w1: Split memcpy() of struct cn_msg flexible array
authorKees Cook <keescook@chromium.org>
Tue, 27 Sep 2022 00:39:27 +0000 (17:39 -0700)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Fri, 30 Sep 2022 12:37:10 +0000 (14:37 +0200)
To work around a misbehavior of the compiler's ability to see into
composite flexible array structs (as detailed in the coming memcpy()
hardening series[1]), split the memcpy() of the header and the payload
so no false positive run-time overflow warning will be generated.

[1] https://lore.kernel.org/linux-hardening/20220901065914.1417829-2-keescook@chromium.org/

Cc: Evgeniy Polyakov <zbr@ioremap.net>
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Link: https://lore.kernel.org/r/20220927003927.1942170-1-keescook@chromium.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/w1/w1_netlink.c

index fa490aa..db110cc 100644 (file)
@@ -611,7 +611,8 @@ static void w1_cn_callback(struct cn_msg *cn, struct netlink_skb_parms *nsp)
                }
                atomic_set(&block->refcnt, 1);
                block->portid = nsp->portid;
-               memcpy(&block->request_cn, cn, sizeof(*cn) + cn->len);
+               block->request_cn = *cn;
+               memcpy(block->request_cn.data, cn->data, cn->len);
                node = (struct w1_cb_node *)(block->request_cn.data + cn->len);
 
                /* Sneeky, when not bundling, reply_size is the allocated space