diff --git a/Documentation/cgroups/unified-hierarchy.txt b/Documentation/cgroups/unified-hierarchy.txt index 4f4563277864f4ce3a2bc17c76c61f7a4002e3bc..71daa35ec2d9201d2a4647e0a088af3ab36bbae3 100644 --- a/Documentation/cgroups/unified-hierarchy.txt +++ b/Documentation/cgroups/unified-hierarchy.txt @@ -327,6 +327,85 @@ supported and the interface files "release_agent" and - use_hierarchy is on by default and the cgroup file for the flag is not created. +- The original lower boundary, the soft limit, is defined as a limit + that is per default unset. As a result, the set of cgroups that + global reclaim prefers is opt-in, rather than opt-out. The costs + for optimizing these mostly negative lookups are so high that the + implementation, despite its enormous size, does not even provide the + basic desirable behavior. First off, the soft limit has no + hierarchical meaning. All configured groups are organized in a + global rbtree and treated like equal peers, regardless where they + are located in the hierarchy. This makes subtree delegation + impossible. Second, the soft limit reclaim pass is so aggressive + that it not just introduces high allocation latencies into the + system, but also impacts system performance due to overreclaim, to + the point where the feature becomes self-defeating. + + The memory.low boundary on the other hand is a top-down allocated + reserve. A cgroup enjoys reclaim protection when it and all its + ancestors are below their low boundaries, which makes delegation of + subtrees possible. Secondly, new cgroups have no reserve per + default and in the common case most cgroups are eligible for the + preferred reclaim pass. This allows the new low boundary to be + efficiently implemented with just a minor addition to the generic + reclaim code, without the need for out-of-band data structures and + reclaim passes. Because the generic reclaim code considers all + cgroups except for the ones running low in the preferred first + reclaim pass, overreclaim of individual groups is eliminated as + well, resulting in much better overall workload performance. + +- The original high boundary, the hard limit, is defined as a strict + limit that can not budge, even if the OOM killer has to be called. + But this generally goes against the goal of making the most out of + the available memory. The memory consumption of workloads varies + during runtime, and that requires users to overcommit. But doing + that with a strict upper limit requires either a fairly accurate + prediction of the working set size or adding slack to the limit. + Since working set size estimation is hard and error prone, and + getting it wrong results in OOM kills, most users tend to err on the + side of a looser limit and end up wasting precious resources. + + The memory.high boundary on the other hand can be set much more + conservatively. When hit, it throttles allocations by forcing them + into direct reclaim to work off the excess, but it never invokes the + OOM killer. As a result, a high boundary that is chosen too + aggressively will not terminate the processes, but instead it will + lead to gradual performance degradation. The user can monitor this + and make corrections until the minimal memory footprint that still + gives acceptable performance is found. + + In extreme cases, with many concurrent allocations and a complete + breakdown of reclaim progress within the group, the high boundary + can be exceeded. But even then it's mostly better to satisfy the + allocation from