Current NFQUEUE target uses a hash, computed over source and
destination address (and other parameters), for steering the packet
to the actual NFQUEUE. This, however forgets about the fact that the
packet eventually is handled by a particular CPU on user request.
If E. g.
1) IRQ affinity is used to handle packets on a particular CPU already
(both single-queue or multi-queue case)
and/or
2) RPS is used to steer packets to a specific softirq
the target easily chooses an NFQUEUE which is not handled by a process
pinned to the same CPU.
The idea is therefore to use the CPU index for determining the
NFQUEUE handling the packet.
E. g. when having a system with 4 CPUs, 4 MQ queues and 4 NFQUEUEs it
looks like this:
+-----+ +-----+ +-----+ +-----+
|NFQ#0| |NFQ#1| |NFQ#2| |NFQ#3|
+-----+ +-----+ +-----+ +-----+
^ ^ ^ ^
| |NFQUEUE | |
+ + + +
+-----+ +-----+ +-----+ +-----+
|rx-0 | |rx-1 | |rx-2 | |rx-3 |
+-----+ +-----+ +-----+ +-----+
The NFQUEUEs not necessarily have to start with number 0, setups with
less NFQUEUEs than packet-handling CPUs are not a problem as well.
This patch extends the NFQUEUE target to accept a new
NFQ_FLAG_CPU_FANOUT flag. If this is specified the target uses the
CPU index for determining the NFQUEUE being used. I have to introduce
rev3 for this. The 'flags' are folded into _v2 'bypass'.
By changing the way which queue is assigned, I'm able to improve the
performance if the processes reading on the NFQUEUs are pinned
correctly.
Signed-off-by: Holger Eitzenberger <holger@eitzenberger.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
For some events it's useful to weight sample with a hardware
provided number. This expresses how expensive the action the
sample represent was. This allows the profiler to scale
the samples to be more informative to the programmer.
There is already the period which is used similarly, but it
means something different, so I chose to not overload it.
Instead a new sample type for WEIGHT is added.
Can be used for multiple things. Initially it is used for TSX
abort costs and profiling by memory latencies (so to make
expensive load appear higher up in the histograms). The concept
is quite generic and can be extended to many other kinds of
events or architectures, as long as the hardware provides
suitable auxillary values. In principle it could be also used
for software tracepoints.
This adds the generic glue. A new optional sample format for a
64-bit weight value.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Stephane Eranian <eranian@google.com>
Cc: peterz@infradead.org
Cc: acme@redhat.com
Cc: jolsa@redhat.com
Cc: namhyung.kim@lge.com
Link: http://lkml.kernel.org/r/1359040242-8269-5-git-send-email-eranian@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Linux 3.9-rc5
* tag 'v3.9-rc5': (1080 commits)
Linux 3.9-rc5
Revert "lockdep: check that no locks held at freeze time"
dw_dmac: adjust slave_id accordingly to request line base
dmaengine: dw_dma: fix endianess for DT xlate function
PNP: List Rafael Wysocki as a maintainer
rbd: don't zero-fill non-image object requests
ia64 idle: delete stale (*idle)() function pointer
Btrfs: don't drop path when printing out tree errors in scrub
target: Fix RESERVATION_CONFLICT status regression for iscsi-target special case
tcm_vhost: Avoid VIRTIO_RING_F_EVENT_IDX feature bit
Revert "mm: introduce VM_POPULATE flag to better deal with racy userspace programs"
usb: ftdi_sio: Add support for Mitsubishi FX-USB-AW/-BD
mg_disk: fix error return code in mg_probe()
Btrfs: fix wrong return value of btrfs_lookup_csum()
Btrfs: fix wrong reservation of csums
Btrfs: fix double free in the btrfs_qgroup_account_ref()
Btrfs: limit the global reserve to 512mb
Btrfs: hold the ordered operations mutex when waiting on ordered extents
Btrfs: fix space accounting for unlink and rename
Btrfs: fix space leak when we fail to reserve metadata space
...
