mm:Add watermark slope for high mark

When tuning the watermark_scale_factor to reduce stalls and compactions the high mark is also changed, it changed a bit too much. So this patch introduces a slope that can reduce this overhead a bit, or increase it if needed.

Signed-off-by Peter Enderborg <[email protected]>

---
 Documentation/sysctl/vm.txt | 15 +++++++++++++++
 include/linux/mm.h          |  1 +
 include/linux/mmzone.h      |  2 ++
 kernel/sysctl.c             |  9 +++++++++
 mm/page_alloc.c             |  6 +++++-
 5 files changed, 32 insertions(+), 1 deletion(-)

Documentation/sysctl/vm.txt +15 -0

index eda628c..aecff6c 100644
@@ -62,6 +62,7 @@ Currently, these files are in /proc/sys/vm:
 - user_reserve_kbytes
 - vfs_cache_pressure
 - watermark_scale_factor
+- watermark_high_factor_slope
 - zone_reclaim_mode
 
 ==============================================================

@@ -857,6 +858,20 @@ that the number of free pages kswapd maintains for latency reasons is
 too small for the allocation bursts occurring in the system. This knob
 can then be used to tune kswapd aggressiveness accordingly.
 
+=============================================================
+
+watermark_high_factor_slope:
+
+This factor is high mark for watermark_scale_factor.
+The unit is in percent.
+Max value is 1000 and min value is 100. (High watermark is the same as
+low water mark) Low watermark is min_wmark_pages + watermark_scale_factor.
+and high watermark is
+min_wmark_pages+(watermark_scale_factor * watermark_high_factor_slope).
+
+The default value is 200.
+
+
 ==============================================================
 
 zone_reclaim_mode:

include/linux/mm.h +1 -0

index 7661156..c89536b 100644
@@ -2094,6 +2094,7 @@ extern void zone_pcp_reset(struct zone *zone);
 /* page_alloc.c */
 extern int min_free_kbytes;
 extern int watermark_scale_factor;
+extern int watermark_high_factor_slope;
 
 /* nommu.c */
 extern atomic_long_t mmap_pages_allocated;

include/linux/mmzone.h +2 -0

index 67f2e3c..91bf842 100644
@@ -886,6 +886,8 @@ int min_free_kbytes_sysctl_handler(struct ctl_table *, int,
 					void __user *, size_t *, loff_t *);
 int watermark_scale_factor_sysctl_handler(struct ctl_table *, int,
 					void __user *, size_t *, loff_t *);
+//int watermark_high_factor_tilt_sysctl_handler(struct ctl_table *, int,
+//					void __user *, size_t *, loff_t *);
 extern int sysctl_lowmem_reserve_ratio[MAX_NR_ZONES-1];
 int lowmem_reserve_ratio_sysctl_handler(struct ctl_table *, int,
 					void __user *, size_t *, loff_t *);

kernel/sysctl.c +9 -0

index 2fb4e27..83c48c9 100644
@@ -1444,6 +1444,15 @@ static struct ctl_table vm_table[] = {
 		.extra2		= &one_thousand,
 	},
 	{
+		.procname	= "watermark_high_factor_slope",
+		.data		= &watermark_high_factor_slope,
+		.maxlen		= sizeof(watermark_high_factor_slope),
+		.mode		= 0644,
+		.proc_handler	= watermark_scale_factor_sysctl_handler,
+		.extra1		= &one_hundred,
+		.extra2		= &one_thousand,
+	},
+	{
 		.procname	= "percpu_pagelist_fraction",
 		.data		= &percpu_pagelist_fraction,
 		.maxlen		= sizeof(percpu_pagelist_fraction),

mm/page_alloc.c +5 -1

index 48b5b01..3dc50ff 100644
@@ -263,6 +263,7 @@ compound_page_dtor * const compound_page_dtors[] = {
 int min_free_kbytes = 1024;
 int user_min_free_kbytes = -1;
 int watermark_scale_factor = 10;
+int watermark_high_factor_slope = 200;
 
 static unsigned long __meminitdata nr_kernel_pages;
 static unsigned long __meminitdata nr_all_pages;

