summaryrefslogtreecommitdiff
path: root/src
diff options
context:
space:
mode:
Diffstat (limited to 'src')
-rw-r--r--src/basic/siphash24.c2
-rw-r--r--src/core/cgroup.c185
-rw-r--r--src/core/cgroup.h1
-rw-r--r--src/core/load-fragment.c99
-rw-r--r--src/resolve/resolved-resolv-conf.c49
-rw-r--r--src/shared/bus-unit-util.c4
-rw-r--r--src/shared/seccomp-util.c216
-rw-r--r--src/shared/seccomp-util.h7
8 files changed, 428 insertions, 135 deletions
diff --git a/src/basic/siphash24.c b/src/basic/siphash24.c
index 060e8ba387..8c1cdc3db6 100644
--- a/src/basic/siphash24.c
+++ b/src/basic/siphash24.c
@@ -17,6 +17,8 @@
coding style)
*/
+#include <stdio.h>
+
#include "macro.h"
#include "siphash24.h"
#include "unaligned.h"
diff --git a/src/core/cgroup.c b/src/core/cgroup.c
index fbe69df4e9..f3e0c54b76 100644
--- a/src/core/cgroup.c
+++ b/src/core/cgroup.c
@@ -36,6 +36,22 @@
#define CGROUP_CPU_QUOTA_PERIOD_USEC ((usec_t) 100 * USEC_PER_MSEC)
+static void cgroup_compat_warn(void)
+{
+ static bool cgroup_compat_warned = false;
+
+ if (cgroup_compat_warned)
+ return;
+
+ log_warning("cgroup compatibility translation between legacy and unified hierarchy settings activated. See cgroup-compat debug messages for details.");
+ cgroup_compat_warned = true;
+}
+
+#define log_cgroup_compat(unit, fmt, ...) do { \
+ cgroup_compat_warn(); \
+ log_unit_debug(unit, "cgroup-compat: " fmt, ##__VA_ARGS__); \
+ } while (0)
+
void cgroup_context_init(CGroupContext *c) {
assert(c);
@@ -413,7 +429,7 @@ static uint64_t cgroup_weight_io_to_blkio(uint64_t io_weight) {
CGROUP_BLKIO_WEIGHT_MIN, CGROUP_BLKIO_WEIGHT_MAX);
}
-static void cgroup_apply_io_device_weight(const char *path, const char *dev_path, uint64_t io_weight) {
+static void cgroup_apply_io_device_weight(Unit *u, const char *dev_path, uint64_t io_weight) {
char buf[DECIMAL_STR_MAX(dev_t)*2+2+DECIMAL_STR_MAX(uint64_t)+1];
dev_t dev;
int r;
@@ -423,13 +439,13 @@ static void cgroup_apply_io_device_weight(const char *path, const char *dev_path
return;
xsprintf(buf, "%u:%u %" PRIu64 "\n", major(dev), minor(dev), io_weight);
- r = cg_set_attribute("io", path, "io.weight", buf);
+ r = cg_set_attribute("io", u->cgroup_path, "io.weight", buf);
if (r < 0)
- log_full_errno(IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
- "Failed to set io.weight on %s: %m", path);
+ log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
+ "Failed to set io.weight: %m");
}
-static void cgroup_apply_blkio_device_weight(const char *path, const char *dev_path, uint64_t blkio_weight) {
+static void cgroup_apply_blkio_device_weight(Unit *u, const char *dev_path, uint64_t blkio_weight) {
char buf[DECIMAL_STR_MAX(dev_t)*2+2+DECIMAL_STR_MAX(uint64_t)+1];
dev_t dev;
int r;
@@ -439,13 +455,13 @@ static void cgroup_apply_blkio_device_weight(const char *path, const char *dev_p
return;
xsprintf(buf, "%u:%u %" PRIu64 "\n", major(dev), minor(dev), blkio_weight);
- r = cg_set_attribute("blkio", path, "blkio.weight_device", buf);
+ r = cg_set_attribute("blkio", u->cgroup_path, "blkio.weight_device", buf);
if (r < 0)
- log_full_errno(IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
- "Failed to set blkio.weight_device on %s: %m", path);
+ log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
+ "Failed to set blkio.weight_device: %m");
}
-static unsigned cgroup_apply_io_device_limit(const char *path, const char *dev_path, uint64_t *limits) {
+static unsigned cgroup_apply_io_device_limit(Unit *u, const char *dev_path, uint64_t *limits) {
