slirp: check pkt_len before reading protocol header

While processing ARP/NCSI packets in 'arp_input' or 'ncsi_input'
routines, ensure that pkt_len is large enough to accommodate the
respective protocol headers, lest it should do an OOB access.
Add check to avoid it.

CVE-2020-29129 CVE-2020-29130
  QEMU: slirp: out-of-bounds access while processing ARP/NCSI packets
 -> https://www.openwall.com/lists/oss-security/2020/11/27/1

Reported-by: Qiuhao Li <Qiuhao.Li@outlook.com>
Signed-off-by: Prasad J Pandit <pjp@fedoraproject.org>
Message-Id: <20201126135706.273950-1-ppandit@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
(cherry-picked from 2e1dcbc0)
Signed-off-by: Alex Chen <alex.chen@huawei.com>
This commit is contained in:
Alex Chen 2020-12-14 18:00:31 +08:00 committed by Ying Fang
parent ff04bde967
commit abe003640c
2 changed files with 66 additions and 1 deletions

View File

@ -1,6 +1,6 @@
Name: qemu Name: qemu
Version: 4.1.0 Version: 4.1.0
Release: 34 Release: 35
Epoch: 2 Epoch: 2
Summary: QEMU is a generic and open source machine emulator and virtualizer Summary: QEMU is a generic and open source machine emulator and virtualizer
License: GPLv2 and BSD and MIT and CC-BY License: GPLv2 and BSD and MIT and CC-BY
@ -280,6 +280,7 @@ Patch0267: multifd-tls-fix-memoryleak-of-the-QIOChannelSocket-o.patch
Patch0268: net-remove-an-assert-call-in-eth_get_gso_type.patch Patch0268: net-remove-an-assert-call-in-eth_get_gso_type.patch
Patch0269: json-Fix-a-memleak-in-parse_pair.patch Patch0269: json-Fix-a-memleak-in-parse_pair.patch
Patch0270: Bugfix-hw-acpi-Use-max_cpus-instead-of-cpus-when-bui.patch Patch0270: Bugfix-hw-acpi-Use-max_cpus-instead-of-cpus-when-bui.patch
Patch0271: slirp-check-pkt_len-before-reading-protocol-header.patch
BuildRequires: flex BuildRequires: flex
BuildRequires: bison BuildRequires: bison
@ -626,6 +627,9 @@ getent passwd qemu >/dev/null || \
%endif %endif
%changelog %changelog
* Fri Dec 11 2020 Huawei Technologies Co., Ltd <alex.chen@huawei.com>
- slirp: check pkt_len before reading protocol header for fixing CVE-2020-29129 and CVE-2020-29130
* Thu Oct 29 2020 Huawei Technologies Co., Ltd <alex.chen@huawei.com> * Thu Oct 29 2020 Huawei Technologies Co., Ltd <alex.chen@huawei.com>
- Bugfix: hw/acpi: Use max_cpus instead of cpus when build PPTT table - Bugfix: hw/acpi: Use max_cpus instead of cpus when build PPTT table

View File

@ -0,0 +1,61 @@
From c2df0d478b2605da10363ab57825cdbc34caa680 Mon Sep 17 00:00:00 2001
From: Alex Chen <alex.chen@huawei.com>
Date: Mon, 14 Dec 2020 15:39:46 +0800
Subject: [PATCH] slirp: check pkt_len before reading protocol header
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
While processing ARP/NCSI packets in 'arp_input' or 'ncsi_input'
routines, ensure that pkt_len is large enough to accommodate the
respective protocol headers, lest it should do an OOB access.
Add check to avoid it.
CVE-2020-29129 CVE-2020-29130
QEMU: slirp: out-of-bounds access while processing ARP/NCSI packets
-> https://www.openwall.com/lists/oss-security/2020/11/27/1
Reported-by: Qiuhao Li <Qiuhao.Li@outlook.com>
Signed-off-by: Prasad J Pandit <pjp@fedoraproject.org>
Message-Id: <20201126135706.273950-1-ppandit@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
(cherry-picked from 2e1dcbc0)
Signed-off-by: Alex Chen <alex.chen@huawei.com>
---
slirp/src/ncsi.c | 4 ++++
slirp/src/slirp.c | 4 ++++
2 files changed, 8 insertions(+)
diff --git a/slirp/src/ncsi.c b/slirp/src/ncsi.c
index 6864b735..251c0d2b 100644
--- a/slirp/src/ncsi.c
+++ b/slirp/src/ncsi.c
@@ -147,6 +147,10 @@ void ncsi_input(Slirp *slirp, const uint8_t *pkt, int pkt_len)
uint32_t checksum;
uint32_t *pchecksum;
+ if (pkt_len < ETH_HLEN + sizeof(struct ncsi_pkt_hdr)) {
+ return; /* packet too short */
+ }
+
memset(ncsi_reply, 0, sizeof(ncsi_reply));
memset(reh->h_dest, 0xff, ETH_ALEN);
diff --git a/slirp/src/slirp.c b/slirp/src/slirp.c
index b0194cb3..86b0f52d 100644
--- a/slirp/src/slirp.c
+++ b/slirp/src/slirp.c
@@ -700,6 +700,10 @@ static void arp_input(Slirp *slirp, const uint8_t *pkt, int pkt_len)
return;
}
+ if (pkt_len < ETH_HLEN + sizeof(struct slirp_arphdr)) {
+ return; /* packet too short */
+ }
+
ar_op = ntohs(ah->ar_op);
switch (ar_op) {
case ARPOP_REQUEST:
--
2.23.0