From: Igor Stoppa <igor.stoppa@gmail.com>
To: Tycho Andersen <tycho@tycho.ws>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>,
Mimi Zohar <zohar@linux.vnet.ibm.com>,
Kees Cook <keescook@chromium.org>,
Matthew Wilcox <willy@infradead.org>,
Dave Chinner <david@fromorbit.com>,
James Morris <jmorris@namei.org>,
Michal Hocko <mhocko@kernel.org>,
kernel-hardening@lists.openwall.com,
linux-integrity@vger.kernel.org,
linux-security-module@vger.kernel.org,
igor stoppa <igor.stoppa@huawei.com>,
Dave Hansen <dave.hansen@linux.intel.com>,
Jonathan Corbet <corbet@lwn.net>,
Laura Abbott <labbott@redhat.com>,
Thomas Gleixner <tglx@linutronix.de>,
Kate Stewart <kstewart@linuxfoundation.org>,
"David S. Miller" <davem@davemloft.net>,
Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
Philippe Ombredanne <pombredanne@nexb.com>,
"Paul E. McKenney" <paulmck@linux.vnet.ibm.com>,
Josh Triplett <josh@joshtriplett.org>,
rostedt <rostedt@goodmis.org>,
Lai Jiangshan <jiangshanlai@gmail.com>,
linux-kernel <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 14/17] prmem: llist, hlist, both plain and rcu
Date: Thu, 25 Oct 2018 01:52:11 +0300 [thread overview]
Message-ID: <d630474b-73a5-c6ae-2141-49c1184381b2@gmail.com> (raw)
In-Reply-To: <20181024145606.GA9019@cisco>
On 24/10/2018 17:56, Tycho Andersen wrote:
> On Wed, Oct 24, 2018 at 05:03:01PM +0300, Igor Stoppa wrote:
>> On 24/10/18 14:37, Mathieu Desnoyers wrote:
>>> Also, is it the right approach to duplicate existing APIs, or should we
>>> rather hook into page fault handlers and let the kernel do those "shadow"
>>> mappings under the hood ?
>>
>> This question is probably a good candidate for the small Q&A section I have
>> in the 00/17.
>>
>>
>>> Adding a new GFP flags for dynamic allocation, and a macro mapping to
>>> a section attribute might suffice for allocation or definition of such
>>> mostly-read-only/seldom-updated data.
>>
>> I think what you are proposing makes sense from a pure hardening standpoint.
>> From a more defensive one, I'd rather minimise the chances of giving a free
>> pass to an attacker.
>>
>> Maybe there is a better implementation of this, than what I have in mind.
>> But, based on my current understanding of what you are describing, there
>> would be few issues:
>>
>> 1) where would the pool go? The pool is a way to manage multiple vmas and
>> express common property they share. Even before a vma is associated to the
>> pool.
>>
>> 2) there would be more code that can seamlessly deal with both protected and
>> regular data. Based on what? Some parameter, I suppose.
>> That parameter would be the new target.
>> If the code is "duplicated", as you say, the actual differences are baked in
>> at compile time. The "duplication" would also allow to have always inlined
>> functions for write-rare and leave more freedom to the compiler for their
>> non-protected version.
>>
>> Besides, I think the separate wr version also makes it very clear, to the
>> user of the API, that there will be a price to pay, in terms of performance.
>> The more seamlessly alternative might make this price less obvious.
>
> What about something in the middle, where we move list to list_impl.h,
> and add a few macros where you have list_set_prev() in prlist now, so
> we could do,
>
> // prlist.h
>
> #define list_set_next(head, next) wr_ptr(&head->next, next)
> #define list_set_prev(head, prev) wr_ptr(&head->prev, prev)
>
> #include <linux/list_impl.h>
>
> // list.h
>
> #define list_set_next(next) (head->next = next)
> #define list_set_next(prev) (head->prev = prev)
>
> #include <linux/list_impl.h>
>
> I wonder then if you can get rid of some of the type punning too? It's
> not clear exactly why that's necessary from the series, but perhaps
> I'm missing something obvious :)
nothing obvious, probably there is only half a reference in the slides I
linked-to in the cover letter :-)
So far I have minimized the number of "intrinsic" write rare functions,
mostly because I would want first to reach an agreement on the
implementation of the core write-rare.
However, once that is done, it might be good to convert also the prlists
to be "intrinsics". A list node is 2 pointers.
If that was the alignment, i.e. __align(sizeof(list_head)), it might be
possible to speed up a lot the list handling even as write rare.
Taking as example the insertion operation, it would be probably
sufficient, in most cases, to have only two remappings:
- one covering the page with the latest two nodes
- one covering the page with the list head
That is 2 vs 8 remappings, and a good deal of memory barriers less.
This would be incompatible with what you are proposing, yet it would be
justifiable, I think, because it would provide better performance to
prlist, potentially widening its adoption, where performance is a concern.
> I also wonder how much the actual differences being baked in at
> compile time makes. Most (all?) of this code is inlined.
If the inlined function expects to receive a prlist_head *, instead of a
list_head *, doesn't it help turning runtime bugs into buildtime bugs?