char limit_bufs[_CGROUP_IO_LIMIT_TYPE_MAX][DECIMAL_STR_MAX(uint64_t)];
char buf[DECIMAL_STR_MAX(dev_t)*2+2+(6+DECIMAL_STR_MAX(uint64_t)+1)*4];
CGroupIOLimitType type;
@@ -469,14 +485,14 @@ static unsigned cgroup_apply_io_device_limit(const char *path, const char *dev_p
xsprintf(buf, "%u:%u rbps=%s wbps=%s riops=%s wiops=%s\n", major(dev), minor(dev),
limit_bufs[CGROUP_IO_RBPS_MAX], limit_bufs[CGROUP_IO_WBPS_MAX],
limit_bufs[CGROUP_IO_RIOPS_MAX], limit_bufs[CGROUP_IO_WIOPS_MAX]);
- r = cg_set_attribute("io", path, "io.max", buf);
+ r = cg_set_attribute("io", u->cgroup_path, "io.max", buf);
if (r < 0)
- log_full_errno(IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
- "Failed to set io.max on %s: %m", path);
+ log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
+ "Failed to set io.max: %m");
return n;
}
-static unsigned cgroup_apply_blkio_device_limit(const char *path, const char *dev_path, uint64_t rbps, uint64_t wbps) {
+static unsigned cgroup_apply_blkio_device_limit(Unit *u, const char *dev_path, uint64_t rbps, uint64_t wbps) {
char buf[DECIMAL_STR_MAX(dev_t)*2+2+DECIMAL_STR_MAX(uint64_t)+1];
dev_t dev;
unsigned n = 0;
@@ -489,18 +505,18 @@ static unsigned cgroup_apply_blkio_device_limit(const char *path, const char *de
if (rbps != CGROUP_LIMIT_MAX)
n++;
sprintf(buf, "%u:%u %" PRIu64 "\n", major(dev), minor(dev), rbps);
- r = cg_set_attribute("blkio", path, "blkio.throttle.read_bps_device", buf);
+ r = cg_set_attribute("blkio", u->cgroup_path, "blkio.throttle.read_bps_device", buf);
if (r < 0)
- log_full_errno(IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
- "Failed to set blkio.throttle.read_bps_device on %s: %m", path);
+ log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
+ "Failed to set blkio.throttle.read_bps_device: %m");
if (wbps != CGROUP_LIMIT_MAX)
n++;
sprintf(buf, "%u:%u %" PRIu64 "\n", major(dev), minor(dev), wbps);
- r = cg_set_attribute("blkio", path, "blkio.throttle.write_bps_device", buf);
+ r = cg_set_attribute("blkio", u->cgroup_path, "blkio.throttle.write_bps_device", buf);
if (r < 0)
- log_full_errno(IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
- "Failed to set blkio.throttle.write_bps_device on %s: %m", path);
+ log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
+ "Failed to set blkio.throttle.write_bps_device: %m");
return n;
}
@@ -509,23 +525,30 @@ 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;
}
-static void cgroup_apply_unified_memory_limit(const char *path, const char *file, uint64_t v) {
+static void cgroup_apply_unified_memory_limit(Unit *u, const char *file, uint64_t v) {
char buf[DECIMAL_STR_MAX(uint64_t) + 1] = "max";
int r;
if (v != CGROUP_LIMIT_MAX)
xsprintf(buf, "%" PRIu64 "\n", v);
- r = cg_set_attribute("memory", path, file, buf);
+ r = cg_set_attribute("memory", u->cgroup_path, file, buf);
if (r < 0)
- log_full_errno(IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
- "Failed to set %s on %s: %m", file, path);
+ log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
+ "Failed to set %s: %m", file);
}
-void cgroup_context_apply(CGroupContext *c, CGroupMask mask, const char *path, ManagerState state) {
+static void cgroup_context_apply(Unit *u, CGroupMask mask, ManagerState state) {
+ const char *path;
+ CGroupContext *c;
bool is_root;
int r;
+ assert(u);
+
+ c = unit_get_cgroup_context(u);
+ path = u->cgroup_path;
+
assert(c);
assert(path);
@@ -551,14 +574,14 @@ void cgroup_context_apply(CGroupContext *c, CGroupMask mask, const char *path, M
c->cpu_shares != CGROUP_CPU_SHARES_INVALID ? c->cpu_shares : CGROUP_CPU_SHARES_DEFAULT);
