linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Niklas Cassel <Niklas.Cassel@wdc.com>
To: Kees Cook <keescook@chromium.org>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>,
	Eric Biederman <ebiederm@xmission.com>,
	Paul Walmsley <paul.walmsley@sifive.com>,
	Palmer Dabbelt <palmer@dabbelt.com>,
	Albert Ou <aou@eecs.berkeley.edu>,
	Greg Ungerer <gerg@linux-m68k.org>,
	Mike Frysinger <vapier@gentoo.org>,
	Damien Le Moal <damien.lemoal@opensource.wdc.com>,
	"stable@vger.kernel.org" <stable@vger.kernel.org>,
	"linux-fsdevel@vger.kernel.org" <linux-fsdevel@vger.kernel.org>,
	"linux-mm@kvack.org" <linux-mm@kvack.org>,
	"linux-riscv@lists.infradead.org"
	<linux-riscv@lists.infradead.org>
Subject: Re: [PATCH v2] binfmt_flat: do not stop relocating GOT entries prematurely on riscv
Date: Fri, 15 Apr 2022 01:24:54 +0000	[thread overview]
Message-ID: <YljJZYNWXeWsT6HW@x1-carbon> (raw)
In-Reply-To: <202204141559.B2A0EB4F7@keescook>

On Thu, Apr 14, 2022 at 04:05:54PM -0700, Kees Cook wrote:
> On Thu, Apr 14, 2022 at 11:10:18AM +0200, Niklas Cassel wrote:

(snip)

> > +static inline u32 __user *skip_got_header(u32 __user *rp)
> > +{
> > +	if (IS_ENABLED(CONFIG_RISCV)) {
> > +		/*
> > +		 * RISC-V has a 16 byte GOT PLT header for elf64-riscv
> > +		 * and 8 byte GOT PLT header for elf32-riscv.
> > +		 * Skip the whole GOT PLT header, since it is reserved
> > +		 * for the dynamic linker (ld.so).
> > +		 */
> > +		u32 rp_val0, rp_val1;
> > +
> > +		if (get_user(rp_val0, rp))
> > +			return rp;
> > +		if (get_user(rp_val1, rp + 1))
> > +			return rp;
> > +
> > +		if (rp_val0 == 0xffffffff && rp_val1 == 0xffffffff)
> > +			rp += 4;
> > +		else if (rp_val0 == 0xffffffff)
> > +			rp += 2;
> 
> Just so I understand; due to the FILL(0) and the ALIGN, val1 will be 0
> (or more specifically, not -1) in all other cases, yes?

For elf64-riscv with a .got.plt header:
rp+0: -1, rp+1: -1, rp+2: 0, rp+3: 0

For elf32-riscv with a .got.plt header:
rp+0: -1, rp+1: 0

At least riscv binutils 2.32, 2.37 and 2.38 all create a .got.plt header
even when there are no .got.plt entries following the header.

Even if the .got.plt section was empty, there will still be data in the
.got section, so rp+0 will still not be -1.

If there is no data in the .got section, then the _GLOBAL_OFFSET_TABLE_
symbol will not be defined, so elf2flt will not set the FLAT_FLAG_GOTPIC
flag. (This code is only executed if that flag is set.)


Kind regards,
Niklas

  reply	other threads:[~2022-04-15  1:25 UTC|newest]

Thread overview: 34+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-04-14  9:10 Niklas Cassel
2022-04-14 23:05 ` Kees Cook
2022-04-15  1:24   ` Niklas Cassel [this message]
2022-04-14 23:27 ` Kees Cook
2022-04-14 23:41   ` Damien Le Moal
2022-04-15  1:26   ` Konstantin Ryabitsev
2022-04-16  0:14     ` Kees Cook
2022-04-14 23:51 ` Damien Le Moal
2022-04-15  0:30   ` Niklas Cassel
2022-04-15  0:56     ` Damien Le Moal
2022-04-15  1:08       ` Niklas Cassel
2022-04-15  1:13         ` Damien Le Moal
2022-04-15  2:11           ` Niklas Cassel
2022-04-15  2:14             ` Damien Le Moal
2022-04-20  4:04     ` Greg Ungerer
2022-04-20 14:58       ` [PATCH] binfmt_flat: Remove shared library support Eric W. Biederman
2022-04-20 16:17         ` Palmer Dabbelt
2022-04-20 16:59           ` Rich Felker
2022-04-20 17:47             ` Kees Cook
2022-04-20 20:04               ` Arnd Bergmann
2022-04-20 20:23                 ` Rich Felker
2022-04-20 23:00                   ` Damien Le Moal
2022-04-25  3:38               ` Rob Landley
2022-04-25  7:40                 ` Greg Ungerer
2022-04-20 23:36         ` Damien Le Moal
2022-04-20 23:53         ` Greg Ungerer
2022-04-21  6:52           ` Geert Uytterhoeven
2022-04-21  7:12             ` Arnd Bergmann
2022-04-22 10:26               ` Vladimir Murzin
2022-04-22 15:18               ` Patrice CHOTARD
2022-04-21 12:43             ` Rich Felker
2022-04-25  3:50               ` Rob Landley
2022-04-21  0:05         ` (subset) " Kees Cook
2022-04-16  4:25 ` [PATCH v2] binfmt_flat: do not stop relocating GOT entries prematurely on riscv Kees Cook

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=YljJZYNWXeWsT6HW@x1-carbon \
    --to=niklas.cassel@wdc.com \
    --cc=aou@eecs.berkeley.edu \
    --cc=damien.lemoal@opensource.wdc.com \
    --cc=ebiederm@xmission.com \
    --cc=gerg@linux-m68k.org \
    --cc=keescook@chromium.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=linux-riscv@lists.infradead.org \
    --cc=palmer@dabbelt.com \
    --cc=paul.walmsley@sifive.com \
    --cc=stable@vger.kernel.org \
    --cc=vapier@gentoo.org \
    --cc=viro@zeniv.linux.org.uk \
    /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