^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1) // SPDX-License-Identifier: GPL-2.0-only
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 3) * Linux VM pressure
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5) * Copyright 2012 Linaro Ltd.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6) * Anton Vorontsov <anton.vorontsov@linaro.org>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8) * Based on ideas from Andrew Morton, David Rientjes, KOSAKI Motohiro,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9) * Leonid Moiseichuk, Mel Gorman, Minchan Kim and Pekka Enberg.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12) #include <linux/cgroup.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13) #include <linux/fs.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14) #include <linux/log2.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) #include <linux/sched.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16) #include <linux/mm.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17) #include <linux/vmstat.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18) #include <linux/eventfd.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) #include <linux/slab.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20) #include <linux/swap.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21) #include <linux/printk.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22) #include <linux/vmpressure.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24) #include <trace/hooks/mm.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27) * The window size (vmpressure_win) is the number of scanned pages before
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) * we try to analyze scanned/reclaimed ratio. So the window is used as a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29) * rate-limit tunable for the "low" level notification, and also for
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30) * averaging the ratio for medium/critical levels. Using small window
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31) * sizes can cause lot of false positives, but too big window size will
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32) * delay the notifications.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) * As the vmscan reclaimer logic works with chunks which are multiple of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35) * SWAP_CLUSTER_MAX, it makes sense to use it for the window size as well.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37) * TODO: Make the window size depend on machine size, as we do for vmstat
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38) * thresholds. Currently we set it to 512 pages (2MB for 4KB pages).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40) static const unsigned long vmpressure_win = SWAP_CLUSTER_MAX * 16;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) * These thresholds are used when we account memory pressure through
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44) * scanned/reclaimed ratio. The current values were chosen empirically. In
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45) * essence, they are percents: the higher the value, the more number
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46) * unsuccessful reclaims there were.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48) static const unsigned int vmpressure_level_med = 60;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49) static const unsigned int vmpressure_level_critical = 95;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52) * When there are too little pages left to scan, vmpressure() may miss the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53) * critical pressure as number of pages will be less than "window size".
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54) * However, in that case the vmscan priority will raise fast as the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55) * reclaimer will try to scan LRUs more deeply.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57) * The vmscan logic considers these special priorities:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59) * prio == DEF_PRIORITY (12): reclaimer starts with that value
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60) * prio <= DEF_PRIORITY - 2 : kswapd becomes somewhat overwhelmed
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61) * prio == 0 : close to OOM, kernel scans every page in an lru
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63) * Any value in this range is acceptable for this tunable (i.e. from 12 to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64) * 0). Current value for the vmpressure_level_critical_prio is chosen
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65) * empirically, but the number, in essence, means that we consider
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66) * critical level when scanning depth is ~10% of the lru size (vmscan
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67) * scans 'lru_size >> prio' pages, so it is actually 12.5%, or one
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68) * eights).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70) static const unsigned int vmpressure_level_critical_prio = ilog2(100 / 10);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72) static struct vmpressure *work_to_vmpressure(struct work_struct *work)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74) return container_of(work, struct vmpressure, work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77) static struct vmpressure *vmpressure_parent(struct vmpressure *vmpr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) struct cgroup_subsys_state *css = vmpressure_to_css(vmpr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80) struct mem_cgroup *memcg = mem_cgroup_from_css(css);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82) memcg = parent_mem_cgroup(memcg);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83) if (!memcg)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 84) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 85) return memcg_to_vmpressure(memcg);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 86) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 87)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 88) enum vmpressure_levels {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 89) VMPRESSURE_LOW = 0,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 90) VMPRESSURE_MEDIUM,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 91) VMPRESSURE_CRITICAL,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 92) VMPRESSURE_NUM_LEVELS,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 93) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 94)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 95) enum vmpressure_modes {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 96) VMPRESSURE_NO_PASSTHROUGH = 0,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 97) VMPRESSURE_HIERARCHY,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 98) VMPRESSURE_LOCAL,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 99) VMPRESSURE_NUM_MODES,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102) static const char * const vmpressure_str_levels[] = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103) [VMPRESSURE_LOW] = "low",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104) [VMPRESSURE_MEDIUM] = "medium",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105) [VMPRESSURE_CRITICAL] = "critical",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) static const char * const vmpressure_str_modes[] = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109) [VMPRESSURE_NO_PASSTHROUGH] = "default",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110) [VMPRESSURE_HIERARCHY] = "hierarchy",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111) [VMPRESSURE_LOCAL] = "local",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114) static enum vmpressure_levels vmpressure_level(unsigned long pressure)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 116) if (pressure >= vmpressure_level_critical)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 117) return VMPRESSURE_CRITICAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) else if (pressure >= vmpressure_level_med)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119) return VMPRESSURE_MEDIUM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120) return VMPRESSURE_LOW;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123) static enum vmpressure_levels vmpressure_calc_level(unsigned long scanned,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124) unsigned long reclaimed)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126) unsigned long scale = scanned + reclaimed;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127) unsigned long pressure = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130) * reclaimed can be greater than scanned for things such as reclaimed
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131) * slab pages. shrink_node() just adds reclaimed pages without a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132) * related increment to scanned pages.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134) if (reclaimed >= scanned)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137) * We calculate the ratio (in percents) of how many pages were