r = cg_set_attribute("cpu", path, "cpu.shares", buf);
if (r < 0)
- log_full_errno(IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
- "Failed to set cpu.shares on %s: %m", path);
+ log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
+ "Failed to set cpu.shares: %m");
sprintf(buf, USEC_FMT "\n", CGROUP_CPU_QUOTA_PERIOD_USEC);
r = cg_set_attribute("cpu", path, "cpu.cfs_period_us", buf);
if (r < 0)
- log_full_errno(IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
- "Failed to set cpu.cfs_period_us on %s: %m", path);
+ log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
+ "Failed to set cpu.cfs_period_us: %m");
if (c->cpu_quota_per_sec_usec != USEC_INFINITY) {
sprintf(buf, USEC_FMT "\n", c->cpu_quota_per_sec_usec * CGROUP_CPU_QUOTA_PERIOD_USEC / USEC_PER_SEC);
@@ -566,8 +589,8 @@ void cgroup_context_apply(CGroupContext *c, CGroupMask mask, const char *path, M
} else
r = cg_set_attribute("cpu", path, "cpu.cfs_quota_us", "-1");
if (r < 0)
- log_full_errno(IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
- "Failed to set cpu.cfs_quota_us on %s: %m", path);
+ log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
+ "Failed to set cpu.cfs_quota_us: %m");
}
if (mask & CGROUP_MASK_IO) {
@@ -580,29 +603,40 @@ void cgroup_context_apply(CGroupContext *c, CGroupMask mask, const char *path, M
if (has_io)
weight = cgroup_context_io_weight(c, state);
- else if (has_blockio)
- weight = cgroup_weight_blkio_to_io(cgroup_context_blkio_weight(c, state));
- else
+ else if (has_blockio) {
+ uint64_t blkio_weight = cgroup_context_blkio_weight(c, state);
+
+ weight = cgroup_weight_blkio_to_io(blkio_weight);
+
+ log_cgroup_compat(u, "Applying [Startup]BlockIOWeight %" PRIu64 " as [Startup]IOWeight %" PRIu64,
+ blkio_weight, weight);
+ } else
weight = CGROUP_WEIGHT_DEFAULT;
xsprintf(buf, "default %" PRIu64 "\n", weight);
r = cg_set_attribute("io", path, "io.weight", buf);
if (r < 0)
- log_full_errno(IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
- "Failed to set io.weight on %s: %m", path);
+ log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
+ "Failed to set io.weight: %m");
if (has_io) {
CGroupIODeviceWeight *w;
/* FIXME: no way to reset this list */
LIST_FOREACH(device_weights, w, c->io_device_weights)
- cgroup_apply_io_device_weight(path, w->path, w->weight);
+ cgroup_apply_io_device_weight(u, w->path, w->weight);
} else if (has_blockio) {
CGroupBlockIODeviceWeight *w;
/* FIXME: no way to reset this list */
- LIST_FOREACH(device_weights, w, c->blockio_device_weights)
- cgroup_apply_io_device_weight(path, w->path, cgroup_weight_blkio_to_io(w->weight));
+ LIST_FOREACH(device_weights, w, c->blockio_device_weights) {
+ weight = cgroup_weight_blkio_to_io(w->weight);
+
+ log_cgroup_compat(u, "Applying BlockIODeviceWeight %" PRIu64 " as IODeviceWeight %" PRIu64 " for %s",
+ w->weight, weight, w->path);
+
+ cgroup_apply_io_device_weight(u, w->path, weight);
+ }
}
}
@@ -611,7 +645,7 @@ void cgroup_context_apply(CGroupContext *c, CGroupMask mask, const char *path, M
CGroupIODeviceLimit *l, *next;
LIST_FOREACH_SAFE(device_limits, l, next, c->io_device_limits) {
- if (!cgroup_apply_io_device_limit(path, l->path, l->limits))
+ if (!cgroup_apply_io_device_limit(u, l->path, l->limits))
cgroup_context_free_io_device_limit(c, l);
}
} else if (has_blockio) {
@@ -627,7 +661,10 @@ void cgroup_context_apply(CGroupContext *c, CGroupMask mask, const char *path, M
limits[CGROUP_IO_RBPS_MAX] = b->rbps;
limits[CGROUP_IO_WBPS_MAX] = b->wbps;
