public inbox for ltp@lists.linux.it
 help / color / mirror / Atom feed
From: Petr Vorel <pvorel@suse.cz>
To: Richard Palethorpe <rpalethorpe@suse.de>
Cc: ltp@lists.linux.it
Subject: Re: [LTP] [PATCH v2 0/3] safe_macros: Fix undefined behaviour in vararg handling
Date: Tue, 29 Nov 2022 14:59:11 +0100	[thread overview]
Message-ID: <Y4YQL2TezOp2dcnD@pevik> (raw)
In-Reply-To: <87a649ubuq.fsf@suse.de>

Hi all,

> Hello,

> So I'm happy with this, but I think Cyril's comment deserves a response:
+1

> > Looking at how glibc handles this, the code looks like:

> > 	int mode = 0;

> > 	if (__OPEN_NEEDS_MODE(oflag)) {
> > 		..
> > 		mode = va_arg(arg, int);
> > 		..
> > 	}

> > That sounds much easier than messing with the macros and should avoid
> > undefined behavior.
+1

> I don't see why, __OPEN_NEEDS_MODE is going to be different between

> functions and libc/kernel versions.

Looking at glibc's __OPEN_NEEDS_MODE definition, the logic is obviously the same
as musl code for open (it just use O_TMPFILE instead of __O_TMPFILE therefore no
need to check for #ifdef __O_TMPFILE).

Kind regards,
Petr

> Reviewed-by: Richard Palethorpe <rpalethorpe@suse.com>

> Tudor Cretu <tudor.cretu@arm.com> writes:

> > Accessing elements in an empty va_list results in undefined behaviour[0]
> > that can include accessing arbitrary stack memory. While typically this
> > doesn't raise a fault, some new more security-oriented architectures
> > (e.g. CHERI[1] or Morello[2]) don't allow it.

> > Therefore, remove the variadicness from safe_* wrappers that always call
> > the functions with the optional argument included.

> > Adapt the respective SAFE_* macros to handle the change by passing a
> > default argument if they're omitted.

> > [0]: [ISO/IEC 9899:2011] Programming Languages—C, 3rd ed, paragraph 7.16.1.1
> > [1]: https://www.cl.cam.ac.uk/research/security/ctsrd/cheri/
> > [2]: https://www.morello-project.org/

> > v2..v1:
> >   - PATCH 1: Remove the NULL argument for mode from SAFE_OPEN instances
> >     to avoid the pointer to int conversion.

> > Tudor Cretu (3):
> >   safe_open: Fix undefined behaviour in vararg handling
> >   safe_openat: Fix undefined behaviour in vararg handling
> >   safe_semctl: Fix undefined behaviour in vararg handling

> >  include/old/safe_macros.h                         |  6 ++++--
> >  include/safe_macros_fn.h                          |  3 ++-
> >  include/tst_safe_file_at.h                        | 10 ++++++----
> >  include/tst_safe_macros.h                         |  6 ++++--
> >  include/tst_safe_sysv_ipc.h                       | 14 +++++++++-----
> >  lib/safe_macros.c                                 | 13 +------------
> >  lib/tst_cgroup.c                                  |  2 +-
> >  lib/tst_safe_file_at.c                            | 11 +++--------
> >  lib/tst_safe_sysv_ipc.c                           | 10 +---------
> >  testcases/kernel/syscalls/fgetxattr/fgetxattr01.c |  2 +-
> >  testcases/kernel/syscalls/fgetxattr/fgetxattr02.c |  2 +-
> >  testcases/kernel/syscalls/fgetxattr/fgetxattr03.c |  2 +-
> >  testcases/kernel/syscalls/fsetxattr/fsetxattr01.c |  2 +-
> >  testcases/kernel/syscalls/fsetxattr/fsetxattr02.c |  2 +-
> >  14 files changed, 36 insertions(+), 49 deletions(-)

> > -- 
> > 2.25.1

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

  reply	other threads:[~2022-11-29 13:59 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-11-29 13:03 [LTP] [PATCH v2 0/3] safe_macros: Fix undefined behaviour in vararg handling Tudor Cretu
2022-11-29 13:03 ` [LTP] [PATCH v2 1/3] safe_open: " Tudor Cretu
2022-11-29 13:03 ` [LTP] [PATCH v2 2/3] safe_openat: " Tudor Cretu
2022-11-29 13:03 ` [LTP] [PATCH v2 3/3] safe_semctl: " Tudor Cretu
2022-11-29 13:23 ` [LTP] [PATCH v2 0/3] safe_macros: " Richard Palethorpe
2022-11-29 13:59   ` Petr Vorel [this message]
2022-11-29 14:04     ` Tudor Cretu
2022-11-29 15:15       ` Petr Vorel
2022-11-30 13:43         ` Tudor Cretu

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=Y4YQL2TezOp2dcnD@pevik \
    --to=pvorel@suse.cz \
    --cc=ltp@lists.linux.it \
    --cc=rpalethorpe@suse.de \
    /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