From: Nicolas Pitre <nico@fluxnic.net>
To: linux-arm-kernel@lists.infradead.org
Subject: Re: [PATCH v3] ARM: xip: Use correct symbol for end of ROM marker
Date: Mon, 16 Nov 2015 20:57:36 +0000 [thread overview]
Message-ID: <alpine.LFD.2.20.1511161550370.5826@knanqh.ubzr> (raw)
In-Reply-To: <20151116181718.GW8644@n2100.arm.linux.org.uk>
On Mon, 16 Nov 2015, Russell King - ARM Linux wrote:
> Nico,
>
> As you originally created the XIP stuff, I hope you can remember
> the details - can you check this patch please?
>
> I'm thinking that we need a new symbol around here:
>
> #ifdef CONFIG_XIP_KERNEL
> __data_loc = ALIGN(4); /* location in binary */
> <== here
> . = PAGE_OFFSET + TEXT_OFFSET;
> #else
>
> to denote the end of the XIP kernel image which must remain
> accessible after boot. We don't need the data sections because
> they will have been copied to RAM, and we probably don't want to
> keep those exposed (it's potentially useful for attackers.)
The _etext symbol is already used for that purpose.
Now we round it up to the next section mapping which might leave quite a
lot of data content exposed in ROM. But given it is more or less the
same data present in RAM except for those bits that are modified at run
time, I don't see what an attacker would gain from the data in ROM
that cannot already be obtained from kernel RAM. If the section mapping
extends to part of the ROM that is no longer kernel data then maybe this
would expose sensitive data. Is that what you're worried about?
Nicolas
WARNING: multiple messages have this Message-ID (diff)
From: nico@fluxnic.net (Nicolas Pitre)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH v3] ARM: xip: Use correct symbol for end of ROM marker
Date: Mon, 16 Nov 2015 15:57:36 -0500 (EST) [thread overview]
Message-ID: <alpine.LFD.2.20.1511161550370.5826@knanqh.ubzr> (raw)
In-Reply-To: <20151116181718.GW8644@n2100.arm.linux.org.uk>
On Mon, 16 Nov 2015, Russell King - ARM Linux wrote:
> Nico,
>
> As you originally created the XIP stuff, I hope you can remember
> the details - can you check this patch please?
>
> I'm thinking that we need a new symbol around here:
>
> #ifdef CONFIG_XIP_KERNEL
> __data_loc = ALIGN(4); /* location in binary */
> <=== here
> . = PAGE_OFFSET + TEXT_OFFSET;
> #else
>
> to denote the end of the XIP kernel image which must remain
> accessible after boot. We don't need the data sections because
> they will have been copied to RAM, and we probably don't want to
> keep those exposed (it's potentially useful for attackers.)
The _etext symbol is already used for that purpose.
Now we round it up to the next section mapping which might leave quite a
lot of data content exposed in ROM. But given it is more or less the
same data present in RAM except for those bits that are modified at run
time, I don't see what an attacker would gain from the data in ROM
that cannot already be obtained from kernel RAM. If the section mapping
extends to part of the ROM that is no longer kernel data then maybe this
would expose sensitive data. Is that what you're worried about?
Nicolas
next prev parent reply other threads:[~2015-11-16 20:57 UTC|newest]
Thread overview: 103+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-11-11 14:17 [PATCH] ARM: xip: Use correct symbol for end of ROM marker Chris Brandt
2015-11-11 14:17 ` Chris Brandt
2015-11-11 14:17 ` Chris Brandt
2015-11-12 12:17 ` Peter Hurley
2015-11-12 12:17 ` Peter Hurley
2015-11-12 12:17 ` Peter Hurley
2015-11-12 13:15 ` Chris Brandt
