diff mbox series

[v2,1/4] mm/memcg: Revert ("mm/memcg: optimize user context object stock access")

Message ID 20220211223537.2175879-2-bigeasy@linutronix.de (mailing list archive)
State New
Headers show
Series mm/memcg: Address PREEMPT_RT problems instead of disabling it. | expand

Commit Message

Sebastian Andrzej Siewior Feb. 11, 2022, 10:35 p.m. UTC
From: Michal Hocko <mhocko@suse.com>

The optimisation is based on a micro benchmark where local_irq_save() is
more expensive than a preempt_disable(). There is no evidence that it is
visible in a real-world workload and there are CPUs where the opposite is
true (local_irq_save() is cheaper than preempt_disable()).

Based on micro benchmarks, the optimisation makes sense on PREEMPT_NONE
where preempt_disable() is optimized away. There is no improvement with
PREEMPT_DYNAMIC since the preemption counter is always available.

The optimization makes also the PREEMPT_RT integration more complicated
since most of the assumption are not true on PREEMPT_RT.

Revert the optimisation since it complicates the PREEMPT_RT integration
and the improvement is hardly visible.

[ bigeasy: Patch body around Michal's diff ]

Link: https://lore.kernel.org/all/YgOGkXXCrD%2F1k+p4@dhcp22.suse.cz
Link: https://lkml.kernel.org/r/YdX+INO9gQje6d0S@linutronix.de
Signed-off-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
---
 mm/memcontrol.c | 94 ++++++++++++++-----------------------------------
 1 file changed, 27 insertions(+), 67 deletions(-)

Comments

Johannes Weiner Feb. 14, 2022, 4:23 p.m. UTC | #1
On Fri, Feb 11, 2022 at 11:35:34PM +0100, Sebastian Andrzej Siewior wrote:
> From: Michal Hocko <mhocko@suse.com>
> 
> The optimisation is based on a micro benchmark where local_irq_save() is
> more expensive than a preempt_disable(). There is no evidence that it is
> visible in a real-world workload and there are CPUs where the opposite is
> true (local_irq_save() is cheaper than preempt_disable()).
> 
> Based on micro benchmarks, the optimisation makes sense on PREEMPT_NONE
> where preempt_disable() is optimized away. There is no improvement with
> PREEMPT_DYNAMIC since the preemption counter is always available.
> 
> The optimization makes also the PREEMPT_RT integration more complicated
> since most of the assumption are not true on PREEMPT_RT.
> 
> Revert the optimisation since it complicates the PREEMPT_RT integration
> and the improvement is hardly visible.
> 
> [ bigeasy: Patch body around Michal's diff ]
> 
> Link: https://lore.kernel.org/all/YgOGkXXCrD%2F1k+p4@dhcp22.suse.cz
> Link: https://lkml.kernel.org/r/YdX+INO9gQje6d0S@linutronix.de
> Signed-off-by: Michal Hocko <mhocko@suse.com>
> Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>

Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Roman Gushchin Feb. 14, 2022, 7:45 p.m. UTC | #2
On Fri, Feb 11, 2022 at 11:35:34PM +0100, Sebastian Andrzej Siewior wrote:
> From: Michal Hocko <mhocko@suse.com>
> 
> The optimisation is based on a micro benchmark where local_irq_save() is
> more expensive than a preempt_disable(). There is no evidence that it is
> visible in a real-world workload and there are CPUs where the opposite is
> true (local_irq_save() is cheaper than preempt_disable()).
> 
> Based on micro benchmarks, the optimisation makes sense on PREEMPT_NONE
> where preempt_disable() is optimized away. There is no improvement with
> PREEMPT_DYNAMIC since the preemption counter is always available.
> 
> The optimization makes also the PREEMPT_RT integration more complicated
> since most of the assumption are not true on PREEMPT_RT.
> 
> Revert the optimisation since it complicates the PREEMPT_RT integration
> and the improvement is hardly visible.
> 
> [ bigeasy: Patch body around Michal's diff ]
> 
> Link: https://lore.kernel.org/all/YgOGkXXCrD%2F1k+p4@dhcp22.suse.cz
> Link: https://lkml.kernel.org/r/YdX+INO9gQje6d0S@linutronix.de
> Signed-off-by: Michal Hocko <mhocko@suse.com>
> Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>