Currently, when a socket receives something on the error queue it only wakes up
the socket on select if it is in the "read" list, that is the socket has
something to read. It is useful also to wake the socket if it is in the error
list, which would enable software to wait on error queue packets without waking
up for regular data on the socket. The main use case is for receiving
timestamped transmit packets which return the timestamp to the socket via the
error queue. This enables an application to select on the socket for the error
queue only instead of for the regular traffic.
-v2-
* Added the SO_SELECT_ERR_QUEUE socket option to every architechture specific file
* Modified every socket poll function that checks error queue
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Cc: Jeffrey Kirsher <jeffrey.t.kirsher@intel.com>
Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Matthew Vick <matthew.vick@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds a menu option to the V4L2_CID_EXPOSURE_METERING
control for multi-zone metering.
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
It contains the public netlink interface bits required by userspace to
make use of the interface.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: Jesse Gross <jesse@nicira.com>
Add a base to be used for allocation of all the SI476X specific
controls in the corresponding driver.
Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Implement functionality to get the TLS pointer for the metag
architecture using regsets.
This provides multi-threaded debug support for GDB.
Signed-off-by: Paul Clothier <Paul.Clothier@imgtec.com>
Signed-off-by: James Hogan <james.hogan@imgtec.com>
NLMSG_HDRLEN is already aligned value. It's for directly reference
without extra alignment.
The redundant alignment here may confuse the API users.
Signed-off-by: Hong Zhiguo <honkiko@gmail.com>
Acked-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Mark writes:
ASoC/extcon: arizona: Fix interaction between HPDET and headphone outputs
This patch series covers both ASoC and extcon subsystems and fixes an
interaction between the HPDET function and the headphone outputs - we
really shouldn't run HPDET while the headphone is active. The first
patch is a refactoring to make the extcon side easier.
Userspace applications need to know the maximum supported message
size.
The cdc-wdm driver translates between a character device stream
and a message based protocol. Each message is transported as a
usb control message with no further encapsulation or syncronization.
Each read or write on the character device should translate to
exactly one usb control message to ensure that message boundaries
are kept intact. That means that the userspace application must
know the maximum message size supported by the device and driver,
making this size a vital part of the cdc-wdm character device API.
CDC WDM and CDC MBIM functions export the maximum supported
message size through CDC functional descriptors. The cdc-wdm and
cdc_mbim drivers will parse these descriptors and use the value
chosen by the device. The only current way for a userspace
application to retrive the value is by duplicating the descriptor
parsing. This is an unnecessary complex task, and application
writers are likely to postpone it, using a fixed value and adding
a "todo" item.
QMI functions have no way to tell the host what message size they
support. The qmi_wwan driver use a fixed value based on protocol
recommendations and observed device behaviour. Userspace
applications must know and hard code the same value. This scheme
will break if we ever encounter a QMI device needing a device
specific message size quirk. We are currently unable to support
such a device because using a non default size would break the
implicit userspace API.
The message size is currently a hidden attribute of the cdc-wdm
userspace API. Retrieving it is unnecessarily complex, increasing
the possibility of drivers and applications using different limits.
The resulting errors are hard to debug, and can only be replicated
on identical hardware.
Exporting the maximum message size from the driver simplifies the
task for the userspace application, and creates a unified
information source independent of device and function class. It also
serves to document that the message size is part of the cdc-wdm
userspace API.
This proposed API extension has been presented for the authors of
userspace applications and libraries using the current API: libmbim,
libqmi, uqmi, oFono and ModemManager. The replies were:
Aleksander Morgado:
"We do really need max message size for MBIM; and as you say, it may be
good to have the max message size info also for QMI, so the new ioctl
seems a good addition. So +1 from my side, for what it's worth."
Dan Williams:
"Yeah, +1 here. I'd prefer the sysfs file, but the fact that that
doesn't work for fd passing pretty much kills it."
No negative replies are so far received.