the slack available in other groups or the rest of + the system than killing the group. Otherwise, memory.max is there + to limit this type of spillover and ultimately contain buggy or even + malicious applications. + +- The original control file names are unwieldy and inconsistent in + many different ways. For example, the upper boundary hit count is + exported in the memory.failcnt file, but an OOM event count has to + be manually counted by listening to memory.oom_control events, and + lower boundary / soft limit events have to be counted by first + setting a threshold for that value and then counting those events. + Also, usage and limit files encode their units in the filename. + That makes the filenames very long, even though this is not + information that a user needs to be reminded of every time they type + out those names. + + To address these naming issues, as well as to signal clearly that + the new interface carries a new configuration model, the naming + conventions in it necessarily differ from the old interface. + +- The original limit files indicate the state of an unset limit with a + Very High Number, and a configured limit can be unset by echoing -1 + into those files. But that very high number is implementation and + architecture dependent and not very descriptive. And while -1 can + be understood as an underflow into the highest possible value, -2 or + -10M etc. do not work, so it's not consistent. + + memory.low, memory.high, and memory.max will use the string + "infinity" to indicate and set the highest possible value. 5. Planned Changes diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h index 353537a5981a6781bf8657eec476f3d735f6bedf..6cfd934c7c9b3ef766fdfc0fb9d75c40238e75d4 100644 --- a/include/linux/memcontrol.h +++ b/include/linux/memcontrol.h @@ -52,7 +52,27 @@ struct mem_cgroup_reclaim_cookie { unsigned int generation; }; +enum mem_cgroup_events_index { + MEM_CGROUP_EVENTS_PGPGIN, /* # of pages paged in */ + MEM_CGROUP_EVENTS_PGPGOUT, /* # of pages paged out */ + MEM_CGROUP_EVENTS_PGFAULT, /* # of page-faults */ + MEM_CGROUP_EVENTS_PGMAJFAULT, /* # of major page-faults */ + MEM_CGROUP_EVENTS_NSTATS, + /* default hierarchy events */ + MEMCG_LOW = MEM_CGROUP_EVENTS_NSTATS, + MEMCG_HIGH, + MEMCG_MAX, + MEMCG_OOM, + MEMCG_NR_EVENTS, +}; + #ifdef CONFIG_MEMCG +void mem_cgroup_events(struct mem_cgroup *memcg, + enum mem_cgroup_events_index idx, + unsigned int nr); + +bool mem_cgroup_low(struct mem_cgroup *root, struct mem_cgroup *memcg); + int mem_cgroup_try_charge(struct page *page, struct mm_struct *mm, gfp_t gfp_mask, struct mem_cgroup **memcgp); void mem_cgroup_commit_charge(struct page *page, struct mem_cgroup *memcg, @@ -175,6 +195,18 @@ void mem_cgroup_split_huge_fixup(struct page *head); #else /* CONFIG_MEMCG */ struct mem_cgroup; +static inline void mem_cgroup_events(struct mem_cgroup *memcg, + enum mem_cgroup_events_index idx, + unsigned int nr) +{ +} + +static inline bool mem_cgroup_low(struct mem_cgroup *root, + struct mem_cgroup *memcg) +{ + return false; +} + static inline int mem_cgroup_try_charge(struct page *page, struct mm_struct *mm, gfp_t gfp_mask, struct mem_cgroup **memcgp) diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 6453ea5a27aace38b9a38814a77455fadf1c105b..ee97c9ac62c08c4475166a6e60fa7c1e48c919a4 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -97,14 +97,6 @@ static const char * const mem_cgroup_stat_names[] = { "swap", }; -enum mem_cgroup_events_index { - MEM_CGROUP_EVENTS_PGPGIN, /* # of