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138) * scanned vs. reclaimed in a given time frame (window). Note that
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139) * time is in VM reclaimer's "ticks", i.e. number of pages
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140) * scanned. This makes it possible to set desired reaction time
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141) * and serves as a ratelimit.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143) pressure = scale - (reclaimed * scale / scanned);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144) pressure = pressure * 100 / scale;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147) pr_debug("%s: %3lu (s: %lu r: %lu)\n", __func__, pressure,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148) scanned, reclaimed);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150) return vmpressure_level(pressure);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153) struct vmpressure_event {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154) struct eventfd_ctx *efd;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155) enum vmpressure_levels level;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156) enum vmpressure_modes mode;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157) struct list_head node;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160) static bool vmpressure_event(struct vmpressure *vmpr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161) const enum vmpressure_levels level,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162) bool ancestor, bool signalled)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164) struct vmpressure_event *ev;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165) bool ret = false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167) mutex_lock(&vmpr->events_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168) list_for_each_entry(ev, &vmpr->events, node) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169) if (ancestor && ev->mode == VMPRESSURE_LOCAL)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171) if (signalled && ev->mode == VMPRESSURE_NO_PASSTHROUGH)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173) if (level < ev->level)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175) eventfd_signal(ev->efd, 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176) ret = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 178) mutex_unlock(&vmpr->events_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 179)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 180) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 181) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 182)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 183) static void vmpressure_work_fn(struct work_struct *work)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 184) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 185) struct vmpressure *vmpr = work_to_vmpressure(work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 186) unsigned long scanned;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 187) unsigned long reclaimed;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 188) enum vmpressure_levels level;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 189) bool ancestor = false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 190) bool signalled = false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 191)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 192) spin_lock(&vmpr->sr_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 193) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 194) * Several contexts might be calling vmpressure(), so it is
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 195) * possible that the work was rescheduled again before the old
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 196) * work context cleared the counters. In that case we will run
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 197) * just after the old work returns, but then scanned might be zero
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 198) * here. No need for any locks here since we don't care if
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 199) * vmpr->reclaimed is in sync.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 200) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 201) scanned = vmpr->tree_scanned;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 202) if (!scanned) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 203) spin_unlock(&vmpr->sr_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 204) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 205) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 206)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 207) reclaimed = vmpr->tree_reclaimed;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 208) vmpr->tree_scanned = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 209) vmpr->tree_reclaimed = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 210) spin_unlock(&vmpr->sr_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 211)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 212) level = vmpressure_calc_level(scanned, reclaimed);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 213)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 214) do {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 215) if (vmpressure_event(vmpr, level, ancestor, signalled))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 216) signalled = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 217) ancestor = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 218) } while ((vmpr = vmpressure_parent(vmpr)));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 219) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 220)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 221) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 222) * vmpressure() - Account memory pressure through scanned/reclaimed ratio
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 223) * @gfp: reclaimer's gfp mask
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 224) * @memcg: cgroup memory controller handle
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 225) * @tree: legacy subtree mode
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 226) * @scanned: number of pages scanned
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 227) * @reclaimed: number of pages reclaimed
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 228) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 229) * This function should be called from the vmscan reclaim path to account
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 230) * "instantaneous" memory pressure (scanned/reclaimed ratio). The raw
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 231) * pressure index is then further refined and averaged over time.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 232) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 233) * If @tree is set, vmpressure is in traditional userspace reporting
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 234) * mode: @memcg is considered the pressure root and userspace is
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 235) * notified of the entire subtree's reclaim efficiency.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 236) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 237) * If @tree is not set, reclaim efficiency is recorded for @memcg, and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 238) * only in-kernel users are notified.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 239) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 240) * This function does not return any value.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 241) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 242) void vmpressure(gfp_t gfp, struct mem_cgroup *memcg, bool tree,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 243) unsigned long scanned, unsigned long reclaimed)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 244) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 245) struct vmpressure *vmpr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 246) bool bypass = false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 247)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 248) if (mem_cgroup_disabled())
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 249) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 250)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 251) vmpr = memcg_to_vmpressure(memcg);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 252)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 253) trace_android_vh_vmpressure(memcg, &bypass);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 254) if (unlikely(bypass))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 255) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 256)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 257) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 258) * Here we only want to account pressure that userland is able to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 259) * help us with. For example, suppose that DMA zone is under
