Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Tue, 25 Feb 2020 11:42:58 -0800
From:   Kees Cook <keescook@chromium.org>
To:     Arvind Sankar <nivedita@alum.mit.edu>
Cc:     Nick Desaulniers <ndesaulniers@google.com>,
        Fangrui Song <maskray@google.com>,
        Borislav Petkov <bp@alien8.de>,
        Nathan Chancellor <natechancellor@gmail.com>,
        Thomas Gleixner <tglx@linutronix.de>,
        "maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT)" <x86@kernel.org>,
        LKML <linux-kernel@vger.kernel.org>,
        clang-built-linux <clang-built-linux@googlegroups.com>,
        Michael Matz <matz@suse.de>,
        Kristen Carlson Accardi <kristen@linux.intel.com>
Subject: Re: --orphan-handling=warn
Message-ID: <202002251140.4F28F0A4F@keescook>
References: <20200109150218.16544-2-nivedita@alum.mit.edu>
 <20200222050845.GA19912@ubuntu-m2-xlarge-x86>
 <20200222065521.GA11284@zn.tnic>
 <20200222070218.GA27571@ubuntu-m2-xlarge-x86>
 <20200222072144.asqaxlv364s6ezbv@google.com>
 <20200222074254.GB11284@zn.tnic>
 <20200222162225.GA3326744@rani.riverdale.lan>
 <CAKwvOdnvMS21s9gLp5nUpDAOu=c7-iWYuKTeFUq+PMhsJOKUgw@mail.gmail.com>
 <202002242122.AA4D1B8@keescook>
 <20200225182951.GA1179890@rani.riverdale.lan>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20200225182951.GA1179890@rani.riverdale.lan>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Tue, Feb 25, 2020 at 01:29:51PM -0500, Arvind Sankar wrote:
> On Mon, Feb 24, 2020 at 09:35:04PM -0800, Kees Cook wrote:
> > Actually, it's rather opposed to the FGKASLR series, as for that, I need
> > some kind of linker script directive like this:
> > 
> > 	/PASSTHRU/ : {
> > 		*(.text.*)
> > 	}
> > 
> > Where "PASSTHRU" would create a 1-to-1 input-section to output-section
> > with the same name, flags, etc.
> > 
> > ld.bfd's handling of orphan sections named .text.* is to put them each
> > as a separate output section, after the existing .text output section.
> > 
> > ld.lld's handling of orphan sections named .text.* is to put them into
> > the .text output section.
> 
> This doesn't match ld's documentation [1] of how orphan sections are to
> be handled, it's supposed to append it into an existing output section
> only if the names match exactly, creating a new one if that isn't so. If
> ld.lld is to be a drop-in replacement for ld.bfd, this probably needs to
> change?

That would be lovely! :P

> Also ld.lld doesn't seem to support the --unique option, I think you'll
> also want that for FGKASLR to avoid merging static functions with the
> same name from unrelated source files.

Right, yes, that seems like something we have to depend on.

> 
> [1] https://sourceware.org/binutils/docs/ld/Orphan-Sections.html
> 
> > 
> > For FGKASLR (as it is currently implemented[2]), the sections need to be
> > individually named output sections (as bfd does it). *However*, with the
> > "warn on orphans" patch, FGKASLR's intentional orphaning will backfire
> > (I guess the warning could be turned off, but I'd like lld to handle
> > FGKASLR at some point.)
> > 
> > Note that cheating and doing the 1-to-1 mapping by handy with a 40,000
> > entry linker script ... made ld.lld take about 15 minutes to do the
> > final link. :(
> 
> Out of curiosity, how long does ld.bfd take on that linker script :)

A single CPU at 100% for 15 minutes. :)

-- 
Kees Cook