@@ -6989,6 +6990,7 @@ static void __setup_per_zone_wmarks(void)
 
 	for_each_zone(zone) {
 		u64 tmp;
+		u64 tmp_high;
 
 		spin_lock_irqsave(&zone->lock, flags);
 		tmp = (u64)pages_min * zone->managed_pages;

@@ -7026,7 +7028,9 @@ static void __setup_per_zone_wmarks(void)
 				      watermark_scale_factor, 10000));
 
 		zone->watermark[WMARK_LOW]  = min_wmark_pages(zone) + tmp;
-		zone->watermark[WMARK_HIGH] = min_wmark_pages(zone) + tmp * 2;
+		tmp_high = mult_frac(tmp, watermark_high_factor_slope, 100);
+		zone->watermark[WMARK_HIGH] = min_wmark_pages(zone) + tmp_high;
+
 
 		spin_unlock_irqrestore(&zone->lock, flags);
 	}

Signature · 1 line

2.7.4

Quoted 4 lines · Peter Enderborg

When tuning the watermark_scale_factor to reduce stalls and compactions the high mark is also changed, it changed a bit too much. So this patch introduces a slope that can reduce this overhead a bit, or increase it if needed.

This doesn't explain what is the problem, why it is a problem and why we need yet another tuning to address it. Users shouldn't really care about internal stuff like watermark tuning for each watermark independently. This looks like a gross hack. Please start over with the problem description and then we can move on to an approapriate fix. Piling up tuning knobs to workaround problems is simply not acceptable.

Quoted 116 lines

Signed-off-by Peter Enderborg <[email protected]>

---
 Documentation/sysctl/vm.txt | 15 +++++++++++++++
 include/linux/mm.h          |  1 +
 include/linux/mmzone.h      |  2 ++
 kernel/sysctl.c             |  9 +++++++++
 mm/page_alloc.c             |  6 +++++-
 5 files changed, 32 insertions(+), 1 deletion(-)

Documentation/sysctl/vm.txt +15 -0

index eda628c..aecff6c 100644
@@ -62,6 +62,7 @@ Currently, these files are in /proc/sys/vm:
 - user_reserve_kbytes
 - vfs_cache_pressure
 - watermark_scale_factor
+- watermark_high_factor_slope
 - zone_reclaim_mode
 
 ==============================================================

@@ -857,6 +858,20 @@ that the number of free pages kswapd maintains for latency reasons is
 too small for the allocation bursts occurring in the system. This knob
 can then be used to tune kswapd aggressiveness accordingly.
 
+=============================================================
+
+watermark_high_factor_slope:
+
+This factor is high mark for watermark_scale_factor.
+The unit is in percent.
+Max value is 1000 and min value is 100. (High watermark is the same as
+low water mark) Low watermark is min_wmark_pages + watermark_scale_factor.
+and high watermark is
+min_wmark_pages+(watermark_scale_factor * watermark_high_factor_slope).
+
+The default value is 200.
+
+
 ==============================================================
 
 zone_reclaim_mode:

include/linux/mm.h +1 -0

index 7661156..c89536b 100644
@@ -2094,6 +2094,7 @@ extern void zone_pcp_reset(struct zone *zone);
 /* page_alloc.c */
 extern int min_free_kbytes;
 extern int watermark_scale_factor;
+extern int watermark_high_factor_slope;
 
 /* nommu.c */
 extern atomic_long_t mmap_pages_allocated;

include/linux/mmzone.h +2 -0

index 67f2e3c..91bf842 100644
@@ -886,6 +886,8 @@ int min_free_kbytes_sysctl_handler(struct ctl_table *, int,
 					void __user *, size_t *, loff_t *);
 int watermark_scale_factor_sysctl_handler(struct ctl_table *, int,
 					void __user *, size_t *, loff_t *);
+//int watermark_high_factor_tilt_sysctl_handler(struct ctl_table *, int,
+//					void __user *, size_t *, loff_t *);
 extern int sysctl_lowmem_reserve_ratio[MAX_NR_ZONES-1];
 int lowmem_reserve_ratio_sysctl_handler(struct ctl_table *, int,
 					void __user *, size_t *, loff_t *);

