Re: [PATCH v3 17/17] mm: add knob to tune lazyfreeing

[Shaohua Li]

On Thu, Nov 12, 2015 at 01:33:13PM +0900, Minchan Kim wrote:
> MADV_FREEed page's hotness is very arguble.
> Someone think it's hot while others are it's cold.
> 
> Quote from Shaohua
> "
> My main concern is the policy how we should treat the FREE pages. Moving it to
> inactive lru is definitionly a good start, I'm wondering if it's enough. The
> MADV_FREE increases memory pressure and cause unnecessary reclaim because of
> the lazy memory free. While MADV_FREE is intended to be a better replacement of
> MADV_DONTNEED, MADV_DONTNEED doesn't have the memory pressure issue as it free
> memory immediately. So I hope the MADV_FREE doesn't have impact on memory
> pressure too. I'm thinking of adding an extra lru list and wartermark for this
> to make sure FREE pages can be freed before system wide page reclaim. As you
> said, this is arguable, but I hope we can discuss about this issue more.
> "
> 
> Quote from me
> "
> It seems the divergence comes from MADV_FREE is *replacement* of MADV_DONTNEED.
> But I don't think so. If we could discard MADV_FREEed page *anytime*, I agree
> but it's not true because the page would be dirty state when VM want to reclaim.
> 
> I'm also against with your's suggestion which let's discard FREEed page before
> system wide page reclaim because system would have lots of clean cold page
> caches or anonymous pages. In such case, reclaiming of them would be better.
> Yeb, it's really workload-dependent so we might need some heuristic which is
> normally what we want to avoid.
> 
> Having said that, I agree with you we could do better than the deactivation
> and frankly speaking, I'm thinking of another LRU list(e.g. tentatively named
> "ezreclaim LRU list"). What I have in mind is to age (anon|file|ez)
> fairly. IOW, I want to percolate ez-LRU list reclaiming into get_scan_count.
> When the MADV_FREE is called, we could move hinted pages from anon-LRU to
> ez-LRU and then If VM find to not be able to discard a page in ez-LRU,
> it could promote it to acive-anon-LRU which would be very natural aging
> concept because it mean someone touches the page recenlty.
> With that, I don't want to bias one side and don't want to add some knob for
> tuning the heuristic but let's rely on common fair aging scheme of VM.
> "
> 
> Quote from Johannes
> "
> thread 1:
> Even if we're wrong about the aging of those MADV_FREE pages, their
> contents are invalidated; they can be discarded freely, and restoring
> them is a mere GFP_ZERO allocation. All other anonymous pages have to
> be written to disk, and potentially be read back.
> 
> [ Arguably, MADV_FREE pages should even be reclaimed before inactive
>   page cache. It's the same cost to discard both types of pages, but
>   restoring page cache involves IO. ]
> 
> It probably makes sense to stop thinking about them as anonymous pages
> entirely at this point when it comes to aging. They're really not. The
> LRU lists are split to differentiate access patterns and cost of page
> stealing (and restoring). From that angle, MADV_FREE pages really have
> nothing in common with in-use anonymous pages, and so they shouldn't
> be on the same LRU list.
> 
> thread:2
> What about them is hot? They contain garbage, you have to write to
> them before you can use them. Granted, you might have to refetch
> cachelines if you don't do cacheline-aligned populating writes, but
> you can do a lot of them before it's more expensive than doing IO.
> 
> "
> 
> Quote from Daniel
> "
> thread:1
> Keep in mind that this is memory the kernel wouldn't be getting back at
> all if the allocator wasn't going out of the way to purge it, and they
> aren't going to go out of their way to purge it if it means the kernel
> is going to steal the pages when there isn't actually memory pressure.
> 
> An allocator would be using MADV_DONTNEED if it didn't expect that the
> pages were going to be used against shortly. MADV_FREE indicates that it
> has time to inform the kernel that they're unused but they could still
> be very hot.
> 
> thread:2
> It's hot because applications churn through memory via the allocator.
> 
> Drop the pages and the application is now churning through page faults
> and zeroing rather than simply reusing memory. It's not something that
> may happen, it *will* happen. A page in the page cache *may* be reused,
> but often won't be, especially when the I/O patterns don't line up well
> with the way it works.
> 
> The whole point of the feature is not requiring the allocator to have
> elaborate mechanisms for aging pages and throttling purging. That ends
> up resulting in lots of memory held by userspace where the kernel can't
> reclaim it under memory pressure. If it's dropped before page cache, it
> isn't going to be able to replace any of that logic in allocators.
