Age | Commit message (Collapse) | Author |
|
Updates #2243
|
|
Software GSO implementation currently has a complicated code path with
implicit assumptions that all packets to WritePackets carry same Data
and it does this to avoid allocations on the path etc. But this makes it
hard to reuse the WritePackets API.
This change breaks all such assumptions by introducing a new Vectorised
View API ReadToVV which can be used to cleanly split a VV into multiple
independent VVs. Further this change also makes packet buffers linkable
to form an intrusive list. This allows us to get rid of the array of
packet buffers that are passed in the WritePackets API call and replace
it with a list of packet buffers.
While this code does introduce some more allocations in the benchmarks
it doesn't cause any degradation.
Updates #231
PiperOrigin-RevId: 304731742
|
|
This allows the link layer endpoints to consistenly hash a TCP
segment to a single underlying queue in case a link layer endpoint
does support multiple underlying queues.
Updates #231
PiperOrigin-RevId: 302760664
|
|
This is a precursor to be being able to build an intrusive list
of PacketBuffers for use in queuing disciplines being implemented.
Updates #2214
PiperOrigin-RevId: 302677662
|
|
PiperOrigin-RevId: 301157950
|
|
PiperOrigin-RevId: 300832988
|
|
PiperOrigin-RevId: 291745021
|
|
* Rename syncutil to sync.
* Add aliases to sync types.
* Replace existing usage of standard library sync package.
This will make it easier to swap out synchronization primitives. For example,
this will allow us to use primitives from github.com/sasha-s/go-deadlock to
check for lock ordering violations.
Updates #1472
PiperOrigin-RevId: 289033387
|
|
PiperOrigin-RevId: 282194656
|
|
PiperOrigin-RevId: 282045221
|
|
PiperOrigin-RevId: 280455453
|
|
PacketBuffers are analogous to Linux's sk_buff. They hold all information about
a packet, headers, and payload. This is important for:
* iptables to access various headers of packets
* Preventing the clutter of passing different net and link headers along with
VectorisedViews to packet handling functions.
This change only affects the incoming packet path, and a future change will
change the outgoing path.
Benchmark Regular PacketBufferPtr PacketBufferConcrete
--------------------------------------------------------------------------------
BM_Recvmsg 400.715MB/s 373.676MB/s 396.276MB/s
BM_Sendmsg 361.832MB/s 333.003MB/s 335.571MB/s
BM_Recvfrom 453.336MB/s 393.321MB/s 381.650MB/s
BM_Sendto 378.052MB/s 372.134MB/s 341.342MB/s
BM_SendmsgTCP/0/1k 353.711MB/s 316.216MB/s 322.747MB/s
BM_SendmsgTCP/0/2k 600.681MB/s 588.776MB/s 565.050MB/s
BM_SendmsgTCP/0/4k 995.301MB/s 888.808MB/s 941.888MB/s
BM_SendmsgTCP/0/8k 1.517GB/s 1.274GB/s 1.345GB/s
BM_SendmsgTCP/0/16k 1.872GB/s 1.586GB/s 1.698GB/s
BM_SendmsgTCP/0/32k 1.017GB/s 1.020GB/s 1.133GB/s
BM_SendmsgTCP/0/64k 475.626MB/s 584.587MB/s 627.027MB/s
BM_SendmsgTCP/0/128k 416.371MB/s 503.434MB/s 409.850MB/s
BM_SendmsgTCP/0/256k 323.449MB/s 449.599MB/s 388.852MB/s
BM_SendmsgTCP/0/512k 243.992MB/s 267.676MB/s 314.474MB/s
BM_SendmsgTCP/0/1M 95.138MB/s 95.874MB/s 95.417MB/s
BM_SendmsgTCP/0/2M 96.261MB/s 94.977MB/s 96.005MB/s
BM_SendmsgTCP/0/4M 96.512MB/s 95.978MB/s 95.370MB/s
BM_SendmsgTCP/0/8M 95.603MB/s 95.541MB/s 94.935MB/s
BM_SendmsgTCP/0/16M 94.598MB/s 94.696MB/s 94.521MB/s
BM_SendmsgTCP/0/32M 94.006MB/s 94.671MB/s 94.768MB/s