- if (!cgroup_apply_io_device_limit(path, b->path, limits))
+ log_cgroup_compat(u, "Applying BlockIO{Read|Write}Bandwidth %" PRIu64 " %" PRIu64 " as IO{Read|Write}BandwidthMax for %s",
+ b->rbps, b->wbps, b->path);
+
+ if (!cgroup_apply_io_device_limit(u, b->path, limits))
cgroup_context_free_blockio_device_bandwidth(c, b);
}
}
@@ -643,29 +680,40 @@ void cgroup_context_apply(CGroupContext *c, CGroupMask mask, const char *path, M
if (has_blockio)
weight = cgroup_context_blkio_weight(c, state);
- else if (has_io)
+ else if (has_io) {
+ uint64_t io_weight = cgroup_context_io_weight(c, state);
+
weight = cgroup_weight_io_to_blkio(cgroup_context_io_weight(c, state));
- else
+
+ log_cgroup_compat(u, "Applying [Startup]IOWeight %" PRIu64 " as [Startup]BlockIOWeight %" PRIu64,
+ io_weight, weight);
+ } else
weight = CGROUP_BLKIO_WEIGHT_DEFAULT;
xsprintf(buf, "%" PRIu64 "\n", weight);
r = cg_set_attribute("blkio", path, "blkio.weight", buf);
if (r < 0)
- log_full_errno(IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
- "Failed to set blkio.weight on %s: %m", path);
+ log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
+ "Failed to set blkio.weight: %m");
if (has_blockio) {
CGroupBlockIODeviceWeight *w;
/* FIXME: no way to reset this list */
LIST_FOREACH(device_weights, w, c->blockio_device_weights)
- cgroup_apply_blkio_device_weight(path, w->path, w->weight);
+ cgroup_apply_blkio_device_weight(u, w->path, w->weight);
} else if (has_io) {
CGroupIODeviceWeight *w;
/* FIXME: no way to reset this list */
- LIST_FOREACH(device_weights, w, c->io_device_weights)
- cgroup_apply_blkio_device_weight(path, w->path, cgroup_weight_io_to_blkio(w->weight));
+ LIST_FOREACH(device_weights, w, c->io_device_weights) {
+ weight = cgroup_weight_io_to_blkio(w->weight);
+
+ log_cgroup_compat(u, "Applying IODeviceWeight %" PRIu64 " as BlockIODeviceWeight %" PRIu64 " for %s",
+ w->weight, weight, w->path);
+
+ cgroup_apply_blkio_device_weight(u, w->path, weight);
+ }
}
}
@@ -674,14 +722,17 @@ void cgroup_context_apply(CGroupContext *c, CGroupMask mask, const char *path, M
CGroupBlockIODeviceBandwidth *b, *next;
LIST_FOREACH_SAFE(device_bandwidths, b, next, c->blockio_device_bandwidths) {
- if (!cgroup_apply_blkio_device_limit(path, b->path, b->rbps, b->wbps))
+ if (!cgroup_apply_blkio_device_limit(u, b->path, b->rbps, b->wbps))
cgroup_context_free_blockio_device_bandwidth(c, b);
}
} else if (has_io) {
CGroupIODeviceLimit *l, *next;
LIST_FOREACH_SAFE(device_limits, l, next, c->io_device_limits) {
- if (!cgroup_apply_blkio_device_limit(path, l->path, l->limits[CGROUP_IO_RBPS_MAX], l->limits[CGROUP_IO_WBPS_MAX]))
+ log_cgroup_compat(u, "Applying IO{Read|Write}Bandwidth %" PRIu64 " %" PRIu64 " as BlockIO{Read|Write}BandwidthMax for %s",
+ l->limits[CGROUP_IO_RBPS_MAX], l->limits[CGROUP_IO_WBPS_MAX], l->path);
+
+ if (!cgroup_apply_blkio_device_limit(u, l->path, l->limits[CGROUP_IO_RBPS_MAX], l->limits[CGROUP_IO_WBPS_MAX]))
cgroup_context_free_io_device_limit(c, l);
}
}
@@ -693,24 +744,32 @@ void cgroup_context_apply(CGroupContext *c, CGroupMask mask, const char *path, M
if (cgroup_context_has_unified_memory_config(c))
max = c->memory_max;
- else
+ else {
max = c->memory_limit;
- cgroup_apply_unified_memory_limit(path, "memory.low", c->memory_low);
- cgroup_apply_unified_memory_limit(path, "memory.high", c->memory_high);
- cgroup_apply_unified_memory_limit(path, "memory.max", max);
+ if (max != CGROUP_LIMIT_MAX)
+ log_cgroup_compat(u, "Applying MemoryLimit %" PRIu64 " as MemoryMax", max);
