Code was checking for NETMAP_{SW,HW}_RING in req->nr_ringid which
had already been masked by NETMAP_RING_MASK. Therefore, the comparisons
always failed and set NR_REG_ALL_NIC. Check against the original nmr
structure.
Submitted by: bpoole@packetforensics.com
Reported by: bpoole@packetforensics.com
Reviewed by: giuseppe.lettieri@unipi.it
Approved by: vmaffione
MFC after: 1 week
The nm_register callback needs to call nm_set_native_flags()
or nm_clear_native_flags() once the device has been stopped.
However, in the current implementation this is not true,
as the device is stopped by vtnet_init_locked(). This causes
race conditions where the driver crashes as soon as it
dequeues netmap buffers assuming they are mbufs (or the other
way around).
To fix the issue, we extend vtnet_init_locked() with a second
argument that, if not zero, will set/clear the netmap flags.
This results in a huge simplification of the nm_register
callback itself.
Also, use netmap_reset() to check if a ring is going to be
re-initialized in netmap mode.
MFC after: 1 week
This function returns NULL if the ring identified by
queue id and direction is in netmap mode. Otherwise
return the corresponding kring.
Use this function to replace vtnet_netmap_queue_on().
MFC after: 1 week
This change prevents a race that happens when rxsync dequeues
N-1 rx packets (with N being the size of the netmap rx ring).
In this situation, the loop exits without re-enabling the
rx interrupts, thus causing the VQ to stall.
MFC after: 1 week
The new index tracks the next netmap slot that is going
to be enqueued into the virtqueue. The index is necessary
to prevent the receive VQ and the netmap rx ring from going
out of sync, considering that we never enqueue N slots, but
at most N-1. This change fixes a bug that causes the VQ
and the netmap ring to go out of sync after N-1 packets
have been received.
MFC after: 1 week
netmap assumes the one "slot" is left unused to distinguish
the empty ring and full ring conditions. This assumption was
violated by vtnet_netmap_rxq_populate().
MFC after: 1 week
The functionality contained in this function is duplicated,
as it is already available in vtnet_txq_free_mbufs()
and vtnet_rxq_free_mbufs().
MFC after: 1 week
The vtnet_netmap_rxq_populate() function erroneously assumed
that kring->nr_hwcur = 0, i.e. the kring was in the initial
state. However, this is not always the case: for example,
when a vtnet reinit is triggered by some changes in the
interface flags or capenable.
This patch changes the behaviour of vtnet_netmap_kring_refill()
so that it always starts publishing the netmap buffers starting
from the current value of kring->nr_hwcur.
MFC after: 1 week
r357614 added CTLFLAG_NEEDGIANT to make it easier to find nodes that are
still not MPSAFE (or already are but aren’t properly marked).
Use it in preparation for a general review of all nodes.
This is non-functional change that adds annotations to SYSCTL_NODE and
SYSCTL_PROC nodes using one of the soon-to-be-required flags.
Mark all obvious cases as MPSAFE. All entries that haven't been marked
as MPSAFE before are by default marked as NEEDGIANT
Approved by: kib (mentor, blanket)
Commented by: kib, gallatin, melifaro
Differential Revision: https://reviews.freebsd.org/D23718
The netmap passthrough subsystem requires proper support in the
hypervisor. In particular, two PCI device ids (from the Red Hat
PCI vendor id 0x1b36) need to be assigned to the two netmap
virtual devices. We then disable these devices until the ids have
not been assigned, in order to avoid conflicts with other
virtual devices emulated by upstream QEMU.
PR: 241774
MFC after: 3 days
removed from objects including calls to free. Pages must not be xbusy
when freed and not on an object. Strengthen assertions to match these
expectations. In practice very little code had to change busy handling
to meet these rules but we can now make stronger guarantees to busy
holders and avoid conditionally dropping busy in free.
Refine vm_page_remove() and vm_page_replace() semantics now that we have
stronger guarantees about busy state. This removes redundant and
potentially problematic code that has proliferated.
Discussed with: markj
Reviewed by: kib
Differential Revision: https://reviews.freebsd.org/D22822
The valectl(4) program is used to manage vale(4) switches.
Add it to the system commands so that it can be used right away.
This program was previously called vale-ctl, and stored in
tools/tools/netmap
Reviewed by: hrs, bcr, lwhsu, kevans
MFC after: 1 week
Differential Revision: https://reviews.freebsd.org/D22146
After r353292, netmap generic adapter on if_vlan interfaces panics on
asserting the NET_EPOCH. In more detail, this happens when
nm_os_generic_xmit_frame() is called, that is in the generic txsync
routine.
Fix the issue by entering the NET_EPOCH during the generic txsync.
We amortize the cost of entering/exiting over a whole batch of
transmissions.
