* [Qemu-devel] [PATCH v2] qtest: virtio: zeroed last vring_desc after allocate
@ 2017-01-17 2:59 Longpeng(Mike)
2017-01-17 10:55 ` Stefan Hajnoczi
0 siblings, 1 reply; 2+ messages in thread
From: Longpeng(Mike) @ 2017-01-17 2:59 UTC (permalink / raw)
To: stefanha, lvivier, eblake, peter.maydell, groug
Cc: arei.gonglei, qemu-devel, Longpeng(Mike)
As qvring_indirect_desc_setup() wouldn't initialize last vring_desc,
so it's filled with dirty data, this might cause virtio backend broken.
For example, the last bit of this vring_desc's flags might be 1, so
virtqueue_read_next_desc() would report "Desc next is ***".
Signed-off-by: Longpeng(Mike) <longpeng2@huawei.com>
---
Changes since v1:
- use writel() for @len as it's 32 bits long. [Stefan]
---
tests/libqos/virtio.c | 8 ++++++++
1 file changed, 8 insertions(+)
diff --git a/tests/libqos/virtio.c b/tests/libqos/virtio.c
index ec30cb9..02e6a43 100644
--- a/tests/libqos/virtio.c
+++ b/tests/libqos/virtio.c
@@ -171,12 +171,20 @@ QVRingIndirectDesc *qvring_indirect_desc_setup(QVirtioDevice *d,
for (i = 0; i < elem - 1; ++i) {
/* indirect->desc[i].addr */
writeq(indirect->desc + (16 * i), 0);
+ /* indirect->desc[i].len */
+ writel(indirect->desc + (16 * i) + 8, 0);
/* indirect->desc[i].flags */
writew(indirect->desc + (16 * i) + 12, VRING_DESC_F_NEXT);
/* indirect->desc[i].next */
writew(indirect->desc + (16 * i) + 14, i + 1);
}
+ /* zeroed last element */
+ writeq(indirect->desc + (16 * i), 0); /* addr */
+ writel(indirect->desc + (16 * i) + 8, 0); /* len */
+ writew(indirect->desc + (16 * i) + 12, 0); /* flags */
+ writew(indirect->desc + (16 * i) + 14, 0); /* next */
+
return indirect;
}
--
1.8.3.1
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [Qemu-devel] [PATCH v2] qtest: virtio: zeroed last vring_desc after allocate
2017-01-17 2:59 [Qemu-devel] [PATCH v2] qtest: virtio: zeroed last vring_desc after allocate Longpeng(Mike)
@ 2017-01-17 10:55 ` Stefan Hajnoczi
0 siblings, 0 replies; 2+ messages in thread
From: Stefan Hajnoczi @ 2017-01-17 10:55 UTC (permalink / raw)
To: Longpeng(Mike)
Cc: stefanha, lvivier, eblake, peter.maydell, groug, arei.gonglei,
qemu-devel
[-- Attachment #1: Type: text/plain, Size: 695 bytes --]
On Tue, Jan 17, 2017 at 10:59:24AM +0800, Longpeng(Mike) wrote:
> As qvring_indirect_desc_setup() wouldn't initialize last vring_desc,
> so it's filled with dirty data, this might cause virtio backend broken.
>
> For example, the last bit of this vring_desc's flags might be 1, so
> virtqueue_read_next_desc() would report "Desc next is ***".
>
> Signed-off-by: Longpeng(Mike) <longpeng2@huawei.com>
>
> ---
> Changes since v1:
> - use writel() for @len as it's 32 bits long. [Stefan]
>
> ---
> tests/libqos/virtio.c | 8 ++++++++
> 1 file changed, 8 insertions(+)
Thanks, applied to my block-next tree:
https://github.com/stefanha/qemu/commits/block-next
Stefan
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 455 bytes --]
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2017-01-17 10:55 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2017-01-17 2:59 [Qemu-devel] [PATCH v2] qtest: virtio: zeroed last vring_desc after allocate Longpeng(Mike)
2017-01-17 10:55 ` Stefan Hajnoczi
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).