[lxc-devel] [PATCH 6/6] Enable lxc_fini() to destroy container on shutdown This works for any bdev-type but is only used for overlayfs and aufs now

[Serge Hallyn]

Quoting Christian Brauner (christianvanbrauner at gmail.com):
> Now we can e.g. implement ephemeral containers in a consistent way.
> 
> Signed-off-by: Christian Brauner <christianvanbrauner at gmail.com>
> 
>  100.0% src/lxc/
> diff --git a/src/lxc/start.c b/src/lxc/start.c
> index ffb8d12..1179d2c 100644
> --- a/src/lxc/start.c
> +++ b/src/lxc/start.c
> @@ -83,6 +83,11 @@ const struct ns_info ns_info[LXC_NS_MAX] = {
>  	[LXC_NS_NET] = {"net", CLONE_NEWNET}
>  };
>  
> +static int bdev_destroy_wrapper(void *data);
> +static int lxc_rmdir_onedev_wrapper(void *data);
> +static void lxc_destroy_container_on_signal(struct lxc_handler *handler,
> +					    const char *name);
> +
>  static void print_top_failing_dir(const char *path)
>  {
>  	size_t len = strlen(path);
> @@ -495,6 +500,13 @@ void lxc_fini(const char *name, struct lxc_handler *handler)
>  		close(handler->ttysock[0]);
>  		close(handler->ttysock[1]);
>  	}
> +	if (handler->conf->ephemeral > 0) {

This doesn't match what was advertised - < 0 should also mean
ephemeral.  (I really think a boolean makes the most sense,
enforced in the conffile as must be 0|1.)

> +		/* Only destroy when rootfs is overlayfs or aufs for now. The
> +		 * function however, works for other bdev types as well. */

What is the reason for that limitation?

> +		if ((strncmp(handler->conf->rootfs.path, "overlayfs:", 10) == 0) ||
> +		   (strncmp(handler->conf->rootfs.path, "aufs:", 5) == 0))
> +			lxc_destroy_container_on_signal(handler, name);
> +	}
>  	cgroup_destroy(handler);
>  	free(handler);
>  }
> @@ -1291,3 +1303,61 @@ int lxc_start(const char *name, char *const argv[], struct lxc_conf *conf,
>  	conf->need_utmp_watch = 1;
>  	return __lxc_start(name, conf, &start_ops, &start_arg, lxcpath, backgrounded);
>  }
> +
> +static void lxc_destroy_container_on_signal(struct lxc_handler *handler,
> +					    const char *name)
> +{
> +	char destroy[MAXPATHLEN];
> +	bool bret = true;
> +	int ret = 0;
> +	if (handler->conf && handler->conf->rootfs.path && handler->conf->rootfs.mount) {
> +		if (am_unpriv())
> +			ret = userns_exec_1(handler->conf, bdev_destroy_wrapper, handler->conf);
> +		else
> +			bret = bdev_destroy(handler->conf);
> +		if (ret < 0 || !bret) {
> +			ERROR("Error destroying rootfs for %s", name);
> +		} else {
> +			INFO("Destroyed rootfs for %s", name);
> +		}
> +	}
> +	ret = snprintf(destroy, MAXPATHLEN, "%s/%s", handler->lxcpath, name);
> +	if (ret < 0 || ret >= MAXPATHLEN)
> +		ERROR("Error creating string");
> +	if (am_unpriv())
> +		ret = userns_exec_1(handler->conf, lxc_rmdir_onedev_wrapper, destroy);
> +	else
> +		ret = lxc_rmdir_onedev(destroy, NULL);
> +	if (ret < 0) {
> +		ERROR("Error destroying container directory for %s", name);
> +	} else {
> +		INFO("Destroyed directory for %s", name);
> +	}
> +}
> +
> +static int bdev_destroy_wrapper(void *data)

The one in lxccontainers.c should somehow be shared.

> +{
> +	struct lxc_conf *conf = data;
> +
> +	if (setgid(0) < 0) {
> +		ERROR("Failed to setgid to 0");
> +		return -1;
> +	}
> +	if (setgroups(0, NULL) < 0)
> +		WARN("Failed to clear groups");
> +	if (setuid(0) < 0) {
> +		ERROR("Failed to setuid to 0");
> +		return -1;
> +	}
> +	if (!bdev_destroy(conf))
> +		return -1;
> +	else
> +		return 0;
> +}
> +
> +static int lxc_rmdir_onedev_wrapper(void *data)

Hm, I guess it's easier just to keep this one static in 
the two places like you're doing.

> +{
> +	char *arg = (char *) data;
> +	return lxc_rmdir_onedev(arg, NULL);
> +}
> +
> -- 
> 2.5.1
>