On 6/21/23 10:20, Alexander Ivanov wrote:
Initially data_end is set to the data_off image header field and must not be
greater than the file size.
Signed-off-by: Alexander Ivanov <alexander.iva...@virtuozzo.com>
---
block/parallels.c | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/block/parallels.c b/block/parallels.c
index 1ec98c722b..4b7eafba1e 100644
--- a/block/parallels.c
+++ b/block/parallels.c
@@ -868,6 +868,11 @@ static int parallels_open(BlockDriverState *bs, QDict
*options, int flags,
and actual data. We can't avoid read-modify-write... */
s->header_size = size;
}
+ if (s->data_end > file_nb_sectors) {
+ error_setg(errp, "Invalid image: incorrect data_off field");
+ ret = -EINVAL;
+ goto fail;
+ }
ret = bdrv_pread(bs->file, 0, s->header_size, s->header, 0);
if (ret < 0) {
for me this seems incorrect. In this case we would not be able to EVER
open such image, even for a check purpose.
This place should have a clause like
if (!(flags & BDRV_O_CHECK)) {
goto fail
}
This error is not fatal from the check point of view.
Den