> 
> The page cache is speculative. Page caching by allocators is not really
> speculative. Using MADV_FREE on the pages at all is speculative. The
> memory is probably going to be reused fairly soon (unless the process
> exits, and then it doesn't matter), but purging will end up reducing
> memory usage for the portions that aren't.
> 
> It would be a different story for a full unpinning/pinning feature since
> that would have other use cases (speculative caches), but this is really
> only useful in allocators.
> "
> You could read all thread from https://lkml.org/lkml/2015/11/4/51
> 
> Yeah, with arguble issue and there is no one decision, I think it
> means we should provide the knob "lazyfreeness"(I hope someone
> give better naming).
> 
> It's similar to swapppiness so higher values will discard MADV_FREE
> pages agreessively. If memory pressure happens and system works with
> DEF_PRIOIRTY(ex, clean cold caches), VM doesn't discard any hinted
> pages until the scanning priority is increased.
> 
> If memory pressure is higher(ie, the priority is not DEF_PRIORITY),
> it scans
> 
> 	nr_to_reclaim * priority * lazyfreensss(def: 20) / 50
> 
> If system has low free memory and file cache, it start to discard
> MADV_FREEed pages unconditionally even though user set lazyfreeness to 0.
> 
> Signed-off-by: Minchan Kim <minchan@xxxxxxxxxx>
> ---
>  Documentation/sysctl/vm.txt | 13 +++++++++
>  drivers/base/node.c         |  4 +--
>  fs/proc/meminfo.c           |  4 +--
>  include/linux/memcontrol.h  |  1 +
>  include/linux/mmzone.h      |  9 +++---
>  include/linux/swap.h        | 15 ++++++++++
>  kernel/sysctl.c             |  9 ++++++
>  mm/memcontrol.c             | 32 +++++++++++++++++++++-
>  mm/vmscan.c                 | 67 ++++++++++++++++++++++++++++-----------------
>  mm/vmstat.c                 |  2 +-
>  10 files changed, 121 insertions(+), 35 deletions(-)
> 
> diff --git a/Documentation/sysctl/vm.txt b/Documentation/sysctl/vm.txt
> index a4482fceacec..c1dc63381f2c 100644
> --- a/Documentation/sysctl/vm.txt
> +++ b/Documentation/sysctl/vm.txt
> @@ -56,6 +56,7 @@ files can be found in mm/swap.c.
>  - percpu_pagelist_fraction
>  - stat_interval
>  - swappiness
> +- lazyfreeness
>  - user_reserve_kbytes
>  - vfs_cache_pressure
>  - zone_reclaim_mode
> @@ -737,6 +738,18 @@ The default value is 60.
>  
>  ==============================================================
>  
> +lazyfreeness
> +
> +This control is used to define how aggressive the kernel will discard
> +MADV_FREE hinted pages.  Higher values will increase agressiveness,
> +lower values decrease the amount of discarding.  A value of 0 instructs
> +the kernel not to initiate discarding until the amount of free and
> +file-backed pages is less than the high water mark in a zone.
> +
> +The default value is 20.
