linux-kernel - Re: [PATCH 0/9] [v3] System Calls for Memory Protection Keys

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]

Date:	Thu, 30 Jun 2016 10:40:17 -0700
From:	Andy Lutomirski <luto@...capital.net>
To:	Dave Hansen <dave@...1.net>
Cc:	Ingo Molnar <mingo@...nel.org>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	X86 ML <x86@...nel.org>, Linux API <linux-api@...r.kernel.org>,
	linux-arch <linux-arch@...r.kernel.org>,
	"linux-mm@...ck.org" <linux-mm@...ck.org>,
	Linus Torvalds <torvalds@...ux-foundation.org>,
	Andrew Morton <akpm@...ux-foundation.org>,
	Arnd Bergmann <arnd@...db.de>,
	Al Viro <viro@...iv.linux.org.uk>,
	Mel Gorman <mgorman@...hsingularity.net>,
	Hugh Dickins <hughd@...gle.com>
Subject: Re: [PATCH 0/9] [v3] System Calls for Memory Protection Keys

On Thu, Jun 30, 2016 at 9:46 AM, Dave Hansen <dave@...1.net> wrote:
> On 06/30/2016 02:41 AM, Ingo Molnar wrote:
>> * Dave Hansen <dave@...1.net> wrote:
>>> Are there any concerns with merging these into the x86 tree so
>>> that they go upstream for 4.8?  The updates here are pretty
>>> minor.
>>
>>>  include/linux/pkeys.h                         |   39 +-
>>>  include/uapi/asm-generic/mman-common.h        |    5 +
>>>  include/uapi/asm-generic/unistd.h             |   12 +-
>>>  mm/mprotect.c                                 |  134 +-
>>
>> So I'd love to have some high level MM review & ack for these syscall ABI
>> extensions.
>
> That's a quite reasonable request, but I'm really surprised by it at
> this point.  The proposed ABI is one very straightforward extension to
> one existing system call, plus four others that you personally suggested.
>

I apologize for the very late review, but (see other thread) I think
we may need to make sure we've defined the signal delivery semantics
in a useful way before enabling these.  I'm not convinced that the
current behavior is helpful.  This may or may not require any change
to the syscall signatures, but I can imagine that doing it right would
involve adding another syscall to *read* the current signal-delivery
state of a pkey or perhaps of all the pkeys.  That could potentially
be achieved by adding an extra pointer parameter to pkey_get so
pkey_get can return both the current state and the state at next
signal delivery.

--Andy