linux-hardening - Re: [PATCH] accessibility: speakup: refactor deprecated strncpy

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <202308251439.36BC33ADB2@keescook>
Date:   Fri, 25 Aug 2023 14:49:39 -0700
From:   Kees Cook <keescook@...omium.org>
To:     Justin Stitt <justinstitt@...gle.com>
Cc:     William Hubbs <w.d.hubbs@...il.com>,
        Chris Brannon <chris@...-brannons.com>,
        Kirk Reiser <kirk@...sers.ca>,
        Samuel Thibault <samuel.thibault@...-lyon.org>,
        speakup@...ux-speakup.org, linux-kernel@...r.kernel.org,
        linux-hardening@...r.kernel.org
Subject: Re: [PATCH] accessibility: speakup: refactor deprecated strncpy

On Thu, Aug 24, 2023 at 09:44:29PM +0000, Justin Stitt wrote:
> `strncpy` is deprecated for use on NUL-terminated destination strings [1].
> 
> Use `strscpy` as it guarantees NUL-termination of its destination buffer [2]
> which allows for simpler and less ambiguous code.
> 
> Also, change `strlen(buf)` to `strlen(ptr)` to be consistent with
> further usage within the scope of the function. Note that these are
> equivalent:
> |419 	const char *ptr = buf;
> 
> Link: www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings[1]
> Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html [2]
> Link: https://github.com/KSPP/linux/issues/90
> Cc: linux-hardening@...r.kernel.org
> Signed-off-by: Justin Stitt <justinstitt@...gle.com>
> ---
> Note: build-tested only.
> ---
>  drivers/accessibility/speakup/kobjects.c | 5 ++---
>  1 file changed, 2 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/accessibility/speakup/kobjects.c b/drivers/accessibility/speakup/kobjects.c
> index a7522d409802..8aa416c5f3fc 100644
> --- a/drivers/accessibility/speakup/kobjects.c
> +++ b/drivers/accessibility/speakup/kobjects.c
> @@ -422,12 +422,11 @@ static ssize_t synth_direct_store(struct kobject *kobj,
>  	if (!synth)
>  		return -EPERM;
>  
> -	len = strlen(buf);
> +	len = strlen(ptr);
>  	spin_lock_irqsave(&speakup_info.spinlock, flags);
>  	while (len > 0) {
>  		bytes = min_t(size_t, len, 250);
> -		strncpy(tmp, ptr, bytes);
> -		tmp[bytes] = '\0';
> +		strscpy(tmp, ptr, bytes);
>  		string_unescape_any_inplace(tmp);
>  		synth_printf("%s", tmp);
>  		ptr += bytes;

Technically, yes, this is fine...

Reviewed-by: Kees Cook <keescook@...omium.org>

But wow do you find the most amazing code. :)

This thing is taking a buffer and chopping it up into at-most 250 byte
chunks (smaller than buf, I might add), and then sending it to
synth_printf() ... which uses a 160 byte buffer and silently
truncates...  and uses "%s" which is just a string copy...
why doesn't this just use synth_write() directly on an unescaped
buf??

I think this entire function should just be:

static ssize_t synth_direct_store(struct kobject *kobj,
                                  struct kobj_attribute *attr,
                                  const char *buf, size_t count)
{
        char *unescaped;
	char *p;

        if (!synth)
                return -EPERM;

	unescaped = kstrdup(buf, GFP_KERNEL);
	if (!unescaped)
		return -ENOMEM;

	string_unescape_any_inplace(unescaped);

        spin_lock_irqsave(&speakup_info.spinlock, flags);
	synth_write(unescaped, strlen(unescaped));
        spin_unlock_irqrestore(&speakup_info.spinlock, flags);

	kfree(unescaped);

        return count;
}

(Though honestly, why does this need unescaping anyway?)

-Kees

-- 
Kees Cook