linux-kernel - Re: [PATCH mm] maccess: fix strncpy_from_user

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 for Android: free password hash cracker in your pocket

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

Message-ID: <CAEf4BzZrQbpp9ZrSj3PTm2P1CAT1YG6O8zN6dc0FdJjHLXTP7w@mail.gmail.com>
Date: Thu, 17 Apr 2025 14:07:04 -0700
From: Andrii Nakryiko <andrii.nakryiko@...il.com>
To: Andrew Morton <akpm@...ux-foundation.org>
Cc: Mykyta Yatsenko <mykyta.yatsenko5@...il.com>, linux-mm@...ck.org, rostedt@...dmis.org, 
	mhiramat@...nel.org, andrii@...nel.org, kernel-team@...a.com, 
	linux-kernel@...r.kernel.org, Mykyta Yatsenko <yatsenko@...a.com>
Subject: Re: [PATCH mm] maccess: fix strncpy_from_user_nofault empty string handling

On Thu, Apr 17, 2025 at 1:40 PM Andrew Morton <akpm@...ux-foundation.org> wrote:
>
> On Thu, 17 Apr 2025 16:28:08 +0100 Mykyta Yatsenko <mykyta.yatsenko5@...il.com> wrote:
>
> > strncpy_from_user_nofault should return the length of the copied string
> > including the trailing NUL, but if the argument unsafe_addr points to
> > an empty string ({'\0'}), the return value is 0.
> >
> > This happens as strncpy_from_user copies terminal symbol into dst
> > and returns 0 (as expected), but strncpy_from_user_nofault does not
> > modify ret as it is not equal to count and not greater than 0, so 0 is
> > returned, which contradicts the contract.
>
> Looks right, I think.
>
> But why do strncpy_from_user() and strncpy_from_user_nofault() have
> different interfaces?
>
> /**
>  * strncpy_from_user: - Copy a NUL terminated string from userspace.
>  * ...
>  * On success, returns the length of the string (not including the trailing
>  * NUL).
>
> /**
>  * strncpy_from_user_nofault: - Copy a NUL terminated string from unsafe user
>  *                              address.
>  * ...
>  * On success, returns the length of the string INCLUDING the trailing NUL.
>
> This is surprising and I'm wondering what led us to do this?

Agreed, this is very surprising and error-prone. strncpy_from_user()
semantics is a bit better, IMO, in that it allows to "detect" empty
string even if buffer size is 1 byte. And there isn't a lot of places
where we use strncpy_from_user_nofault (only 6, it seems). Maybe we
should just change the semantics of strncpy_from_user_nofault?