From: Pedro Alves <palves@redhat.com>
To: Tom Tromey <tom@tromey.com>, gdb-patches@sourceware.org
Subject: Re: [PATCH] Disable the undefined behavior sanitizer by default
Date: Tue, 09 Oct 2018 18:05:00 -0000 [thread overview]
Message-ID: <7f7e560e-d005-04e0-92fb-94d959b3ed6f@redhat.com> (raw)
In-Reply-To: <20181009180231.20321-1-tom@tromey.com>
On 10/09/2018 07:02 PM, Tom Tromey wrote:
> There have been a few undefined behavior failures reported, and Pedro
> suggested that the sanitizer be disabled by default. This patch
> implements this.
Thanks. TBC, I think it'll be good to enable it by default.
But for now, we're already aware of a few issues that need
fixing, and the people fixing them can do --enable-ubsan
explicitly.
Please push.
>
> gdb/ChangeLog
> 2018-10-09 Tom Tromey <tom@tromey.com>
>
> * configure: Rebuild.
> * sanitize.m4 (AM_GDB_UBSAN): Default to no.
> * NEWS: Update --enable-ubsan documentation.
>
> gdb/doc/ChangeLog
> 2018-10-09 Tom Tromey <tom@tromey.com>
>
> * gdb.texinfo (Configure Options): Update --enable-ubsan
> documentation.
> ---
> gdb/ChangeLog | 6 ++++++
> gdb/NEWS | 10 +++++-----
> gdb/configure | 2 +-
> gdb/doc/ChangeLog | 5 +++++
> gdb/doc/gdb.texinfo | 13 +++++++------
> gdb/sanitize.m4 | 2 +-
> 6 files changed, 25 insertions(+), 13 deletions(-)
>
> diff --git a/gdb/NEWS b/gdb/NEWS
> index 8a0dd52fa7..126e61e282 100644
> --- a/gdb/NEWS
> +++ b/gdb/NEWS
> @@ -122,11 +122,11 @@ FreeBSD/riscv riscv*-*-freebsd*
>
> --enable-ubsan
>
> - Enable or disable the undefined behavior sanitizer. Release
> - versions of GDB disable this by default if it is available, but
> - development versions enable it. Enabling this can cause a
> - performance penalty. The undefined behavior sanitizer was first
> - introduced in GCC 4.9.
> + Enable or disable the undefined behavior sanitizer. This is
> + disabled by default, but passing --enable-ubsan=yes or
> + --enable-ubsan=auto to configure will enable it. Enabling this can
> + cause a performance penalty. The undefined behavior sanitizer was
> + first introduced in GCC 4.9.
>
> *** Changes in GDB 8.2
>
> diff --git a/gdb/configure b/gdb/configure
> index 304adc81fe..adb26e5d4d 100755
> --- a/gdb/configure
> +++ b/gdb/configure
> @@ -15614,7 +15614,7 @@ ac_compiler_gnu=$ac_cv_c_compiler_gnu
> if test "${enable_ubsan+set}" = set; then :
> enableval=$enable_ubsan;
> else
> - enable_ubsan=auto
> + enable_ubsan=no
> fi
>
> if test "x$enable_ubsan" = xauto; then
> diff --git a/gdb/doc/gdb.texinfo b/gdb/doc/gdb.texinfo
> index efd6dffb1e..b0dc3bf67c 100644
> --- a/gdb/doc/gdb.texinfo
> +++ b/gdb/doc/gdb.texinfo
> @@ -35771,12 +35771,13 @@ to the compiler, which will fail the compilation if the compiler
> outputs any warning messages.
>
> @item --enable-ubsan
> -Enable the GCC undefined behavior sanitizer. By default this is
> -disabled in @value{GDBN} releases, but enabled, when available, when
> -building from git. The undefined behavior sanitizer checks for
> -C@t{++} undefined behavior. It has a performance cost, so if you are
> -looking at @value{GDBN}'s performance, you should disable it. The
> -undefined behavior sanitizer was first introduced in GCC 4.9.
> +Enable the GCC undefined behavior sanitizer. This is disabled by
> +default, but passing @code{--enable-ubsan=yes} or
> +@code{--enable-ubsan=auto} to @code{configure} will enable it. The
> +undefined behavior sanitizer checks for C@t{++} undefined behavior.
> +It has a performance cost, so if you are looking at @value{GDBN}'s
> +performance, you should disable it. The undefined behavior sanitizer
> +was first introduced in GCC 4.9.
> @end table
>
> @node System-wide configuration
> diff --git a/gdb/sanitize.m4 b/gdb/sanitize.m4
> index 76df44da4f..979f6461c0 100644
> --- a/gdb/sanitize.m4
> +++ b/gdb/sanitize.m4
> @@ -20,7 +20,7 @@ AC_DEFUN([AM_GDB_UBSAN],[
> AC_ARG_ENABLE(ubsan,
> AS_HELP_STRING([--enable-ubsan],
> [enable undefined behavior sanitizer (auto/yes/no)]),
> - [],enable_ubsan=auto)
> + [],enable_ubsan=no)
> if test "x$enable_ubsan" = xauto; then
> if $development; then
> enable_ubsan=yes
>
Thanks,
Pedro Alves
next prev parent reply other threads:[~2018-10-09 18:05 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-10-09 18:02 Tom Tromey
2018-10-09 18:05 ` Pedro Alves [this message]
2018-10-09 18:20 ` Eli Zaretskii
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=7f7e560e-d005-04e0-92fb-94d959b3ed6f@redhat.com \
--to=palves@redhat.com \
--cc=gdb-patches@sourceware.org \
--cc=tom@tromey.com \
/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