Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Kuniyuki Iwashima <kuniyu@amazon.com>
To:     "David S. Miller" <davem@davemloft.net>,
        Eric Dumazet <edumazet@google.com>,
        Jakub Kicinski <kuba@kernel.org>,
        Paolo Abeni <pabeni@redhat.com>,
        Luis Chamberlain <mcgrof@kernel.org>,
        Kees Cook <keescook@chromium.org>,
        Iurii Zaikin <yzaikin@google.com>
CC:     Kuniyuki Iwashima <kuniyu@amazon.com>,
        Kuniyuki Iwashima <kuni1840@gmail.com>,
        <netdev@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
        Subash Abhinov Kasiviswanathan <subashab@codeaurora.org>
Subject: [PATCH v1 net 04/16] sysctl: Add proc_douintvec_lockless().
Date:   Tue, 5 Jul 2022 22:21:18 -0700
Message-ID: <20220706052130.16368-5-kuniyu@amazon.com>
In-Reply-To: <20220706052130.16368-1-kuniyu@amazon.com>
References: <20220706052130.16368-1-kuniyu@amazon.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Content-Type: text/plain
Precedence: bulk

A sysctl variable is accessed concurrently, and there is always a chance of
data-race.  So, all readers and writers need some basic protection to avoid
load/store-tearing.

This patch changes proc_douintvec() to use READ_ONCE()/WRITE_ONCE()
internally to fix a data-race on the sysctl side.  For now,
proc_douintvec() itself is tolerant to a data-race, but we still need to
add annotations on the other subsystem's side.

In case we miss such fixes, this patch converts proc_douintvec() to a
wrapper of proc_douintvec_lockless().  When we fix a data-race in the other
subsystem, we can explicitly set it as a handler.

Also, this patch removes proc_douintvec()'s document and adds
proc_douintvec_lockless()'s one so that no one will use proc_douintvec()
anymore.

Fixes: e7d316a02f68 ("sysctl: handle error writing UINT_MAX to u32 fields")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
---
CC: Subash Abhinov Kasiviswanathan <subashab@codeaurora.org>
---
 include/linux/sysctl.h |  1 +
 kernel/sysctl.c        | 20 +++++++++++++++-----
 2 files changed, 16 insertions(+), 5 deletions(-)

diff --git a/include/linux/sysctl.h b/include/linux/sysctl.h
index cb87919b5508..770ee1833c25 100644
--- a/include/linux/sysctl.h
+++ b/include/linux/sysctl.h
@@ -85,6 +85,7 @@ PROC_HANDLER(proc_do_static_key);
 
 PROC_HANDLER(proc_dobool_lockless);
 PROC_HANDLER(proc_dointvec_lockless);
+PROC_HANDLER(proc_douintvec_lockless);
 
 /*
  * Register a set of sysctl names by calling register_sysctl_table
diff --git a/kernel/sysctl.c b/kernel/sysctl.c
index 50d9b78aa0b3..be8a7d912180 100644
--- a/kernel/sysctl.c
+++ b/kernel/sysctl.c
@@ -474,9 +474,11 @@ static int do_proc_douintvec_conv(unsigned long *lvalp,
 	if (write) {
 		if (*lvalp > UINT_MAX)
 			return -EINVAL;
-		*valp = *lvalp;
+
+		WRITE_ONCE(*valp, *lvalp);
 	} else {
-		unsigned int val = *valp;
+		unsigned int val = READ_ONCE(*valp);
+
 		*lvalp = (unsigned long)val;
 	}
 	return 0;
@@ -775,7 +777,7 @@ static int proc_dointvec_minmax_warn_RT_change(struct ctl_table *table,
 #endif
 
 /**
- * proc_douintvec - read a vector of unsigned integers
+ * proc_douintvec_lockless - read/write a vector of unsigned integers locklessly
  * @table: the sysctl table
  * @write: %TRUE if this is a write to the sysctl file
  * @buffer: the user buffer
@@ -787,13 +789,19 @@ static int proc_dointvec_minmax_warn_RT_change(struct ctl_table *table,
  *
  * Returns 0 on success.
  */
-int proc_douintvec(struct ctl_table *table, int write, void *buffer,
-		size_t *lenp, loff_t *ppos)
+int proc_douintvec_lockless(struct ctl_table *table, int write, void *buffer,
+			    size_t *lenp, loff_t *ppos)
 {
 	return do_proc_douintvec(table, write, buffer, lenp, ppos,
 				 do_proc_douintvec_conv, NULL);
 }
 
+int proc_douintvec(struct ctl_table *table, int write, void *buffer,
+		   size_t *lenp, loff_t *ppos)
+{
+	return proc_douintvec_lockless(table, write, buffer, lenp, ppos);
+}
+
 /*
  * Taint values can only be increased
  * This means we can safely use a temporary.
@@ -1513,6 +1521,7 @@ PROC_HANDLER_ENOSYS(proc_do_large_bitmap);
 
 PROC_HANDLER_ENOSYS(proc_dobool_lockless);
 PROC_HANDLER_ENOSYS(proc_dointvec_lockless);
+PROC_HANDLER_ENOSYS(proc_douintvec_lockless);
 
 #endif /* CONFIG_PROC_SYSCTL */
 
@@ -2425,3 +2434,4 @@ EXPORT_SYMBOL(proc_do_large_bitmap);
 
 EXPORT_SYMBOL(proc_dobool_lockless);
 EXPORT_SYMBOL(proc_dointvec_lockless);
+EXPORT_SYMBOL(proc_douintvec_lockless);
-- 
2.30.2