Cc: Aleksander Morgado <aleksander@lanedo.com>
Cc: Dan Williams <dcbw@redhat.com>
Signed-off-by: Bjørn Mork <bjorn@mork.no>
Acked-by: Oliver Neukum <oliver@neukum.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Simplify the debugging ioctls by creating the VIDIOC_DBG_G_CHIP_NAME ioctl.
This will eventually replace VIDIOC_DBG_G_CHIP_IDENT. Chip matching is done
by the name or index of subdevices or an index to a bridge chip. Most of this
can all be done automatically, so most drivers just need to provide get/set
register ops.
In particular, it is now possible to get/set subdev registers without
requiring assistance of the bridge driver.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Control whether video sequence headers should be repeated.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Pull to get the thermal netlink multicast group name fix, otherwise
the assertion added in net-next to netlink to detect that kind of bug
makes systems unbootable for some folks.
Signed-off-by: David S. Miller <davem@davemloft.net>
A lot of SOCs including Texas Instruments Davinci family mainly use
video decoders as input devices. This patch adds a flag
'MEDIA_ENT_T_V4L2_SUBDEV_DECODER' media entity type for decoder's.
Along side updates the documentation for this media entity type.
Signed-off-by: Manjunath Hadli <manjunath.hadli@ti.com>
Signed-off-by: Lad, Prabhakar <prabhakar.lad@ti.com>
Reviewed-by: Sylwester Nawrocki <s.nawrocki@samsung.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
The netlink_diag can be built as a module, just like it's done in
unix sockets.
The core dumping message carries the basic info about netlink sockets:
family, type and protocol, portis, dst_group, dst_portid, state.
Groups can be received as an optional parameter NETLINK_DIAG_GROUPS.
Netlink sockets cab be filtered by protocols.
The socket inode number and cookie is reserved for future per-socket info
retrieving. The per-protocol filtering is also reserved for future by
requiring the sdiag_protocol to be zero.
The file /proc/net/netlink doesn't provide enough information for
dumping netlink sockets. It doesn't provide dst_group, dst_portid,
groups above 32.
v2: fix NETLINK_DIAG_MAX. Now it's equal to the last constant.
Acked-by: Pavel Emelyanov <xemul@parallels.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Pablo Neira Ayuso <pablo@netfilter.org>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Gao feng <gaofeng@cn.fujitsu.com>
Cc: Thomas Graf <tgraf@suug.ch>
Signed-off-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Merge reason:
From: Alexander Graf <agraf@suse.de>
"Just recently this really important patch got pulled into Linus' tree for 3.9:
commit 1674400aae
Author: Anton Blanchard <anton <at> samba.org>
Date: Tue Mar 12 01:51:51 2013 +0000
Without that commit, I can not boot my G5, thus I can't run automated tests on it against my queue.
Could you please merge kvm/next against linus/master, so that I can base my trees against that?"
* upstream/master: (653 commits)
PCI: Use ROM images from firmware only if no other ROM source available
sparc: remove unused "config BITS"
sparc: delete "if !ULTRA_HAS_POPULATION_COUNT"
KVM: Fix bounds checking in ioapic indirect register reads (CVE-2013-1798)
KVM: x86: Convert MSR_KVM_SYSTEM_TIME to use gfn_to_hva_cache functions (CVE-2013-1797)
KVM: x86: fix for buffer overflow in handling of MSR_KVM_SYSTEM_TIME (CVE-2013-1796)
arm64: Kconfig.debug: Remove unused CONFIG_DEBUG_ERRORS
arm64: Do not select GENERIC_HARDIRQS_NO_DEPRECATED
inet: limit length of fragment queue hash table bucket lists
qeth: Fix scatter-gather regression
qeth: Fix invalid router settings handling
qeth: delay feature trace
sgy-cts1000: Remove __dev* attributes
KVM: x86: fix deadlock in clock-in-progress request handling
KVM: allow host header to be included even for !CONFIG_KVM
hwmon: (lm75) Fix tcn75 prefix
hwmon: (lm75.h) Update header inclusion
MAINTAINERS: Remove Mark M. Hoffman
xfs: ensure we capture IO errors correctly
xfs: fix xfs_iomap_eof_prealloc_initial_size type
...