> +
> +==============================================================
> +
>  - user_reserve_kbytes
>  
>  When overcommit_memory is set to 2, "never overcommit" mode, reserve
> diff --git a/drivers/base/node.c b/drivers/base/node.c
> index f7a1f2107b43..3b0bf1b78b2e 100644
> --- a/drivers/base/node.c
> +++ b/drivers/base/node.c
> @@ -69,8 +69,8 @@ static ssize_t node_read_meminfo(struct device *dev,
>  		       "Node %d Inactive(anon): %8lu kB\n"
>  		       "Node %d Active(file):   %8lu kB\n"
>  		       "Node %d Inactive(file): %8lu kB\n"
> -		       "Node %d Unevictable:    %8lu kB\n"
>  		       "Node %d LazyFree:	%8lu kB\n"
> +		       "Node %d Unevictable:    %8lu kB\n"
>  		       "Node %d Mlocked:        %8lu kB\n",
>  		       nid, K(i.totalram),
>  		       nid, K(i.freeram),
> @@ -83,8 +83,8 @@ static ssize_t node_read_meminfo(struct device *dev,
>  		       nid, K(node_page_state(nid, NR_INACTIVE_ANON)),
>  		       nid, K(node_page_state(nid, NR_ACTIVE_FILE)),
>  		       nid, K(node_page_state(nid, NR_INACTIVE_FILE)),
> -		       nid, K(node_page_state(nid, NR_UNEVICTABLE)),
>  		       nid, K(node_page_state(nid, NR_LZFREE)),
> +		       nid, K(node_page_state(nid, NR_UNEVICTABLE)),
>  		       nid, K(node_page_state(nid, NR_MLOCK)));
>  
>  #ifdef CONFIG_HIGHMEM
> diff --git a/fs/proc/meminfo.c b/fs/proc/meminfo.c
> index 3444f7c4e0b6..f47e6a5aa2e5 100644
> --- a/fs/proc/meminfo.c
> +++ b/fs/proc/meminfo.c
> @@ -101,8 +101,8 @@ static int meminfo_proc_show(struct seq_file *m, void *v)
>  		"Inactive(anon): %8lu kB\n"
>  		"Active(file):   %8lu kB\n"
>  		"Inactive(file): %8lu kB\n"
> -		"Unevictable:    %8lu kB\n"
>  		"LazyFree:	 %8lu kB\n"
> +		"Unevictable:    %8lu kB\n"
>  		"Mlocked:        %8lu kB\n"
>  #ifdef CONFIG_HIGHMEM
>  		"HighTotal:      %8lu kB\n"
> @@ -159,8 +159,8 @@ static int meminfo_proc_show(struct seq_file *m, void *v)
>  		K(pages[LRU_INACTIVE_ANON]),
>  		K(pages[LRU_ACTIVE_FILE]),
>  		K(pages[LRU_INACTIVE_FILE]),
> -		K(pages[LRU_UNEVICTABLE]),
>  		K(pages[LRU_LZFREE]),
> +		K(pages[LRU_UNEVICTABLE]),
>  		K(global_page_state(NR_MLOCK)),
>  #ifdef CONFIG_HIGHMEM
>  		K(i.totalhigh),
> diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
> index 3e3318ddfc0e..5522ff733506 100644
> --- a/include/linux/memcontrol.h
> +++ b/include/linux/memcontrol.h
> @@ -210,6 +210,7 @@ struct mem_cgroup {
>  	int		under_oom;
>  
>  	int	swappiness;
> +	int	lzfreeness;
>  	/* OOM-Killer disable */
>  	int		oom_kill_disable;
>  
> diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
> index 1aaa436da0d5..cca514a9701d 100644
> --- a/include/linux/mmzone.h
> +++ b/include/linux/mmzone.h
> @@ -120,8 +120,8 @@ enum zone_stat_item {
>  	NR_ACTIVE_ANON,		/*  "     "     "   "       "         */
>  	NR_INACTIVE_FILE,	/*  "     "     "   "       "         */
>  	NR_ACTIVE_FILE,		/*  "     "     "   "       "         */
> -	NR_UNEVICTABLE,		/*  "     "     "   "       "         */
>  	NR_LZFREE,		/*  "     "     "   "       "         */
> +	NR_UNEVICTABLE,		/*  "     "     "   "       "         */
>  	NR_MLOCK,		/* mlock()ed pages found and moved off LRU */
>  	NR_ANON_PAGES,	/* Mapped anonymous pages */
>  	NR_FILE_MAPPED,	/* pagecache pages mapped into pagetables.
