linux-kernel - Re: [PATCH 1/2] uaccess: Simplify code pattern for masked user copies

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] [day] [month] [year] [list]

Message-ID: <CAHk-=wjGYtJOrgrjXF6zceePOp5G_F7VTfikjdGOsJ3eqsyfJg@mail.gmail.com>
Date: Sun, 9 Feb 2025 13:38:20 -0800
From: Linus Torvalds <torvalds@...ux-foundation.org>
To: David Laight <david.laight.linux@...il.com>
Cc: linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org, 
	Alexander Viro <viro@...iv.linux.org.uk>, Christian Brauner <brauner@...nel.org>, Jan Kara <jack@...e.cz>, 
	Arnd Bergmann <arnd@...db.de>, Kees Cook <kees@...nel.org>
Subject: Re: [PATCH 1/2] uaccess: Simplify code pattern for masked user copies

On Sun, 9 Feb 2025 at 13:18, David Laight <david.laight.linux@...il.com> wrote:
>
> Except for the ppc? case which needs the size to open a bounded window.

It passes the size down, but I didn't actually see it *use* the size
anywhere outside of the actual range check.

So it has code like

  static __always_inline void allow_user_access(void __user *to, const
void __user *from,
                                              u32 size, unsigned long dir)
  {
        BUILD_BUG_ON(!__builtin_constant_p(dir));

        if (!(dir & KUAP_WRITE))
                return;

        current->thread.kuap = (__force u32)to;
        uaccess_begin_32s((__force u32)to);
  }

but notice how the size is basically not an issue. Same for the 8xx case:

  static __always_inline void allow_user_access(void __user *to, const
void __user *from,
                                              unsigned long size,
unsigned long dir)
  {
        uaccess_begin_8xx(MD_APG_INIT);
  }

or the booke case:

  static __always_inline void allow_user_access(void __user *to, const
void __user *from,
                                              unsigned long size,
unsigned long dir)
  {
        uaccess_begin_booke(current->thread.pid);
  }

but admittedly this is all a maze of small helper functions calling
each other, so I might have missed some path.

             Linus