linux-kernel - [PATCH v2] rbtree: remove unneeded explicit alignment in struct rb

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-ID: <YQ1ToK8EMdAO4CyH@precision>
Date:   Fri, 6 Aug 2021 17:22:08 +0200
From:   Mete Polat <metepolat2000@...il.com>
To:     Michel Lespinasse <michel@...pinasse.org>,
        Davidlohr Bueso <dbueso@...e.de>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Lukas Bulwahn <lukas.bulwahn@...il.com>,
        Peter Zijlstra <peterz@...radead.org>
Cc:     Jesper Nilsson <jesper@....nu>, Arnd Bergmann <arnd@...db.de>,
        David Woodhouse <dwmw2@...radead.org>,
        Ingo Molnar <mingo@...nel.org>,
        Randy Dunlap <rdunlap@...radead.org>,
        Geert Uytterhoeven <geert@...ux-m68k.org>,
        kernel-janitors@...r.kernel.org,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        linux-m68k <linux-m68k@...ts.linux-m68k.org>
Subject: [PATCH v2] rbtree: remove unneeded explicit alignment in struct
 rb_node

Commit e977145aeaad ("[RBTREE] Add explicit alignment to sizeof(long) for
struct rb_node.") adds an explicit alignment to the struct rb_node due to
some speciality of the CRIS architecture.

The support for the CRIS architecture was removed with commit c690eddc2f3b
("CRIS: Drop support for the CRIS port")

So, remove this now unneeded explicit alignment in struct rb_node as well.

This basically reverts commit e977145aeaad ("[RBTREE] Add explicit
alignment to sizeof(long) for struct rb_node.").

The rbtree node color is stored in the LSB of '__rb_parent_color'.
Only mask the first bit in '__rb_parent()', otherwise it modifies the
node's parent address on m68k.

Reported-by: Randy Dunlap <rdunlap@...radead.org>
Reported-by: Mete Polat <metepolat2000@...il.com>
Signed-off-by: Lukas Bulwahn <lukas.bulwahn@...il.com>
Signed-off-by: Mete Polat <metepolat2000@...il.com>
---
I have tested it on x86, but not on m68k. Can you ack that Geert?

 include/linux/rbtree.h           | 3 +--
 include/linux/rbtree_augmented.h | 2 +-
 2 files changed, 2 insertions(+), 3 deletions(-)

diff --git a/include/linux/rbtree.h b/include/linux/rbtree.h
index d31ecaf4fdd3..e9390be1ba67 100644
--- a/include/linux/rbtree.h
+++ b/include/linux/rbtree.h
@@ -25,8 +25,7 @@ struct rb_node {
 	unsigned long  __rb_parent_color;
 	struct rb_node *rb_right;
 	struct rb_node *rb_left;
-} __attribute__((aligned(sizeof(long))));
-    /* The alignment might seem pointless, but allegedly CRIS needs it */
+};
 
 struct rb_root {
 	struct rb_node *rb_node;
diff --git a/include/linux/rbtree_augmented.h b/include/linux/rbtree_augmented.h
index d1c53e9d8c75..94b6a0f4499e 100644
--- a/include/linux/rbtree_augmented.h
+++ b/include/linux/rbtree_augmented.h
@@ -145,7 +145,7 @@ RB_DECLARE_CALLBACKS(RBSTATIC, RBNAME,					      \
 #define	RB_RED		0
 #define	RB_BLACK	1
 
-#define __rb_parent(pc)    ((struct rb_node *)(pc & ~3))
+#define __rb_parent(pc)    ((struct rb_node *)(pc & ~1))
 
 #define __rb_color(pc)     ((pc) & 1)
 #define __rb_is_black(pc)  __rb_color(pc)
-- 
2.32.0