vhost: fix OOB in get_rx_bufs()
Subsystems (7)
PATCH 1/1 vhost: fix OOB in get_rx_bufs()
After batched used ring updating was introduced in commit e2b3b35eb989 ("vhost_net: batch used ring update in rx"). We tend to batch heads in vq->heads for more than one packet. But the quota passed to get_rx_bufs() was not correctly limited, which can result a OOB write in vq->heads.
headcount = get_rx_bufs(vq, vq->heads + nvq->done_idx,
vhost_len, &in, vq_log, &log,
likely(mergeable) ? UIO_MAXIOV : 1);UIO_MAXIOV was still used which is wrong since we could have batched used in vq->heads, this will cause OOB if the next buffer needs more than 960 (1024 (UIO_MAXIOV) - 64 (VHOST_NET_BATCH)) heads after we've batched 64 (VHOST_NET_BATCH) heads:
=============================================================================
BUG kmalloc-8k (Tainted: G B ): Redzone overwritten
-----------------------------------------------------------------------------
INFO: 0x00000000fd93b7a2-0x00000000f0713384. First byte 0xa9 instead of 0xcc
INFO: Allocated in alloc_pd+0x22/0x60 age=3933677 cpu=2 pid=2674
kmem_cache_alloc_trace+0xbb/0x140
alloc_pd+0x22/0x60
gen8_ppgtt_create+0x11d/0x5f0
i915_ppgtt_create+0x16/0x80
i915_gem_create_context+0x248/0x390
i915_gem_context_create_ioctl+0x4b/0xe0
drm_ioctl_kernel+0xa5/0xf0
drm_ioctl+0x2ed/0x3a0
do_vfs_ioctl+0x9f/0x620
ksys_ioctl+0x6b/0x80
__x64_sys_ioctl+0x11/0x20
do_syscall_64+0x43/0xf0
entry_SYSCALL_64_after_hwframe+0x44/0xa9INFO: Slab 0x00000000d13e87af objects=3 used=3 fp=0x (null) flags=0x200000000010201
INFO: Object 0x0000000003278802 @offset=17064 fp=0x00000000e2e6652b
Fixing this by allocating UIO_MAXIOV + VHOST_NET_BATCH iovs for vhost-net. This is done through set the limitation through vhost_dev_init(), then set_owner can allocate the number of iov in a per device manner.
This fixes CVE-2018-16880.
--- drivers/vhost/net.c | 3 ++- drivers/vhost/scsi.c | 2 +- drivers/vhost/vhost.c | 7 ++++--- drivers/vhost/vhost.h | 4 +++- drivers/vhost/vsock.c | 2 +- 5 files changed, 11 insertions(+), 7 deletions(-)
drivers/vhost/net.c +2 -1
@@ -1337,7 +1337,8 @@ static int vhost_net_open(struct inode *inode, struct file *f) n->vqs[i].rx_ring =NULL ;vhost_net_buf_init (&n->vqs[i].rxq); } -vhost_dev_init (dev, vqs,VHOST_NET_VQ_MAX ); +vhost_dev_init (dev, vqs,VHOST_NET_VQ_MAX , +UIO_MAXIOV +VHOST_NET_BATCH );vhost_poll_init (n->poll +VHOST_NET_VQ_TX , handle_tx_net,EPOLLOUT , dev);vhost_poll_init (n->poll +VHOST_NET_VQ_RX , handle_rx_net,EPOLLIN , dev);
drivers/vhost/scsi.c +1 -1
@@ -1627,7 +1627,7 @@ static int vhost_scsi_open(struct inode *inode, struct file *f) vqs[i] = &vs->vqs[i].vq; vs->vqs[i].vq.handle_kick = vhost_scsi_handle_kick; } -vhost_dev_init (&vs->dev, vqs,VHOST_SCSI_MAX_VQ ); +vhost_dev_init (&vs->dev, vqs,VHOST_SCSI_MAX_VQ ,UIO_MAXIOV );vhost_scsi_init_inflight (vs,NULL );
drivers/vhost/vhost.c +4 -3
@@ -390,9 +390,9 @@ static long vhost_dev_alloc_iovecs(struct vhost_dev *dev) vq->indirect =kmalloc_array (UIO_MAXIOV ,sizeof (*vq->indirect),GFP_KERNEL ); - vq->log =kmalloc_array (UIO_MAXIOV ,sizeof (*vq->log), + vq->log =kmalloc_array (dev->iov_limit,sizeof (*vq->log),GFP_KERNEL ); - vq->heads =kmalloc_array (UIO_MAXIOV ,sizeof (*vq->heads), + vq->heads =kmalloc_array (dev->iov_limit,sizeof (*vq->heads),GFP_KERNEL );if (!vq->indirect || !vq->log || !vq->heads) goto err_nomem;
