Re: [PATCH] Kernel: Improvement in code readability when memdup_user_nul() fails.
From: Eric W. Biederman
Date: Fri Nov 11 2016 - 11:58:56 EST
Sachin Shukla <sachin.s5@xxxxxxxxxxx> writes:
> From: "Sachin Shukla" <sachin.s5@xxxxxxxxxxx>
>
> There is no need to call kfree() if memdup_user_nul() fails, as no memory
> was allocated and the error in the error-valued pointer should be
> returned.
Nacked-by: "Eric W. Biederman" <ebiederm@xxxxxxxxxxxx>
This is stupid, unnecesary gratituous change that makes the maintenance
of the code worse.
>
> Signed-off-by: Sachin Shukla <sachin.s5@xxxxxxxxxxx>
> ---
> kernel/user_namespace.c | 25 ++++++++++++++-----------
> 1 file changed, 14 insertions(+), 11 deletions(-)
>
> diff --git a/kernel/user_namespace.c b/kernel/user_namespace.c
> index 86b7854..a0ffbf0 100644
> --- a/kernel/user_namespace.c
> +++ b/kernel/user_namespace.c
> @@ -672,28 +672,31 @@ static ssize_t map_write(struct file *file, const char __user *buf,
> */
> mutex_lock(&userns_state_mutex);
>
> - ret = -EPERM;
> /* Only allow one successful write to the map */
> - if (map->nr_extents != 0)
> - goto out;
> + if (map->nr_extents != 0) {
> + mutex_unlock(&userns_state_mutex);
> + return -EPERM;
> + }
>
> /*
> * Adjusting namespace settings requires capabilities on the target.
> */
> - if (cap_valid(cap_setid) && !file_ns_capable(file, ns, CAP_SYS_ADMIN))
> - goto out;
> + if (cap_valid(cap_setid) && !file_ns_capable(file, ns, CAP_SYS_ADMIN)) {
> + mutex_unlock(&userns_state_mutex);
> + return -EPERM;
> + }
>
> /* Only allow < page size writes at the beginning of the file */
> - ret = -EINVAL;
> - if ((*ppos != 0) || (count >= PAGE_SIZE))
> - goto out;
> + if ((*ppos != 0) || (count >= PAGE_SIZE)) {
> + mutex_unlock(&userns_state_mutex);
> + return -EINVAL;
> + }
>
> /* Slurp in the user data */
> kbuf = memdup_user_nul(buf, count);
> if (IS_ERR(kbuf)) {
> - ret = PTR_ERR(kbuf);
> - kbuf = NULL;
> - goto out;
> + mutex_unlock(&userns_state_mutex);
> + return PTR_ERR(kbuf);
> }
>
> /* Parse the user data */