Re: vhost changes (batched) in linux-next after 12/13 trigger random crashes in KVM guests after reboot

From: Christian Borntraeger
Date: Fri Feb 14 2020 - 02:34:03 EST


I did
ping -c 20 -f ... ; reboot
twice

The ping after the first reboot showed .......E

this was on the host console

[ 55.951885] CPU: 34 PID: 1908 Comm: CPU 0/KVM Not tainted 5.5.0+ #21
[ 55.951891] Hardware name: IBM 3906 M04 704 (LPAR)
[ 55.951892] Call Trace:
[ 55.951902] [<0000001ede114132>] show_stack+0x8a/0xd0
[ 55.951906] [<0000001edeb0672a>] dump_stack+0x8a/0xb8
[ 55.951915] [<000003ff803736a6>] vhost_vring_ioctl+0x6fe/0x858 [vhost]
[ 55.951919] [<000003ff8042a608>] vhost_net_ioctl+0x510/0x570 [vhost_net]
[ 55.951924] [<0000001ede3c4dd8>] do_vfs_ioctl+0x430/0x6f8
[ 55.951926] [<0000001ede3c5124>] ksys_ioctl+0x84/0xb0
[ 55.951927] [<0000001ede3c51ba>] __s390x_sys_ioctl+0x2a/0x38
[ 55.951931] [<0000001edeb27f72>] system_call+0x2a6/0x2c8
[ 55.951949] CPU: 34 PID: 1908 Comm: CPU 0/KVM Not tainted 5.5.0+ #21
[ 55.951950] Hardware name: IBM 3906 M04 704 (LPAR)
[ 55.951951] Call Trace:
[ 55.951952] [<0000001ede114132>] show_stack+0x8a/0xd0
[ 55.951954] [<0000001edeb0672a>] dump_stack+0x8a/0xb8
[ 55.951956] [<000003ff803736a6>] vhost_vring_ioctl+0x6fe/0x858 [vhost]
[ 55.951958] [<000003ff8042a608>] vhost_net_ioctl+0x510/0x570 [vhost_net]
[ 55.951959] [<0000001ede3c4dd8>] do_vfs_ioctl+0x430/0x6f8
[ 55.951961] [<0000001ede3c5124>] ksys_ioctl+0x84/0xb0
[ 55.951962] [<0000001ede3c51ba>] __s390x_sys_ioctl+0x2a/0x38
[ 55.951964] [<0000001edeb27f72>] system_call+0x2a6/0x2c8
[ 55.951997] Guest moved vq 0000000063d896c6 used index from 44 to 0
[ 56.609831] unexpected descriptor format for RX: out 0, in 0
[ 86.540460] CPU: 6 PID: 1908 Comm: CPU 0/KVM Not tainted 5.5.0+ #21
[ 86.540464] Hardware name: IBM 3906 M04 704 (LPAR)
[ 86.540466] Call Trace:
[ 86.540473] [<0000001ede114132>] show_stack+0x8a/0xd0
[ 86.540477] [<0000001edeb0672a>] dump_stack+0x8a/0xb8
[ 86.540486] [<000003ff803736a6>] vhost_vring_ioctl+0x6fe/0x858 [vhost]
[ 86.540490] [<000003ff8042a608>] vhost_net_ioctl+0x510/0x570 [vhost_net]
[ 86.540494] [<0000001ede3c4dd8>] do_vfs_ioctl+0x430/0x6f8
[ 86.540496] [<0000001ede3c5124>] ksys_ioctl+0x84/0xb0
[ 86.540498] [<0000001ede3c51ba>] __s390x_sys_ioctl+0x2a/0x38
[ 86.540501] [<0000001edeb27f72>] system_call+0x2a6/0x2c8
[ 86.540524] CPU: 6 PID: 1908 Comm: CPU 0/KVM Not tainted 5.5.0+ #21
[ 86.540525] Hardware name: IBM 3906 M04 704 (LPAR)
[ 86.540526] Call Trace:
[ 86.540527] [<0000001ede114132>] show_stack+0x8a/0xd0
[ 86.540528] [<0000001edeb0672a>] dump_stack+0x8a/0xb8
[ 86.540531] [<000003ff803736a6>] vhost_vring_ioctl+0x6fe/0x858 [vhost]
[ 86.540532] [<000003ff8042a608>] vhost_net_ioctl+0x510/0x570 [vhost_net]
[ 86.540534] [<0000001ede3c4dd8>] do_vfs_ioctl+0x430/0x6f8
[ 86.540536] [<0000001ede3c5124>] ksys_ioctl+0x84/0xb0
[ 86.540537] [<0000001ede3c51ba>] __s390x_sys_ioctl+0x2a/0x38
[ 86.540538] [<0000001edeb27f72>] system_call+0x2a6/0x2c8
[ 86.540570] unexpected descriptor format for RX: out 0, in 0
[ 86.540577] Unexpected header len for TX: 0 expected 0