+ }
+
+ 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);
} else {
char buf[DECIMAL_STR_MAX(uint64_t) + 1];
if (c->memory_limit != CGROUP_LIMIT_MAX)
xsprintf(buf, "%" PRIu64 "\n", c->memory_limit);
- else
+ else {
xsprintf(buf, "%" PRIu64 "\n", c->memory_max);
+ if (c->memory_max != CGROUP_LIMIT_MAX)
+ log_cgroup_compat(u, "Applying MemoryMax %" PRIu64 " as MemoryLimit", c->memory_max);
+ }
+
r = cg_set_attribute("memory", path, "memory.limit_in_bytes", buf);
if (r < 0)
- log_full_errno(IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
- "Failed to set memory.limit_in_bytes on %s: %m", path);
+ log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
+ "Failed to set memory.limit_in_bytes: %m");
}
}
@@ -726,8 +785,8 @@ void cgroup_context_apply(CGroupContext *c, CGroupMask mask, const char *path, M
else
r = cg_set_attribute("devices", path, "devices.allow", "a");
if (r < 0)
- log_full_errno(IN_SET(r, -ENOENT, -EROFS, -EINVAL, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
- "Failed to reset devices.list on %s: %m", path);
+ log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EINVAL, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
+ "Failed to reset devices.list: %m");
if (c->device_policy == CGROUP_CLOSED ||
(c->device_policy == CGROUP_AUTO && c->device_allow)) {
@@ -773,7 +832,7 @@ void cgroup_context_apply(CGroupContext *c, CGroupMask mask, const char *path, M
else if (startswith(a->path, "char-"))
whitelist_major(path, a->path + 5, 'c', acc);
else
- log_debug("Ignoring device %s while writing cgroup attribute.", a->path);
+ log_unit_debug(u, "Ignoring device %s while writing cgroup attribute.", a->path);
}
}
@@ -788,8 +847,8 @@ void cgroup_context_apply(CGroupContext *c, CGroupMask mask, const char *path, M
r = cg_set_attribute("pids", path, "pids.max", "max");
if (r < 0)
- log_full_errno(IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
- "Failed to set pids.max on %s: %m", path);
+ log_unit_full(u, IN_SET(r, -ENOENT, -EROFS, -EACCES) ? LOG_DEBUG : LOG_WARNING, r,
+ "Failed to set pids.max: %m");
}
}
@@ -1224,7 +1283,7 @@ static int unit_realize_cgroup_now(Unit *u, ManagerState state) {
return r;
/* Finally, apply the necessary attributes. */
- cgroup_context_apply(unit_get_cgroup_context(u), target_mask, u->cgroup_path, state);
+ cgroup_context_apply(u, target_mask, state);
return 0;
}
@@ -1355,7 +1414,7 @@ void unit_prune_cgroup(Unit *u) {
r = cg_trim_everywhere(u->manager->cgroup_supported, u->cgroup_path, !is_root_slice);
if (r < 0) {
- log_debug_errno(r, "Failed to destroy cgroup %s, ignoring: %m", u->cgroup_path);
+ log_unit_debug_errno(u, r, "Failed to destroy cgroup %s, ignoring: %m", u->cgroup_path);
return;
}
diff --git a/src/core/cgroup.h b/src/core/cgroup.h
index ff87adfba1..f21409bd5d 100644
--- a/src/core/cgroup.h
+++ b/src/core/cgroup.h
@@ -125,7 +125,6 @@ struct CGroupContext {
void cgroup_context_init(CGroupContext *c);
void cgroup_context_done(CGroupContext *c);
void cgroup_context_dump(CGroupContext *c, FILE* f, const char *prefix);
-void cgroup_context_apply(CGroupContext *c, CGroupMask mask, const char *path, ManagerState state);
CGroupMask cgroup_context_get_mask(CGroupContext *c);
diff --git a/src/core/load-fragment.c b/src/core/load-fragment.c
index 09d3f65c77..b53301a147 100644
--- a/src/core/load-fragment.c
+++ b/src/core/load-fragment.c
@@ -2396,6 +2396,55 @@ int config_parse_documentation(const char *unit,
}
#ifdef HAVE_SECCOMP
+static int syscall_filter_parse_one(
+ const char *unit,
