[v4] mm: khugepaged: recalculate min_free_kbytes after memory hotplug as expected by khugepaged

[v4] mm: khugepaged: recalculate min_free_kbytes after memory hotplug as expected by khugepaged

[Vijay Balakrishna]

When memory is hotplug added or removed the min_free_kbytes must be
recalculated based on what is expected by khugepaged.  Currently
after hotplug, min_free_kbytes will be set to a lower default and higher
default set when THP enabled is lost.  This change restores min_free_kbytes
as expected for THP consumers.

Fixes: f000565adb77 ("thp: set recommended min free kbytes")

Signed-off-by: Vijay Balakrishna <vijayb@linux.microsoft.com>
Cc: stable@vger.kernel.org
Reviewed-by: Pavel Tatashin <pasha.tatashin@soleen.com>
Acked-by: Michal Hocko <mhocko@suse.com>
---
v3 -> v4
- made changes to move khugepaged_min_free_kbytes_update into
  init_per_zone_wmark_min and rested changes
  [suggestion from Michal Hocko]

[v2 1/2]
- removed symptoms references from changelog

[v2 2/2]
- addressed following issues Michal Hocko raised:
  . nr_free_buffer_pages can oveflow in int on very large machines
  . min_free_kbytes can decrease the size theoretically

v1 -> v2
--------
- addressed issue Kirill A. Shutemov raised:
  . changes would override min_free_kbytes set by user

 include/linux/khugepaged.h |  5 +++++
 mm/khugepaged.c            | 13 +++++++++++--
 mm/page_alloc.c            |  3 +++
 3 files changed, 19 insertions(+), 2 deletions(-)

diff --git a/include/linux/khugepaged.h b/include/linux/khugepaged.h
index bc45ea1efbf7..c941b7377321 100644
--- a/include/linux/khugepaged.h
+++ b/include/linux/khugepaged.h
@@ -15,6 +15,7 @@ extern int __khugepaged_enter(struct mm_struct *mm);
 extern void __khugepaged_exit(struct mm_struct *mm);
 extern int khugepaged_enter_vma_merge(struct vm_area_struct *vma,
 				      unsigned long vm_flags);
+extern void khugepaged_min_free_kbytes_update(void);
 #ifdef CONFIG_SHMEM
 extern void collapse_pte_mapped_thp(struct mm_struct *mm, unsigned long addr);
 #else
@@ -85,6 +86,10 @@ static inline void collapse_pte_mapped_thp(struct mm_struct *mm,
 					   unsigned long addr)
 {
 }
+
+static inline void khugepaged_min_free_kbytes_update(void)
+{
+}
 #endif /* CONFIG_TRANSPARENT_HUGEPAGE */
 
 #endif /* _LINUX_KHUGEPAGED_H */
diff --git a/mm/khugepaged.c b/mm/khugepaged.c
index cfa0dba5fd3b..4f7107476a6f 100644
--- a/mm/khugepaged.c
+++ b/mm/khugepaged.c
@@ -56,6 +56,9 @@ enum scan_result {
 #define CREATE_TRACE_POINTS
 #include <trace/events/huge_memory.h>
 
+static struct task_struct *khugepaged_thread __read_mostly;
+static DEFINE_MUTEX(khugepaged_mutex);
+
 /* default scan 8*512 pte (or vmas) every 30 second */
 static unsigned int khugepaged_pages_to_scan __read_mostly;
 static unsigned int khugepaged_pages_collapsed;
@@ -2292,8 +2295,6 @@ static void set_recommended_min_free_kbytes(void)
 
 int start_stop_khugepaged(void)
 {
-	static struct task_struct *khugepaged_thread __read_mostly;
-	static DEFINE_MUTEX(khugepaged_mutex);
 	int err = 0;
 
 	mutex_lock(&khugepaged_mutex);
@@ -2320,3 +2321,11 @@ int start_stop_khugepaged(void)
 	mutex_unlock(&khugepaged_mutex);
 	return err;
 }
+
+void khugepaged_min_free_kbytes_update(void)
+{
+	mutex_lock(&khugepaged_mutex);
+	if (khugepaged_enabled() && khugepaged_thread)
+		set_recommended_min_free_kbytes();
+	mutex_unlock(&khugepaged_mutex);
+}
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index fab5e97dc9ca..ac25d3526fa5 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -69,6 +69,7 @@
 #include <linux/nmi.h>
 #include <linux/psi.h>
 #include <linux/padata.h>
+#include <linux/khugepaged.h>
 
 #include <asm/sections.h>
 #include <asm/tlbflush.h>
@@ -7891,6 +7892,8 @@ int __meminit init_per_zone_wmark_min(void)
 	setup_min_slab_ratio();
 #endif
 
+	khugepaged_min_free_kbytes_update();
+
 	return 0;
 }
 postcore_initcall(init_per_zone_wmark_min)
-- 
2.28.0

Re: [v4] mm: khugepaged: recalculate min_free_kbytes after memory hotplug as expected by khugepaged

[Michal Hocko]