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Process connector can now also detect coredumping events.
Main aim of patch is get notified at start of coredumping, instead of
having to wait for it to finish and then being notified through EXIT
event.
Could be used for instance by process-managers that want to get
notified as soon as possible about process failures, and not
necessarily beeing notified after coredump, which could be in the
order of minutes depending on size of coredump, piping and so on.
Signed-off-by: Jesper Derehag <jderehag@hotmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
It is very useful to do dynamic truncation of packets. In particular,
we're interested to push the necessary header bytes to the user space and
cut off user payload that should probably not be transferred for some reasons
(e.g. privacy, speed, or others). With the ancillary extension PAY_OFFSET,
we can load it into the accumulator, and return it. E.g. in bpfc syntax ...
ld #poff ; { 0x20, 0, 0, 0xfffff034 },
ret a ; { 0x16, 0, 0, 0x00000000 },
... as a filter will accomplish this without having to do a big hackery in
a BPF filter itself. Follow-up JIT implementations are welcome.
Thanks to Eric Dumazet for suggesting and discussing this during the
Netfilter Workshop in Copenhagen.
Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add the CAIF Virtio shared memory driver for talking
to a modem.
This CAIF Link layer communicates to the modem over
shared memory. It is implemented as a virtio_driver.
The underlying virtio device is managed by the remoteproc
framework. The Virtio queue is used for transmitting data
to the modem, and the new vringh is used for receiving data.
Genalloc is used for managing the shared memory used for TX
data. The default dma-alloc-coherent allocator can only
allocate whole pages, and this wastes too much shared memory.
Flow control is implemented by stopping the TX-queues if the
virtio queues go full or we run out of memory. Queued are
reopened when queues are below the watermark.
NAPI is used in RX path, and a dedicated tasklet is used
for releasing TX buffers.
Signed-off-by: Erwan Yvin <erwan.yvin@stericsson.com>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> (minor fixes)
Changes:
v3->v2: rebase (no other changes)
passes selftest
v2->v1: read f->num_members only once
fix bug: test rollover mode + flag
Minimize packet drop in a fanout group. If one socket is full,
roll over packets to another from the group. Maintain flow
affinity during normal load using an rxhash fanout policy, while
dispersing unexpected traffic storms that hit a single cpu, such
as spoofed-source DoS flows. Rollover breaks affinity for flows
arriving at saturated sockets during those conditions.
The patch adds a fanout policy ROLLOVER that rotates between sockets,
filling each socket before moving to the next. It also adds a fanout
flag ROLLOVER. If passed along with any other fanout policy, the
primary policy is applied until the chosen socket is full. Then,
rollover selects another socket, to delay packet drop until the
entire system is saturated.
Probing sockets is not free. Selecting the last used socket, as
rollover does, is a greedy approach that maximizes chance of
success, at the cost of extreme load imbalance. In practice, with
sufficiently long queues to absorb bursts, sockets are drained in
parallel and load balance looks uniform in `top`.
To avoid contention, scales counters with number of sockets and
accesses them lockfree. Values are bounds checked to ensure
correctness.
Tested using an application with 9 threads pinned to CPUs, one socket
per thread and sufficient busywork per packet operation to limits each
thread to handling 32 Kpps. When sent 500 Kpps single UDP stream
packets, a FANOUT_CPU setup processes 32 Kpps in total without this
patch, 270 Kpps with the patch. Tested with read() and with a packet
ring (V1).
Also, passes psock_fanout.c unit test added to selftests.
Signed-off-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
TCPCT uses option-number 253, reserved for experimental use and should
not be used in production environments.
Further, TCPCT does not fully implement RFC 6013.
