netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Ben Hutchings <bhutchings@solarflare.com>
To: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Cc: <shemminger@vyatta.com>, <netdev@vger.kernel.org>,
	<junwei.zhang@6wind.com>
Subject: Re: [PATCH iproute2 v2] ipbatch: fix use of 'ip netns exec'
Date: Tue, 9 Jul 2013 00:36:45 +0100	[thread overview]
Message-ID: <1373326605.1860.12.camel@bwh-desktop.uk.level5networks.com> (raw)
In-Reply-To: <1373276660-14088-1-git-send-email-nicolas.dichtel@6wind.com>

On Mon, 2013-07-08 at 11:44 +0200, Nicolas Dichtel wrote:
[...]
> --- a/ip/ipnetns.c
> +++ b/ip/ipnetns.c
> @@ -138,6 +138,7 @@ static int netns_exec(int argc, char **argv)
>  	const char *name, *cmd;
>  	char net_path[MAXPATHLEN];
>  	int netns;
> +	int pid, status;
>  
>  	if (argc < 1) {
>  		fprintf(stderr, "No netns name specified\n");
> @@ -185,10 +186,23 @@ static int netns_exec(int argc, char **argv)
>  	/* Setup bind mounts for config files in /etc */
>  	bind_etc(name);
>  
> -	if (execvp(cmd, argv + 1)  < 0)
> -		fprintf(stderr, "exec of \"%s\" failed: %s\n",
> -			cmd, strerror(errno));
> -	return EXIT_FAILURE;
> +	pid = fork();
> +	if (pid < 0)
> +		return EXIT_FAILURE;

On failure, this should emit an error message:
		perror("fork");
and then return -1, surely?

> +	else if (pid > 0)
> +		waitpid(pid, &status, 0);
> +	else {
> +		/* Child */
> +		if (execvp(cmd, argv + 1)  < 0)
> +			fprintf(stderr, "exec of \"%s\" failed: %s\n",
> +				cmd, strerror(errno));
> +		return EXIT_FAILURE;

No, the child must not return from this function.  Otherwise there will
be two processes trying to do the same cleanup on the resources that
they still share.  The child should call _exit(127) if exec*() fails.

> +	}
> +	/* ip must returns the status of the child, but do_cmd() will add a
> +	 * minus to this returned value, so let's add another one here to
> +	 * cancel it.
> +	 */
> +	return -WEXITSTATUS(status);
>  }
>  
>  static int is_pid(const char *str)

WEXITSTATUS() is valid only if waitpid() was successful and WIFEXITED()
yields true.

Ben.

-- 
Ben Hutchings, Staff Engineer, Solarflare
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.

  reply	other threads:[~2013-07-08 23:36 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <20130705141639.56b47e7e@samsung-9>
2013-07-08  9:44 ` [PATCH iproute2 v2] ipbatch: fix use of 'ip netns exec' Nicolas Dichtel
2013-07-08 23:36   ` Ben Hutchings [this message]
2013-07-09 10:26     ` [PATCH iproute2 v3] " Nicolas Dichtel
2013-07-09 13:20       ` Eric Dumazet
2013-07-09 15:55         ` [PATCH iproute2 v4] " Nicolas Dichtel
2013-07-09 16:19           ` Ben Hutchings
2013-07-09 16:59             ` Nicolas Dichtel
2013-07-09 18:13               ` Ben Hutchings
2013-07-10  9:52                 ` Nicolas Dichtel

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1373326605.1860.12.camel@bwh-desktop.uk.level5networks.com \
    --to=bhutchings@solarflare.com \
    --cc=junwei.zhang@6wind.com \
    --cc=netdev@vger.kernel.org \
    --cc=nicolas.dichtel@6wind.com \
    --cc=shemminger@vyatta.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).