Or maybe I miss your point?
--
igor
next prev parent reply other threads:[~2018-10-24 22:52 UTC|newest]
Thread overview: 140+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-10-23 21:34 [RFC v1 PATCH 00/17] prmem: protected memory Igor Stoppa
2018-10-23 21:34 ` [PATCH 01/17] prmem: linker section for static write rare Igor Stoppa
2018-10-23 21:34 ` [PATCH 02/17] prmem: write rare for static allocation Igor Stoppa
2018-10-25 0:24 ` Dave Hansen
2018-10-29 18:03 ` Igor Stoppa
2018-10-26 9:41 ` Peter Zijlstra
2018-10-29 20:01 ` Igor Stoppa
2018-10-23 21:34 ` [PATCH 03/17] prmem: vmalloc support for dynamic allocation Igor Stoppa
2018-10-25 0:26 ` Dave Hansen
2018-10-29 18:07 ` Igor Stoppa
2018-10-23 21:34 ` [PATCH 04/17] prmem: " Igor Stoppa
2018-10-23 21:34 ` [PATCH 05/17] prmem: shorthands for write rare on common types Igor Stoppa
2018-10-25 0:28 ` Dave Hansen
2018-10-29 18:12 ` Igor Stoppa
2018-10-23 21:34 ` [PATCH 06/17] prmem: test cases for memory protection Igor Stoppa
2018-10-24 3:27 ` Randy Dunlap
2018-10-24 14:24 ` Igor Stoppa
2018-10-25 16:43 ` Dave Hansen
2018-10-29 18:16 ` Igor Stoppa
2018-10-23 21:34 ` [PATCH 07/17] prmem: lkdtm tests " Igor Stoppa
2018-10-23 21:34 ` [PATCH 08/17] prmem: struct page: track vmap_area Igor Stoppa
2018-10-24 3:12 ` Matthew Wilcox
2018-10-24 23:01 ` Igor Stoppa
2018-10-25 2:13 ` Matthew Wilcox
2018-10-29 18:21 ` Igor Stoppa
2018-10-23 21:34 ` [PATCH 09/17] prmem: hardened usercopy Igor Stoppa
2018-10-29 11:45 ` Chris von Recklinghausen
2018-10-29 18:24 ` Igor Stoppa
2018-10-23 21:34 ` [PATCH 10/17] prmem: documentation Igor Stoppa
2018-10-24 3:48 ` Randy Dunlap
2018-10-24 14:30 ` Igor Stoppa
2018-10-24 23:04 ` Mike Rapoport
2018-10-29 19:05 ` Igor Stoppa
2018-10-26 9:26 ` Peter Zijlstra
2018-10-26 10:20 ` Matthew Wilcox
2018-10-29 19:28 ` Igor Stoppa
2018-10-26 10:46 ` Kees Cook
2018-10-28 18:31 ` Peter Zijlstra
2018-10-29 21:04 ` Igor Stoppa
2018-10-30 15:26 ` Peter Zijlstra
2018-10-30 16:37 ` Kees Cook
2018-10-30 17:06 ` Andy Lutomirski
2018-10-30 17:58 ` Matthew Wilcox
2018-10-30 18:03 ` Dave Hansen
2018-10-31 9:18 ` Peter Zijlstra
2018-10-30 18:28 ` Tycho Andersen
2018-10-30 19:20 ` Matthew Wilcox
2018-10-30 20:43 ` Igor Stoppa
2018-10-30 21:02 ` Andy Lutomirski
2018-10-30 21:07 ` Kees Cook
2018-10-30 21:25 ` Igor Stoppa
2018-10-30 22:15 ` Igor Stoppa
2018-10-31 10:11 ` Peter Zijlstra
2018-10-31 20:38 ` Andy Lutomirski
2018-10-31 20:53 ` Andy Lutomirski
2018-10-31 9:45 ` Peter Zijlstra
2018-10-30 21:35 ` Matthew Wilcox
2018-10-30 21:49 ` Igor Stoppa
2018-10-31 4:41 ` Andy Lutomirski
2018-10-31 9:08 ` Igor Stoppa
2018-10-31 19:38 ` Igor Stoppa
2018-10-31 10:02 ` Peter Zijlstra
2018-10-31 20:36 ` Andy Lutomirski
2018-10-31 21:00 ` Peter Zijlstra
2018-10-31 22:57 ` Andy Lutomirski
2018-10-31 23:10 ` Igor Stoppa
2018-10-31 23:19 ` Andy Lutomirski
2018-10-31 23:26 ` Igor Stoppa
2018-11-01 8:21 ` Thomas Gleixner
2018-11-01 15:58 ` Igor Stoppa
2018-11-01 17:08 ` Peter Zijlstra
2018-10-30 18:51 ` Andy Lutomirski
2018-10-30 19:14 ` Kees Cook
2018-10-30 21:25 ` Matthew Wilcox
2018-10-30 21:55 ` Igor Stoppa
2018-10-30 22:08 ` Matthew Wilcox
2018-10-31 9:29 ` Peter Zijlstra
2018-10-30 23:18 ` Nadav Amit
2018-10-31 9:08 ` Peter Zijlstra
2018-11-01 16:31 ` Nadav Amit
2018-11-02 21:11 ` Nadav Amit
