[lxc-devel] [PATCH RFC] do_mount_entry: add nexec, nosuid, nodev, rdonly flags if needed at remount

[Serge Hallyn]

See http://lkml.org/lkml/2014/8/13/746 and its history.  The kernel now refuses
mounts if we don't add ro,nosuid,nodev,noexec flags if they were already there.

Also use the newly found info to skip remount if unneeded.  For background, if
you want to create a read-only bind mount, then you must first mount(2) with
MS_BIND to create the bind mount, then re-mount(2) again to get the new mount
options to apply.  So if this wasn't a bind mount, or no new mount options were
introduced, then we don't do the second mount(2).

null_endofword() and get_field() were not changed, only moved up in
the file.

(Note, while I can start containers inside a privileged container with
this patch, most of the lxc tests still fail with the kernel in question;
Andy's patch seems to still be needed - a kernel with which is available
at https://launchpad.net/~serge-hallyn/+archive/ubuntu/userns-natty
ppa:serge-hallyn/userns-natty)

Signed-off-by: Serge Hallyn <serge.hallyn at ubuntu.com>
---
 src/lxc/conf.c | 107 ++++++++++++++++++++++++++++++++++++++++++---------------
 1 file changed, 80 insertions(+), 27 deletions(-)

diff --git a/src/lxc/conf.c b/src/lxc/conf.c
index 2b3aff1..01ace43 100644
--- a/src/lxc/conf.c
+++ b/src/lxc/conf.c
@@ -1927,10 +1927,74 @@ int parse_mntopts(const char *mntopts, unsigned long *mntflags,
 	return 0;
 }
 
+static void null_endofword(char *word)
+{
+	while (*word && *word != ' ' && *word != '\t')
+		word++;
+	*word = '\0';
+}
+
+/*
+ * skip @nfields spaces in @src
+ */
+static char *get_field(char *src, int nfields)
+{
+	char *p = src;
+	int i;
+
+	for (i = 0; i < nfields; i++) {
+		while (*p && *p != ' ' && *p != '\t')
+			p++;
+		if (!*p)
+			break;
+		p++;
+	}
+	return p;
+}
+
+static unsigned long get_mount_flags(const char *path)
+{
+	FILE *f = fopen("/proc/self/mountinfo", "r");
+	char *line = NULL;
+	size_t len = 0;
+	unsigned long flags = 0;
+
+	if (!f) {
+		WARN("Failed to open /proc/self/mountinfo");
+		return 0;
+	}
+	while (getline(&line, &len, f) != -1) {
+		char *target, *opts, *p, *saveptr = NULL;
+		target = get_field(line, 4);
+		if (!target)
+			continue;
+		opts = get_field(target, 2);
+		if (!opts)
+			continue;
+		null_endofword(opts);
+		for (p = strtok_r(opts, ",", &saveptr); p;
+			p = strtok_r(NULL, ",", &saveptr)) {
+			if (strcmp(p, "ro") == 0)
+				flags |= MS_RDONLY;
+			else if (strcmp(p, "nodev") == 0)
+				flags |= MS_NODEV;
+			else if (strcmp(p, "nosuid") == 0)
+				flags |= MS_NOSUID;
+			else if (strcmp(p, "noexec") == 0)
+				flags |= MS_NOEXEC;
+			/* XXX todo - we'll have to deal with atime? */
+		}
+	}
+	free(line);
+	fclose(f);
+	return flags;
+}
+
 static int mount_entry(const char *fsname, const char *target,
 		       const char *fstype, unsigned long mountflags,
 		       const char *data, int optional)
 {
+	unsigned long extraflags;
 	if (mount(fsname, target, fstype, mountflags & ~MS_REMOUNT, data)) {
 		if (optional) {
 			INFO("failed to mount '%s' on '%s' (optional): %s", fsname,
@@ -1944,9 +2008,22 @@ static int mount_entry(const char *fsname, const char *target,
 	}
 
 	if ((mountflags & MS_REMOUNT) || (mountflags & MS_BIND)) {
+		extraflags = get_mount_flags(target);
+		DEBUG("flags was %lu, extraflags set to %lu\n", mountflags, extraflags);
+		if (!(mountflags & MS_REMOUNT) && (mountflags & MS_BIND)) {
+			if (!(extraflags & ~mountflags))
+				DEBUG("Skipping remount");
+				goto skipremount;
+		}
+		DEBUG("proceeding with remont");
+		mountflags |= extraflags;
+	}
 
-		DEBUG("remounting %s on %s to respect bind or remount options",
-		      fsname, target);
+
+	if ((mountflags & MS_REMOUNT) || (mountflags & MS_BIND)) {
+		DEBUG("remounting %s on %s to respect bind or remount options %lu (extra %lu)",
+		      fsname ? fsname : "(none)",
+		      target ? target : "(none)", mountflags, extraflags);
 
 		if (mount(fsname, target, fstype,
 			  mountflags | MS_REMOUNT, data)) {
@@ -1963,6 +2040,7 @@ static int mount_entry(const char *fsname, const char *target,
 		}
 	}
 
+skipremount:
 	DEBUG("mounted '%s' on '%s', type '%s'", fsname, target, fstype);
 
 	return 0;
@@ -3888,31 +3966,6 @@ void tmp_proc_unmount(struct lxc_conf *lxc_conf)
 	}
 }
 
-static void null_endofword(char *word)
-{
-	while (*word && *word != ' ' && *word != '\t')
-		word++;
-	*word = '\0';
-}
-
-/*
- * skip @nfields spaces in @src
- */
-static char *get_field(char *src, int nfields)
-{
-	char *p = src;
-	int i;
-
-	for (i = 0; i < nfields; i++) {
-		while (*p && *p != ' ' && *p != '\t')
-			p++;
-		if (!*p)
-			break;
-		p++;
-	}
-	return p;
-}
-
 static void remount_all_slave(void)
 {
 	/* walk /proc/mounts and change any shared entries to slave */
-- 
2.1.0.rc1