| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Wed, 30 Jan 2019 20:40:23 +0100
From: Greg KH <gregkh@...uxfoundation.org>
To: Vincent Whitchurch <vincent.whitchurch@...s.com>
Cc: arnd@...db.de, sudeep.dutt@...el.com, ashutosh.dixit@...el.com,
linux-kernel@...r.kernel.org,
virtualization@...ts.linux-foundation.org, tiwei.bie@...el.com,
luto@...nel.org, Vincent Whitchurch <rabinv@...s.com>
Subject: Re: [PATCH v2] mic: vop: Fix broken virtqueues
On Wed, Jan 30, 2019 at 05:28:09PM +0100, Vincent Whitchurch wrote:
> VOP is broken in mainline since commit 1ce9e6055fa0a9043 ("virtio_ring:
> introduce packed ring support"); attempting to use the virtqueues leads
> to various kernel crashes. I'm testing it with my not-yet-merged
> loopback patches, but even the in-tree MIC hardware cannot work.
>
> The problem is not in the referenced commit per se, but is due to the
> following hack in vop_find_vq() which depends on the layout of private
> structures in other source files, which that commit happened to change:
>
> /*
> * To reassign the used ring here we are directly accessing
> * struct vring_virtqueue which is a private data structure
> * in virtio_ring.c. At the minimum, a BUILD_BUG_ON() in
> * vring_new_virtqueue() would ensure that
> * (&vq->vring == (struct vring *) (&vq->vq + 1));
> */
> vr = (struct vring *)(vq + 1);
> vr->used = used;
>
> Fix vop by using __vring_new_virtqueue() to create the needed vring
> layout from the start, instead of attempting to patch in the used ring
> later. __vring_new_virtqueue() was added way back in commit
> 2a2d1382fe9dcc ("virtio: Add improved queue allocation API") in order to
> address mic's usecase, according to the commit message.
>
> The removal patch also has the hack and we need to save the used ring
> pointer separately to remove it.
>
> Fixes: 1ce9e6055fa0 ("virtio_ring: introduce packed ring support")
> Signed-off-by: Vincent Whitchurch <vincent.whitchurch@...s.com>
> ---
> v2: fix removal path also
>
> drivers/misc/mic/vop/vop_main.c | 69 +++++++++++++++++++--------------
> 1 file changed, 40 insertions(+), 29 deletions(-)
>
> diff --git a/drivers/misc/mic/vop/vop_main.c b/drivers/misc/mic/vop/vop_main.c
> index b60b5ec4a28a..c4517703dc1d 100644
> --- a/drivers/misc/mic/vop/vop_main.c
> +++ b/drivers/misc/mic/vop/vop_main.c
> @@ -48,7 +48,8 @@
> * @dc: Virtio device control
> * @vpdev: VOP device which is the parent for this virtio device
> * @vr: Buffer for accessing the VRING
> - * @used: Buffer for used
> + * @used_virt: Virtual address of used ring
> + * @used: DMA address of used ring
> * @used_size: Size of the used buffer
> * @reset_done: Track whether VOP reset is complete
> * @virtio_cookie: Cookie returned upon requesting a interrupt
> @@ -62,6 +63,7 @@ struct _vop_vdev {
> struct mic_device_ctrl __iomem *dc;
> struct vop_device *vpdev;
> void __iomem *vr[VOP_MAX_VRINGS];
> + void *used_virt[VOP_MAX_VRINGS];
> dma_addr_t used[VOP_MAX_VRINGS];
> int used_size[VOP_MAX_VRINGS];
> struct completion reset_done;
> @@ -261,12 +263,12 @@ static bool vop_notify(struct virtqueue *vq)
> static void vop_del_vq(struct virtqueue *vq, int n)
> {
> struct _vop_vdev *vdev = to_vopvdev(vq->vdev);
> - struct vring *vr = (struct vring *)(vq + 1);
> struct vop_device *vpdev = vdev->vpdev;
>
> dma_unmap_single(&vpdev->dev, vdev->used[n],
> vdev->used_size[n], DMA_BIDIRECTIONAL);
> - free_pages((unsigned long)vr->used, get_order(vdev->used_size[n]));
> + free_pages((unsigned long)vdev->used_virt[n],
> + get_order(vdev->used_size[n]));
> vring_del_virtqueue(vq);