On 14.02.20 08:06, Eugenio PÃrez wrote:
> Hi Christian.
>
> Sorry, that was meant to be applied over previous debug patch.
>
> Here I inline the one meant to be applied over eccb852f1fe6bede630e2e4f1a121a81e34354ab.
>
> Thanks!
>
> From d978ace99e4844b49b794d768385db3d128a4cc0 Mon Sep 17 00:00:00 2001
> From: =?UTF-8?q?Eugenio=20P=C3=A9rez?= <eperezma@xxxxxxxxxx>
> Date: Fri, 14 Feb 2020 08:02:26 +0100
> Subject: [PATCH] vhost: disable all features and trace last_avail_idx and
> ioctl calls
>
> ---
> drivers/vhost/net.c | 20 +++++++++++++++++---
> drivers/vhost/vhost.c | 25 +++++++++++++++++++++++--
> drivers/vhost/vhost.h | 10 +++++-----
> 3 files changed, 45 insertions(+), 10 deletions(-)
>
> diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
> index e158159671fa..e4d5f843f9c0 100644
> --- a/drivers/vhost/net.c
> +++ b/drivers/vhost/net.c
> @@ -1505,10 +1505,13 @@ static long vhost_net_set_backend(struct vhost_net *n, unsigned index, int fd)
>
> mutex_lock(&n->dev.mutex);
> r = vhost_dev_check_owner(&n->dev);
> - if (r)
> + if (r) {
> + pr_debug("vhost_dev_check_owner index=%u fd=%d rc r=%d", index, fd, r);
> goto err;
> + }
>
> if (index >= VHOST_NET_VQ_MAX) {
> + pr_debug("vhost_dev_check_owner index=%u fd=%d MAX=%d", index, fd, VHOST_NET_VQ_MAX);
> r = -ENOBUFS;
> goto err;
> }
> @@ -1518,22 +1521,26 @@ static long vhost_net_set_backend(struct vhost_net *n, unsigned index, int fd)
>
> /* Verify that ring has been setup correctly. */
> if (!vhost_vq_access_ok(vq)) {
> + pr_debug("vhost_net_set_backend index=%u fd=%d !vhost_vq_access_ok", index, fd);
> r = -EFAULT;
> goto err_vq;
> }
> sock = get_socket(fd);
> if (IS_ERR(sock)) {
> r = PTR_ERR(sock);
> + pr_debug("vhost_net_set_backend index=%u fd=%d get_socket err r=%d", index, fd, r);
> goto err_vq;
> }
>
> /* start polling new socket */
> oldsock = vq->private_data;
> if (sock != oldsock) {
> + pr_debug("sock=%p != oldsock=%p index=%u fd=%d vq=%p", sock, oldsock, index, fd, vq);
> ubufs = vhost_net_ubuf_alloc(vq,
> sock && vhost_sock_zcopy(sock));
> if (IS_ERR(ubufs)) {
> r = PTR_ERR(ubufs);
> + pr_debug("ubufs index=%u fd=%d err r=%d vq=%p", index, fd, r, vq);
> goto err_ubufs;
> }
>
> @@ -1541,11 +1548,15 @@ static long vhost_net_set_backend(struct vhost_net *n, unsigned index, int fd)
> vq->private_data = sock;
> vhost_net_buf_unproduce(nvq);
> r = vhost_vq_init_access(vq);
> - if (r)
> + if (r) {
> + pr_debug("init_access index=%u fd=%d r=%d vq=%p", index, fd, r, vq);
> goto err_used;
> + }
> r = vhost_net_enable_vq(n, vq);
> - if (r)
> + if (r) {
> + pr_debug("enable_vq index=%u fd=%d r=%d vq=%p", index, fd, r, vq);
> goto err_used;
> + }
> if (index == VHOST_NET_VQ_RX)
> nvq->rx_ring = get_tap_ptr_ring(fd);
>
> @@ -1559,6 +1570,8 @@ static long vhost_net_set_backend(struct vhost_net *n, unsigned index, int fd)
>
> mutex_unlock(&vq->mutex);
>
> + pr_debug("sock=%p", sock);
> +
> if (oldubufs) {
> vhost_net_ubuf_put_wait_and_free(oldubufs);
> mutex_lock(&vq->mutex);
> @@ -1710,6 +1723,7 @@ static long vhost_net_ioctl(struct file *f, unsigned int ioctl,
>
> switch (ioctl) {
> case VHOST_NET_SET_BACKEND:
> + pr_debug("VHOST_NET_SET_BACKEND");
> if (copy_from_user(&backend, argp, sizeof backend))