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 260) * pressure; if we notify userland about that kind of pressure,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 261) * then it will be mostly a waste as it will trigger unnecessary
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 262) * freeing of memory by userland (since userland is more likely to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 263) * have HIGHMEM/MOVABLE pages instead of the DMA fallback). That
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 264) * is why we include only movable, highmem and FS/IO pages.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 265) * Indirect reclaim (kswapd) sets sc->gfp_mask to GFP_KERNEL, so
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 266) * we account it too.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 267) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 268) if (!(gfp & (__GFP_HIGHMEM | __GFP_MOVABLE | __GFP_IO | __GFP_FS)))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 269) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 270)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 271) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 272) * If we got here with no pages scanned, then that is an indicator
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 273) * that reclaimer was unable to find any shrinkable LRUs at the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 274) * current scanning depth. But it does not mean that we should
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 275) * report the critical pressure, yet. If the scanning priority
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 276) * (scanning depth) goes too high (deep), we will be notified
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 277) * through vmpressure_prio(). But so far, keep calm.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 278) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 279) if (!scanned)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 280) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 281)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 282) if (tree) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 283) spin_lock(&vmpr->sr_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 284) scanned = vmpr->tree_scanned += scanned;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 285) vmpr->tree_reclaimed += reclaimed;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 286) spin_unlock(&vmpr->sr_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 287)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 288) if (scanned < vmpressure_win)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 289) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 290) schedule_work(&vmpr->work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 291) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 292) enum vmpressure_levels level;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 293)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 294) /* For now, no users for root-level efficiency */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 295) if (!memcg || mem_cgroup_is_root(memcg))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 296) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 297)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 298) spin_lock(&vmpr->sr_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 299) scanned = vmpr->scanned += scanned;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 300) reclaimed = vmpr->reclaimed += reclaimed;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 301) if (scanned < vmpressure_win) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 302) spin_unlock(&vmpr->sr_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 303) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 304) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 305) vmpr->scanned = vmpr->reclaimed = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 306) spin_unlock(&vmpr->sr_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 307)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 308) level = vmpressure_calc_level(scanned, reclaimed);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 309)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 310) if (level > VMPRESSURE_LOW) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 311) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 312) * Let the socket buffer allocator know that
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 313) * we are having trouble reclaiming LRU pages.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 314) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 315) * For hysteresis keep the pressure state
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 316) * asserted for a second in which subsequent
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 317) * pressure events can occur.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 318) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 319) memcg->socket_pressure = jiffies + HZ;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 320) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 321) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 322) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 323)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 324) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 325) * vmpressure_prio() - Account memory pressure through reclaimer priority level
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 326) * @gfp: reclaimer's gfp mask
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 327) * @memcg: cgroup memory controller handle
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 328) * @prio: reclaimer's priority
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 329) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 330) * This function should be called from the reclaim path every time when
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 331) * the vmscan's reclaiming priority (scanning depth) changes.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 332) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 333) * This function does not return any value.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 334) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 335) void vmpressure_prio(gfp_t gfp, struct mem_cgroup *memcg, int prio)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 336) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 337) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 338) * We only use prio for accounting critical level. For more info
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 339) * see comment for vmpressure_level_critical_prio variable above.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 340) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 341) if (prio > vmpressure_level_critical_prio)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 342) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 343)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 344) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 345) * OK, the prio is below the threshold, updating vmpressure
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 346) * information before shrinker dives into long shrinking of long
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 347) * range vmscan. Passing scanned = vmpressure_win, reclaimed = 0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 348) * to the vmpressure() basically means that we signal 'critical'
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 349) * level.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 350) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 351) vmpressure(gfp, memcg, true, vmpressure_win, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 352) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 353)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 354) #define MAX_VMPRESSURE_ARGS_LEN (strlen("critical") + strlen("hierarchy") + 2)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 355)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 356) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 357) * vmpressure_register_event() - Bind vmpressure notifications to an eventfd
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 358) * @memcg: memcg that is interested in vmpressure notifications
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 359) * @eventfd: eventfd context to link notifications with
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 360) * @args: event arguments (pressure level threshold, optional mode)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 361) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 362) * This function associates eventfd context with the vmpressure
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 363) * infrastructure, so that the notifications will be delivered to the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 364) * @eventfd. The @args parameter is a comma-delimited string that denotes a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 365) * pressure level threshold (one of vmpressure_str_levels, i.e. "low", "medium",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 366) * or "critical") and an optional mode (one of vmpressure_str_modes, i.e.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 367) * "hierarchy" or "local").