On Mon 28-09-20 17:07:27, Vijay Balakrishna wrote:
> When memory is hotplug added or removed the min_free_kbytes must be

s@must@should@

> recalculated based on what is expected by khugepaged.  Currently
> after hotplug, min_free_kbytes will be set to a lower default and higher
> default set when THP enabled is lost.  This change restores min_free_kbytes
> as expected for THP consumers.
> 
> Fixes: f000565adb77 ("thp: set recommended min free kbytes")
> 
> Signed-off-by: Vijay Balakrishna <vijayb@linux.microsoft.com>
> Cc: stable@vger.kernel.org
> Reviewed-by: Pavel Tatashin <pasha.tatashin@soleen.com>
> Acked-by: Michal Hocko <mhocko@suse.com>
> ---
> v3 -> v4
> - made changes to move khugepaged_min_free_kbytes_update into
>   init_per_zone_wmark_min and rested changes
>   [suggestion from Michal Hocko]
> 
> [v2 1/2]
> - removed symptoms references from changelog
> 
> [v2 2/2]
> - addressed following issues Michal Hocko raised:
>   . nr_free_buffer_pages can oveflow in int on very large machines
>   . min_free_kbytes can decrease the size theoretically
> 
> v1 -> v2
> --------
> - addressed issue Kirill A. Shutemov raised:
>   . changes would override min_free_kbytes set by user
> 
>  include/linux/khugepaged.h |  5 +++++
>  mm/khugepaged.c            | 13 +++++++++++--
>  mm/page_alloc.c            |  3 +++
>  3 files changed, 19 insertions(+), 2 deletions(-)
> 
> diff --git a/include/linux/khugepaged.h b/include/linux/khugepaged.h
> index bc45ea1efbf7..c941b7377321 100644
> --- a/include/linux/khugepaged.h
> +++ b/include/linux/khugepaged.h
> @@ -15,6 +15,7 @@ extern int __khugepaged_enter(struct mm_struct *mm);
>  extern void __khugepaged_exit(struct mm_struct *mm);
>  extern int khugepaged_enter_vma_merge(struct vm_area_struct *vma,
>  				      unsigned long vm_flags);
> +extern void khugepaged_min_free_kbytes_update(void);
>  #ifdef CONFIG_SHMEM
>  extern void collapse_pte_mapped_thp(struct mm_struct *mm, unsigned long addr);
>  #else
> @@ -85,6 +86,10 @@ static inline void collapse_pte_mapped_thp(struct mm_struct *mm,
>  					   unsigned long addr)
>  {
>  }
> +
> +static inline void khugepaged_min_free_kbytes_update(void)
> +{
> +}
>  #endif /* CONFIG_TRANSPARENT_HUGEPAGE */
>  
>  #endif /* _LINUX_KHUGEPAGED_H */
> diff --git a/mm/khugepaged.c b/mm/khugepaged.c
> index cfa0dba5fd3b..4f7107476a6f 100644
> --- a/mm/khugepaged.c
> +++ b/mm/khugepaged.c
> @@ -56,6 +56,9 @@ enum scan_result {
>  #define CREATE_TRACE_POINTS
>  #include <trace/events/huge_memory.h>
>  
> +static struct task_struct *khugepaged_thread __read_mostly;
> +static DEFINE_MUTEX(khugepaged_mutex);
> +
>  /* default scan 8*512 pte (or vmas) every 30 second */
>  static unsigned int khugepaged_pages_to_scan __read_mostly;
>  static unsigned int khugepaged_pages_collapsed;
> @@ -2292,8 +2295,6 @@ static void set_recommended_min_free_kbytes(void)
>  
>  int start_stop_khugepaged(void)
>  {
> -	static struct task_struct *khugepaged_thread __read_mostly;
> -	static DEFINE_MUTEX(khugepaged_mutex);
>  	int err = 0;
>  
>  	mutex_lock(&khugepaged_mutex);
> @@ -2320,3 +2321,11 @@ int start_stop_khugepaged(void)
>  	mutex_unlock(&khugepaged_mutex);
>  	return err;
>  }
> +
> +void khugepaged_min_free_kbytes_update(void)
> +{
> +	mutex_lock(&khugepaged_mutex);
> +	if (khugepaged_enabled() && khugepaged_thread)
> +		set_recommended_min_free_kbytes();
> +	mutex_unlock(&khugepaged_mutex);
> +}
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index fab5e97dc9ca..ac25d3526fa5 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -69,6 +69,7 @@
>  #include <linux/nmi.h>
>  #include <linux/psi.h>
>  #include <linux/padata.h>
> +#include <linux/khugepaged.h>
>  
>  #include <asm/sections.h>
>  #include <asm/tlbflush.h>
> @@ -7891,6 +7892,8 @@ int __meminit init_per_zone_wmark_min(void)
>  	setup_min_slab_ratio();
>  #endif
>  
> +	khugepaged_min_free_kbytes_update();
> +
>  	return 0;
>  }
>  postcore_initcall(init_per_zone_wmark_min)
> -- 
> 2.28.0

-- 
Michal Hocko
SUSE Labs