core: introduce MemorySwapMax= (#3659)

Similar to MemoryMax=, MemorySwapMax= limits swap usage. This controls
controls "memory.swap.max" attribute in unified cgroup.

7 files changed, 48 insertions, 6 deletions

diff --git a/man/systemd.resource-control.xml b/man/systemd.resource-control.xml
index 84dbfa2ff3..c11f420fe5 100644
--- a/man/systemd.resource-control.xml
+++ b/man/systemd.resource-control.xml
@@ -326,6 +326,24 @@
       </varlistentry>
 
       <varlistentry>
+        <term><varname>MemorySwapMax=<replaceable>bytes</replaceable></varname></term>
+
+        <listitem>
+          <para>Specify the absolute limit on swap usage of the executed processes in this unit.</para>
+
+          <para>Takes a swap size in bytes. If the value is suffixed with K, M, G or T, the specified swap size is
+          parsed as Kilobytes, Megabytes, Gigabytes, or Terabytes (with the base 1024), respectively. If assigned the
+          special value <literal>infinity</literal>, no swap limit is applied. This controls the
+          <literal>memory.swap.max</literal> control group attribute. For details about this control group attribute,
+          see <ulink url="https://www.kernel.org/doc/Documentation/cgroup-v2.txt">cgroup-v2.txt</ulink>.</para>
+
+          <para>Implies <literal>MemoryAccounting=true</literal>.</para>
+
+          <para>This setting is supported only if the unified control group hierarchy is used.</para>
+        </listitem>
+      </varlistentry>
+
+      <varlistentry>
         <term><varname>MemoryLimit=<replaceable>bytes</replaceable></varname></term>
 
         <listitem>
diff --git a/src/core/cgroup.c b/src/core/cgroup.c
index 20f9f9fc7e..7873f88785 100644
--- a/src/core/cgroup.c
+++ b/src/core/cgroup.c
@@ -66,6 +66,7 @@ void cgroup_context_init(CGroupContext *c) {
 
         c->memory_high = CGROUP_LIMIT_MAX;
         c->memory_max = CGROUP_LIMIT_MAX;
+        c->memory_swap_max = CGROUP_LIMIT_MAX;
 
         c->memory_limit = CGROUP_LIMIT_MAX;
 
@@ -173,6 +174,7 @@ void cgroup_context_dump(CGroupContext *c, FILE* f, const char *prefix) {
                 "%sMemoryLow=%" PRIu64 "\n"
                 "%sMemoryHigh=%" PRIu64 "\n"
                 "%sMemoryMax=%" PRIu64 "\n"
+                "%sMemorySwapMax=%" PRIu64 "\n"
                 "%sMemoryLimit=%" PRIu64 "\n"
                 "%sTasksMax=%" PRIu64 "\n"
                 "%sDevicePolicy=%s\n"
@@ -194,6 +196,7 @@ void cgroup_context_dump(CGroupContext *c, FILE* f, const char *prefix) {
                 prefix, c->memory_low,
                 prefix, c->memory_high,
                 prefix, c->memory_max,
+                prefix, c->memory_swap_max,
                 prefix, c->memory_limit,
                 prefix, c->tasks_max,
                 prefix, cgroup_device_policy_to_string(c->device_policy),
@@ -617,7 +620,7 @@ static unsigned cgroup_apply_blkio_device_limit(Unit *u, const char *dev_path, u
 }
 
 static bool cgroup_context_has_unified_memory_config(CGroupContext *c) {
-        return c->memory_low > 0 || c->memory_high != CGROUP_LIMIT_MAX || c->memory_max != CGROUP_LIMIT_MAX;
+        return c->memory_low > 0 || c->memory_high != CGROUP_LIMIT_MAX || c->memory_max != CGROUP_LIMIT_MAX || c->memory_swap_max != CGROUP_LIMIT_MAX;
 }
 
 static void cgroup_apply_unified_memory_limit(Unit *u, const char *file, uint64_t v) {
@@ -848,10 +851,12 @@ static void cgroup_context_apply(Unit *u, CGroupMask mask, ManagerState state) {
         if ((mask & CGROUP_MASK_MEMORY) && !is_root) {
                 if (cg_all_unified() > 0) {
                         uint64_t max = c->memory_max;
+                        uint64_t swap_max = c->memory_swap_max;
 