+ const char *filename,
+ unsigned line,
+ ExecContext *c,
+ bool invert,
+ const char *t,
+ bool warn) {
+ int r;
+
+ if (*t == '@') {
+ const SystemCallFilterSet *set;
+
+ for (set = syscall_filter_sets; set->set_name; set++)
+ if (streq(set->set_name, t)) {
+ const char *sys;
+
+ NULSTR_FOREACH(sys, set->value) {
+ r = syscall_filter_parse_one(unit, filename, line, c, invert, sys, false);
+ if (r < 0)
+ return r;
+ }
+ break;
+ }
+ } else {
+ int id;
+
+ id = seccomp_syscall_resolve_name(t);
+ if (id < 0) {
+ if (warn)
+ log_syntax(unit, LOG_ERR, filename, line, 0, "Failed to parse system call, ignoring: %s", t);
+ return 0;
+ }
+
+ /* If we previously wanted to forbid a syscall and now
+ * we want to allow it, then remove it from the list
+ */
+ if (!invert == c->syscall_whitelist) {
+ r = set_put(c->syscall_filter, INT_TO_PTR(id + 1));
+ if (r == 0)
+ return 0;
+ if (r < 0)
+ return log_oom();
+ } else
+ set_remove(c->syscall_filter, INT_TO_PTR(id + 1));
+ }
+ return 0;
+}
+
int config_parse_syscall_filter(
const char *unit,
const char *filename,
@@ -2408,13 +2457,6 @@ int config_parse_syscall_filter(
void *data,
void *userdata) {
- static const char default_syscalls[] =
- "execve\0"
- "exit\0"
- "exit_group\0"
- "rt_sigreturn\0"
- "sigreturn\0";
-
ExecContext *c = data;
Unit *u = userdata;
bool invert = false;
@@ -2448,53 +2490,26 @@ int config_parse_syscall_filter(
/* Allow everything but the ones listed */
c->syscall_whitelist = false;
else {
- const char *i;
-
/* Allow nothing but the ones listed */
c->syscall_whitelist = true;
/* Accept default syscalls if we are on a whitelist */
- NULSTR_FOREACH(i, default_syscalls) {
- int id;
-
- id = seccomp_syscall_resolve_name(i);
- if (id < 0)
- continue;
-
- r = set_put(c->syscall_filter, INT_TO_PTR(id + 1));
- if (r == 0)
- continue;
- if (r < 0)
- return log_oom();
- }
+ r = syscall_filter_parse_one(unit, filename, line, c, false, "@default", false);
+ if (r < 0)
+ return r;
}
}
FOREACH_WORD_QUOTED(word, l, rvalue, state) {
_cleanup_free_ char *t = NULL;
- int id;
t = strndup(word, l);
if (!t)
return log_oom();
- id = seccomp_syscall_resolve_name(t);
- if (id < 0) {
- log_syntax(unit, LOG_ERR, filename, line, 0, "Failed to parse system call, ignoring: %s", t);
- continue;
- }
-
- /* If we previously wanted to forbid a syscall and now
- * we want to allow it, then remove it from the list
- */
- if (!invert == c->syscall_whitelist) {
- r = set_put(c->syscall_filter, INT_TO_PTR(id + 1));
- if (r == 0)
- continue;
- if (r < 0)
- return log_oom();
- } else
- set_remove(c->syscall_filter, INT_TO_PTR(id + 1));
+ r = syscall_filter_parse_one(unit, filename, line, c, invert, t, true);
+ if (r < 0)
+ return r;
}
if (!isempty(state))
log_syntax(unit, LOG_ERR, filename, line, 0, "Trailing garbage, ignoring.");
@@ -2796,7 +2811,7 @@ int config_parse_memory_limit(
uint64_t bytes = CGROUP_LIMIT_MAX;
int r;
- if (!isempty(rvalue) && !streq(rvalue, "infinity") && !streq(rvalue, "max")) {
+ if (!isempty(rvalue) && !streq(rvalue, "infinity")) {
r = parse_size(rvalue, 1024, &bytes);
if (r < 0 || bytes < 1) {
log_syntax(unit, LOG_ERR, filename, line, r, "Memory limit '%s' invalid. Ignoring.", rvalue);
@@ -3065,7 +3080,7 @@ int config_parse_io_limit(
return 0;
}
- if (streq("max", limit)) {
+ if (streq("infinity", limit)) {
num = CGROUP_LIMIT_MAX;
} else {
r = parse_size(limit, 1000, &num);
diff --git a/src/resolve/resolved-resolv-conf.c b/src/resolve/resolved-resolv-conf.c
index ff03acc772..fa89de4c21 100644