@@ -414,7 +414,7 @@ static void vhost_dev_free_iovecs(struct vhost_dev *dev) }void vhost_dev_init (struct vhost_dev *dev, -struct vhost_virtqueue **vqs,int nvqs) +struct vhost_virtqueue **vqs,int nvqs,int iov_limit) {struct vhost_virtqueue *vq;int i;
@@ -427,6 +427,7 @@ void vhost_dev_init(struct vhost_dev *dev, dev->iotlb =NULL ; dev->mm =NULL ; dev->worker =NULL ; + dev->iov_limit = iov_limit;init_llist_head (&dev->work_list);init_waitqueue_head (&dev->wait);INIT_LIST_HEAD (&dev->read_list);
drivers/vhost/vhost.h +3 -1
@@ -170,9 +170,11 @@ struct vhost_dev {struct list_head read_list;struct list_head pending_list;wait_queue_head_t wait; +int iov_limit; }; -void vhost_dev_init (struct vhost_dev *,struct vhost_virtqueue **vqs,int nvqs); +void vhost_dev_init (struct vhost_dev *,struct vhost_virtqueue **vqs, +int nvqs,int iov_limit);long vhost_dev_set_owner (struct vhost_dev *dev);bool vhost_dev_has_owner (struct vhost_dev *dev);long vhost_dev_check_owner (struct vhost_dev *);
drivers/vhost/vsock.c +1 -1
@@ -531,7 +531,7 @@ static int vhost_vsock_dev_open(struct inode *inode, struct file *file) vsock->vqs[VSOCK_VQ_TX ].handle_kick = vhost_vsock_handle_tx_kick; vsock->vqs[VSOCK_VQ_RX ].handle_kick = vhost_vsock_handle_rx_kick; -vhost_dev_init (&vsock->dev, vqs,ARRAY_SIZE (vsock->vqs)); +vhost_dev_init (&vsock->dev, vqs,ARRAY_SIZE (vsock->vqs),UIO_MAXIOV ); file->private_data = vsock;spin_lock_init (&vsock->send_pkt_list_lock);
Signature · 1 line
2.17.1
Quoted 46 lines · Jason Wang
After batched used ring updating was introduced in commit
e2b3b35eb989("vhost_net: batch used ring update in rx"). We tend to batch heads invq->headsfor more than one packet. But the quota passed toget_rx_bufs()was not correctly limited, which can result a OOB write invq->heads.headcount = get_rx_bufs(vq, vq->heads + nvq->done_idx, vhost_len, &in, vq_log, &log, likely(mergeable) ? UIO_MAXIOV : 1);
UIO_MAXIOVwas still used which is wrong since we could have batched used invq->heads, this will cause OOB if the next buffer needs more than 960 (1024 (UIO_MAXIOV) - 64 (VHOST_NET_BATCH)) heads after we've batched 64 (VHOST_NET_BATCH) heads:=============================================================================
BUG kmalloc-8k (Tainted: G B ): Redzone overwritten
-----------------------------------------------------------------------------INFO: 0x00000000fd93b7a2-0x00000000f0713384. First byte
0xa9instead of0xcc
INFO: Allocated inalloc_pd+0x22/0x60age=3933677 cpu=2 pid=2674kmem_cache_alloc_trace+0xbb/0x140 alloc_pd+0x22/0x60 gen8_ppgtt_create+0x11d/0x5f0 i915_ppgtt_create+0x16/0x80 i915_gem_create_context+0x248/0x390 i915_gem_context_create_ioctl+0x4b/0xe0 drm_ioctl_kernel+0xa5/0xf0 drm_ioctl+0x2ed/0x3a0 do_vfs_ioctl+0x9f/0x620 ksys_ioctl+0x6b/0x80 __x64_sys_ioctl+0x11/0x20 do_syscall_64+0x43/0xf0 entry_SYSCALL_64_after_hwframe+0x44/0xa9INFO: Slab
0x00000000d13e87afobjects=3 used=3 fp=0x (null) flags=0x200000000010201
INFO: Object0x0000000003278802@offset=17064 fp=0x00000000e2e6652bFixing this by allocating
UIO_MAXIOV+VHOST_NET_BATCHiovs for vhost-net. This is done through set the limitation throughvhost_dev_init(), thenset_ownercan allocate the number of iov in a per device manner.This fixes CVE-2018-16880.