-                        if (cgroup_context_has_unified_memory_config(c))
+                        if (cgroup_context_has_unified_memory_config(c)) {
                                 max = c->memory_max;
-                        else {
+                                swap_max = c->memory_swap_max;
+                        } else {
                                 max = c->memory_limit;
 
                                 if (max != CGROUP_LIMIT_MAX)
@@ -861,6 +866,7 @@ static void cgroup_context_apply(Unit *u, CGroupMask mask, ManagerState state) {
                         cgroup_apply_unified_memory_limit(u, "memory.low", c->memory_low);
                         cgroup_apply_unified_memory_limit(u, "memory.high", c->memory_high);
                         cgroup_apply_unified_memory_limit(u, "memory.max", max);
+                        cgroup_apply_unified_memory_limit(u, "memory.swap.max", swap_max);
                 } else {
                         char buf[DECIMAL_STR_MAX(uint64_t) + 1];
                         uint64_t val = c->memory_limit;
diff --git a/src/core/cgroup.h b/src/core/cgroup.h
index 2fe9cc4039..4cd168f63e 100644
--- a/src/core/cgroup.h
+++ b/src/core/cgroup.h
@@ -101,6 +101,7 @@ struct CGroupContext {
         uint64_t memory_low;
         uint64_t memory_high;
         uint64_t memory_max;
+        uint64_t memory_swap_max;
 
         /* For legacy hierarchies */
         uint64_t cpu_shares;
diff --git a/src/core/dbus-cgroup.c b/src/core/dbus-cgroup.c
index 2ca80d2996..c4067a95bf 100644
--- a/src/core/dbus-cgroup.c
+++ b/src/core/dbus-cgroup.c
@@ -233,6 +233,7 @@ const sd_bus_vtable bus_cgroup_vtable[] = {
         SD_BUS_PROPERTY("MemoryLow", "t", NULL, offsetof(CGroupContext, memory_low), 0),
         SD_BUS_PROPERTY("MemoryHigh", "t", NULL, offsetof(CGroupContext, memory_high), 0),
         SD_BUS_PROPERTY("MemoryMax", "t", NULL, offsetof(CGroupContext, memory_max), 0),
+        SD_BUS_PROPERTY("MemorySwapMax", "t", NULL, offsetof(CGroupContext, memory_swap_max), 0),
         SD_BUS_PROPERTY("MemoryLimit", "t", NULL, offsetof(CGroupContext, memory_limit), 0),
         SD_BUS_PROPERTY("DevicePolicy", "s", property_get_cgroup_device_policy, offsetof(CGroupContext, device_policy), 0),
         SD_BUS_PROPERTY("DeviceAllow", "a(ss)", property_get_device_allow, 0, 0),
@@ -875,7 +876,7 @@ int bus_cgroup_set_property(
 
                 return 1;
 
-        } else if (STR_IN_SET(name, "MemoryLow", "MemoryHigh", "MemoryMax")) {
+        } else if (STR_IN_SET(name, "MemoryLow", "MemoryHigh", "MemoryMax", "MemorySwapMax")) {
                 uint64_t v;
 
                 r = sd_bus_message_read(message, "t", &v);
@@ -889,6 +890,8 @@ int bus_cgroup_set_property(
                                 c->memory_low = v;
                         else if (streq(name, "MemoryHigh"))
                                 c->memory_high = v;
+                        else if (streq(name, "MemorySwapMax"))
+                                c->memory_swap_max = v;
                         else
                                 c->memory_max = v;
 