--- a/src/resolve/resolved-resolv-conf.c
+++ b/src/resolve/resolved-resolv-conf.c
@@ -164,30 +164,32 @@ static void write_resolv_conf_server(DnsServer *s, FILE *f, unsigned *count) {
}
static void write_resolv_conf_search(
- const char *domain,
- FILE *f,
- unsigned *count,
- unsigned *length) {
+ OrderedSet *domains,
+ FILE *f) {
+ unsigned length = 0, count = 0;
+ Iterator i;
+ char *domain;
- assert(domain);
+ assert(domains);
assert(f);
- assert(length);
- if (*count >= MAXDNSRCH ||
- *length + strlen(domain) > 256) {
- if (*count == MAXDNSRCH)
- fputs(" # Too many search domains configured, remaining ones ignored.", f);
- if (*length <= 256)
- fputs(" # Total length of all search domains is too long, remaining ones ignored.", f);
+ fputs("search", f);
- return;
+ ORDERED_SET_FOREACH(domain, domains, i) {
+ if (++count > MAXDNSRCH) {
+ fputs("\n# Too many search domains configured, remaining ones ignored.", f);
+ break;
+ }
+ length += strlen(domain) + 1;
+ if (length > 256) {
+ fputs("\n# Total length of all search domains is too long, remaining ones ignored.", f);
+ break;
+ }
+ fputc(' ', f);
+ fputs(domain, f);
}
- (*length) += strlen(domain);
- (*count)++;
-
- fputc(' ', f);
- fputs(domain, f);
+ fputs("\n", f);
}
static int write_resolv_conf_contents(FILE *f, OrderedSet *dns, OrderedSet *domains) {
@@ -209,15 +211,8 @@ static int write_resolv_conf_contents(FILE *f, OrderedSet *dns, OrderedSet *doma
write_resolv_conf_server(s, f, &count);
}
- if (!ordered_set_isempty(domains)) {
- unsigned length = 0, count = 0;
- char *domain;
-
- fputs("search", f);
- ORDERED_SET_FOREACH(domain, domains, i)
- write_resolv_conf_search(domain, f, &count, &length);
- fputs("\n", f);
- }
+ if (!ordered_set_isempty(domains))
+ write_resolv_conf_search(domains, f);
return fflush_and_check(f);
}
diff --git a/src/shared/bus-unit-util.c b/src/shared/bus-unit-util.c
index 502e98d9dc..bf0b2e89e3 100644
--- a/src/shared/bus-unit-util.c
+++ b/src/shared/bus-unit-util.c
@@ -169,7 +169,7 @@ int bus_append_unit_property_assignment(sd_bus_message *m, const char *assignmen
} else if (STR_IN_SET(field, "MemoryLow", "MemoryHigh", "MemoryMax", "MemoryLimit")) {
uint64_t bytes;
- if (isempty(eq) || streq(eq, "max") || streq(eq, "infinity"))
+ if (isempty(eq) || streq(eq, "infinity"))
bytes = CGROUP_LIMIT_MAX;
else {
r = parse_size(eq, 1024, &bytes);
@@ -306,7 +306,7 @@ int bus_append_unit_property_assignment(sd_bus_message *m, const char *assignmen
return -EINVAL;
}
- if (streq(bandwidth, "max")) {
+ if (streq(bandwidth, "infinity")) {
bytes = CGROUP_LIMIT_MAX;
} else {
r = parse_size(bandwidth, 1000, &bytes);
diff --git a/src/shared/seccomp-util.c b/src/shared/seccomp-util.c
index cebe0fce2a..30d22d2242 100644
--- a/src/shared/seccomp-util.c
+++ b/src/shared/seccomp-util.c
@@ -88,3 +88,219 @@ int seccomp_add_secondary_archs(scmp_filter_ctx *c) {
return 0;
}
+
+const SystemCallFilterSet syscall_filter_sets[] = {
+ {
+ /* Clock */
+ .set_name = "@clock",
+ .value =
+ "adjtimex\0"
+ "settimeofday\0"
+ }, {
+ /* Default list */
+ .set_name = "@default",
+ .value =
+ "execve\0"
+ "exit\0"
+ "exit_group\0"
+ "rt_sigreturn\0"
+ "sigreturn\0"
+ }, {
+ /* Event loop use */
+ .set_name = "@io-event",
+ .value =
+ "_newselect\0"
+ "epoll_create1\0"
+ "epoll_create\0"
+ "epoll_ctl\0"
+ "epoll_ctl_old\0"
+ "epoll_pwait\0"
+ "epoll_wait\0"
+ "epoll_wait_old\0"
+ "eventfd2\0"
+ "eventfd\0"
+ "poll\0"
+ "ppoll\0"
+ "pselect6\0"
+ "select\0"
+ }, {
+ /* Message queues, SYSV IPC or other IPC: unusual */
