netdev - [PATCH net-next] tls: cap the output scatter list to something reasonable

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]

Date:   Wed,  2 Feb 2022 14:20:31 -0800
From:   Jakub Kicinski <kuba@...nel.org>
To:     davem@...emloft.net
Cc:     netdev@...r.kernel.org, viro@...iv.linux.org.uk, borisp@...dia.com,
        john.fastabend@...il.com, daniel@...earbox.net,
        vfedorenko@...ek.ru, kernel-team@...com, axboe@...nel.dk,
        Jakub Kicinski <kuba@...nel.org>
Subject: [PATCH net-next] tls: cap the output scatter list to something reasonable

TLS recvmsg() passes user pages as destination for decrypt.
The decrypt operation is repeated record by record, each
record being 16kB, max. TLS allocates an sg_table and uses
iov_iter_get_pages() to populate it with enough pages to
fit the decrypted record.

Even though we decrypt a single message at a time we size
the sg_table based on the entire length of the iovec.
This leads to unnecessarily large allocations, risking
triggering OOM conditions.

Use iov_iter_truncate() / iov_iter_reexpand() to construct
a "capped" version of iov_iter_npages(). Alternatively we
could parametrize iov_iter_npages() to take the size as
arg instead of using i->count, or do something else..

Signed-off-by: Jakub Kicinski <kuba@...nel.org>
---
 include/linux/uio.h | 17 +++++++++++++++++
 net/tls/tls_sw.c    |  3 ++-
 2 files changed, 19 insertions(+), 1 deletion(-)

diff --git a/include/linux/uio.h b/include/linux/uio.h
index 1198a2bfc9bf..739285fe5a2f 100644
--- a/include/linux/uio.h
+++ b/include/linux/uio.h
@@ -273,6 +273,23 @@ static inline void iov_iter_reexpand(struct iov_iter *i, size_t count)
 	i->count = count;
 }
 
+static inline int
+iov_iter_npages_cap(struct iov_iter *i, int maxpages, size_t max_bytes)
+{
+	size_t shorted = 0;
+	int npages;
+
+	if (iov_iter_count(i) > max_bytes) {
+		shorted = iov_iter_count(i) - max_bytes;
+		iov_iter_truncate(i, max_bytes);
+	}
+	npages = iov_iter_npages(i, INT_MAX);
+	if (shorted)
+		iov_iter_reexpand(i, iov_iter_count(i) + shorted);
+
+	return npages;
+}
+
 struct csum_state {
 	__wsum csum;
 	size_t off;
diff --git a/net/tls/tls_sw.c b/net/tls/tls_sw.c
index efc84845bb6b..0024a692f0f8 100644
--- a/net/tls/tls_sw.c
+++ b/net/tls/tls_sw.c
@@ -1433,7 +1433,8 @@ static int decrypt_internal(struct sock *sk, struct sk_buff *skb,
 
 	if (*zc && (out_iov || out_sg)) {
 		if (out_iov)
-			n_sgout = iov_iter_npages(out_iov, INT_MAX) + 1;
+			n_sgout = 1 +
+				iov_iter_npages_cap(out_iov, INT_MAX, data_len);
 		else
 			n_sgout = sg_nents(out_sg);
 		n_sgin = skb_nsg(skb, rxm->offset + prot->prepend_size,
-- 
2.34.1