As a nice side-effect, removing TCPCT increases TCP's performance for
very short flows:
Doing an apache-benchmark with -c 100 -n 100000, sending HTTP-requests
for files of 1KB size.
before this patch:
average (among 7 runs) of 20845.5 Requests/Second
after:
average (among 7 runs) of 21403.6 Requests/Second
Signed-off-by: Christoph Paasch <christoph.paasch@uclouvain.be>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch generalizes VXLAN forwarding table entries allowing an administrator
to:
1) specify multiple destinations for a given MAC
2) specify alternate vni's in the VXLAN header
3) specify alternate destination UDP ports
4) use multicast MAC addresses as fdb lookup keys
5) specify multicast destinations
6) specify the outgoing interface for forwarded packets
The combination allows configuration of more complex topologies using VXLAN
encapsulation.
Changes since v1: rebase to 3.9.0-rc2
Signed-Off-By: David L Stevens <dlstevens@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This driver supports host initiated backup of the guest. On Windows guests,
the host can generate application consistent backups using the Windows VSS
framework. On Linux, we ensure that the backup will be file system consistent.
This driver allows the host to initiate a "Freeze" operation on all the mounted
file systems in the guest. Once the mounted file systems in the guest are frozen,
the host snapshots the guest's file systems. Once this is done, the guest's file
systems are "thawed".
This driver has a user-level component (daemon) that invokes the appropriate
operation on all the mounted file systems in response to the requests from
the host. The duration for which the guest is frozen is very short - a few seconds.
During this interval, the diff disk is comitted.
In this version of the patch I have addressed the feedback from Olaf Herring.
Also, some of the connector related issues have been fixed.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Reviewed-by: Haiyang Zhang <haiyangz@microsoft.com>
Cc: Evgeniy Polyakov <zbr@ioremap.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Merge misc fixes from Andrew Morton:
- A bunch of fixes
- Finish off the idr API conversions before someone starts to use the
old interfaces again.
* emailed patches from Andrew Morton <akpm@linux-foundation.org>:
idr: idr_alloc() shouldn't trigger lowmem warning when preloaded
UAPI: fix endianness conditionals in M32R's asm/stat.h
UAPI: fix endianness conditionals in linux/raid/md_p.h
UAPI: fix endianness conditionals in linux/acct.h
UAPI: fix endianness conditionals in linux/aio_abi.h
decompressors: fix typo "POWERPC"
mm/fremap.c: fix oops on error path
idr: deprecate idr_pre_get() and idr_get_new[_above]()
tidspbridge: convert to idr_alloc()
zcache: convert to idr_alloc()
mlx4: remove leftover idr_pre_get() call
workqueue: convert to idr_alloc()
nfsd: convert to idr_alloc()
nfsd: remove unused get_new_stid()
kernel/signal.c: use __ARCH_HAS_SA_RESTORER instead of SA_RESTORER
signal: always clear sa_restorer on execve
mm: remove_memory(): fix end_pfn setting
include/linux/res_counter.h needs errno.h
In the UAPI header files, __BIG_ENDIAN and __LITTLE_ENDIAN must be
compared against __BYTE_ORDER in preprocessor conditionals where these are
exposed to userspace (that is they're not inside __KERNEL__ conditionals).
However, in the main kernel the norm is to check for
"defined(__XXX_ENDIAN)" rather than comparing against __BYTE_ORDER and
this has incorrectly leaked into the userspace headers.
The definition of struct mdp_superblock_s in linux/raid/md_p.h is wrong in
this way. Note that userspace will likely interpret the ordering of the
fields incorrectly as the big-endian variant on a little-endian machines -
depending on header inclusion order.
[!!!] NOTE [!!!] This patch may adversely change the userspace API. It might
be better to fix the ordering of events_hi, events_lo, cp_events_hi and
cp_events_lo in struct mdp_superblock_s / typedef mdp_super_t.
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: NeilBrown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
In the UAPI header files, __BIG_ENDIAN and __LITTLE_ENDIAN must be
compared against __BYTE_ORDER in preprocessor conditionals where these are
exposed to userspace (that is they're not inside __KERNEL__ conditionals).