diff --git a/src/core/load-fragment-gperf.gperf.m4 b/src/core/load-fragment-gperf.gperf.m4
index 057b92bfa3..2e6c965aec 100644
--- a/src/core/load-fragment-gperf.gperf.m4
+++ b/src/core/load-fragment-gperf.gperf.m4
@@ -131,6 +131,7 @@ $1.MemoryAccounting,             config_parse_bool,                  0,
 $1.MemoryLow,                    config_parse_memory_limit,          0,                             offsetof($1, cgroup_context)
 $1.MemoryHigh,                   config_parse_memory_limit,          0,                             offsetof($1, cgroup_context)
 $1.MemoryMax,                    config_parse_memory_limit,          0,                             offsetof($1, cgroup_context)
+$1.MemorySwapMax,                config_parse_memory_limit,          0,                             offsetof($1, cgroup_context)
 $1.MemoryLimit,                  config_parse_memory_limit,          0,                             offsetof($1, cgroup_context)
 $1.DeviceAllow,                  config_parse_device_allow,          0,                             offsetof($1, cgroup_context)
 $1.DevicePolicy,                 config_parse_device_policy,         0,                             offsetof($1, cgroup_context.device_policy)
diff --git a/src/core/load-fragment.c b/src/core/load-fragment.c
index d3fc3e36d9..8f067b5586 100644
--- a/src/core/load-fragment.c
+++ b/src/core/load-fragment.c
@@ -2990,8 +2990,12 @@ int config_parse_memory_limit(
                 c->memory_high = bytes;
         else if (streq(lvalue, "MemoryMax"))
                 c->memory_max = bytes;
-        else
+        else if (streq(lvalue, "MemorySwapMax"))
+                c->memory_swap_max = bytes;
+        else if (streq(lvalue, "MemoryLimit"))
                 c->memory_limit = bytes;
+        else
+                return -EINVAL;
 
         return 0;
 }
diff --git a/src/systemctl/systemctl.c b/src/systemctl/systemctl.c
index 4b87bdceb2..682805045d 100644
--- a/src/systemctl/systemctl.c
+++ b/src/systemctl/systemctl.c
@@ -3571,6 +3571,7 @@ typedef struct UnitStatusInfo {
         uint64_t memory_low;
         uint64_t memory_high;
         uint64_t memory_max;
+        uint64_t memory_swap_max;
         uint64_t memory_limit;
         uint64_t cpu_usage_nsec;
         uint64_t tasks_current;
@@ -3883,7 +3884,8 @@ static void print_status_info(
 
                 printf("   Memory: %s", format_bytes(buf, sizeof(buf), i->memory_current));
 
-                if (i->memory_low > 0 || i->memory_high != CGROUP_LIMIT_MAX || i->memory_max != CGROUP_LIMIT_MAX ||
+                if (i->memory_low > 0 || i->memory_high != CGROUP_LIMIT_MAX ||
+                    i->memory_max != CGROUP_LIMIT_MAX || i->memory_swap_max != CGROUP_LIMIT_MAX ||
                     i->memory_limit != CGROUP_LIMIT_MAX) {
                         const char *prefix = "";
 
@@ -3900,6 +3902,10 @@ static void print_status_info(
                                 printf("%smax: %s", prefix, format_bytes(buf, sizeof(buf), i->memory_max));
                                 prefix = " ";
                         }
+                        if (i->memory_swap_max != CGROUP_LIMIT_MAX) {
+                                printf("%sswap max: %s", prefix, format_bytes(buf, sizeof(buf), i->memory_swap_max));
+                                prefix = " ";
+                        }
                         if (i->memory_limit != CGROUP_LIMIT_MAX) {
                                 printf("%slimit: %s", prefix, format_bytes(buf, sizeof(buf), i->memory_limit));
                                 prefix = " ";
@@ -4140,6 +4146,8 @@ static int status_property(const char *name, sd_bus_message *m, UnitStatusInfo *
                         i->memory_high = u;
                 else if (streq(name, "MemoryMax"))
                         i->memory_max = u;
+                else if (streq(name, "MemorySwapMax"))
+                        i->memory_swap_max = u;
                 else if (streq(name, "MemoryLimit"))
                         i->memory_limit = u;
                 else if (streq(name, "TasksCurrent"))
@@ -4655,6 +4663,7 @@ static int show_one(
                 .memory_current = (uint64_t) -1,
                 .memory_high = CGROUP_LIMIT_MAX,
                 .memory_max = CGROUP_LIMIT_MAX,
+                .memory_swap_max = CGROUP_LIMIT_MAX,
                 .memory_limit = (uint64_t) -1,
                 .cpu_usage_nsec = (uint64_t) -1,
                 .tasks_current = (uint64_t) -1,