169b9a
From 9ea892729b9c77eaf7b923da8a3e370dbb022e6e Mon Sep 17 00:00:00 2001
6e672b
From: Kevin Wolf <kwolf@redhat.com>
6e672b
Date: Thu, 26 Jul 2018 16:24:48 +0200
6e672b
Subject: [PATCH 6/8] multiboot: Check validity of mh_header_addr
6e672b
6e672b
RH-Author: Kevin Wolf <kwolf@redhat.com>
6e672b
Message-id: <20180726162448.22072-7-kwolf@redhat.com>
6e672b
Patchwork-id: 81515
6e672b
O-Subject: [RHEL-7.6/7.5.z qemu-kvm PATCH 6/6] multiboot: Check validity of mh_header_addr
169b9a
Bugzilla: 1549822
6e672b
RH-Acked-by: John Snow <jsnow@redhat.com>
6e672b
RH-Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
6e672b
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
6e672b
6e672b
I couldn't find a case where this prevents something bad from happening
6e672b
that isn't already caught by other checks, but let's err on the safe
6e672b
side and check that mh_header_addr is as expected.
6e672b
6e672b
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
6e672b
Reviewed-by: Jack Schwartz <jack.schwartz@oracle.com>
6e672b
(cherry picked from commit dbf2dce7aabb7723542bd182175904846d70b0f9)
6e672b
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
6e672b
Signed-off-by: Miroslav Rezanina <mrezanin@redhat.com>
6e672b
---
6e672b
 hw/i386/multiboot.c | 4 ++++
6e672b
 1 file changed, 4 insertions(+)
6e672b
6e672b
diff --git a/hw/i386/multiboot.c b/hw/i386/multiboot.c
6e672b
index 6f489a5..e1f6e56 100644
6e672b
--- a/hw/i386/multiboot.c
6e672b
+++ b/hw/i386/multiboot.c
6e672b
@@ -208,6 +208,10 @@ int load_multiboot(FWCfgState *fw_cfg,
6e672b
             error_report("invalid load_addr address");
6e672b
             exit(1);
6e672b
         }
6e672b
+        if (mh_header_addr - mh_load_addr > i) {
6e672b
+            error_report("invalid header_addr address");
6e672b
+            exit(1);
6e672b
+        }
6e672b
 
6e672b
         uint32_t mb_kernel_text_offset = i - (mh_header_addr - mh_load_addr);
6e672b
         uint32_t mb_load_size = 0;
6e672b
-- 
6e672b
1.8.3.1
6e672b