However, in the main kernel the norm is to check for
"defined(__XXX_ENDIAN)" rather than comparing against __BYTE_ORDER and
this has incorrectly leaked into the userspace headers.
The definition of ACCT_BYTEORDER in linux/acct.h is wrong in this way.
Note that userspace will likely interpret this incorrectly as the
big-endian variant on little-endian machines - depending on header
inclusion order.
[!!!] NOTE [!!!] This patch may adversely change the userspace API. It might
be better to fix the value of ACCT_BYTEORDER.
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
In the UAPI header files, __BIG_ENDIAN and __LITTLE_ENDIAN must be
compared against __BYTE_ORDER in preprocessor conditionals where these are
exposed to userspace (that is they're not inside __KERNEL__ conditionals).
However, in the main kernel the norm is to check for
"defined(__XXX_ENDIAN)" rather than comparing against __BYTE_ORDER and
this has incorrectly leaked into the userspace headers.
The definition of PADDED() in linux/aio_abi.h is wrong in this way. Note
that userspace will likely interpret this and thus the order of fields in
struct iocb incorrectly as the little-endian variant on big-endian
machines - depending on header inclusion order.
[!!!] NOTE [!!!] This patch may adversely change the userspace API. It might
be better to fix the ordering of aio_key and aio_reserved1 in struct iocb.
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Benjamin LaHaise <bcrl@kvack.org>
Acked-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Entertainment systems used in aircraft need additional keycodes for their
Passenger Control Units, so let's add them.
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
This is the second of the TLP patch series; it augments the basic TLP
algorithm with a loss detection scheme.
This patch implements a mechanism for loss detection when a Tail
loss probe retransmission plugs a hole thereby masking packet loss
from the sender. The loss detection algorithm relies on counting
TLP dupacks as outlined in Sec. 3 of:
http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01
The basic idea is: Sender keeps track of TLP "episode" upon
retransmission of a TLP packet. An episode ends when the sender receives
an ACK above the SND.NXT (tracked by tlp_high_seq) at the time of the
episode. We want to make sure that before the episode ends the sender
receives a "TLP dupack", indicating that the TLP retransmission was
unnecessary, so there was no loss/hole that needed plugging. If the
sender gets no TLP dupack before the end of the episode, then it reduces
ssthresh and the congestion window, because the TLP packet arriving at
the receiver probably plugged a hole.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
- New VFIO_SET_IRQ ioctl option to pass the eventfd that is signaled when
an error occurs in the vfio_pci_device
- Register pci_error_handler for the vfio_pci driver
- When the device encounters an error, the error handler registered by
the vfio_pci driver gets invoked by the AER infrastructure
- In the error handler, signal the eventfd registered for the device.
- This results in the qemu eventfd handler getting invoked and
appropriate action taken for the guest.
Signed-off-by: Vijay Mohan Pandarathil <vijaymohan.pandarathil@hp.com>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
This adds a netlink interface for service name lookup support.
Multiple URIs can be passed nested into the NFC_ATTR_LLC_SDP attribute
using the NFC_CMD_LLC_SDREQ netlink command.
When the SNL reply is received, a NFC_EVENT_LLC_SDRES event is sent to
the user space. URI and SAP tuples are passed back, nested into
NFC_ATTR_LLC_SDP attribute.
Signed-off-by: Thierry Escande <thierry.escande@linux.intel.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Some LLCP services (e.g. the validation ones) require some control over
the LLCP link parameters like the receive window (RW) or the MIU extension
(MIUX). This can only be done through socket options.
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
Split the vSockets header into kernel and UAPI parts. The former gets the bits
that used to be in __KERNEL__ guards, while the latter gets everything that is
user-visible. Tested by compiling vsock (+transport) and a simple user-mode
vSockets application.
Reported-by: David Howells <dhowells@redhat.com>
Acked-by: Dmitry Torokhov <dtor@vmware.com>
Signed-off-by: Andy King <acking@vmware.com>
Acked-by: David Howells <dhowells@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