pages paged in */ - MEM_CGROUP_EVENTS_PGPGOUT, /* # of pages paged out */ - MEM_CGROUP_EVENTS_PGFAULT, /* # of page-faults */ - MEM_CGROUP_EVENTS_PGMAJFAULT, /* # of major page-faults */ - MEM_CGROUP_EVENTS_NSTATS, -}; - static const char * const mem_cgroup_events_names[] = { "pgpgin", "pgpgout", @@ -138,7 +130,7 @@ enum mem_cgroup_events_target { struct mem_cgroup_stat_cpu { long count[MEM_CGROUP_STAT_NSTATS]; - unsigned long events[MEM_CGROUP_EVENTS_NSTATS]; + unsigned long events[MEMCG_NR_EVENTS]; unsigned long nr_page_events; unsigned long targets[MEM_CGROUP_NTARGETS]; }; @@ -284,6 +276,10 @@ struct mem_cgroup { struct page_counter memsw; struct page_counter kmem; + /* Normal memory consumption range */ + unsigned long low; + unsigned long high; + unsigned long soft_limit; /* vmpressure notifications */ @@ -2315,6 +2311,8 @@ static int try_charge(struct mem_cgroup *memcg, gfp_t gfp_mask, if (!(gfp_mask & __GFP_WAIT)) goto nomem; + mem_cgroup_events(mem_over_limit, MEMCG_MAX, 1); + nr_reclaimed = try_to_free_mem_cgroup_pages(mem_over_limit, nr_pages, gfp_mask, may_swap); @@ -2356,6 +2354,8 @@ static int try_charge(struct mem_cgroup *memcg, gfp_t gfp_mask, if (fatal_signal_pending(current)) goto bypass; + mem_cgroup_events(mem_over_limit, MEMCG_OOM, 1); + mem_cgroup_oom(mem_over_limit, gfp_mask, get_order(nr_pages)); nomem: if (!(gfp_mask & __GFP_NOFAIL)) @@ -2367,6 +2367,16 @@ static int try_charge(struct mem_cgroup *memcg, gfp_t gfp_mask, css_get_many(&memcg->css, batch); if (batch > nr_pages) refill_stock(memcg, batch - nr_pages); + /* + * If the hierarchy is above the normal consumption range, + * make the charging task trim their excess contribution. + */ + do { + if (page_counter_read(&memcg->memory) <= memcg->high) + continue; + mem_cgroup_events(memcg, MEMCG_HIGH, 1); + try_to_free_mem_cgroup_pages(memcg, nr_pages, gfp_mask, true); + } while ((memcg = parent_mem_cgroup(memcg))); done: return ret; } @@ -4276,7 +4286,7 @@ static ssize_t memcg_write_event_control(struct kernfs_open_file *of, return ret; } -static struct cftype mem_cgroup_files[] = { +static struct cftype mem_cgroup_legacy_files[] = { { .name = "usage_in_bytes", .private = MEMFILE_PRIVATE(_MEM, RES_USAGE), @@ -4552,6 +4562,7 @@ mem_cgroup_css_alloc(struct cgroup_subsys_state *parent_css) if (parent_css == NULL) { root_mem_cgroup = memcg; page_counter_init(&memcg->memory, NULL); + memcg->high = PAGE_COUNTER_MAX; memcg->soft_limit = PAGE_COUNTER_MAX; page_counter_init(&memcg->memsw, NULL); page_counter_init(&memcg->kmem, NULL); @@ -4597,6 +4608,7 @@ mem_cgroup_css_online(struct cgroup_subsys_state *css) if (parent->use_hierarchy) { page_counter_init(&memcg->memory, &parent->memory); + memcg->high = PAGE_COUNTER_MAX; memcg->soft_limit = PAGE_COUNTER_MAX; page_counter_init(&memcg->memsw, &parent->memsw); page_counter_init(&memcg->kmem, &parent->kmem); @@ -4607,6 +4619,7 @@ mem_cgroup_css_online(struct cgroup_subsys_state *css) */ } else { page_counter_init(&memcg->memory, NULL); + memcg->high = PAGE_COUNTER_MAX; memcg->soft_limit = PAGE_COUNTER_MAX; page_counter_init(&memcg->memsw, NULL); page_counter_init(&memcg->kmem, NULL); @@ -4682,6 +4695,8 @@ static void mem_cgroup_css_reset(struct cgroup_subsys_state *css) mem_cgroup_resize_limit(memcg, PAGE_COUNTER_MAX); mem_cgroup_resize_memsw_limit(memcg, PAGE_COUNTER_MAX); memcg_update_kmem_limit(memcg, PAGE_COUNTER_MAX); + memcg->low = 0; + memcg->high = PAGE_COUNTER_MAX; memcg->soft_limit = PAGE_COUNTER_MAX; } @@ -5267,6 +5282,147 @@ static void