The function vmdk_read_cid() can fail if the read on the underlying
block device fails, or if there's a format error in the VMDK file.
However its API doesn't provide a mechanism to report these errors,
and in some cases we were returning a CID of 0 and in some cases a
CID of 0xffffffff, either of which might potentially be valid values.
Change the function to return 0 on success or a negative errno, and
return the CID via a uint32_t* argument. Update the callsites to
handle and propagate the error appropriately.
This fixes in passing a Coverity-spotted issue (CID 1350038) where
we weren't checking the return value from sscanf().
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
block/vmdk.c | 44 ++++++++++++++++++++++++++++++++------------
1 file changed, 32 insertions(+), 12 deletions(-)
diff --git a/block/vmdk.c b/block/vmdk.c
index 55581b03fe..0c9949fc0c 100644
--- a/block/vmdk.c
+++ b/block/vmdk.c
@@ -242,10 +242,11 @@ static void vmdk_free_last_extent(BlockDriverState *bs)
s->extents = g_renew(VmdkExtent, s->extents, s->num_extents);
}
-static uint32_t vmdk_read_cid(BlockDriverState *bs, int parent)
+/* Return -ve errno, or 0 on success and write CID into *pcid. */
+static int vmdk_read_cid(BlockDriverState *bs, int parent, uint32_t *pcid)
{
char *desc;
- uint32_t cid = 0xffffffff;
+ uint32_t cid;
const char *p_name, *cid_str;
size_t cid_str_size;
BDRVVmdkState *s = bs->opaque;
@@ -254,8 +255,7 @@ static uint32_t vmdk_read_cid(BlockDriverState *bs, int parent)
desc = g_malloc0(DESC_SIZE);
ret = bdrv_pread(bs->file, s->desc_offset, desc, DESC_SIZE);
if (ret < 0) {
- g_free(desc);
- return 0;
+ goto out;
}
if (parent) {
@@ -268,13 +268,21 @@ static uint32_t vmdk_read_cid(BlockDriverState *bs, int parent)
desc[DESC_SIZE - 1] = '\0';
p_name = strstr(desc, cid_str);
- if (p_name != NULL) {
- p_name += cid_str_size;
- sscanf(p_name, "%" SCNx32, &cid);
+ if (p_name == NULL) {
+ ret = -EINVAL;
+ goto out;
}
+ p_name += cid_str_size;
+ if (sscanf(p_name, "%" SCNx32, &cid) != 1) {
+ ret = -EINVAL;
+ goto out;
+ }
+ *pcid = cid;
+ ret = 0;
+out:
g_free(desc);
- return cid;
+ return ret;
}
static int vmdk_write_cid(BlockDriverState *bs, uint32_t cid)
@@ -322,7 +330,10 @@ static int vmdk_is_cid_valid(BlockDriverState *bs)
if (!s->cid_checked && bs->backing) {
BlockDriverState *p_bs = bs->backing->bs;
- cur_pcid = vmdk_read_cid(p_bs, 0);
+ if (vmdk_read_cid(p_bs, 0, &cur_pcid) != 0) {
+ /* read failure: report as not valid */
+ return 0;
+ }
if (s->parent_cid != cur_pcid) {
/* CID not valid */
return 0;
@@ -975,8 +986,14 @@ static int vmdk_open(BlockDriverState *bs, QDict *options, int flags,
if (ret) {
goto fail;
}
- s->cid = vmdk_read_cid(bs, 0);
- s->parent_cid = vmdk_read_cid(bs, 1);
+ ret = vmdk_read_cid(bs, 0, &s->cid);
+ if (ret) {
+ goto fail;
+ }
+ ret = vmdk_read_cid(bs, 1, &s->parent_cid);
+ if (ret) {
+ goto fail;
+ }
qemu_co_mutex_init(&s->lock);
/* Disable migration when VMDK images are used */
@@ -2007,8 +2024,11 @@ static int vmdk_create(const char *filename, QemuOpts *opts, Error **errp)
ret = -EINVAL;
goto exit;
}
- parent_cid = vmdk_read_cid(blk_bs(blk), 0);
+ ret = vmdk_read_cid(blk_bs(blk), 0, &parent_cid);
blk_unref(blk);
+ if (ret) {
+ goto exit;
+ }
snprintf(parent_desc_line, BUF_SIZE,
"parentFileNameHint=\"%s\"", backing_file);
}
--
2.11.0
On Sun, 07/09 18:06, Peter Maydell wrote: > The function vmdk_read_cid() can fail if the read on the underlying > block device fails, or if there's a format error in the VMDK file. > However its API doesn't provide a mechanism to report these errors, > and in some cases we were returning a CID of 0 and in some cases a > CID of 0xffffffff, either of which might potentially be valid values. > > Change the function to return 0 on success or a negative errno, and > return the CID via a uint32_t* argument. Update the callsites to > handle and propagate the error appropriately. > > This fixes in passing a Coverity-spotted issue (CID 1350038) where > we weren't checking the return value from sscanf(). > > Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Fam Zheng <famz@redhat.com>