> @@ -179,14 +179,15 @@ enum lru_list {
>  	LRU_ACTIVE_ANON = LRU_BASE + LRU_ACTIVE,
>  	LRU_INACTIVE_FILE = LRU_BASE + LRU_FILE,
>  	LRU_ACTIVE_FILE = LRU_BASE + LRU_FILE + LRU_ACTIVE,
> -	LRU_UNEVICTABLE,
>  	LRU_LZFREE,
> +	LRU_UNEVICTABLE,
>  	NR_LRU_LISTS
>  };
>  
>  #define for_each_lru(lru) for (lru = 0; lru < NR_LRU_LISTS; lru++)
> -
> -#define for_each_evictable_lru(lru) for (lru = 0; lru <= LRU_ACTIVE_FILE; lru++)
> +#define for_each_anon_file_lru(lru) \
> +		for (lru = 0; lru <= LRU_ACTIVE_FILE; lru++)
> +#define for_each_evictable_lru(lru) for (lru = 0; lru <= LRU_LZFREE; lru++)
>  
>  static inline int is_file_lru(enum lru_list lru)
>  {
> diff --git a/include/linux/swap.h b/include/linux/swap.h
> index f0310eeab3ec..73bcdc9d0e88 100644
> --- a/include/linux/swap.h
> +++ b/include/linux/swap.h
> @@ -330,6 +330,7 @@ extern unsigned long mem_cgroup_shrink_node_zone(struct mem_cgroup *mem,
>  						unsigned long *nr_scanned);
>  extern unsigned long shrink_all_memory(unsigned long nr_pages);
>  extern int vm_swappiness;
> +extern int vm_lazyfreeness;
>  extern int remove_mapping(struct address_space *mapping, struct page *page);
>  extern unsigned long vm_total_pages;
>  
> @@ -361,11 +362,25 @@ static inline int mem_cgroup_swappiness(struct mem_cgroup *memcg)
>  	return memcg->swappiness;
>  }
>  
> +static inline int mem_cgroup_lzfreeness(struct mem_cgroup *memcg)
> +{
> +	/* root ? */
> +	if (mem_cgroup_disabled() || !memcg->css.parent)
> +		return vm_lazyfreeness;
> +
> +	return memcg->lzfreeness;
> +}
> +
>  #else
>  static inline int mem_cgroup_swappiness(struct mem_cgroup *mem)
>  {
>  	return vm_swappiness;
>  }
> +
> +static inline int mem_cgroup_lzfreeness(struct mem_cgroup *mem)
> +{
> +	return vm_lazyfreeness;
> +}
>  #endif
>  #ifdef CONFIG_MEMCG_SWAP
>  extern void mem_cgroup_swapout(struct page *page, swp_entry_t entry);
> diff --git a/kernel/sysctl.c b/kernel/sysctl.c
> index e69201d8094e..2496b10c08e9 100644
> --- a/kernel/sysctl.c
> +++ b/kernel/sysctl.c
> @@ -1268,6 +1268,15 @@ static struct ctl_table vm_table[] = {
>  		.extra1		= &zero,
>  		.extra2		= &one_hundred,
>  	},
> +	{
> +		.procname	= "lazyfreeness",
> +		.data		= &vm_lazyfreeness,
> +		.maxlen		= sizeof(vm_lazyfreeness),
> +		.mode		= 0644,
> +		.proc_handler	= proc_dointvec_minmax,
> +		.extra1		= &zero,
> +		.extra2		= &one_hundred,
> +	},
>  #ifdef CONFIG_HUGETLB_PAGE
>  	{
>  		.procname	= "nr_hugepages",
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index 1dc599ce1bcb..5bdbe2a20dc0 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -108,8 +108,8 @@ static const char * const mem_cgroup_lru_names[] = {
>  	"active_anon",
>  	"inactive_file",
>  	"active_file",
> -	"unevictable",
>  	"lazyfree",
> +	"unevictable",
>  };
>  
>  #define THRESHOLDS_EVENTS_TARGET 128
> @@ -3288,6 +3288,30 @@ static int mem_cgroup_swappiness_write(struct cgroup_subsys_state *css,
>  	return 0;
>  }
>  
> +static u64 mem_cgroup_lzfreeness_read(struct cgroup_subsys_state *css,
> +				      struct cftype *cft)
> +{
> +	struct mem_cgroup *memcg = mem_cgroup_from_css(css);
> +
> +	return mem_cgroup_lzfreeness(memcg);
> +}
> +
> +static int mem_cgroup_lzfreeness_write(struct cgroup_subsys_state *css,
> +				       struct cftype *cft, u64 val)
> +{
> +	struct mem_cgroup *memcg = mem_cgroup_from_css(css);
> +
> +	if (val > 100)
> +		return -EINVAL;
> +
> +	if (css->parent)
> +		memcg->lzfreeness = val;
> +	else
> +		vm_lazyfreeness = val;
> +
> +	return 0;
> +}
> +
>  static void __mem_cgroup_threshold(struct mem_cgroup *memcg, bool swap)
>  {
>  	struct mem_cgroup_threshold_ary *t;
> @@ -4085,6 +4109,11 @@ static struct cftype mem_cgroup_legacy_files[] = {
>  		.write_u64 = mem_cgroup_swappiness_write,
>  	},
>  	{
> +		.name = "lazyfreeness",
> +		.read_u64 = mem_cgroup_lzfreeness_read,
> +		.write_u64 = mem_cgroup_lzfreeness_write,
> +	},
> +	{
>  		.name = "move_charge_at_immigrate",
>  		.read_u64 = mem_cgroup_move_charge_read,
>  		.write_u64 = mem_cgroup_move_charge_write,
> @@ -4305,6 +4334,7 @@ mem_cgroup_css_online(struct cgroup_subsys_state *css)
>  	memcg->use_hierarchy = parent->use_hierarchy;
>  	memcg->oom_kill_disable = parent->oom_kill_disable;
>  	memcg->swappiness = mem_cgroup_swappiness(parent);
> +	memcg->lzfreeness = mem_cgroup_lzfreeness(parent);
>  
>  	if (parent->use_hierarchy) {
>  		page_counter_init(&memcg->memory, &parent->memory);
> diff --git a/mm/vmscan.c b/mm/vmscan.c
> index cd65db9d3004..f1abc8a6ca31 100644
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c
> @@ -141,6 +141,10 @@ struct scan_control {
>   */
>  int vm_swappiness = 60;
>  /*
> + * From 0 .. 100.  Higher means more lazy freeing.