> vpdev->hw_ops->unmap(vpdev, vdev->vr[n]);
> vdev->vr[n] = NULL;
> @@ -284,6 +286,26 @@ static void vop_del_vqs(struct virtio_device *dev)
> vop_del_vq(vq, idx++);
> }
>
> +static struct virtqueue *vop_new_virtqueue(unsigned int index,
> + unsigned int num,
> + struct virtio_device *vdev,
> + bool context,
> + void *pages,
> + bool (*notify)(struct virtqueue *vq),
> + void (*callback)(struct virtqueue *vq),
> + const char *name,
> + void *used)
> +{
> + bool weak_barriers = false;
> + struct vring vring;
> +
> + vring_init(&vring, num, pages, MIC_VIRTIO_RING_ALIGN);
> + vring.used = used;
> +
> + return __vring_new_virtqueue(index, vring, vdev, weak_barriers, context,
> + notify, callback, name);
> +}
> +
> /*
> * This routine will assign vring's allocated in host/io memory. Code in
> * virtio_ring.c however continues to access this io memory as if it were local
> @@ -303,7 +325,6 @@ static struct virtqueue *vop_find_vq(struct virtio_device *dev,
> struct _mic_vring_info __iomem *info;
> void *used;
> int vr_size, _vr_size, err, magic;
> - struct vring *vr;
> u8 type = ioread8(&vdev->desc->type);
>
> if (index >= ioread8(&vdev->desc->num_vq))
> @@ -322,17 +343,7 @@ static struct virtqueue *vop_find_vq(struct virtio_device *dev,
> return ERR_PTR(-ENOMEM);
> vdev->vr[index] = va;
> memset_io(va, 0x0, _vr_size);
> - vq = vring_new_virtqueue(
> - index,
> - le16_to_cpu(config.num), MIC_VIRTIO_RING_ALIGN,
> - dev,
> - false,
> - ctx,
> - (void __force *)va, vop_notify, callback, name);
> - if (!vq) {
> - err = -ENOMEM;
> - goto unmap;
> - }
> +
> info = va + _vr_size;
> magic = ioread32(&info->magic);
>
> @@ -341,18 +352,27 @@ static struct virtqueue *vop_find_vq(struct virtio_device *dev,
> goto unmap;
> }
>
> - /* Allocate and reassign used ring now */
> vdev->used_size[index] = PAGE_ALIGN(sizeof(__u16) * 3 +
> sizeof(struct vring_used_elem) *
> le16_to_cpu(config.num));
> used = (void *)__get_free_pages(GFP_KERNEL | __GFP_ZERO,
> get_order(vdev->used_size[index]));
> + vdev->used_virt[index] = used;
> if (!used) {
> err = -ENOMEM;
> dev_err(_vop_dev(vdev), "%s %d err %d\n",
> __func__, __LINE__, err);
> - goto del_vq;
> + goto unmap;
> + }
> +
> + vq = vop_new_virtqueue(index, le16_to_cpu(config.num), dev, ctx,
> + (void __force *)va, vop_notify, callback,
> + name, used);
> + if (!vq) {
> + err = -ENOMEM;
> + goto free_used;
> }
> +
> vdev->used[index] = dma_map_single(&vpdev->dev, used,
> vdev->used_size[index],
> DMA_BIDIRECTIONAL);
> @@ -360,26 +380,17 @@ static struct virtqueue *vop_find_vq(struct virtio_device *dev,
> err = -ENOMEM;
> dev_err(_vop_dev(vdev), "%s %d err %d\n",
> __func__, __LINE__, err);
> - goto free_used;
> + goto del_vq;
> }
> writeq(vdev->used[index], &vqconfig->used_address);
> - /*
> - * To reassign the used ring here we are directly accessing
> - * struct vring_virtqueue which is a private data structure
> - * in virtio_ring.c. At the minimum, a BUILD_BUG_ON() in
> - * vring_new_virtqueue() would ensure that
> - * (&vq->vring == (struct vring *) (&vq->vq + 1));
> - */
> - vr = (struct vring *)(vq + 1);
> - vr->used = used;
>
> vq->priv = vdev;
> return vq;
> +del_vq:
> + vring_del_virtqueue(vq);
> free_used:
> free_pages((unsigned long)used,
> get_order(vdev->used_size[index]));
> -del_vq:
> - vring_del_virtqueue(vq);
> unmap:
> vpdev->hw_ops->unmap(vpdev, vdev->vr[index]);
> return ERR_PTR(err);
> --
> 2.20.0
This patch is already in my tree, can you just send a "fixup" patch
instead of me having to revert the old one and then adding this one
again?
thanks,
greg k-h
Powered by blists - more mailing lists