From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <vries@sourceware.org>
Received: by sourceware.org (Postfix, from userid 2205)
 id 46F263857C42; Thu, 14 Jul 2022 06:19:09 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 46F263857C42
Content-Type: text/plain; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
From: Tom de Vries <vries@sourceware.org>
To: gdb-cvs@sourceware.org
Subject: [binutils-gdb] [gdb/symtab] Make per_cu->unit_type atomic
X-Act-Checkin: binutils-gdb
X-Git-Author: Tom de Vries <tdevries@suse.de>
X-Git-Refname: refs/heads/master
X-Git-Oldrev: 4f92e10cda142fd1f213e01a53ca687e38cddf22
X-Git-Newrev: b35bd7d552f80518ad90e81f592b73ee2ef736d7
Message-Id: <20220714061909.46F263857C42@sourceware.org>
Date: Thu, 14 Jul 2022 06:19:09 +0000 (GMT)
X-BeenThere: gdb-cvs@sourceware.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Gdb-cvs mailing list <gdb-cvs.sourceware.org>
List-Unsubscribe: <https://sourceware.org/mailman/options/gdb-cvs>,
 <mailto:gdb-cvs-request@sourceware.org?subject=unsubscribe>
List-Archive: <https://sourceware.org/pipermail/gdb-cvs/>
List-Help: <mailto:gdb-cvs-request@sourceware.org?subject=help>
List-Subscribe: <https://sourceware.org/mailman/listinfo/gdb-cvs>,
 <mailto:gdb-cvs-request@sourceware.org?subject=subscribe>
X-List-Received-Date: Thu, 14 Jul 2022 06:19:09 -0000

https://sourceware.org/git/gitweb.cgi?p=3Dbinutils-gdb.git;h=3Db35bd7d552f8=
0518ad90e81f592b73ee2ef736d7

commit b35bd7d552f80518ad90e81f592b73ee2ef736d7
Author: Tom de Vries <tdevries@suse.de>
Date:   Thu Jul 14 08:19:00 2022 +0200

    [gdb/symtab] Make per_cu->unit_type atomic
   =20
    With gdb with -fsanitize=3Dthread and test-case gdb.ada/array_bounds.ex=
p, I run
    into a data race between:
    ...
      Read of size 1 at 0x7b2000025f0f by main thread:
        #0 packed<dwarf_unit_type, 1ul>::operator dwarf_unit_type() const p=
acked.h:54
        #1 dwarf2_per_cu_data::set_unit_type(dwarf_unit_type) read.h:339
    ...
    and:
    ...
      Previous write of size 1 at 0x7b2000025f0f by thread T3:
        #0 dwarf2_per_cu_data::set_unit_type(dwarf_unit_type) read.h:341
    ...
   =20
    Fix this by making per_cu->unit_type atomic.
   =20
    Tested on x86_64-linux.
   =20
    Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=3D29286

Diff:
---
 gdb/dwarf2/read.h | 23 ++++++++++++++---------
 1 file changed, 14 insertions(+), 9 deletions(-)

diff --git a/gdb/dwarf2/read.h b/gdb/dwarf2/read.h
index fcf0cd70530..acbe298b310 100644
--- a/gdb/dwarf2/read.h
+++ b/gdb/dwarf2/read.h
@@ -178,7 +178,7 @@ public:
=20
 private:
   /* The unit type of this CU.  */
-  packed<dwarf_unit_type, 1> m_unit_type =3D (dwarf_unit_type) 0;
+  std::atomic<packed<dwarf_unit_type, 1>> m_unit_type {(dwarf_unit_type)0};
=20
   /* The language of this CU.  */
   packed<language, LANGUAGE_BYTES> m_lang =3D language_unknown;
@@ -329,19 +329,24 @@ public:
=20
   dwarf_unit_type unit_type (bool strict_p =3D true) const
   {
+    dwarf_unit_type ut =3D m_unit_type.load ();
     if (strict_p)
-      gdb_assert (m_unit_type !=3D 0);
-    return m_unit_type;
+      gdb_assert (ut !=3D 0);
+    return ut;
   }
=20
   void set_unit_type (dwarf_unit_type unit_type)
   {
-    if (m_unit_type =3D=3D 0)
-      /* Set if not set already.  */
-      m_unit_type =3D unit_type;
-    else
-      /* If already set, verify that it's the same value.  */
-      gdb_assert (m_unit_type =3D=3D unit_type);
+    /* Set if not set already.  */
+    packed<dwarf_unit_type, 1> nope =3D (dwarf_unit_type)0;
+    if (m_unit_type.compare_exchange_strong (nope, unit_type))
+      return;
+
+    /* If already set, verify that it's the same value.  */
+    nope =3D unit_type;
+    if (m_unit_type.compare_exchange_strong (nope, unit_type))
+      return;
+    gdb_assert_not_reached ();
   }
=20
   enum language lang (bool strict_p =3D true) const