Zero page disabled\n");
 	}
 
@@ -90,7 +99,15 @@ static int restart_kzeropaged(void)
 		zero_page_dev_info.batch_size = batch_size;
 		zero_page_dev_info.delay_jiffies = msecs_to_jiffies(delay_millisecs);
 
+		hugepage_reporting_unregister(&zero_hugepage_dev_info);
+

mm/page_prezero.c static int restart_kzeropaged(void) · line 113

Quoted 10 lines
mm/page_prezero.c +6 -0
+		zero_hugepage_dev_info.report = zero_free_pages;
+		zero_hugepage_dev_info.mini_order = mini_page_order;
+		zero_hugepage_dev_info.batch_size = batch_size;
+		zero_hugepage_dev_info.delay_jiffies = msecs_to_jiffies(delay_millisecs);
+
 		err = page_reporting_register(&zero_page_dev_info);
+		err |= hugepage_reporting_register(&zero_hugepage_dev_info);
 		pr_info("Zero page enabled\n");
 	}
 

Free page reporting in virtio-balloon doesn't give you any guarantees regarding zeroing of pages. Take a look at the QEMU implementation - e.g., with vfio all reports are simply ignored.

Quoted 3 lines

Also, I am not sure if mangling such details ("zeroing of pages") into the page reporting infrastructure is a good idea.

Oh, now I get what you are doing here, you rely on zero_free_pages of your other patch series and are not relying on virtio-balloon free page reporting to do the zeroing.

You really should have mentioned that this patch series relies on the other one and in which way.

Signature · 3 lines
Thanks,

David / dhildenb
#3 lore raw
Quoted 8 lines
Quoted 7 lines

Free page reporting in virtio-balloon doesn't give you any guarantees regarding zeroing of pages. Take a look at the QEMU implementation - e.g., with vfio all reports are simply ignored.

Also, I am not sure if mangling such details ("zeroing of pages") into the page reporting infrastructure is a good idea.

Quoted 6 lines

Oh, now I get what you are doing here, you rely on zero_free_pages of your other patch series and are not relying on virtio-balloon free page reporting to do the zeroing.

You really should have mentioned that this patch series relies on the other one and in which way.

I am sorry for that. After I sent out the patch, I realized I should mention that, so I sent out an updated version which added the information you mentioned :)

Thanks !
Liang