Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Thu, 19 Jan 2017 22:17:12 +0100
From: Helge Deller <deller@....de>
To: Heiko Carstens <heiko.carstens@...ibm.com>,
 Mark Rutland <mark.rutland@....com>
Cc: Laura Abbott <labbott@...hat.com>, Kees Cook <keescook@...omium.org>,
 Jason Wessel <jason.wessel@...driver.com>, Jonathan Corbet <corbet@....net>,
 Russell King <linux@...linux.org.uk>,
 Catalin Marinas <catalin.marinas@....com>, Will Deacon
 <will.deacon@....com>, "James E.J. Bottomley" <jejb@...isc-linux.org>,
 Martin Schwidefsky <schwidefsky@...ibm.com>,
 Thomas Gleixner <tglx@...utronix.de>, Ingo Molnar <mingo@...hat.com>,
 "H. Peter Anvin" <hpa@...or.com>, x86@...nel.org,
 Rob Herring <robh@...nel.org>, "Rafael J. Wysocki" <rjw@...ysocki.net>,
 Len Brown <len.brown@...el.com>, Pavel Machek <pavel@....cz>,
 Jessica Yu <jeyu@...hat.com>, linux-doc@...r.kernel.org,
 linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
 linux-parisc@...r.kernel.org, linux-s390@...r.kernel.org,
 linux-pm@...r.kernel.org, kernel-hardening@...ts.openwall.com
Subject: Re: [PATCH 1/2] security: Change name of CONFIG_DEBUG_RODATA

On 19.01.2017 12:33, Heiko Carstens wrote:
> On Thu, Jan 19, 2017 at 10:56:46AM +0000, Mark Rutland wrote:
>>> +config HARDENED_PAGE_MAPPINGS
>>> +	bool "Mark kernel mappings with stricter permissions (RO/W^X)"
>>> +	default y
>>> +	depends on ARCH_HAS_HARDENED_MAPPINGS
>>> +	help
>>> +          If this is set, kernel text and rodata memory will be made read-only,
>>> +	  and non-text memory will be made non-executable. This provides
>>> +	  protection against certain security attacks (e.g. executing the heap
>>> +	  or modifying text).
>>> +
>>> +	  Unless your system has known restrictions or performance issues, it
>>> +	  is recommended to say Y here.
>>
>> It's somewhat unfortunate that this means the feature is no longer
>> mandatory on arm64 (and s390+x86). We have a boot-time switch to turn
>> the protections off, so I was hoping we could make this mandatory on all
>> architectures with support.
>>
>> It would be good to see if we could make this mandatory for arm and
>> parisc, or if it really needs to be optional for either of those.
> 
> Looks like the config option is a no-op on parisc just like it is on
> s390. Irrelavant of the config option at least on s390 the page tables for
> kernel text and rodata will be read-only anyway.

Right, that's true at the moment for parisc as well.
I do have unfinished patches which will add runtime kernel patching
for ftrace to parisc, and those patches will need to add code to 
enable/disable ro text/data like x86.
 
> This works since ages and I don't see a reason why this should be
> changed. Also trying to disable this with the "rodata=" command line option
> does not work at least on s390, and I guess this is true for parisc as
> well.

I never tried the option itself, but it should work on parisc to disable
the ro protection by this option.
 
> The only thing implemented with CONFIG_DEBUG_RODATA on both architectures
> is to emit a message that states memory has been protected
> (mark_rodata_ro).
> This just avoids a wrong "Kernel memory protection disabled." message.
> 
> So yes, I'd really like to keep this option mandatory.
 
I'd be fine with a rename of the config option to ARCH_HAS_HARDENED_MAPPINGS
and keeping the "rodata=" command line.

Helge

Powered by blists - more mailing lists

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.