Fixes e2b3b35eb989 ("vhost_net: batch used ring update in rx")Signed-off-by Jason Wang <[email protected]>
Quoted 7 lines
--- drivers/vhost/net.c | 3 ++- drivers/vhost/scsi.c | 2 +- drivers/vhost/vhost.c | 7 ++++--- drivers/vhost/vhost.h | 4 +++- drivers/vhost/vsock.c | 2 +- 5 files changed, 11 insertions(+), 7 deletions(-)
No change in the scsi and vsock cases. I haven't reviewed the net case.
Quoted 46 lines · Jason Wang
After batched used ring updating was introduced in commit
e2b3b35eb989("vhost_net: batch used ring update in rx"). We tend to batch heads invq->headsfor more than one packet. But the quota passed toget_rx_bufs()was not correctly limited, which can result a OOB write invq->heads.headcount = get_rx_bufs(vq, vq->heads + nvq->done_idx, vhost_len, &in, vq_log, &log, likely(mergeable) ? UIO_MAXIOV : 1);
UIO_MAXIOVwas still used which is wrong since we could have batched used invq->heads, this will cause OOB if the next buffer needs more than 960 (1024 (UIO_MAXIOV) - 64 (VHOST_NET_BATCH)) heads after we've batched 64 (VHOST_NET_BATCH) heads:=============================================================================
BUG kmalloc-8k (Tainted: G B ): Redzone overwritten
-----------------------------------------------------------------------------INFO: 0x00000000fd93b7a2-0x00000000f0713384. First byte
0xa9instead of0xcc
INFO: Allocated inalloc_pd+0x22/0x60age=3933677 cpu=2 pid=2674kmem_cache_alloc_trace+0xbb/0x140 alloc_pd+0x22/0x60 gen8_ppgtt_create+0x11d/0x5f0 i915_ppgtt_create+0x16/0x80 i915_gem_create_context+0x248/0x390 i915_gem_context_create_ioctl+0x4b/0xe0 drm_ioctl_kernel+0xa5/0xf0 drm_ioctl+0x2ed/0x3a0 do_vfs_ioctl+0x9f/0x620 ksys_ioctl+0x6b/0x80 __x64_sys_ioctl+0x11/0x20 do_syscall_64+0x43/0xf0 entry_SYSCALL_64_after_hwframe+0x44/0xa9INFO: Slab
0x00000000d13e87afobjects=3 used=3 fp=0x (null) flags=0x200000000010201
INFO: Object0x0000000003278802@offset=17064 fp=0x00000000e2e6652bFixing this by allocating
UIO_MAXIOV+VHOST_NET_BATCHiovs for vhost-net. This is done through set the limitation throughvhost_dev_init(), thenset_ownercan allocate the number of iov in a per device manner.This fixes CVE-2018-16880.
Fixes e2b3b35eb989 ("vhost_net: batch used ring update in rx")Signed-off-by Jason Wang <[email protected]>
Quoted 7 lines
--- drivers/vhost/net.c | 3 ++- drivers/vhost/scsi.c | 2 +- drivers/vhost/vhost.c | 7 ++++--- drivers/vhost/vhost.h | 4 +++- drivers/vhost/vsock.c | 2 +- 5 files changed, 11 insertions(+), 7 deletions(-)
No change in the scsi and vsock cases. I haven't reviewed the net case.
_______________________________________________
Virtualization mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/virtualization
David Miller
maintainerFrom: Jason Wang <[email protected]>
Date: Mon, 28 Jan 2019 15:05:05 +0800
Quoted 14 lines
After batched used ring updating was introduced in commit
e2b3b35eb989("vhost_net: batch used ring update in rx"). We tend to batch heads invq->headsfor more than one packet. But the quota passed toget_rx_bufs()was not correctly limited, which can result a OOB write invq->heads.headcount = get_rx_bufs(vq, vq->heads + nvq->done_idx, vhost_len, &in, vq_log, &log, likely(mergeable) ? UIO_MAXIOV : 1);
UIO_MAXIOVwas still used which is wrong since we could have batched used invq->heads, this will cause OOB if the next buffer needs more than 960 (1024 (UIO_MAXIOV) - 64 (VHOST_NET_BATCH)) heads after we've batched 64 (VHOST_NET_BATCH) heads:
...