PR: 241489
Reported by: Aleksandr Fedorov <aleksandr.fedorov@itglobal.com>
This patch is part of an effort to make bhyve networking (in particular TCP)
faster. The key strategy to enhance TCP throughput is to let the whole packet
datapath work with TSO/LRO packets (up to 64KB each), so that the per-packet
overhead is amortized over a large number of bytes.
This capability is supported in the guest by means of the vtnet(4) driver,
which is able to handle TSO/LRO packets leveraging the virtio-net header
(see struct virtio_net_hdr and struct virtio_net_hdr_mrg_rxbuf).
A bhyve VM exchanges packets with the host through a network backend,
which can be vale(4) or if_tap(4).
While vale(4) supports TSO/LRO packets, if_tap(4) does not.
This patch extends if_tap(4) with the ability to understand the virtio-net
header, so that a tapX interface can process TSO/LRO packets.
A couple of ioctl commands have been added to configure and probe the
virtio-net header. Once the virtio-net header is set, the tapX interface
acquires all the IFCAP capabilities necessary for TSO/LRO.
Reviewed by: kevans
Differential Revision: https://reviews.freebsd.org/D21263
Atomics are used for page busy and valid state when the shared busy is
held. The details of the locking protocol and valid and dirty
synchronization are in the updated vm_page.h comments.
Reviewed by: kib, markj
Tested by: pho
Sponsored by: Netflix, Intel
Differential Revision: https://reviews.freebsd.org/D21594
There are several mechanisms by which a vm_page reference is held,
preventing the page from being freed back to the page allocator. In
particular, holding the page's object lock is sufficient to prevent the
page from being freed; holding the busy lock or a wiring is sufficent as
well. These references are protected by the page lock, which must
therefore be acquired for many per-page operations. This results in
false sharing since the page locks are external to the vm_page
structures themselves and each lock protects multiple structures.
Transition to using an atomically updated per-page reference counter.
The object's reference is counted using a flag bit in the counter. A
second flag bit is used to atomically block new references via
pmap_extract_and_hold() while removing managed mappings of a page.
Thus, the reference count of a page is guaranteed not to increase if the
page is unbusied, unmapped, and the object's write lock is held. As
a consequence of this, the page lock no longer protects a page's
identity; operations which move pages between objects are now
synchronized solely by the objects' locks.
The vm_page_wire() and vm_page_unwire() KPIs are changed. The former
requires that either the object lock or the busy lock is held. The
latter no longer has a return value and may free the page if it releases
the last reference to that page. vm_page_unwire_noq() behaves the same
as before; the caller is responsible for checking its return value and
freeing or enqueuing the page as appropriate. vm_page_wire_mapped() is
introduced for use in pmap_extract_and_hold(). It fails if the page is
concurrently being unmapped, typically triggering a fallback to the
fault handler. vm_page_wire() no longer requires the page lock and
vm_page_unwire() now internally acquires the page lock when releasing
the last wiring of a page (since the page lock still protects a page's
queue state). In particular, synchronization details are no longer
leaked into the caller.
The change excises the page lock from several frequently executed code
paths. In particular, vm_object_terminate() no longer bounces between
page locks as it releases an object's pages, and direct I/O and
sendfile(SF_NOCACHE) completions no longer require the page lock. In
these latter cases we now get linear scalability in the common scenario
where different threads are operating on different files.
__FreeBSD_version is bumped. The DRM ports have been updated to
accomodate the KPI changes.
Reviewed by: jeff (earlier version)
Tested by: gallatin (earlier version), pho
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D20486
The netmap_pt.c module has become obsolete after
the refactoring that added netmap_kloop.c.
Remove it and unlink it from the build system.
MFC after: 1 week
Print the adapter name rather than the address of the adapter
to avoid kernel address leakage.
PR: Bug 238642
Submitted by: Fuqian Huang <huangfq.daxian@gmail.com>
Reviewed by: vmaffione
MFC after: 1 week
This patch fixes 2 panics. The first one is due to the current VNET not
being set in the emulated adapter transmission path. The second one
is caused by the M_PKTHDR flag not being set when preallocated mbufs
are recycled in the transmit path.
Submitted by: aleksandr.fedorov@itglobal.com
Reviewed by: vmaffione
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D20824
These are obviously missing from the .c files, but don't show up in any
tinderbox configuration (due to latent header pollution of some kind). It
seems some configurations don't have this pollution, and the includes are
obviously missing, so go ahead and add them.
Reported by: Peter Jeremy <peter AT rulingia.com>
X-MFC-With: r347984
Some applications forward from/to host rings most or all the
traffic received or sent on a physical interface. In this
cases it is desirable to have more than a pair of RX/TX host
rings, and use multiple threads to speed up forwarding.
This change adds support for multiple host rings. On registering
a netmap port, the user can specify the number of desired receive
and transmit host rings in the nr_host_tx_rings and nr_host_rx_rings
fields of the nmreq_register structure.
MFC after: 2 weeks
This change adds a counter (kqueue_users) to keep track of how many
kqueue users are referencing a given struct nm_selinfo.
