Re: [PATCH] daemon: plug memory leak on overlong path

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Sat, Dec 18, 2021 at 10:47:03AM +0100, René Scharfe wrote:

> Release the strbuf containing the interpolated path after copying it to
> a stack buffer and before erroring out if it's too long.

Thanks, this looks obviously correct. The problem was introduced by my
6bdb0083be (daemon: detect and reject too-long paths, 2016-10-22).

I don't think it's that big a deal in practice, since we'd always be in
a worker process handling a single request, and will exit immediately
after returning from the function. So you could not, say, convince a
long-running git-daemon to leak a bunch of memory over time.

But definitely still worth fixing.

> diff --git a/daemon.c b/daemon.c
> index 4a000ee4af..94a5b8a364 100644
> --- a/daemon.c
> +++ b/daemon.c
> @@ -232,13 +232,13 @@ static const char *path_ok(const char *directory, struct hostinfo *hi)
> 
>  		rlen = strlcpy(interp_path, expanded_path.buf,
>  			       sizeof(interp_path));
> +		strbuf_release(&expanded_path);
>  		if (rlen >= sizeof(interp_path)) {
>  			logerror("interpolated path too large: %s",
>  				 interp_path);
>  			return NULL;
>  		}
> 
> -		strbuf_release(&expanded_path);
>  		loginfo("Interpolated dir '%s'", interp_path);

A common mistake in these kinds of fixes is that the variable to be
freed is used to generate an error message in the early-return path.
Here we put "interp_path" in the message instead. That does mean the
error message shows the truncated name rather than the full one. That
may be a hidden feature, though. :)

-Peff



[Index of Archives]     [Linux Kernel Development]     [Gcc Help]     [IETF Annouce]     [DCCP]     [Netdev]     [Networking]     [Security]     [V4L]     [Bugtraq]     [Yosemite]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Linux SCSI]     [Fedora Users]

  Powered by Linux