Quoted 9 lines
Fixing this by allocating
UIO_MAXIOV+VHOST_NET_BATCHiovs for vhost-net. This is done through set the limitation throughvhost_dev_init(), thenset_ownercan allocate the number of iov in a per device manner.This fixes CVE-2018-16880.
Fixes e2b3b35eb989 ("vhost_net: batch used ring update in rx")Signed-off-by Jason Wang <[email protected]>
Applied and queued up for -stable, thanks!
Michael S. Tsirkin
maintainerQuoted 23 lines · David Miller
From: Jason Wang <[email protected]>
Date: Mon, 28 Jan 2019 15:05:05 +0800Quoted 14 lines
After batched used ring updating was introduced in commit
e2b3b35eb989("vhost_net: batch used ring update in rx"). We tend to batch heads invq->headsfor more than one packet. But the quota passed toget_rx_bufs()was not correctly limited, which can result a OOB write invq->heads.headcount = get_rx_bufs(vq, vq->heads + nvq->done_idx, vhost_len, &in, vq_log, &log, likely(mergeable) ? UIO_MAXIOV : 1);
UIO_MAXIOVwas still used which is wrong since we could have batched used invq->heads, this will cause OOB if the next buffer needs more than 960 (1024 (UIO_MAXIOV) - 64 (VHOST_NET_BATCH)) heads after we've batched 64 (VHOST_NET_BATCH) heads:...
Quoted 5 lines
Fixing this by allocating
UIO_MAXIOV+VHOST_NET_BATCHiovs for vhost-net. This is done through set the limitation throughvhost_dev_init(), thenset_ownercan allocate the number of iov in a per device manner.
Quoted 6 lines
Quoted 4 lines
This fixes CVE-2018-16880.
Fixes e2b3b35eb989 ("vhost_net: batch used ring update in rx")Signed-off-by Jason Wang <[email protected]>Applied and queued up for -stable, thanks!
Wow it seems we are down to hours round time post to queue. It would be hard to keep up that rate generally.
However, I am guessing this was already in downstreams, and it's a CVE, so I guess it's a no brainer and review wasn't really necessary - was that the idea? Just checking.
Signature · 1 line
MST
David Miller
maintainerFrom: "Michael S. Tsirkin" <[email protected]>
Date: Tue, 29 Jan 2019 17:54:44 -0500
Quoted 29 lines · David Miller
Quoted 28 lines
From: Jason Wang <[email protected]>
Date: Mon, 28 Jan 2019 15:05:05 +0800Quoted 14 lines
After batched used ring updating was introduced in commit
e2b3b35eb989("vhost_net: batch used ring update in rx"). We tend to batch heads invq->headsfor more than one packet. But the quota passed toget_rx_bufs()was not correctly limited, which can result a OOB write invq->heads.headcount = get_rx_bufs(vq, vq->heads + nvq->done_idx, vhost_len, &in, vq_log, &log, likely(mergeable) ? UIO_MAXIOV : 1);
UIO_MAXIOVwas still used which is wrong since we could have batched used invq->heads, this will cause OOB if the next buffer needs more than 960 (1024 (UIO_MAXIOV) - 64 (VHOST_NET_BATCH)) heads after we've batched 64 (VHOST_NET_BATCH) heads:...
Quoted 9 lines
Fixing this by allocating
UIO_MAXIOV+VHOST_NET_BATCHiovs for vhost-net. This is done through set the limitation throughvhost_dev_init(), thenset_ownercan allocate the number of iov in a per device manner.This fixes CVE-2018-16880.
Fixes e2b3b35eb989 ("vhost_net: batch used ring update in rx")Signed-off-by Jason Wang <[email protected]>
Quoted 7 lines
Quoted 1 line
Applied and queued up for -stable, thanks!
Wow it seems we are down to hours round time post to queue. It would be hard to keep up that rate generally.
However, I am guessing this was already in downstreams, and it's a CVE, so I guess it's a no brainer and review wasn't really necessary - was that the idea? Just checking.
Yeah the CVE pushed my hand a little bit, and I knew I was going to send Linus a pull request today because David Watson needs some TLS changes in net-next.
David Miller
maintainerFrom: David Miller <[email protected]>
Date: Tue, 29 Jan 2019 15:10:26 -0800 (PST)
Quoted 3 lines
Yeah the CVE pushed my hand a little bit, and I knew I was going to send Linus a pull request today because David Watson needs some TLS changes in net-next.
