From: Duncan Sands <baldrick@wanadoo.fr>
To: Linux Kernel <linux-kernel@vger.kernel.org>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Subject: [PATCH] UP locking fix for net/socket.c
Date: Thu, 31 Oct 2002 12:57:20 +0100 [thread overview]
Message-ID: <200210311257.20535.baldrick@wanadoo.fr> (raw)
[-- Attachment #1: Type: text/plain, Size: 1123 bytes --]
This race occurs on UP (not SMP) systems. socket.c must
not continue using net_family after sock_unregister(net_family)
has returned. Here is a scenario for this to occur: go to sleep in
the following call:
if ((i = net_families[family]->create(sock, protocol)) < 0)
While sleeping, a task calls sock_unregister(family), which
succeeds since on UP there is currently no locking of any kind.
Duncan.
Patches against 2.4.19 and 2.5.45 attached. 2.4.19 patch:
--- linux/net/socket.c.orig 2002-08-03 02:39:46.000000000 +0200
+++ linux/net/socket.c 2002-10-31 09:16:50.000000000 +0100
@@ -132,7 +132,6 @@
static struct net_proto_family *net_families[NPROTO];
-#ifdef CONFIG_SMP
static atomic_t net_family_lockct = ATOMIC_INIT(0);
static spinlock_t net_family_lock = SPIN_LOCK_UNLOCKED;
@@ -170,13 +169,6 @@
atomic_dec(&net_family_lockct);
}
-#else
-#define net_family_write_lock() do { } while(0)
-#define net_family_write_unlock() do { } while(0)
-#define net_family_read_lock() do { } while(0)
-#define net_family_read_unlock() do { } while(0)
-#endif
-
/*
* Statistics counters of the socket lists
[-- Attachment #2: patch_no_race_2.5 --]
[-- Type: text/x-diff, Size: 691 bytes --]
--- Yield/net/socket.c.orig 2002-10-31 09:10:07.000000000 +0100
+++ Yield/net/socket.c 2002-10-31 09:13:48.000000000 +0100
@@ -138,7 +138,6 @@
static struct net_proto_family *net_families[NPROTO];
-#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT)
static atomic_t net_family_lockct = ATOMIC_INIT(0);
static spinlock_t net_family_lock = SPIN_LOCK_UNLOCKED;
@@ -175,13 +174,6 @@
atomic_dec(&net_family_lockct);
}
-#else
-#define net_family_write_lock() do { } while(0)
-#define net_family_write_unlock() do { } while(0)
-#define net_family_read_lock() do { } while(0)
-#define net_family_read_unlock() do { } while(0)
-#endif
-
/*
* Statistics counters of the socket lists
[-- Attachment #3: patch_no_race_2.4 --]
[-- Type: text/x-diff, Size: 658 bytes --]
--- linux/net/socket.c.orig 2002-08-03 02:39:46.000000000 +0200
+++ linux/net/socket.c 2002-10-31 09:16:50.000000000 +0100
@@ -132,7 +132,6 @@
static struct net_proto_family *net_families[NPROTO];
-#ifdef CONFIG_SMP
static atomic_t net_family_lockct = ATOMIC_INIT(0);
static spinlock_t net_family_lock = SPIN_LOCK_UNLOCKED;
@@ -170,13 +169,6 @@
atomic_dec(&net_family_lockct);
}
-#else
-#define net_family_write_lock() do { } while(0)
-#define net_family_write_unlock() do { } while(0)
-#define net_family_read_lock() do { } while(0)
-#define net_family_read_unlock() do { } while(0)
-#endif
-
/*
* Statistics counters of the socket lists
reply other threads:[~2002-10-31 11:50 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
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=200210311257.20535.baldrick@wanadoo.fr \
--to=baldrick@wanadoo.fr \
--cc=alan@lxorguk.ukuu.org.uk \
--cc=linux-kernel@vger.kernel.org \
/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
all inboxes | Powered by JetHome®