platform: warn about growing sysctl logging cache and clear it

When debug-logging for platform is enabled, every access to sysctl
is cached (to log the last values).

This cache can grow quite large if the system has a large number of
interfaces (e.g. docker creating veth pairs for each container).

We already used to clear the cache, when we were about to access
sysctl *and* logging was disabled in the meantime.

Now, when logging setup changes, immediately clear the cache.
Having "nm-logging.c" call into platform code is a bit of a hack
and a better design would be to have logging code emit a signal to
which platform would subscribe. But that seems to involve much
more code (especially, as no other users care about such a signal
and because nm-logging is not a GObject).

Also, log a warning when the cache grows large to inform the user
about the cache and what he can do to clear it. The extra effort to
clear the cache when changing logging setup is done so that we do
what we tell the user: changing the logging level, will clear the
cache -- right away, not some time later when the next message is
logged.

3 files changed, 44 insertions, 12 deletions

diff --git a/src/nm-logging.c b/src/nm-logging.c
index e58a5de384..6476d7664f 100644
--- a/src/nm-logging.c
+++ b/src/nm-logging.c
@@ -40,6 +40,7 @@
 #include "nm-default.h"
 #include "nm-errors.h"
 #include "NetworkManagerUtils.h"
+#include "nm-linux-platform.h"
 
 static void
 nm_log_handler (const gchar *log_domain,
@@ -169,6 +170,7 @@ nm_logging_setup (const char  *level,
 	NMLogLevel new_log_level = log_level;
 	char **tmp, **iter;
 	int i;
+	gboolean had_platform_debug;
 
 	g_return_val_if_fail (!bad_domains || !*bad_domains, FALSE);
 	g_return_val_if_fail (!error || !*error, FALSE);
@@ -276,10 +278,20 @@ nm_logging_setup (const char  *level,
 
 	g_clear_pointer (&logging_domains_to_string, g_free);
 
+	had_platform_debug = nm_logging_enabled (LOGL_DEBUG, LOGD_PLATFORM);
+
 	log_level = new_log_level;
 	for (i = 0; i < G_N_ELEMENTS (new_logging); i++)
 		logging[i] = new_logging[i];
 
+	if (   had_platform_debug
+	    && !nm_logging_enabled (LOGL_DEBUG, LOGD_PLATFORM)) {
+		/* when debug logging is enabled, platform will cache all access to
+		 * sysctl. When the user disables debug-logging, we want to clear that
+		 * cache right away. */
+		_nm_linux_platform_sysctl_clear_cache ();
+	}
+
 	if (unrecognized)
 		*bad_domains = g_string_free (unrecognized, FALSE);
 
diff --git a/src/platform/nm-linux-platform.c b/src/platform/nm-linux-platform.c
index 2daa1aec46..76d74da0bd 100644
--- a/src/platform/nm-linux-platform.c
+++ b/src/platform/nm-linux-platform.c
@@ -731,6 +731,7 @@ struct _NMLinuxPlatformPrivate {
 	GIOChannel *event_channel;
 	guint event_id;
 
+	gboolean sysctl_get_warned;
 	GHashTable *sysctl_get_prev_values;
 
 	GUdevClient *udev_client;
@@ -2546,15 +2547,32 @@ sysctl_set (NMPlatform *platform, const char *path, const char *value)
 	return (nwrote == len);
 }
 
+static GSList *sysctl_clear_cache_list;
+
+void
+_nm_linux_platform_sysctl_clear_cache (void)
+{
+	while (sysctl_clear_cache_list) {
+		NMLinuxPlatformPrivate *priv = NM_LINUX_PLATFORM_GET_PRIVATE (sysctl_clear_cache_list->data);
+
+		sysctl_clear_cache_list = g_slist_delete_link (sysctl_clear_cache_list, sysctl_clear_cache_list);
+
+		g_hash_table_destroy (priv->sysctl_get_prev_values);
+		priv->sysctl_get_prev_values = NULL;
+		priv->sysctl_get_warned = FALSE;
+	}
+}
+
 static void
 _log_dbg_sysctl_get_impl (NMPlatform *platform, const char *path, const char *contents)
 {
 	NMLinuxPlatformPrivate *priv = NM_LINUX_PLATFORM_GET_PRIVATE (platform);
 	const char *prev_value = NULL;
 
-	if (!priv->sysctl_get_prev_values)
+	if (!priv->sysctl_get_prev_values) {
+		sysctl_clear_cache_list = g_slist_prepend (sysctl_clear_cache_list, platform);
 		priv->sysctl_get_prev_values = g_hash_table_new_full (g_str_hash, g_str_equal, g_free, g_free);
-	else
+	} else
 		prev_value = g_hash_table_lookup (priv->sysctl_get_prev_values, path);
 
 	if (prev_value) {
@@ -2574,20 +2592,18 @@ _log_dbg_sysctl_get_impl (NMPlatform *platform, const char *path, const char *co
 		g_free (contents_escaped);
 		g_hash_table_insert (priv->sysctl_get_prev_values, g_strdup (path), g_strdup (contents));
 	}
+
+	if (   !priv->sysctl_get_warned
+	    && g_hash_table_size (priv->sysctl_get_prev_values) > 50000) {
+		_LOGW ("sysctl: the internal cache for debug-logging of sysctl values grew pretty large. You can clear it by disabling debug-logging: `nmcli general logging level KEEP domains PLATFORM:INFO`.");
+		priv->sysctl_get_warned = TRUE;
+	}
 }
 
 #define _log_dbg_sysctl_get(platform, path, contents) \
 	G_STMT_START { \
-		if (_LOGD_ENABLED ()) { \
+		if (_LOGD_ENABLED ()) \
 			_log_dbg_sysctl_get_impl (platform, path, contents); \
-		} else { \
-			NMLinuxPlatformPrivate *priv = NM_LINUX_PLATFORM_GET_PRIVATE (platform); \
-			\
-			if (priv->sysctl_get_prev_values) { \
-				g_hash_table_destroy (priv->sysctl_get_prev_values); \
-				priv->sysctl_get_prev_values = NULL; \
-			} \
-		} \
 	} G_STMT_END
 
 static char *
@@ -5029,8 +5045,10 @@ nm_linux_platform_finalize (GObject *object)
 	g_object_unref (priv->udev_client);
 	g_hash_table_unref (priv->wifi_data);
 
-	if (priv->sysctl_get_prev_values)
+	if (priv->sysctl_get_prev_values) {
+		sysctl_clear_cache_list = g_slist_remove (sysctl_clear_cache_list, object);
 		g_hash_table_destroy (priv->sysctl_get_prev_values);
+	}
 
 	G_OBJECT_CLASS (nm_linux_platform_parent_class)->finalize (object);
 }
diff --git a/src/platform/nm-linux-platform.h b/src/platform/nm-linux-platform.h
index a9e2cd82f9..81f9c73871 100644
--- a/src/platform/nm-linux-platform.h
+++ b/src/platform/nm-linux-platform.h
@@ -50,4 +50,6 @@ GType nm_linux_platform_get_type (void);
 
 void nm_linux_platform_setup (void);
 
+void _nm_linux_platform_sysctl_clear_cache (void);
+
 #endif /* __NETWORKMANAGER_LINUX_PLATFORM_H__ */