From: Serge Hallyn <serge.hallyn@ubuntu.com>
Date: Fri, 28 Feb 2014 03:49:27 +0000 (-0600)
Subject: clone: don't ever mark the clone's rootfs as being the old, on disk
X-Git-Tag: lxc-1.1.0.alpha1~252
X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=8d2efe40a32700eaec6a8f3057b3b9bb16721b3d;p=thirdparty%2Flxc.git

clone: don't ever mark the clone's rootfs as being the old, on disk

Otherwise an interrupted clone can lead to the original rootfs
being delete.

There is a period during lxcapi_clone during which we have written down
a temporary configuration file on disk, for the new container, using the
old rootfs. Interruption of clone doesn't allow us to do the cleanup we
do in error paths, so a subsequent lxc-destroy removes the old rootfs.

Fix this by doing the copy_storage as early as possible, and not
writing down the rootfs when we write down the temporary configuration
file.

(note - I tested this by putting a series of
'if (strcmp(newname, "u%d") == 0) exit(1)' inline to trigger
interruption between most blocks.  If someone has a good idea
for a generic way to regression-test this henceforth that'd be
great)

See https://bugs.launchpad.net/lxc/+bug/1285850

Signed-off-by: Serge Hallyn <serge.hallyn@ubuntu.com>
Acked-by: Stéphane Graber <stgraber@ubuntu.com>
---

diff --git a/src/lxc/lxccontainer.c b/src/lxc/lxccontainer.c
index d9aa973f4..c60f92717 100644
--- a/src/lxc/lxccontainer.c
+++ b/src/lxc/lxccontainer.c
@@ -2610,6 +2610,7 @@ static struct lxc_container *lxcapi_clone(struct lxc_container *c, const char *n
 	char newpath[MAXPATHLEN];
 	int ret, storage_copied = 0;
 	const char *n, *l;
+	char *origroot = NULL;
 	struct clone_update_data data;
 	FILE *fout;
 	pid_t pid;
@@ -2645,6 +2646,10 @@ static struct lxc_container *lxcapi_clone(struct lxc_container *c, const char *n
 	}
 
 	// copy the configuration, tweak it as needed,
+	if (c->lxc_conf->rootfs.path) {
+		origroot = c->lxc_conf->rootfs.path;
+		c->lxc_conf->rootfs.path = NULL;
+	}
 	fout = fopen(newpath, "w");
 	if (!fout) {
 		SYSERROR("open %s", newpath);
@@ -2652,6 +2657,7 @@ static struct lxc_container *lxcapi_clone(struct lxc_container *c, const char *n
 	}
 	write_config(fout, c->lxc_conf);
 	fclose(fout);
+	c->lxc_conf->rootfs.path = origroot;
 
 	sprintf(newpath, "%s/%s/rootfs", l, n);
 	if (mkdir(newpath, 0755) < 0) {
@@ -2671,6 +2677,12 @@ static struct lxc_container *lxcapi_clone(struct lxc_container *c, const char *n
 		ERROR("clone: failed to create new container (%s %s)", n, l);
 		goto out;
 	}
+	c2->lxc_conf->rootfs.path = origroot;
+
+	// copy/snapshot rootfs's
+	ret = copy_storage(c, c2, bdevtype, flags, bdevdata, newsize);
+	if (ret < 0)
+		goto out;
 
 	// update utsname
 	if (!set_config_item_locked(c2, "lxc.utsname", newname)) {
@@ -2694,11 +2706,6 @@ static struct lxc_container *lxcapi_clone(struct lxc_container *c, const char *n
 	if (!(flags & LXC_CLONE_KEEPMACADDR))
 		network_new_hwaddrs(c2);
 
-	// copy/snapshot rootfs's
-	ret = copy_storage(c, c2, bdevtype, flags, bdevdata, newsize);
-	if (ret < 0)
-		goto out;
-
 	// We've now successfully created c2's storage, so clear it out if we
 	// fail after this
 	storage_copied = 1;