I also want to make a general comment.... for the record.
If I let patches slip consistently past 24 hours my backlog is unmanageable. Even with aggressively applying things quickly I'm right now at 70-75. If I do not do what I am doing, then it's in the 100-150 range.
So I am at the point where I often must move forward with patches that I think I personally can verify and vet on my own.
Michael S. Tsirkin
maintainerQuoted 9 lines · David Miller
From: David Miller <[email protected]>
Date: Tue, 29 Jan 2019 15:10:26 -0800 (PST)Quoted 3 lines
Yeah the CVE pushed my hand a little bit, and I knew I was going to send Linus a pull request today because David Watson needs some TLS changes in net-next.
I also want to make a general comment.... for the record.
Quoted 7 lines
If I let patches slip consistently past 24 hours my backlog is unmanageable. Even with aggressively applying things quickly I'm right now at 70-75. If I do not do what I am doing, then it's in the 100-150 range.
So I am at the point where I often must move forward with patches that I think I personally can verify and vet on my own.
If it helps I can include most virtio stuff in my pull requests instead. Or if that can't work since there's too often a dependency on net-next, maybe Jason wants to create a tree and send pull requests to you. Let us know if that will help, and which of the options looks better from your POV.
Signature · 1 line
MST
David Miller
maintainerFrom: "Michael S. Tsirkin" <[email protected]>
Date: Tue, 29 Jan 2019 20:36:31 -0500
Quoted 5 lines
If it helps I can include most virtio stuff in my pull requests instead. Or if that can't work since there's too often a dependency on net-next, maybe Jason wants to create a tree and send pull requests to you. Let us know if that will help, and which of the options looks better from your POV.
Thanks for offering Michael, I really appreciate it.
Let me think about the logistics of that and how it may or may not help me with my backlog.
David Miller
maintainerFrom: "Michael S. Tsirkin" <[email protected]>
Date: Tue, 29 Jan 2019 20:36:31 -0500
Quoted 5 lines
If it helps I can include most virtio stuff in my pull requests instead. Or if that can't work since there's too often a dependency on net-next, maybe Jason wants to create a tree and send pull requests to you. Let us know if that will help, and which of the options looks better from your POV.
Thanks for offering Michael, I really appreciate it.
Let me think about the logistics of that and how it may or may not help me with my backlog.
Michael S. Tsirkin
maintainerQuoted 9 lines · David Miller
From: David Miller <[email protected]>
Date: Tue, 29 Jan 2019 15:10:26 -0800 (PST)Quoted 3 lines
Yeah the CVE pushed my hand a little bit, and I knew I was going to send Linus a pull request today because David Watson needs some TLS changes in net-next.
I also want to make a general comment.... for the record.
Quoted 7 lines
If I let patches slip consistently past 24 hours my backlog is unmanageable. Even with aggressively applying things quickly I'm right now at 70-75. If I do not do what I am doing, then it's in the 100-150 range.
So I am at the point where I often must move forward with patches that I think I personally can verify and vet on my own.
If it helps I can include most virtio stuff in my pull requests instead. Or if that can't work since there's too often a dependency on net-next, maybe Jason wants to create a tree and send pull requests to you. Let us know if that will help, and which of the options looks better from your POV.
Signature · 1 line
MST
David Miller
maintainerFrom: David Miller <[email protected]>
Date: Tue, 29 Jan 2019 15:10:26 -0800 (PST)
Quoted 3 lines
Yeah the CVE pushed my hand a little bit, and I knew I was going to send Linus a pull request today because David Watson needs some TLS changes in net-next.
I also want to make a general comment.... for the record.
If I let patches slip consistently past 24 hours my backlog is unmanageable. Even with aggressively applying things quickly I'm right now at 70-75. If I do not do what I am doing, then it's in the 100-150 range.
So I am at the point where I often must move forward with patches that I think I personally can verify and vet on my own.
David Miller
maintainerFrom: "Michael S. Tsirkin" <[email protected]>
Date: Tue, 29 Jan 2019 17:54:44 -0500
Quoted 29 lines · David Miller
Quoted 28 lines
From: Jason Wang <[email protected]>
Date: Mon, 28 Jan 2019 15:05:05 +0800Quoted 14 lines
After batched used ring updating was introduced in commit
e2b3b35eb989("vhost_net: batch used ring update in rx"). We tend to batch heads invq->headsfor more than one packet. But the quota passed toget_rx_bufs()was not correctly limited, which can result a OOB write invq->heads.headcount = get_rx_bufs(vq, vq->heads + nvq->done_idx, vhost_len, &in, vq_log, &log, likely(mergeable) ? UIO_MAXIOV : 1);
UIO_MAXIOVwas still used which is wrong since we could have batched used invq->heads, this will cause OOB if the next buffer needs more than 960 (1024 (UIO_MAXIOV) - 64 (VHOST_NET_BATCH)) heads after we've batched 64 (VHOST_NET_BATCH) heads:...