+ .set_name = "@ipc",
+ .value = "ipc\0"
+ "mq_getsetattr\0"
+ "mq_notify\0"
+ "mq_open\0"
+ "mq_timedreceive\0"
+ "mq_timedsend\0"
+ "mq_unlink\0"
+ "msgctl\0"
+ "msgget\0"
+ "msgrcv\0"
+ "msgsnd\0"
+ "process_vm_readv\0"
+ "process_vm_writev\0"
+ "semctl\0"
+ "semget\0"
+ "semop\0"
+ "semtimedop\0"
+ "shmat\0"
+ "shmctl\0"
+ "shmdt\0"
+ "shmget\0"
+ }, {
+ /* Kernel module control */
+ .set_name = "@module",
+ .value =
+ "create_module\0"
+ "delete_module\0"
+ "finit_module\0"
+ "init_module\0"
+ }, {
+ /* Mounting */
+ .set_name = "@mount",
+ .value =
+ "chroot\0"
+ "mount\0"
+ "oldumount\0"
+ "pivot_root\0"
+ "umount2\0"
+ "umount\0"
+ }, {
+ /* Network or Unix socket IO, should not be needed if not network facing */
+ .set_name = "@network-io",
+ .value =
+ "accept4\0"
+ "accept\0"
+ "bind\0"
+ "connect\0"
+ "getpeername\0"
+ "getsockname\0"
+ "getsockopt\0"
+ "listen\0"
+ "recv\0"
+ "recvfrom\0"
+ "recvmmsg\0"
+ "recvmsg\0"
+ "send\0"
+ "sendmmsg\0"
+ "sendmsg\0"
+ "sendto\0"
+ "setsockopt\0"
+ "shutdown\0"
+ "socket\0"
+ "socketcall\0"
+ "socketpair\0"
+ }, {
+ /* Unusual, obsolete or unimplemented, some unknown even to libseccomp */
+ .set_name = "@obsolete",
+ .value =
+ "_sysctl\0"
+ "afs_syscall\0"
+ "break\0"
+ "fattach\0"
+ "fdetach\0"
+ "ftime\0"
+ "get_kernel_syms\0"
+ "get_mempolicy\0"
+ "getmsg\0"
+ "getpmsg\0"
+ "gtty\0"
+ "isastream\0"
+ "lock\0"
+ "madvise1\0"
+ "modify_ldt\0"
+ "mpx\0"
+ "pciconfig_iobase\0"
+ "perf_event_open\0"
+ "prof\0"
+ "profil\0"
+ "putmsg\0"
+ "putpmsg\0"
+ "query_module\0"
+ "rtas\0"
+ "s390_runtime_instr\0"
+ "security\0"
+ "sgetmask\0"
+ "ssetmask\0"
+ "stty\0"
+ "subpage_prot\0"
+ "switch_endian\0"
+ "sys_debug_setcontext\0"
+ "tuxcall\0"
+ "ulimit\0"
+ "uselib\0"
+ "vm86\0"
+ "vm86old\0"
+ "vserver\0"
+ }, {
+ /* Nice grab-bag of all system calls which need superuser capabilities */
+ .set_name = "@privileged",
+ .value =
+ "@clock\0"
+ "@module\0"
+ "@raw-io\0"
+ "acct\0"
+ "bdflush\0"
+ "bpf\0"
+ "chown32\0"
+ "chown\0"
+ "chroot\0"
+ "fchown32\0"
+ "fchown\0"
+ "fchownat\0"
+ "kexec_file_load\0"
+ "kexec_load\0"
+ "lchown32\0"
+ "lchown\0"
+ "nfsservctl\0"
+ "pivot_root\0"
+ "quotactl\0"
+ "reboot\0"
+ "setdomainname\0"
+ "setfsuid32\0"
+ "setfsuid\0"
+ "setgroups32\0"
+ "setgroups\0"
+ "sethostname\0"
+ "setresuid32\0"
+ "setresuid\0"
+ "setreuid32\0"
+ "setreuid\0"
+ "setuid32\0"
+ "setuid\0"
+ "stime\0"
+ "swapoff\0"
+ "swapon\0"
+ "sysctl\0"
+ "vhangup\0"
+ }, {
+ /* Process control, execution, namespaces */
+ .set_name = "@process",
+ .value =
+ "arch_prctl\0"
+ "clone\0"
+ "execve\0"
+ "execveat\0"
+ "fork\0"
+ "kill\0"
+ "prctl\0"
+ "setns\0"
+ "tgkill\0"
+ "tkill\0"
+ "unshare\0"
+ "vfork\0"
+ }, {
+ /* Raw I/O ports */
+ .set_name = "@raw-io",
+ .value =
+ "ioperm\0"
+ "iopl\0"
+ "pciconfig_read\0"
+ "pciconfig_write\0"
+ "s390_pci_mmio_read\0"
+ "s390_pci_mmio_write\0"
+ }, {
+ .set_name = NULL,
+ .value = NULL
+ }
+};
diff --git a/src/shared/seccomp-util.h b/src/shared/seccomp-util.h
index 4ed2afc1b2..be33eecb85 100644
--- a/src/shared/seccomp-util.h
+++ b/src/shared/seccomp-util.h
@@ -26,3 +26,10 @@ const char* seccomp_arch_to_string(uint32_t c);
int seccomp_arch_from_string(const char *n, uint32_t *ret);
int seccomp_add_secondary_archs(scmp_filter_ctx *c);
+
+typedef struct SystemCallFilterSet {
+ const char *set_name;
+ const char *value;
+} SystemCallFilterSet;
+
+extern const SystemCallFilterSet syscall_filter_sets[];
generated by cgit v1.2.3-54-g00ecf (git 2.45.2) at 2024-12-26 14:57:34 +0000