Am 10.07.2017 um 03:58 hat Fam Zheng geschrieben: > On Sun, 07/09 18:06, Peter Maydell wrote: > > The function vmdk_read_cid() can fail if the read on the underlying > > block device fails, or if there's a format error in the VMDK file. > > However its API doesn't provide a mechanism to report these errors, > > and in some cases we were returning a CID of 0 and in some cases a > > CID of 0xffffffff, either of which might potentially be valid values. > > > > Change the function to return 0 on success or a negative errno, and > > return the CID via a uint32_t* argument. Update the callsites to > > handle and propagate the error appropriately. > > > > This fixes in passing a Coverity-spotted issue (CID 1350038) where > > we weren't checking the return value from sscanf(). > > > > Signed-off-by: Peter Maydell <peter.maydell@linaro.org> > > Reviewed-by: Fam Zheng <famz@redhat.com> Thanks, applied to the block branch. Kevin
Am 09.07.2017 um 19:06 hat Peter Maydell geschrieben: > The function vmdk_read_cid() can fail if the read on the underlying > block device fails, or if there's a format error in the VMDK file. > However its API doesn't provide a mechanism to report these errors, > and in some cases we were returning a CID of 0 and in some cases a > CID of 0xffffffff, either of which might potentially be valid values. > > Change the function to return 0 on success or a negative errno, and > return the CID via a uint32_t* argument. Update the callsites to > handle and propagate the error appropriately. > > This fixes in passing a Coverity-spotted issue (CID 1350038) where > we weren't checking the return value from sscanf(). > > Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Fam, this is the commit that introduced the qemu-iotests 059 failure for vmdk. I think what's happening is that we use an image produced by a fuzzer, and with the additional checks introduced in this patch, we now fail earlier and don't test the condition any more that we wanted to test. So do we need a new version of sample_images/afl9.vmdk.bz2 that has a valid CID? Kevin
On 28/07/2017 14:54, Kevin Wolf wrote: > Am 09.07.2017 um 19:06 hat Peter Maydell geschrieben: >> The function vmdk_read_cid() can fail if the read on the underlying >> block device fails, or if there's a format error in the VMDK file. >> However its API doesn't provide a mechanism to report these errors, >> and in some cases we were returning a CID of 0 and in some cases a >> CID of 0xffffffff, either of which might potentially be valid values. >> >> Change the function to return 0 on success or a negative errno, and >> return the CID via a uint32_t* argument. Update the callsites to >> handle and propagate the error appropriately. >> >> This fixes in passing a Coverity-spotted issue (CID 1350038) where >> we weren't checking the return value from sscanf(). >> >> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> > > Fam, this is the commit that introduced the qemu-iotests 059 failure for > vmdk. I think what's happening is that we use an image produced by a > fuzzer, and with the additional checks introduced in this patch, we now > fail earlier and don't test the condition any more that we wanted to > test. > > So do we need a new version of sample_images/afl9.vmdk.bz2 that has a > valid CID? This has never been fixed, has it? I still see the failure. Paolo
On Fri, Jan 19, 2018 at 7:35 PM, Paolo Bonzini <pbonzini@redhat.com> wrote: > On 28/07/2017 14:54, Kevin Wolf wrote: >> Am 09.07.2017 um 19:06 hat Peter Maydell geschrieben: >>> The function vmdk_read_cid() can fail if the read on the underlying >>> block device fails, or if there's a format error in the VMDK file. >>> However its API doesn't provide a mechanism to report these errors, >>> and in some cases we were returning a CID of 0 and in some cases a >>> CID of 0xffffffff, either of which might potentially be valid values. >>> >>> Change the function to return 0 on success or a negative errno, and >>> return the CID via a uint32_t* argument. Update the callsites to >>> handle and propagate the error appropriately. >>> >>> This fixes in passing a Coverity-spotted issue (CID 1350038) where >>> we weren't checking the return value from sscanf(). >>> >>> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> >> >> Fam, this is the commit that introduced the qemu-iotests 059 failure for >> vmdk. I think what's happening is that we use an image produced by a >> fuzzer, and with the additional checks introduced in this patch, we now >> fail earlier and don't test the condition any more that we wanted to >> test. >> >> So do we need a new version of sample_images/afl9.vmdk.bz2 that has a >> valid CID? I'll send a patch to fix the CID today. Fam > > This has never been fixed, has it? I still see the failure. > > Paolo
© 2016 - 2025 Red Hat, Inc.