linux-kernel - [PATCH] relay: Fix signedness issues in relay_page_release() and relay_file_splice

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>] [day] [month] [year] [list]

Message-ID: <4B27FF64.6070602@gmail.com>
Date:	Tue, 15 Dec 2009 22:28:04 +0100
From:	Roel Kluin <roel.kluin@...il.com>
To:	Tom Zanussi <tzanussi@...il.com>,
	Andrew Morton <akpm@...ux-foundation.org>,
	LKML <linux-kernel@...r.kernel.org>
Subject: [PATCH] relay: Fix signedness issues in relay_page_release() and
 relay_file_splice_read()

In subbuf_splice_actor() `ret' is unsigned, so the `ret < 0'
test does not work and splice_to_pipe() errors are not acted upon.
It appears the cause is confusion between the signedness of the
return values `ret' and `nonpad_ret'.

Signed-off-by: Roel Kluin <roel.kluin@...il.com>
---
Found using coccinelle: http://coccinelle.lip6.fr/

I tried to solve this in the patch below, A remaining problem may be
that `spliced' wraps, any comments?

The rationale for the patch below:

padding (size_t) is only added to `ret' and not to `*nonpad_ret', so
I think that ret should be unsigned, or even size_t, whereas `*nonpad_ret'
should be int and carry the error.

the caller relay_file_splice_read() should therefore check the error in
`nonpad_ret', which should be int, rather than in `ret' - unsigned.

checkpatch likes it and it appears to build, but maybe there are
comments?

diff --git a/kernel/relay.c b/kernel/relay.c
index 760c262..632fca2 100644
--- a/kernel/relay.c
+++ b/kernel/relay.c
@@ -1215,7 +1215,7 @@ static void relay_page_release(struct splice_pipe_desc *spd, unsigned int i)
 /*
  *	subbuf_splice_actor - splice up to one subbuf's worth of data
  */
-static int subbuf_splice_actor(struct file *in,
+static unsigned int subbuf_splice_actor(struct file *in,
 			       loff_t *ppos,
 			       struct pipe_inode_info *pipe,
 			       size_t len,
@@ -1292,7 +1292,7 @@ static int subbuf_splice_actor(struct file *in,
 		return 0;
 
 	ret = *nonpad_ret = splice_to_pipe(pipe, &spd);
-	if (ret < 0 || ret < total_len)
+	if (*nonpad_ret < 0 || ret < total_len)
 		return ret;
 
         if (read_start + ret == nonpad_end)
@@ -1308,7 +1308,7 @@ static ssize_t relay_file_splice_read(struct file *in,
 				      unsigned int flags)
 {
 	ssize_t spliced;
-	int ret;
+	unsigned ret;
 	int nonpad_ret = 0;
 
 	ret = 0;
@@ -1316,11 +1316,11 @@ static ssize_t relay_file_splice_read(struct file *in,
 
 	while (len && !spliced) {
 		ret = subbuf_splice_actor(in, ppos, pipe, len, flags, &nonpad_ret);
-		if (ret < 0)
+		if (nonpad_ret < 0)
 			break;
 		else if (!ret) {
 			if (flags & SPLICE_F_NONBLOCK)
-				ret = -EAGAIN;
+				nonpad_ret = -EAGAIN;
 			break;
 		}
 
@@ -1336,7 +1336,7 @@ static ssize_t relay_file_splice_read(struct file *in,
 	if (spliced)
 		return spliced;
 
-	return ret;
+	return nonpad_ret;
 }
 
 const struct file_operations relay_file_operations = {
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/