Acked-by: Roman Gushchin <guro@fb.com>

Thanks!
diff mbox series

Patch

diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 09d342c7cbd0d..4b1572ae990d8 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -2085,23 +2085,17 @@  void unlock_page_memcg(struct page *page)
 	folio_memcg_unlock(page_folio(page));
 }
 
-struct obj_stock {
+struct memcg_stock_pcp {
+	struct mem_cgroup *cached; /* this never be root cgroup */
+	unsigned int nr_pages;
+
 #ifdef CONFIG_MEMCG_KMEM
 	struct obj_cgroup *cached_objcg;
 	struct pglist_data *cached_pgdat;
 	unsigned int nr_bytes;
 	int nr_slab_reclaimable_b;
 	int nr_slab_unreclaimable_b;
-#else
-	int dummy[0];
 #endif
-};
-
-struct memcg_stock_pcp {
-	struct mem_cgroup *cached; /* this never be root cgroup */
-	unsigned int nr_pages;
-	struct obj_stock task_obj;
-	struct obj_stock irq_obj;
 
 	struct work_struct work;
 	unsigned long flags;
@@ -2111,12 +2105,12 @@  static DEFINE_PER_CPU(struct memcg_stock_pcp, memcg_stock);
 static DEFINE_MUTEX(percpu_charge_mutex);
 
 #ifdef CONFIG_MEMCG_KMEM
-static void drain_obj_stock(struct obj_stock *stock);
+static void drain_obj_stock(struct memcg_stock_pcp *stock);
 static bool obj_stock_flush_required(struct memcg_stock_pcp *stock,
 				     struct mem_cgroup *root_memcg);
 
 #else
-static inline void drain_obj_stock(struct obj_stock *stock)
+static inline void drain_obj_stock(struct memcg_stock_pcp *stock)
 {
 }
 static bool obj_stock_flush_required(struct memcg_stock_pcp *stock,
@@ -2193,9 +2187,7 @@  static void drain_local_stock(struct work_struct *dummy)
 	local_irq_save(flags);
 
 	stock = this_cpu_ptr(&memcg_stock);
-	drain_obj_stock(&stock->irq_obj);
-	if (in_task())
-		drain_obj_stock(&stock->task_obj);
+	drain_obj_stock(stock);
 	drain_stock(stock);
 	clear_bit(FLUSHING_CACHED_CHARGE, &stock->flags);
 
@@ -2770,41 +2762,6 @@  static struct mem_cgroup *get_mem_cgroup_from_objcg(struct obj_cgroup *objcg)
  */
 #define OBJCGS_CLEAR_MASK	(__GFP_DMA | __GFP_RECLAIMABLE | __GFP_ACCOUNT)
 
-/*
- * Most kmem_cache_alloc() calls are from user context. The irq disable/enable
- * sequence used in this case to access content from object stock is slow.
- * To optimize for user context access, there are now two object stocks for
- * task context and interrupt context access respectively.
- *
- * The task context object stock can be accessed by disabling preemption only
- * which is cheap in non-preempt kernel. The interrupt context object stock
- * can only be accessed after disabling interrupt. User context code can
- * access interrupt object stock, but not vice versa.
- */
-static inline struct obj_stock *get_obj_stock(unsigned long *pflags)
-{
-	struct memcg_stock_pcp *stock;
-
-	if (likely(in_task())) {
-		*pflags = 0UL;
-		preempt_disable();
-		stock = this_cpu_ptr(&memcg_stock);
-		return &stock->task_obj;
-	}
-
-	local_irq_save(*pflags);
-	stock = this_cpu_ptr(&memcg_stock);
-	return &stock->irq_obj;
-}
-
-static inline void put_obj_stock(unsigned long flags)
-{
-	if (likely(in_task()))
-		preempt_enable();
-	else
-		local_irq_restore(flags);
-}
-
 /*
  * mod_objcg_mlstate() may be called with irq enabled, so
  * mod_memcg_lruvec_state() should be used.
@@ -3075,10 +3032,13 @@  void __memcg_kmem_uncharge_page(struct page *page, int order)
 void mod_objcg_state(struct obj_cgroup *objcg, struct pglist_data *pgdat,
 		     enum node_stat_item idx, int nr)
 {
+	struct memcg_stock_pcp *stock;
 	unsigned long flags;
-	struct obj_stock *stock = get_obj_stock(&flags);
 	int *bytes;
 
