We currently use hdr_len as a hint of head length which is advertised by
guest. But when guest advertise a very big value, it can lead to an 64K+
allocating of kmalloc() which has a very high possibility of failure when host
memory is fragmented or under heavy stress. The huge hdr_len also reduce the
effect of zerocopy or even disable if a gso skb is linearized in guest.
To solves those issues, this patch introduces an upper limit (PAGE_SIZE) of the
head, which guarantees an order 0 allocation each time.
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Conflicts:
drivers/net/wireless/iwlwifi/pcie/trans.c
include/linux/inetdevice.h
The inetdevice.h conflict involves moving the IPV4_DEVCONF values
into a UAPI header, overlapping additions of some new entries.
The iwlwifi conflict is a context overlap.
Signed-off-by: David S. Miller <davem@davemloft.net>
When the user turns off VNET_HDR support on the
macvtap device, there is no way to provide any
offload information to the user. So, it's safer
to ignore offload setting then depend on the user
setting them correctly.
Signed-off-by: Vlad Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When the user turns off IFF_VNET_HDR flag, attempts to change
offload features via TUNSETOFFLOAD do not work. This could cause
GSO packets to be delivered to the user when the user is
not prepared to handle them.
To solve, allow processing of TUNSETOFFLOAD when IFF_VNET_HDR is
disabled.
Signed-off-by: Vlad Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In macvtap, tap_features specific the features of that the user
has specified via ioctl(). If we treat macvtap as a macvlan+tap
then we could all the tap a pseudo-device and give it other features
like SG and GSO. Then we can stop using the features of lower
device (macvlan) when forwarding the traffic the tap.
This solves the issue of possible checksum offload mismatch between
tap feature and macvlan features.
Signed-off-by: Vlad Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since commit ac4e4af1e5 ("macvtap: Consistently use rcu functions"),
Thomas gets two different warnings :
BUG: using smp_processor_id() in preemptible [00000000] code: vhost-45891/45892
caller is macvtap_do_read+0x45c/0x600 [macvtap]
CPU: 1 PID: 45892 Comm: vhost-45891 Not tainted 3.11.0-bisecttest #13
Call Trace:
([<00000000001126ee>] show_trace+0x126/0x144)
[<00000000001127d2>] show_stack+0xc6/0xd4
[<000000000068bcec>] dump_stack+0x74/0xd8
[<0000000000481066>] debug_smp_processor_id+0xf6/0x114
[<000003ff802e9a18>] macvtap_do_read+0x45c/0x600 [macvtap]
[<000003ff802e9c1c>] macvtap_recvmsg+0x60/0x88 [macvtap]
[<000003ff80318c5e>] handle_rx+0x5b2/0x800 [vhost_net]
[<000003ff8028f77c>] vhost_worker+0x15c/0x1c4 [vhost]
[<000000000015f3ac>] kthread+0xd8/0xe4
[<00000000006934a6>] kernel_thread_starter+0x6/0xc
[<00000000006934a0>] kernel_thread_starter+0x0/0xc
And
BUG: using smp_processor_id() in preemptible [00000000] code: vhost-45897/45898
caller is macvlan_start_xmit+0x10a/0x1b4 [macvlan]
CPU: 1 PID: 45898 Comm: vhost-45897 Not tainted 3.11.0-bisecttest #16
Call Trace:
([<00000000001126ee>] show_trace+0x126/0x144)
[<00000000001127d2>] show_stack+0xc6/0xd4
[<000000000068bdb8>] dump_stack+0x74/0xd4
[<0000000000481132>] debug_smp_processor_id+0xf6/0x114
[<000003ff802b72ca>] macvlan_start_xmit+0x10a/0x1b4 [macvlan]
[<000003ff802ea69a>] macvtap_get_user+0x982/0xbc4 [macvtap]
[<000003ff802ea92a>] macvtap_sendmsg+0x4e/0x60 [macvtap]
[<000003ff8031947c>] handle_tx+0x494/0x5ec [vhost_net]
[<000003ff8028f77c>] vhost_worker+0x15c/0x1c4 [vhost]
[<000000000015f3ac>] kthread+0xd8/0xe4
[<000000000069356e>] kernel_thread_starter+0x6/0xc
[<0000000000693568>] kernel_thread_starter+0x0/0xc
2 locks held by vhost-45897/45898:
#0: (&vq->mutex){+.+.+.}, at: [<000003ff8031903c>] handle_tx+0x54/0x5ec [vhost_net]
#1: (rcu_read_lock){.+.+..}, at: [<000003ff802ea53c>] macvtap_get_user+0x824/0xbc4 [macvtap]
In the first case, macvtap_put_user() calls macvlan_count_rx()
in a preempt-able context, and this is not allowed.
