| Message ID | 20180410052630.11270-2-stefanha@redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Tue, Apr 10, 2018 at 01:26:29PM +0800, Stefan Hajnoczi wrote: > Commit d65026c6c62e7d9616c8ceb5a53b68bcdc050525 ("vhost: validate log > when IOTLB is enabled") introduced a regression. The logic was > originally: > > if (vq->iotlb) > return 1; > return A && B; > > After the patch the short-circuit logic for A was inverted: > > if (A || vq->iotlb) > return A; > return B; > > This patch fixes the regression by rewriting the checks in the obvious > way, no longer returning A when vq->iotlb is non-NULL (which is hard to > understand). > > Reported-by: syzbot+65a84dde0214b0387ccd@syzkaller.appspotmail.com > Cc: Jason Wang <jasowang@redhat.com> > Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> This patch only makes sense after patch 2/2 is applied. Otherwise the logic seems reversed below. Can you pls squash these two? > --- > drivers/vhost/vhost.c | 8 +++++--- > 1 file changed, 5 insertions(+), 3 deletions(-) > > diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c > index 5320039671b7..93fd0c75b0d8 100644 > --- a/drivers/vhost/vhost.c > +++ b/drivers/vhost/vhost.c > @@ -1244,10 +1244,12 @@ static int vq_log_access_ok(struct vhost_virtqueue *vq, > /* Caller should have vq mutex and device mutex */ > int vhost_vq_access_ok(struct vhost_virtqueue *vq) > { > - int ret = vq_log_access_ok(vq, vq->log_base); > + if (!vq_log_access_ok(vq, vq->log_base)) > + return 0; > > - if (ret || vq->iotlb) > - return ret; > + /* Access validation occurs at prefetch time with IOTLB */ > + if (vq->iotlb) > + return 1; > > return vq_access_ok(vq, vq->num, vq->desc, vq->avail, vq->used); > } > -- > 2.14.3
On Tue, Apr 10, 2018 at 04:22:35PM +0300, Michael S. Tsirkin wrote: > On Tue, Apr 10, 2018 at 01:26:29PM +0800, Stefan Hajnoczi wrote: > > Commit d65026c6c62e7d9616c8ceb5a53b68bcdc050525 ("vhost: validate log > > when IOTLB is enabled") introduced a regression. The logic was > > originally: > > > > if (vq->iotlb) > > return 1; > > return A && B; > > > > After the patch the short-circuit logic for A was inverted: > > > > if (A || vq->iotlb) > > return A; > > return B; > > > > This patch fixes the regression by rewriting the checks in the obvious > > way, no longer returning A when vq->iotlb is non-NULL (which is hard to > > understand). > > > > Reported-by: syzbot+65a84dde0214b0387ccd@syzkaller.appspotmail.com > > Cc: Jason Wang <jasowang@redhat.com> > > Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> > > This patch only makes sense after patch 2/2 is applied. > Otherwise the logic seems reversed below. > > Can you pls squash these two? Oh, in fact the patch is ok. This just goes to show that patch 2/2 is useful. But squashing is not required. Sorry about the noise. Acked-by: Michael S. Tsirkin <mst@redhat.com> Fixes: d65026c6c ("vhost: validate log when IOTLB is enabled") probably stable material since patch it fixes was queued to stable? > > --- > > drivers/vhost/vhost.c | 8 +++++--- > > 1 file changed, 5 insertions(+), 3 deletions(-) > > > > diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c > > index 5320039671b7..93fd0c75b0d8 100644 > > --- a/drivers/vhost/vhost.c > > +++ b/drivers/vhost/vhost.c > > @@ -1244,10 +1244,12 @@ static int vq_log_access_ok(struct vhost_virtqueue *vq, > > /* Caller should have vq mutex and device mutex */ > > int vhost_vq_access_ok(struct vhost_virtqueue *vq) > > { > > - int ret = vq_log_access_ok(vq, vq->log_base); > > + if (!vq_log_access_ok(vq, vq->log_base)) > > + return 0; > > > > - if (ret || vq->iotlb) > > - return ret; > > + /* Access validation occurs at prefetch time with IOTLB */ > > + if (vq->iotlb) > > + return 1; > > > > return vq_access_ok(vq, vq->num, vq->desc, vq->avail, vq->used); > > } > > -- > > 2.14.3
diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c index 5320039671b7..93fd0c75b0d8 100644 --- a/drivers/vhost/vhost.c +++ b/drivers/vhost/vhost.c @@ -1244,10 +1244,12 @@ static int vq_log_access_ok(struct vhost_virtqueue *vq, /* Caller should have vq mutex and device mutex */ int vhost_vq_access_ok(struct vhost_virtqueue *vq) { - int ret = vq_log_access_ok(vq, vq->log_base); + if (!vq_log_access_ok(vq, vq->log_base)) + return 0; - if (ret || vq->iotlb) - return ret; + /* Access validation occurs at prefetch time with IOTLB */ + if (vq->iotlb) + return 1; return vq_access_ok(vq, vq->num, vq->desc, vq->avail, vq->used); }
Commit d65026c6c62e7d9616c8ceb5a53b68bcdc050525 ("vhost: validate log when IOTLB is enabled") introduced a regression. The logic was originally: if (vq->iotlb) return 1; return A && B; After the patch the short-circuit logic for A was inverted: if (A || vq->iotlb) return A; return B; This patch fixes the regression by rewriting the checks in the obvious way, no longer returning A when vq->iotlb is non-NULL (which is hard to understand). Reported-by: syzbot+65a84dde0214b0387ccd@syzkaller.appspotmail.com Cc: Jason Wang <jasowang@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> --- drivers/vhost/vhost.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-)