2018-10-31 9:36 ` Peter Zijlstra
2018-10-31 11:33 ` Matthew Wilcox
2018-11-13 14:25 ` Igor Stoppa
2018-11-13 17:16 ` Andy Lutomirski
2018-11-13 17:43 ` Nadav Amit
2018-11-13 17:47 ` Andy Lutomirski
2018-11-13 18:06 ` Nadav Amit
2018-11-13 18:31 ` Igor Stoppa
2018-11-13 18:33 ` Igor Stoppa
2018-11-13 18:36 ` Andy Lutomirski
2018-11-13 19:03 ` Igor Stoppa
2018-11-21 16:34 ` Igor Stoppa
2018-11-21 17:36 ` Nadav Amit
2018-11-21 18:01 ` Igor Stoppa
2018-11-21 18:15 ` Andy Lutomirski
2018-11-22 19:27 ` Igor Stoppa
2018-11-22 20:04 ` Matthew Wilcox
2018-11-22 20:53 ` Andy Lutomirski
2018-12-04 12:34 ` Igor Stoppa
2018-11-13 18:48 ` Andy Lutomirski
2018-11-13 19:35 ` Igor Stoppa
2018-11-13 18:26 ` Igor Stoppa
2018-11-13 18:35 ` Andy Lutomirski
2018-11-13 19:01 ` Igor Stoppa
2018-10-31 9:27 ` Igor Stoppa
2018-10-26 11:09 ` Markus Heiser
2018-10-29 19:35 ` Igor Stoppa
2018-10-26 15:05 ` Jonathan Corbet
2018-10-29 19:38 ` Igor Stoppa
2018-10-29 20:35 ` Igor Stoppa
2018-10-23 21:34 ` [PATCH 11/17] prmem: llist: use designated initializer Igor Stoppa
2018-10-23 21:34 ` [PATCH 12/17] prmem: linked list: set alignment Igor Stoppa
2018-10-26 9:31 ` Peter Zijlstra
2018-10-23 21:35 ` [PATCH 13/17] prmem: linked list: disable layout randomization Igor Stoppa
2018-10-24 13:43 ` Alexey Dobriyan
2018-10-29 19:40 ` Igor Stoppa
2018-10-26 9:32 ` Peter Zijlstra
2018-10-26 10:17 ` Matthew Wilcox
2018-10-30 15:39 ` Peter Zijlstra
2018-10-23 21:35 ` [PATCH 14/17] prmem: llist, hlist, both plain and rcu Igor Stoppa
2018-10-24 11:37 ` Mathieu Desnoyers
2018-10-24 14:03 ` Igor Stoppa
2018-10-24 14:56 ` Tycho Andersen
2018-10-24 22:52 ` Igor Stoppa [this message]
2018-10-25 8:11 ` Tycho Andersen
2018-10-28 9:52 ` Steven Rostedt
2018-10-29 19:43 ` Igor Stoppa
2018-10-26 9:38 ` Peter Zijlstra
2018-10-23 21:35 ` [PATCH 15/17] prmem: test cases for prlist and prhlist Igor Stoppa
2018-10-23 21:35 ` [PATCH 16/17] prmem: pratomic-long Igor Stoppa
2018-10-25 0:13 ` Peter Zijlstra
2018-10-29 21:17 ` Igor Stoppa
2018-10-30 15:58 ` Peter Zijlstra
2018-10-30 16:28 ` Will Deacon
2018-10-31 9:10 ` Peter Zijlstra
2018-11-01 3:28 ` Kees Cook
2018-10-23 21:35 ` [PATCH 17/17] prmem: ima: turn the measurements list write rare Igor Stoppa
2018-10-24 23:03 ` [RFC v1 PATCH 00/17] prmem: protected memory Dave Chinner
2018-10-29 19:47 ` Igor Stoppa
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=d630474b-73a5-c6ae-2141-49c1184381b2@gmail.com \
--to=igor.stoppa@gmail.com \
--cc=corbet@lwn.net \
--cc=dave.hansen@linux.intel.com \
--cc=davem@davemloft.net \
--cc=david@fromorbit.com \
--cc=gregkh@linuxfoundation.org \
--cc=igor.stoppa@huawei.com \
--cc=jiangshanlai@gmail.com \
--cc=jmorris@namei.org \
--cc=josh@joshtriplett.org \
--cc=keescook@chromium.org \
--cc=kernel-hardening@lists.openwall.com \
--cc=kstewart@linuxfoundation.org \
--cc=labbott@redhat.com \
--cc=linux-integrity@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-security-module@vger.kernel.org \
--cc=mathieu.desnoyers@efficios.com \
--cc=mhocko@kernel.org \
--cc=paulmck@linux.vnet.ibm.com \
--cc=pombredanne@nexb.com \
--cc=rostedt@goodmis.org \
--cc=tglx@linutronix.de \
--cc=tycho@tycho.ws \
--cc=willy@infradead.org \
--cc=zohar@linux.vnet.ibm.com \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).