public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
From: Jakub Jelinek <jakub@redhat.com>
To: "Martin Liška" <mliska@suse.cz>
Cc: gcc-patches@gcc.gnu.org, arnd@linaro.org
Subject: Re: [PATCH] Come up with --param asan-stack-small-redzone (PR sanitizer/81715).
Date: Tue, 25 Sep 2018 09:49:00 -0000	[thread overview]
Message-ID: <20180925092415.GC8250@tucnak> (raw)
In-Reply-To: <e603cb5f-c598-f599-90df-33773a1bb357@suse.cz>

On Tue, Sep 25, 2018 at 11:05:30AM +0200, Martin Liška wrote:
> As requested in PR81715, GCC emits bigger middle redzones for small variables.
> It's analyzed in following comment: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81715#c28

First of all, does LLVM make the variable sized red zone size only for
automatic variables, or also for global/local statics, or for alloca?

Have you considered also making the red zones larger for very large
variables?

> For now I'm suggesting to shrink shadow memory for variables <= 16B to 32B (including variable storage).
> LLVM is more aggressive as they allocate just 16B of shadow memory for variables <= 4B. That would
> require bigger code refactoring in asan.c and I would like to avoid that.

What exactly would need changing to support the 12-15 bytes long red zones
for 4-1 bytes long automatic vars?
Just asan_emit_stack_protection or something other?

> +	      poly_uint64 size = stack_vars[i].size;
> +	      /* For small variables shrink middle redzone (including
> +	       * variable store) just to ASAN_RED_ZONE_SIZE.  */

We don't use this comment style (* at start of comment continuation lines).
Otherwise it looks reasonable, but I wouldn't stop here.

	Jakub

  reply	other threads:[~2018-09-25  9:25 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-09-25  9:05 Martin Liška
2018-09-25  9:49 ` Jakub Jelinek [this message]
2018-09-25 10:17   ` Martin Liška
2018-09-25 10:25     ` Martin Liška
2018-09-25 12:19     ` Jakub Jelinek
2018-09-25 15:40       ` Martin Liška
2018-09-25 15:55         ` Jakub Jelinek
2018-09-26  9:34           ` Martin Liška
2018-10-09  8:53             ` Martin Liška
2018-10-23  9:17               ` Martin Liška
2018-11-13  7:50                 ` Martin Liška
2018-11-28 12:01             ` Jakub Jelinek
2018-11-29 15:03               ` Martin Liška
2018-11-29 15:17                 ` Jakub Jelinek
2018-11-29 16:37                   ` Martin Liška
2018-11-29 16:50                     ` Jakub Jelinek
2018-11-30 11:44                       ` Martin Liška
2018-11-30 12:01                         ` Jakub Jelinek
2018-12-01  7:36                         ` [PATCH] Partial fix for asan on big endian targets (PR sanitizer/88289) Jakub Jelinek
2018-12-02 12:46                           ` [committed] Fix ICE in asan_clear_shadow (PR sanitizer/88291) Jakub Jelinek

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=20180925092415.GC8250@tucnak \
    --to=jakub@redhat.com \
    --cc=arnd@linaro.org \
    --cc=gcc-patches@gcc.gnu.org \
    --cc=mliska@suse.cz \
    /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 read-only IMAP folder(s) and NNTP newsgroup(s).