mem_cgroup_bind(struct cgroup_subsys_state *root_css) mem_cgroup_from_css(root_css)->use_hierarchy = true; } +static u64 memory_current_read(struct cgroup_subsys_state *css, + struct cftype *cft) +{ + return mem_cgroup_usage(mem_cgroup_from_css(css), false); +} + +static int memory_low_show(struct seq_file *m, void *v) +{ + struct mem_cgroup *memcg = mem_cgroup_from_css(seq_css(m)); + unsigned long low = ACCESS_ONCE(memcg->low); + + if (low == PAGE_COUNTER_MAX) + seq_puts(m, "infinity\n"); + else + seq_printf(m, "%llu\n", (u64)low * PAGE_SIZE); + + return 0; +} + +static ssize_t memory_low_write(struct kernfs_open_file *of, + char *buf, size_t nbytes, loff_t off) +{ + struct mem_cgroup *memcg = mem_cgroup_from_css(of_css(of)); + unsigned long low; + int err; + + buf = strstrip(buf); + err = page_counter_memparse(buf, "infinity", &low); + if (err) + return err; + + memcg->low = low; + + return nbytes; +} + +static int memory_high_show(struct seq_file *m, void *v) +{ + struct mem_cgroup *memcg = mem_cgroup_from_css(seq_css(m)); + unsigned long high = ACCESS_ONCE(memcg->high); + + if (high == PAGE_COUNTER_MAX) + seq_puts(m, "infinity\n"); + else + seq_printf(m, "%llu\n", (u64)high * PAGE_SIZE); + + return 0; +} + +static ssize_t memory_high_write(struct kernfs_open_file *of, + char *buf, size_t nbytes, loff_t off) +{ + struct mem_cgroup *memcg = mem_cgroup_from_css(of_css(of)); + unsigned long high; + int err; + + buf = strstrip(buf); + err = page_counter_memparse(buf, "infinity", &high); + if (err) + return err; + + memcg->high = high; + + return nbytes; +} + +static int memory_max_show(struct seq_file *m, void *v) +{ + struct mem_cgroup *memcg = mem_cgroup_from_css(seq_css(m)); + unsigned long max = ACCESS_ONCE(memcg->memory.limit); + + if (max == PAGE_COUNTER_MAX) + seq_puts(m, "infinity\n"); + else + seq_printf(m, "%llu\n", (u64)max * PAGE_SIZE); + + return 0; +} + +static ssize_t memory_max_write(struct kernfs_open_file *of, + char *buf, size_t nbytes, loff_t off) +{ + struct mem_cgroup *memcg = mem_cgroup_from_css(of_css(of)); + unsigned long max; + int err; + + buf = strstrip(buf); + err = page_counter_memparse(buf, "infinity", &max); + if (err) + return err; + + err = mem_cgroup_resize_limit(memcg, max); + if (err) + return err; + + return nbytes; +} + +static int memory_events_show(struct seq_file *m, void *v) +{ + struct mem_cgroup *memcg = mem_cgroup_from_css(seq_css(m)); + + seq_printf(m, "low %lu\n", mem_cgroup_read_events(memcg, MEMCG_LOW)); + seq_printf(m, "high %lu\n", mem_cgroup_read_events(memcg, MEMCG_HIGH)); + seq_printf(m, "max %lu\n", mem_cgroup_read_events(memcg, MEMCG_MAX)); + seq_printf(m, "oom %lu\n", mem_cgroup_read_events(memcg, MEMCG_OOM)); + + return 0; +} + +static struct cftype memory_files[] = { + { + .name = "current", + .read_u64 = memory_current_read, + }, + { + .name = "low", + .flags = CFTYPE_NOT_ON_ROOT, + .seq_show = memory_low_show, + .write = memory_low_write, + }, + { + .name = "high", + .flags = CFTYPE_NOT_ON_ROOT, + .seq_show = memory_high_show, + .write = memory_high_write, + }, + { + .name = "max", + .flags = CFTYPE_NOT_ON_ROOT, + .seq_show = memory_max_show, + .write = memory_max_write, + }, + { + .name = "events", + .flags = CFTYPE_NOT_ON_ROOT, + .seq_show = memory_events_show, + }, + { } /* terminate */ +}; + struct cgroup_subsys memory_cgrp_subsys = { .css_alloc = mem_cgroup_css_alloc, .css_online = mem_cgroup_css_online, @@ -5277,7 +5433,8 @@ struct cgroup_subsys memory_cgrp_subsys = { .cancel_attach = mem_cgroup_cancel_attach, .attach = mem_cgroup_move_task, .bind = mem_cgroup_bind, - .legacy_cftypes = mem_cgroup_files, + .dfl_cftypes = memory_files, + .legacy_cftypes = mem_cgroup_legacy_files, .early_init = 0, }; @@ -5312,6 +5469,56 @@ static void __init enable_swap_cgroup(void) } #endif +/** + * mem_cgroup_events - count memory events against a cgroup + * @memcg: the memory cgroup + * @idx: the event index + * @nr: the number of events to account for + */ +void mem_cgroup_events(struct mem_cgroup *memcg, + enum mem_cgroup_events_index idx, + unsigned int nr) +{ + this_cpu_add(memcg->stat->events[idx], nr); +} + +/** + * mem_cgroup_low - check if memory consumption is below the normal range + * @root: the highest ancestor to consider + * @memcg: the memory cgroup to check + * + * Returns %true if memory consumption of @memcg, and that of all + * configurable ancestors up to @root, is below the normal range. + */ +bool mem_cgroup_low(struct mem_cgroup *root, struct mem_cgroup *memcg) +{ + if (mem_cgroup_disabled()) + return false; + + /* + * The toplevel group doesn't have a configurable range, so + * it's never low when looked at directly, and it is not + * considered an ancestor when assessing the hierarchy. + */ + + if (memcg == root_mem_cgroup) + return false; + + if (page_counter_read(&memcg->memory) > memcg->low) + return false; + + while (memcg != root) { + memcg = parent_mem_cgroup(memcg); + + if (memcg == root_mem_cgroup) + break; + + if (page_counter_read(&memcg->memory) > memcg->low) + return false; + } + return true; +} + #ifdef CONFIG_MEMCG_SWAP /** * mem_cgroup_swapout - transfer a memsw charge to swap diff --git a/mm/vmscan.c b/mm/vmscan.c index b6dfa0081a8ecb6d24555085bdd21b2c4c50722d..8e645ee520455432e583deb2614199d84e422066 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -91,6 +91,9 @@ struct scan_control { /* Can pages be swapped as part of reclaim? */ unsigned int may_swap:1; + /* Can cgroups be reclaimed below their normal consumption range? */ + unsigned int may_thrash:1; + unsigned int hibernation_mode:1; /* One of the zones is ready for compaction */ @@ -2294,6 +2297,12 @@ static bool shrink_zone(struct zone *zone, struct scan_control *sc, struct lruvec *lruvec; int swappiness; + if (mem_cgroup_low(root, memcg)) { + if (!sc->may_thrash) + continue; + mem_cgroup_events(memcg, MEMCG_LOW, 1); + } + lruvec = mem_cgroup_zone_lruvec(zone, memcg); swappiness = mem_cgroup_swappiness(memcg); @@ -2315,8 +2324,7 @@ static bool shrink_zone(struct zone *zone, struct scan_control *sc, mem_cgroup_iter_break(root, memcg); break; } - memcg = mem_cgroup_iter(root, memcg, &reclaim); - } while (memcg); + } while ((memcg = mem_cgroup_iter(root, memcg, &reclaim))); /* * Shrink the slab caches in the same proportion that @@ -2519,10 +2527,11 @@ static bool shrink_zones(struct zonelist *zonelist, struct scan_control *sc) static unsigned long do_try_to_free_pages(struct zonelist *zonelist, struct scan_control *sc) { + int initial_priority = sc->priority; unsigned long total_scanned = 0; unsigned long writeback_threshold; bool zones_reclaimable; - +retry: delayacct_freepages_start(); if (global_reclaim(sc)) @@ -2572,6 +2581,13 @@ static unsigned long do_try_to_free_pages(struct zonelist *zonelist, if (sc->compaction_ready) return 1; + /* Untapped cgroup reserves? Don't OOM, retry. */ + if (!sc->may_thrash) { + sc->priority = initial_priority; + sc->may_thrash = 1; + goto retry; + } + /* Any of the zones still reclaimable? Don't OOM. */ if (zones_reclaimable) return 1;