peterdelevoryas / rpms / qemu

Forked from rpms/qemu 2 years ago
Clone

Blame 0214-scsi-disk-fix-check-for-out-of-range-LBA.patch

5544c1
From 2fd88dcb7fdb90508a85c30a32516ff57f081cba Mon Sep 17 00:00:00 2001
5544c1
From: Paolo Bonzini <pbonzini@redhat.com>
5544c1
Date: Wed, 5 Sep 2012 17:54:36 +0200
5544c1
Subject: [PATCH] scsi-disk: fix check for out-of-range LBA
5544c1
5544c1
This fix is needed to correctly handle 0-block read and writes.
5544c1
Without it, a 0-block access at LBA 0 would underflow.
5544c1
5544c1
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
5544c1
(cherry picked from commit 12ca76fc48081b3a0ad1a70546abfcf198aedfc4)
5544c1
5544c1
Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com>
5544c1
---
5544c1
 hw/scsi-disk.c | 6 +++++-
5544c1
 1 file changed, 5 insertions(+), 1 deletion(-)
5544c1
5544c1
diff --git a/hw/scsi-disk.c b/hw/scsi-disk.c
5544c1
index 3959603..d621852 100644
5544c1
--- a/hw/scsi-disk.c
5544c1
+++ b/hw/scsi-disk.c
5544c1
@@ -1456,9 +1456,13 @@ static inline bool check_lba_range(SCSIDiskState *s,
5544c1
      * The first line tests that no overflow happens when computing the last
5544c1
      * sector.  The second line tests that the last accessed sector is in
5544c1
      * range.
5544c1
+     *
5544c1
+     * Careful, the computations should not underflow for nb_sectors == 0,
5544c1
+     * and a 0-block read to the first LBA beyond the end of device is
5544c1
+     * valid.
5544c1
      */
5544c1
     return (sector_num <= sector_num + nb_sectors &&
5544c1
-            sector_num + nb_sectors - 1 <= s->qdev.max_lba);
5544c1
+            sector_num + nb_sectors <= s->qdev.max_lba + 1);
5544c1
 }
5544c1
 
5544c1
 typedef struct UnmapCBData {
5544c1
-- 
5544c1
1.7.12.1
5544c1