BM_SendmsgTCP/0/64M 94.133MB/s 94.333MB/s 94.746MB/s
BM_SendmsgTCP/0/128M 93.615MB/s 93.497MB/s 93.573MB/s
BM_SendmsgTCP/0/256M 93.241MB/s 95.100MB/s 93.272MB/s
BM_SendmsgTCP/1/1k 303.644MB/s 316.074MB/s 308.430MB/s
BM_SendmsgTCP/1/2k 537.093MB/s 584.962MB/s 529.020MB/s
BM_SendmsgTCP/1/4k 882.362MB/s 939.087MB/s 892.285MB/s
BM_SendmsgTCP/1/8k 1.272GB/s 1.394GB/s 1.296GB/s
BM_SendmsgTCP/1/16k 1.802GB/s 2.019GB/s 1.830GB/s
BM_SendmsgTCP/1/32k 2.084GB/s 2.173GB/s 2.156GB/s
BM_SendmsgTCP/1/64k 2.515GB/s 2.463GB/s 2.473GB/s
BM_SendmsgTCP/1/128k 2.811GB/s 3.004GB/s 2.946GB/s
BM_SendmsgTCP/1/256k 3.008GB/s 3.159GB/s 3.171GB/s
BM_SendmsgTCP/1/512k 2.980GB/s 3.150GB/s 3.126GB/s
BM_SendmsgTCP/1/1M 2.165GB/s 2.233GB/s 2.163GB/s
BM_SendmsgTCP/1/2M 2.370GB/s 2.219GB/s 2.453GB/s
BM_SendmsgTCP/1/4M 2.005GB/s 2.091GB/s 2.214GB/s
BM_SendmsgTCP/1/8M 2.111GB/s 2.013GB/s 2.109GB/s
BM_SendmsgTCP/1/16M 1.902GB/s 1.868GB/s 1.897GB/s
BM_SendmsgTCP/1/32M 1.655GB/s 1.665GB/s 1.635GB/s
BM_SendmsgTCP/1/64M 1.575GB/s 1.547GB/s 1.575GB/s
BM_SendmsgTCP/1/128M 1.524GB/s 1.584GB/s 1.580GB/s
BM_SendmsgTCP/1/256M 1.579GB/s 1.607GB/s 1.593GB/s
PiperOrigin-RevId: 278940079
|
|
When VectorisedViews were passed up the stack from packet_dispatchers, we were
passing a sub-slice of the dispatcher's views fields. The dispatchers then
immediately set those views to nil.
This wasn't caught before because every implementer copied the data in these
views before returning.
PiperOrigin-RevId: 277615351
|
|
Right now, we send each tcp packet separately, we call one system
call per-packet. This patch allows to generate multiple tcp packets
and send them by sendmmsg.
The arguable part of this CL is a way how to handle multiple headers.
This CL adds the next field to the Prepandable buffer.
Nginx test results:
Server Software: nginx/1.15.9
Server Hostname: 10.138.0.2
Server Port: 8080
Document Path: /10m.txt
Document Length: 10485760 bytes
w/o gso:
Concurrency Level: 5
Time taken for tests: 5.491 seconds
Complete requests: 100
Failed requests: 0
Total transferred: 1048600200 bytes
HTML transferred: 1048576000 bytes
Requests per second: 18.21 [#/sec] (mean)
Time per request: 274.525 [ms] (mean)
Time per request: 54.905 [ms] (mean, across all concurrent requests)
Transfer rate: 186508.03 [Kbytes/sec] received
sw-gso:
Concurrency Level: 5
Time taken for tests: 3.852 seconds
Complete requests: 100
Failed requests: 0
Total transferred: 1048600200 bytes
HTML transferred: 1048576000 bytes
Requests per second: 25.96 [#/sec] (mean)
Time per request: 192.576 [ms] (mean)
Time per request: 38.515 [ms] (mean, across all concurrent requests)
Transfer rate: 265874.92 [Kbytes/sec] received
w/o gso:
$ ./tcp_benchmark --client --duration 15 --ideal
[SUM] 0.0-15.1 sec 2.20 GBytes 1.25 Gbits/sec
software gso:
$ tcp_benchmark --client --duration 15 --ideal --gso $((1<<16)) --swgso
[SUM] 0.0-15.1 sec 3.99 GBytes 2.26 Gbits/sec
PiperOrigin-RevId: 276112677
|
|
Like (AF_INET, SOCK_RAW) sockets, AF_PACKET sockets require CAP_NET_RAW. With
runsc, you'll need to pass `--net-raw=true` to enable them.
Binding isn't supported yet.
PiperOrigin-RevId: 275909366
|
|
PiperOrigin-RevId: 274638272
|
|
PiperOrigin-RevId: 272083936
|
|
Previously, the only safe way to use an fdbased endpoint was to leak the FD.
This change makes it possible to safely close the FD.
This is the first step towards having stoppable stacks.