+	local_irq_save(flags);
+	stock = this_cpu_ptr(&memcg_stock);
+
 	/*
 	 * Save vmstat data in stock and skip vmstat array update unless
 	 * accumulating over a page of vmstat data or when pgdat or idx
@@ -3129,26 +3089,29 @@  void mod_objcg_state(struct obj_cgroup *objcg, struct pglist_data *pgdat,
 	if (nr)
 		mod_objcg_mlstate(objcg, pgdat, idx, nr);
 
-	put_obj_stock(flags);
+	local_irq_restore(flags);
 }
 
 static bool consume_obj_stock(struct obj_cgroup *objcg, unsigned int nr_bytes)
 {
+	struct memcg_stock_pcp *stock;
 	unsigned long flags;
-	struct obj_stock *stock = get_obj_stock(&flags);
 	bool ret = false;
 
+	local_irq_save(flags);
+
+	stock = this_cpu_ptr(&memcg_stock);
 	if (objcg == stock->cached_objcg && stock->nr_bytes >= nr_bytes) {
 		stock->nr_bytes -= nr_bytes;
 		ret = true;
 	}
 
-	put_obj_stock(flags);
+	local_irq_restore(flags);
 
 	return ret;
 }
 
-static void drain_obj_stock(struct obj_stock *stock)
+static void drain_obj_stock(struct memcg_stock_pcp *stock)
 {
 	struct obj_cgroup *old = stock->cached_objcg;
 
@@ -3204,13 +3167,8 @@  static bool obj_stock_flush_required(struct memcg_stock_pcp *stock,
 {
 	struct mem_cgroup *memcg;
 
-	if (in_task() && stock->task_obj.cached_objcg) {
-		memcg = obj_cgroup_memcg(stock->task_obj.cached_objcg);
-		if (memcg && mem_cgroup_is_descendant(memcg, root_memcg))
-			return true;
-	}
-	if (stock->irq_obj.cached_objcg) {
-		memcg = obj_cgroup_memcg(stock->irq_obj.cached_objcg);
+	if (stock->cached_objcg) {
+		memcg = obj_cgroup_memcg(stock->cached_objcg);
 		if (memcg && mem_cgroup_is_descendant(memcg, root_memcg))
 			return true;
 	}
@@ -3221,10 +3179,13 @@  static bool obj_stock_flush_required(struct memcg_stock_pcp *stock,
 static void refill_obj_stock(struct obj_cgroup *objcg, unsigned int nr_bytes,
 			     bool allow_uncharge)
 {
+	struct memcg_stock_pcp *stock;
 	unsigned long flags;
-	struct obj_stock *stock = get_obj_stock(&flags);
 	unsigned int nr_pages = 0;
 
+	local_irq_save(flags);
+
+	stock = this_cpu_ptr(&memcg_stock);
 	if (stock->cached_objcg != objcg) { /* reset if necessary */
 		drain_obj_stock(stock);
 		obj_cgroup_get(objcg);
@@ -3240,7 +3201,7 @@  static void refill_obj_stock(struct obj_cgroup *objcg, unsigned int nr_bytes,
 		stock->nr_bytes &= (PAGE_SIZE - 1);
 	}
 
-	put_obj_stock(flags);
+	local_irq_restore(flags);
 
 	if (nr_pages)
 		obj_cgroup_uncharge_pages(objcg, nr_pages);
@@ -6821,7 +6782,6 @@  static void uncharge_folio(struct folio *folio, struct uncharge_gather *ug)
 	long nr_pages;
 	struct mem_cgroup *memcg;
 	struct obj_cgroup *objcg;
-	bool use_objcg = folio_memcg_kmem(folio);
 
 	VM_BUG_ON_FOLIO(folio_test_lru(folio), folio);
 
@@ -6830,7 +6790,7 @@  static void uncharge_folio(struct folio *folio, struct uncharge_gather *ug)
 	 * folio memcg or objcg at this point, we have fully
 	 * exclusive access to the folio.
 	 */
-	if (use_objcg) {
+	if (folio_memcg_kmem(folio)) {
 		objcg = __folio_objcg(folio);
 		/*
 		 * This get matches the put at the end of the function and
@@ -6858,7 +6818,7 @@  static void uncharge_folio(struct folio *folio, struct uncharge_gather *ug)
 
 	nr_pages = folio_nr_pages(folio);
 
-	if (use_objcg) {
+	if (folio_memcg_kmem(folio)) {
 		ug->nr_memory += nr_pages;
 		ug->nr_kmem += nr_pages;