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 368) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 369) * To be used as memcg event method.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 370) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 371) * Return: 0 on success, -ENOMEM on memory failure or -EINVAL if @args could
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 372) * not be parsed.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 373) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 374) int vmpressure_register_event(struct mem_cgroup *memcg,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 375) struct eventfd_ctx *eventfd, const char *args)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 376) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 377) struct vmpressure *vmpr = memcg_to_vmpressure(memcg);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 378) struct vmpressure_event *ev;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 379) enum vmpressure_modes mode = VMPRESSURE_NO_PASSTHROUGH;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 380) enum vmpressure_levels level;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 381) char *spec, *spec_orig;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 382) char *token;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 383) int ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 384)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 385) spec_orig = spec = kstrndup(args, MAX_VMPRESSURE_ARGS_LEN, GFP_KERNEL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 386) if (!spec)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 387) return -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 388)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 389) /* Find required level */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 390) token = strsep(&spec, ",");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 391) ret = match_string(vmpressure_str_levels, VMPRESSURE_NUM_LEVELS, token);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 392) if (ret < 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 393) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 394) level = ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 395)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 396) /* Find optional mode */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 397) token = strsep(&spec, ",");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 398) if (token) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 399) ret = match_string(vmpressure_str_modes, VMPRESSURE_NUM_MODES, token);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 400) if (ret < 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 401) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 402) mode = ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 403) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 404)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 405) ev = kzalloc(sizeof(*ev), GFP_KERNEL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 406) if (!ev) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 407) ret = -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 408) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 409) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 410)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 411) ev->efd = eventfd;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 412) ev->level = level;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 413) ev->mode = mode;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 414)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 415) mutex_lock(&vmpr->events_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 416) list_add(&ev->node, &vmpr->events);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 417) mutex_unlock(&vmpr->events_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 418) ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 419) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 420) kfree(spec_orig);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 421) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 422) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 423)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 424) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 425) * vmpressure_unregister_event() - Unbind eventfd from vmpressure
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 426) * @memcg: memcg handle
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 427) * @eventfd: eventfd context that was used to link vmpressure with the @cg
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 428) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 429) * This function does internal manipulations to detach the @eventfd from
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 430) * the vmpressure notifications, and then frees internal resources
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 431) * associated with the @eventfd (but the @eventfd itself is not freed).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 432) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 433) * To be used as memcg event method.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 434) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 435) void vmpressure_unregister_event(struct mem_cgroup *memcg,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 436) struct eventfd_ctx *eventfd)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 437) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 438) struct vmpressure *vmpr = memcg_to_vmpressure(memcg);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 439) struct vmpressure_event *ev;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 440)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 441) mutex_lock(&vmpr->events_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 442) list_for_each_entry(ev, &vmpr->events, node) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 443) if (ev->efd != eventfd)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 444) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 445) list_del(&ev->node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 446) kfree(ev);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 447) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 448) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 449) mutex_unlock(&vmpr->events_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 450) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 451)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 452) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 453) * vmpressure_init() - Initialize vmpressure control structure
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 454) * @vmpr: Structure to be initialized
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 455) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 456) * This function should be called on every allocated vmpressure structure
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 457) * before any usage.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 458) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 459) void vmpressure_init(struct vmpressure *vmpr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 460) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 461) spin_lock_init(&vmpr->sr_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 462) mutex_init(&vmpr->events_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 463) INIT_LIST_HEAD(&vmpr->events);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 464) INIT_WORK(&vmpr->work, vmpressure_work_fn);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 465) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 466)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 467) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 468) * vmpressure_cleanup() - shuts down vmpressure control structure
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 469) * @vmpr: Structure to be cleaned up
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 470) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 471) * This function should be called before the structure in which it is
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 472) * embedded is cleaned up.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 473) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 474) void vmpressure_cleanup(struct vmpressure *vmpr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 475) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 476) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 477) * Make sure there is no pending work before eventfd infrastructure
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 478) * goes away.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 479) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 480) flush_work(&vmpr->work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 481) }
Orange Pi5 kernel
Deprecated Linux kernel 5.10.110 for OrangePi 5/5B/5+ boards
3 Commits
0 Branches
0 Tags