linux-kernel - Re: [PATCH] seccomp: Use current_pt

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: <fbf5bceb-c2fe-300f-fad9-7bcf9abd96d2@linux.com>
Date:   Mon, 24 Aug 2020 21:03:33 +0300
From:   Denis Efremov <efremov@...ux.com>
To:     Kees Cook <keescook@...omium.org>
Cc:     Andy Lutomirski <luto@...capital.net>,
        Will Drewry <wad@...omium.org>, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] seccomp: Use current_pt_regs()



On 8/24/20 7:34 PM, Kees Cook wrote:
> On Mon, Aug 24, 2020 at 03:59:21PM +0300, Denis Efremov wrote:
>> Modify seccomp_do_user_notification(), __seccomp_filter(),
>> __secure_computing() to use current_pt_regs().
> 
> This looks okay. It seems some architectures have a separate
> define for current_pt_regs(), though it's overlapped directly with
> task_pt_regs(). I'm curious what the benefit of the change is?
> 

Generally, it's just a shorthand.
From: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=a3460a59747cfddfa7be4758e5ef08bf5d751d59
- arch versions are "optimized versions"
- some architectures have task_pt_regs() working only
  for traced tasks blocked on signal delivery. current_pt_regs()
  needs to work for *all* processes

My motivation:
I'm going to add cocci rule for using current_uid(), current_xxx(), ...
instead of raw accesses to current->cred, current->cred->uid
https://elixir.bootlin.com/linux/latest/source/include/linux/cred.h#L379
These interfaces use rcu_dereference_protected() internally for access
check.

I though that adding current_pt_regs(), current_user_stack_pointer() to
this cocci rule will be a good idea.

Thanks,
Denis