Updates #837
PiperOrigin-RevId: 270346582
|
|
They are no-ops, so the standard rule works fine.
PiperOrigin-RevId: 268776264
|
|
PiperOrigin-RevId: 267709597
|
|
Signed-off-by: Haibo Xu haibo.xu@arm.com
Change-Id: Ib6b4aa2db19032e58bf0395f714e6883caee460a
|
|
Signed-off-by: Haibo Xu haibo.xu@arm.com
Change-Id: Id4489554b9caa332695df8793d361f8332f6a13b
|
|
This stub had the wrong function signature.
PiperOrigin-RevId: 262992682
|
|
syscall.POLL is not supported on arm64, using syscall.PPOLL
to support both the x86 and arm64. refs #63
Signed-off-by: Haibo Xu <haibo.xu@arm.com>
Change-Id: I2c81a063d3ec4e7e6b38fe62f17a0924977f505e
COPYBARA_INTEGRATE_REVIEW=https://github.com/google/gvisor/pull/543 from xiaobo55x:master ba598263fd3748d1addd48e4194080aa12085164
PiperOrigin-RevId: 260752049
|
|
This can be merged after:
https://github.com/google/gvisor-website/pull/77
or
https://github.com/google/gvisor-website/pull/78
PiperOrigin-RevId: 253132620
|
|
This allows an fdbased endpoint to have multiple underlying fd's from which
packets can be read and dispatched/written to.
This should allow for higher throughput as well as better scalability of the
network stack as number of connections increases.
Updates #231
PiperOrigin-RevId: 251852825
|
|
This is in preparation to support an fdbased endpoint that can read/dispatch
packets from multiple underlying fds.
Updates #231
PiperOrigin-RevId: 249337074
Change-Id: Id7d375186cffcf55ae5e38986e7d605a96916d35
|
|
Based on the guidelines at
https://opensource.google.com/docs/releasing/authors/.
1. $ rg -l "Google LLC" | xargs sed -i 's/Google LLC.*/The gVisor Authors./'
2. Manual fixup of "Google Inc" references.
3. Add AUTHORS file. Authors may request to be added to this file.
4. Point netstack AUTHORS to gVisor AUTHORS. Drop CONTRIBUTORS.
Fixes #209
PiperOrigin-RevId: 245823212
Change-Id: I64530b24ad021a7d683137459cafc510f5ee1de9
|
|
Packet socket receive buffers default to the sysctl value of
net.core.rmem_default and are capped by net.core.rmem_max both
which are usually set to 208KB on most systems.
Since we can't expect every gVisor user to bump these we use
SO_RCVBUFFORCE to exceed the limit. This is possible as runsc runs
with CAP_NET_ADMIN outside the sandbox and can do this before
the FD is passed to the sentry inside the sandbox.
Updates #211
iperf output w/ 4MB buffer.
iperf3 -c 172.17.0.2 -t 100
Connecting to host 172.17.0.2, port 5201
[ 4] local 172.17.0.1 port 40378 connected to 172.17.0.2 port 5201
[ ID] Interval Transfer Bandwidth Retr Cwnd
[ 4] 0.00-1.00 sec 1.15 GBytes 9.89 Gbits/sec 0 1.02 MBytes
[ 4] 1.00-2.00 sec 1.18 GBytes 10.2 Gbits/sec 0 1.02 MBytes
[ 4] 2.00-3.00 sec 965 MBytes 8.09 Gbits/sec 0 1.02 MBytes
[ 4] 3.00-4.00 sec 942 MBytes 7.90 Gbits/sec 0 1.02 MBytes
[ 4] 4.00-5.00 sec 952 MBytes 7.99 Gbits/sec 0 1.02 MBytes
[ 4] 5.00-6.00 sec 1.14 GBytes 9.81 Gbits/sec 0 1.02 MBytes
[ 4] 6.00-7.00 sec 1.13 GBytes 9.68 Gbits/sec 0 1.02 MBytes
[ 4] 7.00-8.00 sec 930 MBytes 7.80 Gbits/sec 0 1.02 MBytes
[ 4] 8.00-9.00 sec 1.15 GBytes 9.91 Gbits/sec 0 1.02 MBytes
[ 4] 9.00-10.00 sec 938 MBytes 7.87 Gbits/sec 0 1.02 MBytes
[ 4] 10.00-11.00 sec 737 MBytes 6.18 Gbits/sec 0 1.02 MBytes
[ 4] 11.00-12.00 sec 1.16 GBytes 9.93 Gbits/sec 0 1.02 MBytes
[ 4] 12.00-13.00 sec 917 MBytes 7.69 Gbits/sec 0 1.02 MBytes
[ 4] 13.00-14.00 sec 1.19 GBytes 10.2 Gbits/sec 0 1.02 MBytes
[ 4] 14.00-15.00 sec 1.01 GBytes 8.70 Gbits/sec 0 1.02 MBytes
[ 4] 15.00-16.00 sec 1.20 GBytes 10.3 Gbits/sec 0 1.02 MBytes
[ 4] 16.00-17.00 sec 1.14 GBytes 9.80 Gbits/sec 0 1.02 MBytes
^C[ 4] 17.00-17.60 sec 718 MBytes 10.1 Gbits/sec 0 1.02 MBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval Transfer Bandwidth Retr
[ 4] 0.00-17.60 sec 18.4 GBytes 8.98 Gbits/sec 0 sender
[ 4] 0.00-17.60 sec 0.00 Bytes 0.00 bits/sec receiver
PiperOrigin-RevId: 245470590
Change-Id: I1c08c5ee8345de6ac070513656a4703312dc3c00
|
|
This CL fixes the following bugs:
- Uses atomic to set/read status instead of binary.LittleEndian.PutUint32 etc
which are not atomic.
