Skip to contentSkip to conversation

mm: filemap: clear idle flag for writes

PATCH Applied Yang Shi 3 participants 6 messages Jun 24, 2020

Subsystems (2)
lore
Share & tools
lore

Yang Shi

author
#0 lore raw

PATCH 1/1 mm: filemap: clear idle flag for writes

Since commit bbddabe2e436aa7869b3ac5248df5c14ddde0cbf ("mm: filemap:
only do access activations on reads"), mark_page_accessed() is called for reads only. But the idle flag is cleared by mark_page_accessed() so the idle flag won't get cleared if the page is write accessed only.

Basically idle page tracking is used to estimate workingset size of workload, noticeable size of workingset might be missed if the idle flag is not maintained correctly.

It seems good enough to just clear idle flag for write operations.

Fixes bbddabe2e436 ("mm: filemap: only do access activations on reads")
Cc Johannes Weiner <[email protected]>
Cc Rik van Riel <[email protected]>
Cc Shakeel Butt <[email protected]>
---
 mm/filemap.c | 6 ++++++
 1 file changed, 6 insertions(+)
mm/filemap.c +6 -0
index f0ae9a6..0589aef 100644
@@ -41,6 +41,7 @@
 #include <linux/delayacct.h>
 #include <linux/psi.h>
 #include <linux/ramfs.h>
+#include <linux/page_idle.h>
 #include "internal.h"
 
 #define CREATE_TRACE_POINTS
@@ -1630,6 +1631,11 @@ struct page *pagecache_get_page(struct address_space *mapping, pgoff_t index,
 
 	if (fgp_flags & FGP_ACCESSED)
 		mark_page_accessed(page);
+	else if (fgp_flags & FGP_WRITE) {
+		/* Clear idle flag for buffer write */
+		if (page_is_idle(page))
+			clear_page_idle(page);
+	}
 
 no_page:
 	if (!page && (fgp_flags & FGP_CREAT)) {
Signature · 1 line
1.8.3.1
maintainer
#1 lore raw
Quoted 25 lines · On Thu, 25 Jun 2020 01:43:32 +0800 Yang Shi <[email protected]> wrote:

Since commit bbddabe2e436aa7869b3ac5248df5c14ddde0cbf ("mm: filemap:
only do access activations on reads"), mark_page_accessed() is called for reads only. But the idle flag is cleared by mark_page_accessed() so the idle flag won't get cleared if the page is write accessed only.

Basically idle page tracking is used to estimate workingset size of workload, noticeable size of workingset might be missed if the idle flag is not maintained correctly.

It seems good enough to just clear idle flag for write operations.

...

mm/filemap.c +1 -0
@@ -41,6 +41,7 @@
 #include <linux/delayacct.h>
 #include <linux/psi.h>
 #include <linux/ramfs.h>
+#include <linux/page_idle.h>
 #include "internal.h"
 
 #define CREATE_TRACE_POINTS
@@ -1630,6 +1631,11 @@ struct page *pagecache_get_page(struct address_space *mapping, pgoff_t index,
 

mm/filemap.c struct page *pagecache_get_page(struct address_space *mapping, pgoff_t index, · line 1641

Quoted 10 lines
mm/filemap.c +5 -0
 	if (fgp_flags & FGP_ACCESSED)
 		mark_page_accessed(page);
+	else if (fgp_flags & FGP_WRITE) {
+		/* Clear idle flag for buffer write */
+		if (page_is_idle(page))
+			clear_page_idle(page);
+	}
 
 no_page:
 	if (!page && (fgp_flags & FGP_CREAT)) {

The kerneldoc comment for pagecache_get_page() could do with some updating - it fails to mention FGP_WRITE, FGP_NOFS and FGP_NOWAIT.

This change seems correct but also will have runtime effects. What are they?

author
#2 lore raw
Quoted 35 lines · Andrew Morton
Quoted 33 lines · On Thu, 25 Jun 2020 01:43:32 +0800 Yang Shi <[email protected]> wrote:

Since commit bbddabe2e436aa7869b3ac5248df5c14ddde0cbf ("mm: filemap:
only do access activations on reads"), mark_page_accessed() is called for reads only. But the idle flag is cleared by mark_page_accessed() so the idle flag won't get cleared if the page is write accessed only.

Basically idle page tracking is used to estimate workingset size of workload, noticeable size of workingset might be missed if the idle flag is not maintained correctly.

It seems good enough to just clear idle flag for write operations.

...

mm/filemap.c +6 -0
@@ -41,6 +41,7 @@
  #include <linux/delayacct.h>
  #include <linux/psi.h>
  #include <linux/ramfs.h>
+#include <linux/page_idle.h>
  #include "internal.h"
 
  #define CREATE_TRACE_POINTS
@@ -1630,6 +1631,11 @@ struct page *pagecache_get_page(struct address_space *mapping, pgoff_t index,
 
  	if (fgp_flags & FGP_ACCESSED)
  		mark_page_accessed(page);
+	else if (fgp_flags & FGP_WRITE) {
+		/* Clear idle flag for buffer write */
+		if (page_is_idle(page))
+			clear_page_idle(page);
+	}
 
Quoted 4 lines

mm/filemap.c struct page *pagecache_get_page(struct address_space *mapping, pgoff_t index, · line 1641

Quoted 2 lines
mm/filemap.c +0 -0
  no_page:
  	if (!page && (fgp_flags & FGP_CREAT)) {

The kerneldoc comment for pagecache_get_page() could do with some updating - it fails to mention FGP_WRITE, FGP_NOFS and FGP_NOWAIT.

Yes, will propose a separate patch later on.

Quoted 3 lines

This change seems correct but also will have runtime effects. What are they?