2015-11-12 13:15 ` Chris Brandt
2015-11-12 13:15 ` Chris Brandt
2015-11-12 16:32 ` Russell King - ARM Linux
2015-11-12 16:32 ` Russell King - ARM Linux
2015-11-12 16:32 ` Russell King - ARM Linux
2015-11-12 21:01 ` [PATCH v2] " Chris Brandt
2015-11-12 21:01 ` Chris Brandt
2015-11-13 7:46 ` Geert Uytterhoeven
2015-11-13 7:46 ` Geert Uytterhoeven
2015-11-13 20:03 ` Chris Brandt
2015-11-13 20:03 ` Chris Brandt
2015-11-16 18:05 ` [PATCH v3] " Chris Brandt
2015-11-16 18:05 ` Chris Brandt
2015-11-16 18:17 ` Russell King - ARM Linux
2015-11-16 18:17 ` Russell King - ARM Linux
2015-11-16 19:46 ` Chris Brandt
2015-11-16 19:46 ` Chris Brandt
2015-11-16 19:53 ` Russell King - ARM Linux
2015-11-16 19:53 ` Russell King - ARM Linux
2015-11-16 20:18 ` Chris Brandt
2015-11-16 20:18 ` Chris Brandt
2015-11-16 20:30 ` Russell King - ARM Linux
2015-11-16 20:30 ` Russell King - ARM Linux
2015-11-16 20:57 ` Nicolas Pitre [this message]
2015-11-16 20:57 ` Nicolas Pitre
2015-11-16 21:09 ` Chris Brandt
2015-11-16 21:09 ` Chris Brandt
2015-11-16 20:27 ` Nicolas Pitre
2015-11-16 20:27 ` Nicolas Pitre
2015-11-16 21:02 ` Chris Brandt
2015-11-16 21:02 ` Chris Brandt
2015-11-16 21:47 ` Nicolas Pitre
2015-11-16 21:47 ` Nicolas Pitre
2015-11-16 22:19 ` Chris Brandt
2015-11-16 22:19 ` Chris Brandt
2015-11-17 0:48 ` Nicolas Pitre
2015-11-17 0:48 ` Nicolas Pitre
2015-11-17 2:11 ` Chris Brandt
2015-11-17 2:11 ` Chris Brandt
2015-11-17 2:37 ` Nicolas Pitre
2015-11-17 2:37 ` Nicolas Pitre
2015-11-17 16:56 ` Chris Brandt
2015-11-17 16:56 ` Chris Brandt
2015-11-17 17:24 ` Nicolas Pitre
2015-11-17 17:24 ` Nicolas Pitre
2015-11-18 3:58 ` Nicolas Pitre
2015-11-18 3:58 ` Nicolas Pitre
2015-11-18 5:12 ` Magnus Damm
2015-11-18 5:12 ` Magnus Damm
2015-11-18 13:45 ` Nicolas Pitre
2015-11-18 13:45 ` Nicolas Pitre
2015-11-18 17:01 ` Nicolas Pitre
2015-11-18 17:01 ` Nicolas Pitre
2015-11-18 19:12 ` Chris Brandt
2015-11-18 19:12 ` Chris Brandt
2015-11-18 20:23 ` Nicolas Pitre
2015-11-18 20:23 ` Nicolas Pitre
2015-11-18 20:51 ` Chris Brandt
2015-11-18 20:51 ` Chris Brandt
2015-11-18 21:36 ` Nicolas Pitre
2015-11-18 21:36 ` Nicolas Pitre
2016-01-29 21:12 ` Chris Brandt
2016-01-29 21:12 ` Chris Brandt
2016-01-29 21:17 ` Nicolas Pitre
2016-01-29 21:17 ` Nicolas Pitre
2015-11-17 16:45 ` Chris Brandt
2015-11-17 16:45 ` Chris Brandt
2015-11-17 16:57 ` Nicolas Pitre
2015-11-17 16:57 ` Nicolas Pitre
2015-11-18 2:09 ` Chris Brandt
2015-11-18 2:09 ` Chris Brandt
2015-11-18 3:17 ` Nicolas Pitre
2015-11-18 3:17 ` Nicolas Pitre
2015-11-18 8:30 ` Arnd Bergmann
2015-11-18 8:30 ` Arnd Bergmann
2015-11-18 15:28 ` Chris Brandt
2015-11-18 15:28 ` Chris Brandt
2015-11-18 15:16 ` Chris Brandt
2015-11-18 15:16 ` Chris Brandt
2015-11-18 17:07 ` Nicolas Pitre
2015-11-18 17:07 ` Nicolas Pitre
2015-11-18 19:36 ` Chris Brandt
2015-11-18 19:36 ` Chris Brandt
2015-11-18 19:44 ` Nicolas Pitre
2015-11-18 19:44 ` Nicolas Pitre
2015-11-18 20:00 ` Chris Brandt
2015-11-18 20:00 ` Chris Brandt
2015-11-17 17:33 ` Russell King - ARM Linux
2015-11-17 17:33 ` Russell King - ARM Linux
2016-02-01 17:52 ` [PATCH v4] " Chris Brandt
2016-02-01 19:12 ` Nicolas Pitre
2016-02-01 19:41 ` Chris Brandt
2016-02-01 20:23 ` Nicolas Pitre
2016-02-02 17:05 ` Chris Brandt
2016-02-02 17:19 ` [PATCH v5] " Chris Brandt
2016-02-02 17:35 ` Nicolas Pitre
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=alpine.LFD.2.20.1511161550370.5826@knanqh.ubzr \
--to=nico@fluxnic.net \
--cc=linux-arm-kernel@lists.infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.