> + */
> +int vm_lazyfreeness = 20;
> +/*
>   * The total number of pages which are beyond the high watermark within all
>   * zones.
>   */
> @@ -2012,10 +2016,11 @@ enum scan_balance {
>   *
>   * nr[0] = anon inactive pages to scan; nr[1] = anon active pages to scan
>   * nr[2] = file inactive pages to scan; nr[3] = file active pages to scan
> + * nr[4] = lazy free pages to scan;
>   */
>  static void get_scan_count(struct lruvec *lruvec, int swappiness,
> -			   struct scan_control *sc, unsigned long *nr,
> -			   unsigned long *lru_pages)
> +			int lzfreeness, struct scan_control *sc,
> +			unsigned long *nr, unsigned long *lru_pages)
>  {
>  	struct zone_reclaim_stat *reclaim_stat = &lruvec->reclaim_stat;
>  	u64 fraction[2];
> @@ -2023,12 +2028,13 @@ static void get_scan_count(struct lruvec *lruvec, int swappiness,
>  	struct zone *zone = lruvec_zone(lruvec);
>  	unsigned long anon_prio, file_prio;
>  	enum scan_balance scan_balance;
> -	unsigned long anon, file;
> +	unsigned long anon, file, lzfree;
>  	bool force_scan = false;
>  	unsigned long ap, fp;
>  	enum lru_list lru;
>  	bool some_scanned;
>  	int pass;
> +	unsigned long scan_lzfree = 0;
>  
>  	/*
>  	 * If the zone or memcg is small, nr[l] can be 0.  This
> @@ -2166,7 +2172,7 @@ static void get_scan_count(struct lruvec *lruvec, int swappiness,
>  	/* Only use force_scan on second pass. */
>  	for (pass = 0; !some_scanned && pass < 2; pass++) {
>  		*lru_pages = 0;
> -		for_each_evictable_lru(lru) {
> +		for_each_anon_file_lru(lru) {
>  			int file = is_file_lru(lru);
>  			unsigned long size;
>  			unsigned long scan;
> @@ -2212,6 +2218,28 @@ static void get_scan_count(struct lruvec *lruvec, int swappiness,
>  			some_scanned |= !!scan;
>  		}
>  	}
> +
> +	lzfree = get_lru_size(lruvec, LRU_LZFREE);
> +	if (lzfree) {
> +		scan_lzfree = sc->nr_to_reclaim *
> +				(DEF_PRIORITY - sc->priority);

scan_lzfree == 0 if sc->priority == DEF_PRIORITY, is this intended?
> +		scan_lzfree = div64_u64(scan_lzfree *
> +					lzfreeness, 50);
> +		if (!scan_lzfree) {
> +			unsigned long zonefile, zonefree;
> +
> +			zonefree = zone_page_state(zone, NR_FREE_PAGES);
> +			zonefile = zone_page_state(zone, NR_ACTIVE_FILE) +
> +				zone_page_state(zone, NR_INACTIVE_FILE);
> +			if (unlikely(zonefile + zonefree <=
> +					high_wmark_pages(zone))) {
> +				scan_lzfree = get_lru_size(lruvec,
> +						LRU_LZFREE) >> sc->priority;
> +			}
> +		}
> +	}
> +
> +	nr[LRU_LZFREE] = min(scan_lzfree, lzfree);
>  }

Looks there is no setting to only reclaim lazyfree pages. Could we have an
option for this? It's legit we don't want to trash page cache because of
lazyfree memory.

Thanks,
Shaohua
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/