> return -EFAULT;
> return vhost_net_set_backend(n, backend.index, backend.fd);
> diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c
> index b5a51b1f2e79..ec25ba32fe81 100644
> --- a/drivers/vhost/vhost.c
> +++ b/drivers/vhost/vhost.c
> @@ -1642,15 +1642,30 @@ long vhost_vring_ioctl(struct vhost_dev *d, unsigned int ioctl, void __user *arg
> r = -EINVAL;
> break;
> }
> +
> + if (vq->last_avail_idx || vq->avail_idx) {
> + pr_debug(
> + "strange VHOST_SET_VRING_BASE [vq=%p][s.index=%u][s.num=%u]",
> + vq, s.index, s.num);
> + dump_stack();
> + r = 0;
> + break;
> + }
> vq->last_avail_idx = s.num;
> /* Forget the cached index value. */
> vq->avail_idx = vq->last_avail_idx;
> + pr_debug(
> + "VHOST_SET_VRING_BASE [vq=%p][vq->last_avail_idx=%u][vq->avail_idx=%u][s.index=%u][s.num=%u]",
> + vq, vq->last_avail_idx, vq->avail_idx, s.index, s.num);
> break;
> case VHOST_GET_VRING_BASE:
> s.index = idx;
> s.num = vq->last_avail_idx;
> if (copy_to_user(argp, &s, sizeof s))
> r = -EFAULT;
> + pr_debug(
> + "VHOST_GET_VRING_BASE [vq=%p][vq->last_avail_idx=%u][vq->avail_idx=%u][s.index=%u][s.num=%u]",
> + vq, vq->last_avail_idx, vq->avail_idx, s.index, s.num);
> break;
> case VHOST_SET_VRING_KICK:
> if (copy_from_user(&f, argp, sizeof f)) {
> @@ -2239,8 +2254,8 @@ static int fetch_buf(struct vhost_virtqueue *vq)
> vq->avail_idx = vhost16_to_cpu(vq, avail_idx);
>
> if (unlikely((u16)(vq->avail_idx - last_avail_idx) > vq->num)) {
> - vq_err(vq, "Guest moved used index from %u to %u",
> - last_avail_idx, vq->avail_idx);
> + vq_err(vq, "Guest moved vq %p used index from %u to %u",
> + vq, last_avail_idx, vq->avail_idx);
> return -EFAULT;
> }
>
> @@ -2316,6 +2331,9 @@ static int fetch_buf(struct vhost_virtqueue *vq)
> BUG_ON(!(vq->used_flags & VRING_USED_F_NO_NOTIFY));
>
> /* On success, increment avail index. */
> + pr_debug(
> + "[vq=%p][vq->last_avail_idx=%u][vq->avail_idx=%u][vq->ndescs=%d][vq->first_desc=%d]",
> + vq, vq->last_avail_idx, vq->avail_idx, vq->ndescs, vq->first_desc);
> vq->last_avail_idx++;
>
> return 0;
> @@ -2432,6 +2450,9 @@ EXPORT_SYMBOL_GPL(vhost_get_vq_desc);
> /* Reverse the effect of vhost_get_vq_desc. Useful for error handling. */
> void vhost_discard_vq_desc(struct vhost_virtqueue *vq, int n)
> {
> + pr_debug(
> + "DISCARD [vq=%p][vq->last_avail_idx=%u][vq->avail_idx=%u][n=%d]",
> + vq, vq->last_avail_idx, vq->avail_idx, n);
> vq->last_avail_idx -= n;
> }
> EXPORT_SYMBOL_GPL(vhost_discard_vq_desc);
> diff --git a/drivers/vhost/vhost.h b/drivers/vhost/vhost.h
> index 661088ae6dc7..08f6d2ccb697 100644
> --- a/drivers/vhost/vhost.h
> +++ b/drivers/vhost/vhost.h
> @@ -250,11 +250,11 @@ int vhost_init_device_iotlb(struct vhost_dev *d, bool enabled);
> } while (0)
>
> enum {
> - VHOST_FEATURES = (1ULL << VIRTIO_F_NOTIFY_ON_EMPTY) |
> - (1ULL << VIRTIO_RING_F_INDIRECT_DESC) |
> - (1ULL << VIRTIO_RING_F_EVENT_IDX) |
> - (1ULL << VHOST_F_LOG_ALL) |
> - (1ULL << VIRTIO_F_ANY_LAYOUT) |
> + VHOST_FEATURES = /* (1ULL << VIRTIO_F_NOTIFY_ON_EMPTY) | */
> + /* (1ULL << VIRTIO_RING_F_INDIRECT_DESC) | */
> + /* (1ULL << VIRTIO_RING_F_EVENT_IDX) | */
> + /* (1ULL << VHOST_F_LOG_ALL) | */
> + /* (1ULL << VIRTIO_F_ANY_LAYOUT) | */
> (1ULL << VIRTIO_F_VERSION_1)
> };
>
>