Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Mon, 21 Oct 2019 20:29:23 +0300
From:   Alexey Dobriyan <adobriyan@gmail.com>
To:     Joel Savitz <jsavitz@redhat.com>
Cc:     linux-kernel@vger.kernel.org,
        Fabrizio D'Angelo <Fabrizio_Dangelo@student.uml.edu>,
        Andrew Morton <akpm@linux-foundation.org>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        Thomas Gleixner <tglx@linutronix.de>,
        linux-fsdevel@vger.kernel.org, fedora-rpi@googlegroups.com
Subject: Re: [PATCH] fs: proc: Clarify warnings for invalid proc dir names
Message-ID: <20191021172923.GA5355@avx2>
References: <20191020221742.5728-1-jsavitz@redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <20191020221742.5728-1-jsavitz@redhat.com>
User-Agent: Mutt/1.10.1 (2018-07-13)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Sun, Oct 20, 2019 at 06:17:42PM -0400, Joel Savitz wrote:
> When one attempts to create a directory in /proc with an invalid name,
> such as one in a subdirectory that doesn't exist, one with a name beyond
> 256 characters, or a reserved name such as '.' or '..', the kernel
> throws a warning message that looks like this:
> 
> 	[ 7913.252558] name 'invalid_name'

Yes, the important part is filename:line which uniquely identifies
the issue.

> This warning message is nearly the same for all invalid cases, including
> the removal of a nonexistent directory. This patch clarifies the warning
> message and differentiates the invalid creation/removal cases so as to
> allow the user to more quickly understand their mistake.

> --- a/fs/proc/generic.c
> +++ b/fs/proc/generic.c
> @@ -173,7 +173,7 @@ static int __xlate_proc_name(const char *name, struct proc_dir_entry **ret,
>  		len = next - cp;
>  		de = pde_subdir_find(de, cp, len);
>  		if (!de) {
> -			WARN(1, "name '%s'\n", name);
> +			WARN(1, "invalid proc dir name '%s'\n", name);

Wrong string anyway, this is nonexistent name, directory or not.

> -		WARN(1, "name len %u\n", qstr.len);
> +		WARN(1, "invalid proc dir name len %u\n", qstr.len);
>  		return NULL;
>  	}
>  	if (qstr.len == 1 && fn[0] == '.') {
> -		WARN(1, "name '.'\n");
> +		WARN(1, "invalid proc dir name '.'\n");
>  		return NULL;
>  	}
>  	if (qstr.len == 2 && fn[0] == '.' && fn[1] == '.') {
> -		WARN(1, "name '..'\n");
> +		WARN(1, "invalid proc dir name '..'\n");
>  		return NULL;
>  	}
>  	if (*parent == &proc_root && name_to_int(&qstr) != ~0U) {
> @@ -402,7 +402,7 @@ static struct proc_dir_entry *__proc_create(struct proc_dir_entry **parent,
>  		return NULL;
>  	}
>  	if (is_empty_pde(*parent)) {
> -		WARN(1, "attempt to add to permanently empty directory");
> +		WARN(1, "attempt to add to permanently empty directory in proc");

"proc" will be spilled over all backtrace.

> @@ -670,7 +670,7 @@ void remove_proc_entry(const char *name, struct proc_dir_entry *parent)
>  		rb_erase(&de->subdir_node, &parent->subdir);
>  	write_unlock(&proc_subdir_lock);
>  	if (!de) {
> -		WARN(1, "name '%s'\n", name);
> +		WARN(1, "unable to remove nonexistent proc dir '%s'\n", name);

I'm not sure if such chatty strings are necessary.