Quoted 9 lines
Fixing this by allocating
UIO_MAXIOV+VHOST_NET_BATCHiovs for vhost-net. This is done through set the limitation throughvhost_dev_init(), thenset_ownercan allocate the number of iov in a per device manner.This fixes CVE-2018-16880.
Fixes e2b3b35eb989 ("vhost_net: batch used ring update in rx")Signed-off-by Jason Wang <[email protected]>
Quoted 7 lines
Quoted 1 line
Applied and queued up for -stable, thanks!
Wow it seems we are down to hours round time post to queue. It would be hard to keep up that rate generally.
However, I am guessing this was already in downstreams, and it's a CVE, so I guess it's a no brainer and review wasn't really necessary - was that the idea? Just checking.
Yeah the CVE pushed my hand a little bit, and I knew I was going to send Linus a pull request today because David Watson needs some TLS changes in net-next.
Michael S. Tsirkin
maintainerQuoted 23 lines · David Miller
From: Jason Wang <[email protected]>
Date: Mon, 28 Jan 2019 15:05:05 +0800Quoted 14 lines
After batched used ring updating was introduced in commit
e2b3b35eb989("vhost_net: batch used ring update in rx"). We tend to batch heads invq->headsfor more than one packet. But the quota passed toget_rx_bufs()was not correctly limited, which can result a OOB write invq->heads.headcount = get_rx_bufs(vq, vq->heads + nvq->done_idx, vhost_len, &in, vq_log, &log, likely(mergeable) ? UIO_MAXIOV : 1);
UIO_MAXIOVwas still used which is wrong since we could have batched used invq->heads, this will cause OOB if the next buffer needs more than 960 (1024 (UIO_MAXIOV) - 64 (VHOST_NET_BATCH)) heads after we've batched 64 (VHOST_NET_BATCH) heads:...
Quoted 5 lines
Fixing this by allocating
UIO_MAXIOV+VHOST_NET_BATCHiovs for vhost-net. This is done through set the limitation throughvhost_dev_init(), thenset_ownercan allocate the number of iov in a per device manner.
Quoted 6 lines
Quoted 4 lines
This fixes CVE-2018-16880.
Fixes e2b3b35eb989 ("vhost_net: batch used ring update in rx")Signed-off-by Jason Wang <[email protected]>Applied and queued up for -stable, thanks!
Wow it seems we are down to hours round time post to queue. It would be hard to keep up that rate generally.
However, I am guessing this was already in downstreams, and it's a CVE, so I guess it's a no brainer and review wasn't really necessary - was that the idea? Just checking.
Signature · 1 line
MST
David Miller
maintainerFrom: Jason Wang <[email protected]>
Date: Mon, 28 Jan 2019 15:05:05 +0800
Quoted 14 lines
After batched used ring updating was introduced in commit
e2b3b35eb989("vhost_net: batch used ring update in rx"). We tend to batch heads invq->headsfor more than one packet. But the quota passed toget_rx_bufs()was not correctly limited, which can result a OOB write invq->heads.headcount = get_rx_bufs(vq, vq->heads + nvq->done_idx, vhost_len, &in, vq_log, &log, likely(mergeable) ? UIO_MAXIOV : 1);
UIO_MAXIOVwas still used which is wrong since we could have batched used invq->heads, this will cause OOB if the next buffer needs more than 960 (1024 (UIO_MAXIOV) - 64 (VHOST_NET_BATCH)) heads after we've batched 64 (VHOST_NET_BATCH) heads:
...
Quoted 9 lines
Fixing this by allocating
UIO_MAXIOV+VHOST_NET_BATCHiovs for vhost-net. This is done through set the limitation throughvhost_dev_init(), thenset_ownercan allocate the number of iov in a per device manner.This fixes CVE-2018-16880.
Fixes e2b3b35eb989 ("vhost_net: batch used ring update in rx")Signed-off-by Jason Wang <[email protected]>
Applied and queued up for -stable, thanks!