In the second case, macvtap_get_user() calls
macvlan_start_xmit() with BH enabled, and this is not allowed.
Reported-by: Thomas Huth <thuth@linux.vnet.ibm.com>
Bisected-by: Thomas Huth <thuth@linux.vnet.ibm.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Tested-by: Thomas Huth <thuth@linux.vnet.ibm.com>
Cc: Vlad Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Adding paged frags skbs to af_unix sockets introduced a performance
regression on large sends because of additional page allocations, even
if each skb could carry at least 100% more payload than before.
We can instruct sock_alloc_send_pskb() to attempt high order
allocations.
Most of the time, it does a single page allocation instead of 8.
I added an additional parameter to sock_alloc_send_pskb() to
let other users to opt-in for this new feature on followup patches.
Tested:
Before patch :
$ netperf -t STREAM_STREAM
STREAM STREAM TEST
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
2304 212992 212992 10.00 46861.15
After patch :
$ netperf -t STREAM_STREAM
STREAM STREAM TEST
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
2304 212992 212992 10.00 57981.11
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: David Rientjes <rientjes@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
To let it be reused and reduce code duplication. Also document this function.
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We try to linearize part of the skb when the number of iov is greater than
MAX_SKB_FRAGS. This is not enough since each single vector may occupy more than
one pages, so zerocopy_sg_fromiovec() may still fail and may break the guest
network.
Solve this problem by calculate the pages needed for iov before trying to do
zerocopy and switch to use copy instead of zerocopy if it needs more than
MAX_SKB_FRAGS.
This is done through introducing a new helper to count the pages for iov, and
call uarg->callback() manually when switching from zerocopy to copy to notify
vhost.
We can do further optimization on top.
This bug were introduced from b92946e291
(macvtap: zerocopy: validate vectors before building skb).
Cc: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The hard-coded 8021.q proto will break 802.1ad traffic. So switch to use
vlan->proto.
Cc: Basil Gor <basil.gor@gmail.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 441ac0fcaa
(macvtap: Convert to using rtnl lock) forget to return what
macvtap_ioctl_set_queue() returns to its caller. This may break multiqueue API
by always falling through to TUNGETFEATURES.
Cc: Vlad Yasevich <vyasevic@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Userspace may produce vectors greater than MAX_SKB_FRAGS. When we try to
linearize parts of the skb to let the rest of iov to be fit in
the frags, we need count copylen into linear when calling macvtap_alloc_skb()
instead of partly counting it into data_len. Since this breaks
zerocopy_sg_from_iovec() since its inner counter assumes nr_frags should
be zero at beginning. This cause nr_frags to be increased wrongly without
setting the correct frags.
This bug were introduced from b92946e291
(macvtap: zerocopy: validate vectors before building skb).
Cc: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Conflicts:
drivers/net/ethernet/freescale/fec_main.c
drivers/net/ethernet/renesas/sh_eth.c
net/ipv4/gre.c
The GRE conflict is between a bug fix (kfree_skb --> kfree_skb_list)
and the splitting of the gre.c code into seperate files.
The FEC conflict was two sets of changes adding ethtool support code
in an "!CONFIG_M5272" CPP protected block.
Finally the sh_eth.c conflict was between one commit add bits set
in the .eesr_err_check mask whilst another commit removed the
.tx_error_check member and assignments.
Signed-off-by: David S. Miller <davem@davemloft.net>
When macvtap forwards skb to its tap, it needs to check
if GSO needs to be performed. This is sometimes necessary
when the HW device performed GRO, but the guest reading
from the tap does not support it (ex: Windows 7).
Signed-off-by: Vlad Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When the user issues TUNSETOFFLOAD ioctl, macvtap does not do
anything other then to verify arguments. This patch adds
functionality to allow users to actually control offload features.
NETIF_F_GSO and NETIF_F_GRO are always on, but the rest of the
features can be controlled.
