linux-kernel - Re: [PATCH v2 1/2] livepatch: introduce shadow variable API

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: <alpine.LSU.2.20.1707171732150.22628@pobox.suse.cz>
Date:   Mon, 17 Jul 2017 17:35:38 +0200 (CEST)
From:   Miroslav Benes <mbenes@...e.cz>
To:     Josh Poimboeuf <jpoimboe@...hat.com>
cc:     Joe Lawrence <joe.lawrence@...hat.com>,
        live-patching@...r.kernel.org, linux-kernel@...r.kernel.org,
        Jessica Yu <jeyu@...hat.com>, Jiri Kosina <jikos@...nel.org>,
        Petr Mladek <pmladek@...e.com>
Subject: Re: [PATCH v2 1/2] livepatch: introduce shadow variable API

On Thu, 13 Jul 2017, Josh Poimboeuf wrote:

> On Wed, Jun 28, 2017 at 11:37:26AM -0400, Joe Lawrence wrote:
> 
> > +Brief API summary
> > +-----------------
> > +
> > +See the full API usage docbook notes in the livepatch/shadow.c
> > +implementation.
> > +
> > +An in-kernel hashtable references all of the shadow variables.  These
> > +references are stored/retrieved through a <obj, num> key pair.
> 
> "num" is rather vague, how about "key"?
> 
> (And note, this and some of the other comments also apply to the code as
> well)
> 
> > +* The klp_shadow variable data structure encapsulates both tracking
> > +meta-data and shadow-data:
> > +  - meta-data
> > +    - obj - pointer to original data
> 
> Instead of "original data", how about calling it the "parent object"?
> That describes it better to me at least.  "Original data" sounds like
> some of the data might be replaced.

I agree that "original data" does not sound right. However, we use "parent 
object" for vmlinux or a module in our code. But I don't have a better 
name and "parent object" sounds good.
 
> > +    - num - numerical description of new data
> 
> "numerical description of new data" sounds a little confusing, how about
> "unique identifier for new data"?
> 
> I'm also not sure about the phrase "new data".  Maybe something like
> "new data field" would be more descriptive?  Or just "new field"?  I
> view it kind of like adding a field to a struct.  Not a big deal either
> way.
>
> > +void *klp_shadow_attach(void *obj, unsigned long num, void *new_data,
> > +			size_t new_size, gfp_t gfp_flags);
> 
> It could be just me, but the "new_" prefixes threw me off a little bit.
> The new is implied anyway.  How about just "data" and "size"?
> 
> And the same comment for the klp_shadow struct.

I agree with Josh on all of this.

Miroslav