Re: [RFC v5][PATCH 8/8] Dump open file descriptors

[Bastian Blank <bastian@waldi.eu.org>]

On Sat, Sep 13, 2008 at 07:06:06PM -0400, Oren Laadan wrote:
> +int cr_scan_fds(struct files_struct *files, int **fdtable)
> +{
> +	struct fdtable *fdt;
> +	int *fds;
> +	int i, n, tot;
> +
> +	n = 0;
> +	tot = CR_DEFAULT_FDTABLE;

Why not?
| int i;
| int n = 0;
| int tot = CR_DEFAULT_FDTABLE;

IHMO easier readable.

> +	spin_lock(&files->file_lock);
> +	fdt = files_fdtable(files);
> +	for (i = 0; i < fdt->max_fds; i++) {

The process is suspended at this state?

> +		if (n == tot) {
> +			/*
> +			 * fcheck_files() is safe with drop/re-acquire
> +			 * of the lock, because it tests:  fd < max_fds
> +			 */
> +			spin_unlock(&files->file_lock);
> +			tot *= 2;
> +			if (tot < 0) {		/* overflow ? */

_NO_. tot is signed, this does not have documented overflow behaviour.
You need to restrict this to a sane number.

> +				kfree(fds);
> +				return -EMFILE;
> +			}
> +			fds = krealloc(fds, tot * sizeof(*fds), GFP_KERNEL);
> +			if (!fds)

krealloc does not free the memory on error, so this is a leak.

Bastian

-- 
The more complex the mind, the greater the need for the simplicity of play.
		-- Kirk, "Shore Leave", stardate 3025.8