Signed-off-by: Vlad Yasevich <vyasevic@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently macvtap uses rcu_bh functions in its
user facing fuction macvtap_get_user() and macvtap_put_user().
However, its packet handlers use normal rcu as the rcu_read_lock()
is taken in netif_receive_skb(). We can safely discontinue
the usage or rcu with bh disabled.
Signed-off-by: Vlad Yasevich <vyasevic@redhat.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Macvtap uses a private lock to protect the relationship between
macvtap_queue and macvlan_dev. The private lock is not needed
since the relationship is managed by user via open(), release(),
and dellink() calls. dellink() already happens under rtnl, so
we can safely convert open() and release(), and use it in ioctl()
as well.
Suggested by Eric Dumazet.
Signed-off-by: Vlad Yasevich <vyasevic@redhat.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
get user pages might fail partially in macvtap zero copy
mode. To recover we need to put all pages that we got,
but code used a wrong index resulting in double-free
errors.
Reported-by: Brad Hubbard <bhubbard@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Return -EINVAL on illegal flag instead of uninitialized value. This fixes the
kbuild test warning.
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
To notify the userspace about our capability of multiqueue.
Signed-off-by: Jason Wang <jasowang@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds TUNSETQUEUE ioctl to let userspace can temporarily disable or
enable a queue of macvtap. This is used to be compatible at API layer of tuntap
to simplify the userspace to manage the queues. This is done through introducing
a linked list to track all taps while using vlan->taps array to only track
active taps.
Signed-off-by: Jason Wang <jasowang@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Linear search were used in both get_slot() and macvtap_get_queue(), this is
because:
- macvtap didn't reshuffle the array of taps when create or destroy a queue, so
when adding a new queue, macvtap must do linear search to find a location for
the new queue. This will also complicate the TUNSETQUEUE implementation for
multiqueue API.
- the queue itself didn't track the queue index, so the we must do a linear
search in the array to find the location of a existed queue.
The solution is straightforward: reshuffle the array and introduce a queue_index
to macvtap_queue.
Signed-off-by: Jason Wang <jasowang@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Factor out the device holding logic to a macvtap_get_vlan(), this will be also
used by multiqueue API.
Signed-off-by: Jason Wang <jasowang@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There's no need to add self to waitqueue if doing a nonblock read. This could
help to avoid the spinlock contention.
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Complier may generate codes that re-read the vlan->numvtaps during
macvtap_get_queue(). This may lead a race if vlan->numvtaps were changed in the
same time and which can lead unexpected result (e.g. very huge value).
We need prevent the compiler from generating such codes by adding an
ACCESS_ONCE() to make sure vlan->numvtaps were only read once.
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
So far, only net_device * could be passed along with netdevice notifier
event. This patch provides a possibility to pass custom structure
able to provide info that event listener needs to know.
Signed-off-by: Jiri Pirko <jiri@resnulli.us>
v2->v3: fix typo on simeth
shortened dev_getter
shortened notifier_info struct name
v1->v2: fix notifier_call parameter in call_netdevice_notifier()
Signed-off-by: David S. Miller <davem@davemloft.net>
Switch to use the new help skb_probe_transport_header() to do the l4 header
probing for untrusted sources. For packets with partial csum, the header should
already been set by skb_partial_csum_set().
Cc: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Set the transport header for 1) some drivers (e.g ixgbe) needs l4 header 2)
precise packet length estimation (introduced in 1def9238) needs l4 header to
compute header length.
For the packets with partial checksum, the patch just set the transport header
to csum_start. Otherwise tries to use skb_flow_dissect() to get l4 offset, if it
fails, just pretend no l4 header.
Cc: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Convert to the much saner new idr interface.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Jason Wang <jasowang@redhat.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Patch cef401de7b (net: fix possible wrong checksum
generation) fixed wrong checksum calculation but it broke TSO by
defining new GSO type but not a netdev feature for that type.
net_gso_ok() would not allow hardware checksum/segmentation
offload of such packets without the feature.
Following patch fixes TSO and wrong checksum. This patch uses
same logic that Eric Dumazet used. Patch introduces new flag
SKBTX_SHARED_FRAG if at least one frag can be modified by
the user. but SKBTX_SHARED_FRAG flag is kept in skb shared
info tx_flags rather than gso_type.
tx_flags is better compared to gso_type since we can have skb with
shared frag without gso packet. It does not link SHARED_FRAG to
GSO, So there is no need to define netdev feature for this.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pravin Shelar mentioned that GSO could potentially generate
wrong TX checksum if skb has fragments that are overwritten
by the user between the checksum computation and transmit.