kernel/sysctl.c +9 -0

index 2fb4e27..83c48c9 100644
@@ -1444,6 +1444,15 @@ static struct ctl_table vm_table[] = {
 		.extra2		= &one_thousand,
 	},
 	{
+		.procname	= "watermark_high_factor_slope",
+		.data		= &watermark_high_factor_slope,
+		.maxlen		= sizeof(watermark_high_factor_slope),
+		.mode		= 0644,
+		.proc_handler	= watermark_scale_factor_sysctl_handler,
+		.extra1		= &one_hundred,
+		.extra2		= &one_thousand,
+	},
+	{
 		.procname	= "percpu_pagelist_fraction",
 		.data		= &percpu_pagelist_fraction,
 		.maxlen		= sizeof(percpu_pagelist_fraction),

mm/page_alloc.c +5 -1

index 48b5b01..3dc50ff 100644
@@ -263,6 +263,7 @@ compound_page_dtor * const compound_page_dtors[] = {
 int min_free_kbytes = 1024;
 int user_min_free_kbytes = -1;
 int watermark_scale_factor = 10;
+int watermark_high_factor_slope = 200;
 
 static unsigned long __meminitdata nr_kernel_pages;
 static unsigned long __meminitdata nr_all_pages;

@@ -6989,6 +6990,7 @@ static void __setup_per_zone_wmarks(void)
 
 	for_each_zone(zone) {
 		u64 tmp;
+		u64 tmp_high;
 
 		spin_lock_irqsave(&zone->lock, flags);
 		tmp = (u64)pages_min * zone->managed_pages;

@@ -7026,7 +7028,9 @@ static void __setup_per_zone_wmarks(void)
 				      watermark_scale_factor, 10000));
 
 		zone->watermark[WMARK_LOW]  = min_wmark_pages(zone) + tmp;
-		zone->watermark[WMARK_HIGH] = min_wmark_pages(zone) + tmp * 2;
+		tmp_high = mult_frac(tmp, watermark_high_factor_slope, 100);
+		zone->watermark[WMARK_HIGH] = min_wmark_pages(zone) + tmp_high;
+
 

Quoted 5 lines

mm/page_alloc.c +0 -0

 		spin_unlock_irqrestore(&zone->lock, flags);
 	}

Signature · 1 line

2.7.4

Agreed. Also if you send a patch adding userspace API or a tuning knob, please CC linux-api mailing list (did that for this reply).

Quoted 4 lines · Peter Enderborg

When tuning the watermark_scale_factor to reduce stalls and compactions the high mark is also changed, it changed a bit too much. So this patch introduces a slope that can reduce this overhead a bit, or increase it if needed.

This doesn't explain what is the problem, why it is a problem and why we need yet another tuning to address it. Users shouldn't really care about internal stuff like watermark tuning for each watermark independently. This looks like a gross hack. Please start over with the problem description and then we can move on to an approapriate fix. Piling up tuning knobs to workaround problems is simply not acceptable.

Quoted 5 lines · Michal Hocko

Quoted 4 lines · Peter Enderborg

When tuning the watermark_scale_factor to reduce stalls and compactions the high mark is also changed, it changed a bit too much. So this patch introduces a slope that can reduce this overhead a bit, or increase it if needed.

Quoted 7 lines

This doesn't explain what is the problem, why it is a problem and why we need yet another tuning to address it. Users shouldn't really care about internal stuff like watermark tuning for each watermark independently. This looks like a gross hack. Please start over with the problem description and then we can move on to an approapriate fix. Piling up tuning knobs to workaround problems is simply not acceptable.

In the original patch - https://lkml.org/lkml/2016/2/18/498 - had a

discussion about small systems with 8GB RAM. In the handheld world, that's a lot of RAM. However, the magic number 2 used in the present algorithm is out of the blue. Compaction problems are the same for both small and big. So small devices also need to increase watermark to
get compaction to work and reduce direct reclaims. Changing the low watermark makes direct reclaim rate drop a lot. But it will cause kswap to work more, and that has a negative impact. Lowering the gap will smooth out the kswap workload to suite embedded devices a lot better. This can be addressed by reducing the high watermark using the slope patch herein. Im sort of understand your opinion on user knobs, but hard-coded magic numbers are even worse.