Re: [PATCH v2] platform/x86: think-lmi: Fix current password length check

[David Laight]

On Tue, 18 Aug 2026 15:39:35 +0200
Thorsten Blum <thorsten.blum@xxxxxxxxx> wrote:

> On Tue, Aug 18, 2026 at 04:20:10PM +0300, Ilpo Järvinen wrote:
> > On Tue, 18 Aug 2026, Thorsten Blum wrote:  
> > > On Tue, Aug 18, 2026 at 02:28:50PM +0300, Ilpo Järvinen wrote:  
> > > > On Thu, 13 Aug 2026, Thorsten Blum wrote:  
> > > > > current_password_store() checks the password length before removing the
> > > > > trailing newline, which can reject valid passwords that are exactly  
> > > > > ->maxlen bytes long.  
> > > > > 
> > > > > It also passes ->maxlen to strscpy(), which truncates passwords without
> > > > > a newline.
> > > > > 
> > > > > Use strchrnul() to measure the password length up to the newline, then
> > > > > copy that many bytes and add a trailing NUL terminator.
> > > > > 
> > > > > Fixes: a40cd7ef22fb ("platform/x86: think-lmi: Add WMI interface support on Lenovo platforms")
> > > > > Cc: stable@xxxxxxxxxxxxxxx
> > > > > Signed-off-by: Thorsten Blum <thorsten.blum@xxxxxxxxx>
> > > > > ---
> > > > > Changes in v2:
> > > > > - Keep and reword the newline comment
> > > > > - v1: https://lore.kernel.org/r/20260810132018.156868-3-thorsten.blum@xxxxxxxxx/
> > > > > ---
> > > > >  drivers/platform/x86/lenovo/think-lmi.c | 8 ++++----
> > > > >  1 file changed, 4 insertions(+), 4 deletions(-)
> > > > > 
> > > > > diff --git a/drivers/platform/x86/lenovo/think-lmi.c b/drivers/platform/x86/lenovo/think-lmi.c
> > > > > index e215e86e3db7..d0ceb6aaa69e 100644
> > > > > --- a/drivers/platform/x86/lenovo/think-lmi.c
> > > > > +++ b/drivers/platform/x86/lenovo/think-lmi.c
> > > > > @@ -438,14 +438,14 @@ static ssize_t current_password_store(struct kobject *kobj,
> > > > >  	struct tlmi_pwd_setting *setting = to_tlmi_pwd_setting(kobj);
> > > > >  	size_t pwdlen;
> > > > >  
> > > > > -	pwdlen = strlen(buf);
> > > > > +	/* Strip newline; setting password won't work if one is present. */
> > > > > +	pwdlen = strchrnul(buf, '\n') - buf;
> > > > >  	/* pwdlen == 0 is allowed to clear the password */
> > > > >  	if (pwdlen && ((pwdlen < setting->minlen) || (pwdlen > setting->maxlen)))
> > > > >  		return -EINVAL;
> > > > >  
> > > > > -	strscpy(setting->password, buf, setting->maxlen);
> > > > > -	/* Strip out CR if one is present, setting password won't work if it is present */
> > > > > -	strreplace(setting->password, '\n', '\0');
> > > > > +	memcpy(setting->password, buf, pwdlen);
> > > > > +	setting->password[pwdlen] = '\0';  
> > > > 
> > > > Hi,
> > > > 
> > > > I don't understand why is this strscpy() -> memcpy() conversion required?   
> > > 
> > > It's not required, strscpy() would also work.
> > >   
> > > > Wouldn't it work with:
> > > > 
> > > > 	strscpy(..., pwdlen);
> > > > 
> > > > ?  
> > > 
> > > However, strscpy(setting->password, buf, pwdlen) wouldn't work because
> > > pwdlen is the number of characters to copy, but the destination buffer
> > > size also needs to include room for the NUL terminator. Since there is
> > > no NUL before pwdlen, it would copy only pwdlen - 1 characters and set
> > > setting->password[pwdlen - 1] = '\0'.  
> > 
> > Okay, I was thinking in my mind if I've a off-by-one error in my 
> > suggestion but didn't want to spend too much time on figuring it out.
> >   
> > > strscpy(setting->password, buf, pwdlen + 1) would work, but since we
> > > already know that exactly pwdlen bytes need to be copied, memcpy() is
> > > sufficient.  

No that overwrites past the end out the output buffer.

> > 
> > It may work, but since you then go to nul terminate it yourself, I think 
> > using the existing function is way better than memcpy() + custom nul 
> > termination code.  
> 
> I used memcpy() because we already determined the string length pwdlen,
> which strscpy() would have to determine again internally.
> 
> The performance difference shouldn't matter here, so it's mostly a
> matter of style or personal preference. I'm fine either way.

Personally I'd always use memcpy() if the size is known.

Whoever added the strscpy() should have used the size of the array not
the soft bound for the password length (which is limited by the array size).

In any case the code could be:
	pwdlen = count;
	if (pwdlen && buf[pwdlen - 1] == '\n')
		pwdlen--;
	...
	memcpy(setting->password, buf, pwdlen);
	setting->password[pwdlen] = 0;
	return count;
}

I'm sure there could be (might even be) a helper to copy some
characters and append a '\0'.

David



> 
> Thanks,
> Thorsten
>