He suggested to linearize skbs but this extra copy can be
avoided for normal tcp skbs cooked by tcp_sendmsg().
This patch introduces a new SKB_GSO_SHARED_FRAG flag, set
in skb_shinfo(skb)->gso_type if at least one frag can be
modified by the user.
Typical sources of such possible overwrites are {vm}splice(),
sendfile(), and macvtap/tun/virtio_net drivers.
Tested:
$ netperf -H 7.7.8.84
MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to
7.7.8.84 () port 0 AF_INET
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
87380 16384 16384 10.00 3959.52
$ netperf -H 7.7.8.84 -t TCP_SENDFILE
TCP SENDFILE TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 7.7.8.84 ()
port 0 AF_INET
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
87380 16384 16384 10.00 3216.80
Performance of the SENDFILE is impacted by the extra allocation and
copy, and because we use order-0 pages, while the TCP_STREAM uses
bigger pages.
Reported-by: Pravin Shelar <pshelar@nicira.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
rcu_dereference occurs in update section. Replacement by
rcu_dereference_protected in order to prevent lockdep
complaint.
Found by Linux Driver Verification project (linuxtesting.org)
Signed-off-by: Denis Efremov <yefremov.denis@gmail.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Ethernet vlan header is not on the packet and kept in the skb->vlan_tci
when it comes from lower dev. This patch inserts vlan header in user
buffer during skb copy on user read.
Signed-off-by: Basil Gor <basil.gor@gmail.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There're several reasons that the vectors need to be validated:
- Return error when caller provides vectors whose num is greater than UIO_MAXIOV.
- Linearize part of skb when userspace provides vectors grater than MAX_SKB_FRAGS.
- Return error when userspace provides vectors whose total length may exceed
- MAX_SKB_FRAGS * PAGE_SIZE.
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Current the SKBTX_DEV_ZEROCOPY is set unconditionally after
zerocopy_sg_from_iovec(), this would lead NULL pointer when macvtap
fails to build zerocopy skb because destructor_arg was not
initialized. Solve this by set this flag after the skb were built
successfully.
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
When get_user_pages_fast() fails to get all requested pages, we could not use
kfree_skb() to free it as it has not been put in the skb fragments. So we need
to call put_page() instead.
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
As the skb fragment were pinned/built from user pages, we should
account the page instead of length for truesize.
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
This patch fixes the offset calculation when building skb:
- offset1 were used as skb data offset not vector offset
- reset offset to zero only when we advance to next vector
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
rcu_assign_pointer(ptr, NULL) can be safely replaced by
RCU_INIT_POINTER(ptr, NULL)
(old rcu_assign_pointer() macro was testing the NULL value and could
omit the smp_wmb(), but this had to be removed because of compiler
warnings)
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
On systems that create and delete lots of dynamic devices the
31bit linux ifindex fails to fit in the 16bit macvtap minor,
resulting in unusable macvtap devices. I have systems running
automated tests that that hit this condition in just a few days.
Use a linux idr allocator to track which mavtap minor numbers
are available and and to track the association between macvtap
minor numbers and macvtap network devices.
Remove the unnecessary unneccessary check to see if the network
device we have found is indeed a macvtap device. With macvtap
specific data structures it is impossible to find any other
kind of networking device.
Increase the macvtap minor range from 65536 to the full 20 bits
that is supported by linux device numbers. It doesn't solve the
original problem but there is no penalty for a larger minor
device range.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Place macvlan_common_newlink at the end of macvtap_newlink because
failing in newlink after registering your network device is not
supported.
Move device_create into a netdevice creation notifier. The network device
notifier is the only hook that is called after the network device has been
registered with the device layer and before register_network_device returns
success.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
To avoid leaking packets in the receive queue. Add a socket destructor
that will run whenever destroy a macvtap socket.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
To see if it is appropriate to enable the macvtap zero copy feature
don't test the lowerdev network device flags. Instead test the
macvtap network device flags which are a direct copy of the lowerdev
flags. This is important because nothing holds a reference to lowerdev
and on a very bad day we lowerdev could be a pointer to stale memory.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jason Wang