Other than a couple of extra cycles when idle page tracking is enabled, I didn't think of other effects. It should be negligible. The idle flag doesn't play a role in page reclaim algorithm, so it won't have impact on that.

#3 lore raw
Quoted 47 lines · Yang Shi
Quoted 35 lines · Andrew Morton
Quoted 33 lines · On Thu, 25 Jun 2020 01:43:32 +0800 Yang Shi <[email protected]> wrote:

Since commit bbddabe2e436aa7869b3ac5248df5c14ddde0cbf ("mm: filemap:
only do access activations on reads"), mark_page_accessed() is called for reads only. But the idle flag is cleared by mark_page_accessed() so the idle flag won't get cleared if the page is write accessed only.

Basically idle page tracking is used to estimate workingset size of workload, noticeable size of workingset might be missed if the idle flag is not maintained correctly.

It seems good enough to just clear idle flag for write operations.

...

mm/filemap.c +6 -0
@@ -41,6 +41,7 @@
  #include <linux/delayacct.h>
  #include <linux/psi.h>
  #include <linux/ramfs.h>
+#include <linux/page_idle.h>
  #include "internal.h"
 
  #define CREATE_TRACE_POINTS
@@ -1630,6 +1631,11 @@ struct page *pagecache_get_page(struct address_space *mapping, pgoff_t index,
 
     if (fgp_flags & FGP_ACCESSED)
             mark_page_accessed(page);
+    else if (fgp_flags & FGP_WRITE) {
+            /* Clear idle flag for buffer write */
+            if (page_is_idle(page))
+                    clear_page_idle(page);
+    }
 
Quoted 4 lines

mm/filemap.c struct page *pagecache_get_page(struct address_space *mapping, pgoff_t index, · line 1641

Quoted 2 lines
mm/filemap.c +0 -0
  no_page:
     if (!page && (fgp_flags & FGP_CREAT)) {

The kerneldoc comment for pagecache_get_page() could do with some updating - it fails to mention FGP_WRITE, FGP_NOFS and FGP_NOWAIT.

Yes, will propose a separate patch later on.

Quoted 1 line
Quoted 9 lines
Quoted 2 lines

This change seems correct but also will have runtime effects. What are they?

Other than a couple of extra cycles when idle page tracking is enabled, I didn't think of other effects. It should be negligible. The idle flag doesn't play a role in page reclaim algorithm, so it won't have impact on that.

The only user visible impact will be on idle page tracking users. They will get more accurate data.

author
#4 lore raw
Quoted 47 lines · Shakeel Butt
Quoted 46 lines · Yang Shi
Quoted 35 lines · Andrew Morton
Quoted 33 lines · On Thu, 25 Jun 2020 01:43:32 +0800 Yang Shi <[email protected]> wrote:

Since commit bbddabe2e436aa7869b3ac5248df5c14ddde0cbf ("mm: filemap:
only do access activations on reads"), mark_page_accessed() is called for reads only. But the idle flag is cleared by mark_page_accessed() so the idle flag won't get cleared if the page is write accessed only.

Basically idle page tracking is used to estimate workingset size of workload, noticeable size of workingset might be missed if the idle flag is not maintained correctly.

It seems good enough to just clear idle flag for write operations.

...

mm/filemap.c +6 -0
@@ -41,6 +41,7 @@
   #include <linux/delayacct.h>
   #include <linux/psi.h>
   #include <linux/ramfs.h>
+#include <linux/page_idle.h>
   #include "internal.h"
 
   #define CREATE_TRACE_POINTS
@@ -1630,6 +1631,11 @@ struct page *pagecache_get_page(struct address_space *mapping, pgoff_t index,
 
      if (fgp_flags & FGP_ACCESSED)
              mark_page_accessed(page);
+    else if (fgp_flags & FGP_WRITE) {
+            /* Clear idle flag for buffer write */
+            if (page_is_idle(page))
+                    clear_page_idle(page);
+    }
 
Quoted 4 lines

mm/filemap.c struct page *pagecache_get_page(struct address_space *mapping, pgoff_t index, · line 1641

Quoted 2 lines
mm/filemap.c +0 -0
   no_page:
      if (!page && (fgp_flags & FGP_CREAT)) {

The kerneldoc comment for pagecache_get_page() could do with some updating - it fails to mention FGP_WRITE, FGP_NOFS and FGP_NOWAIT.

Yes, will propose a separate patch later on.

Quoted 2 lines

This change seems correct but also will have runtime effects. What are they?

Quoted 8 lines
Quoted 6 lines

Other than a couple of extra cycles when idle page tracking is enabled, I didn't think of other effects. It should be negligible. The idle flag doesn't play a role in page reclaim algorithm, so it won't have impact on that.

The only user visible impact will be on idle page tracking users. They will get more accurate data.

Thanks for elaborating this.

Reviewed-by on v1 1/1
#5 lore raw
In reply to Yang Shi
Quoted 10 lines · Yang Shi

Since commit bbddabe2e436aa7869b3ac5248df5c14ddde0cbf ("mm: filemap:
only do access activations on reads"), mark_page_accessed() is called for reads only. But the idle flag is cleared by mark_page_accessed() so the idle flag won't get cleared if the page is write accessed only.

Basically idle page tracking is used to estimate workingset size of workload, noticeable size of workingset might be missed if the idle flag is not maintained correctly.

Quoted 8 lines

It seems good enough to just clear idle flag for write operations.

Fixes bbddabe2e436 ("mm: filemap: only do access activations on reads")
Cc Johannes Weiner <[email protected]>
Cc Rik van Riel <[email protected]>
Cc Shakeel Butt <[email protected]>