In this way, nm_os_selwakeup() can schedule the kevent notification
task only when kqueue is actually being used.
This is important to avoid wasting CPU in the common case where
kqueue is not used.
Reviewed by: Aleksandr Fedorov <aleksandr.fedorov@itglobal.com>
MFC after: 1 week
Differential Revision: https://reviews.freebsd.org/D19177
Changelist:
- Replace ND, D and RD macros with nm_prdis, nm_prinf, nm_prerr
and nm_prlim, to avoid possible naming conflicts.
- Add netmap_krings_mode_commit() helper function and use that
to reduce code duplication.
- Refactor pipes control code to export some functions that
can be reused by the veth driver (on Linux) and epair(4).
- Add check to reject API requests with version less than 11.
- Small code refactoring for the null adapter.
MFC after: 1 week
Add SYNC_KLOOP_MODE option, and add support for direct mode, where application
executes the TXSYNC and RXSYNC in the context of the ioeventfd wake up callback.
MFC after: 5 days
When using poll(), select() or kevent() on netmap file descriptors,
netmap executes the equivalent of NIOCTXSYNC and NIOCRXSYNC commands,
before collecting the events that are ready. In other words, the
poll/kevent callback has side effects. This is done to avoid the
overhead of two system call per iteration (e.g., poll() + ioctl(NIOC*XSYNC)).
When the kqueue subsystem invokes the kqueue(9) f_event callback
(netmap_knrw), it holds the lock of the struct knlist object associated
to the netmap port (the lock is provided at initialization, by calling
knlist_init_mtx).
However, netmap_knrw() may need to wake up another netmap port (or even
the same one), which means that it may need to call knote().
Since knote() needs the lock of the struct knlist object associated to
the to-be-wake-up netmap port, it is possible to have a lock order reversal
problem (AB/BA deadlock).
This change prevents the deadlock by executing the knote() call in a
per-selinfo taskqueue, where it is possible to hold a mutex.
Reviewed by: aleksandr.fedorov_itglobal.com
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D18956
On sync-kloop stop, send a wake-up signal to the kloop, so that
waiting for the timeout is not needed.
Also, improve logging in netmap_freebsd.c.
MFC after: 3 days
Changelist:
- Add the proper memory barriers in the kloop ring processing
functions.
- Fix memory barriers usage in the user helpers (nm_sync_kloop_appl_write,
nm_sync_kloop_appl_read).
- Fix nm_kr_txempty() helper to look at rhead rather than rcur. This
is important since the kloop can read a value of rcur which is ahead
of the value of rhead (see explanation in nm_sync_kloop_appl_write)
- Remove obsolete ptnetmap_guest_write_kring_csb() and
ptnet_guest_read_kring_csb(), and update if_ptnet(4) to use those.
- Prepare in advance the arguments for netmap_sync_kloop_[tr]x_ring(),
to make the kloop faster.
- Provide kernel and user implementation for nm_ldld_barrier() and
nm_ldst_barrier()
MFC after: 2 weeks
The nm_os_selwakeup function needs to call knote() to wake up kqueue(9)
users. However, this function can be called from different code paths,
with different lock requirements.
This patch fixes the knote() call argument to match the relavant lock state.
Also, comments have been updated to reflect current code.
PR: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=219846
Reported by: Aleksandr Fedorov <aleksandr.fedorov@itglobal.com>
Reviewed by: markj
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D18876
To check if txsync can be skipped, it is necessary to look for
unseen TX space. However, this means comparing ring->cur
against ring->tail, rather than ring->head against ring->tail
(like nm_ring_empty() does).
This change also adds some more comments to explain the optimization
performed at the beginning of netmap_poll().
MFC after: 3 days
Sponsored by: Sunny Valley Networks
The bug was introduced by r339639, although it is present in the upstream
netmap code since 2015. It is due to resetting the want_rx variable to
POLLIN, rather than resetting it to POLLIN|POLLRDNORM.
It only affects select(), which uses POLLRDNORM. poll() is not affected,
because it uses POLLIN.
Also, it only affects FreeBSD, because Linux skips the optimization
implemented by the piece of code where the bug occurs.
MFC after: 3 days
Sponsored by: Sunny Valley Networks
This code validates the netmap buf_size against the interface MTU
and maximum descriptor size, to make sure the values are consistent.
Moving this functionality to its own function is needed because this
function is also called by Linux-specific code.
MFC after: 3 days
The iflib subsystem implements netmap support in a driver-independent
way (sys/net/iflib.c). We can therefore remove the headers that
used to implement netmap support for all the drivers now supported
by iflib (em, igb, ixl, ixgbe, lem).
MFC after: 1 week
This allows tcpdump to capture outbound kernel packets while
in netmap mode
Submitted by: Marc de la Gueronniere <mdelagueronniere@verisign.com>
Reviewed by: vmaffione
MFC after: 1 week
Sponsored by: Verisign, Inc.
Differential Revision: https://reviews.freebsd.org/D17896