- Increments ringOffsets for frames that are truncated (i.e status is
tpStatusCopy)
- Does not ignore frames with tpStatusLost bit set as they are valid frames and
only indicate that there some frames were lost before this one and metrics can
be retrieved with a getsockopt call.
- Adds checks to make sure blockSize is a multiple of page size. This is
required as the kernel allocates in pages per block and rejects sizes that are
not page aligned with an EINVAL.
Updates #210
PiperOrigin-RevId: 244959464
Change-Id: I5d61337b7e4c0f8a3063dcfc07791d4c4521ba1f
|
|
It is possible to create a listening socket which will accept
IPv4 and IPv6 connections. In this case, we set IPv6ProtocolNumber
for all accepted endpoints, even if they handle IPv4 connections.
This means that we can't use endpoint.netProto to set gso.L3HdrLen.
PiperOrigin-RevId: 244227948
Change-Id: I5e1863596cb9f3d216febacdb7dc75651882eef1
|
|
RELNOTES: n/a
PiperOrigin-RevId: 244031742
Change-Id: Id0cdb73194018fb5979e67b58510ead19b5a2b81
|
|
PiperOrigin-RevId: 242704699
Change-Id: I87db368ca343b3b4bf4f969b17d3aa4ce2f8bd4f
|
|
The linux packet socket can handle GSO packets, so we can segment packets to
64K instead of the MTU which is usually 1500.
Here are numbers for the nginx-1m test:
runsc: 579330.01 [Kbytes/sec] received
runsc-gso: 1794121.66 [Kbytes/sec] received
runc: 2122139.06 [Kbytes/sec] received
and for tcp_benchmark:
$ tcp_benchmark --duration 15 --ideal
[ 4] 0.0-15.0 sec 86647 MBytes 48456 Mbits/sec
$ tcp_benchmark --client --duration 15 --ideal
[ 4] 0.0-15.0 sec 2173 MBytes 1214 Mbits/sec
$ tcp_benchmark --client --duration 15 --ideal --gso 65536
[ 4] 0.0-15.0 sec 19357 MBytes 10825 Mbits/sec
PiperOrigin-RevId: 240809103
Change-Id: I2637f104db28b5d4c64e1e766c610162a195775a
|
|
Previous memory allocation was excessive (80 MB). Changed
it to use 2 MB instead. There is no drop in perfomance due
to this change:
ab -n 100 -c 10 http://server/latin10m.txt ==> 10 MB file
80 MB: 178 MB/s
2 MB: 181 MB/s
PiperOrigin-RevId: 238321594
Change-Id: I1c8aed13cad5d75f4506d2b406b305117055fbe5
|
|
HandleLocal is very similar conceptually to MULTICAST_LOOP, so we can unify
the implementations. This has the benefit of making HandleLocal apply even when
the fdbased link endpoint isn't in use.
In addition, move looping logic to route creation so that it doesn't need to be
run for each packet. This should improve performance.
PiperOrigin-RevId: 238099480
Change-Id: I72839f16f25310471453bc9d3fb8544815b25c23
|
|
Also exposes ipv4.MaxTotalSize since it is a generally useful constant.
PiperOrigin-RevId: 235799755
Change-Id: I1fa8d5294bf355acf5527cfdf274b3687d3c8b13
|
|
PACKET_RX_RING allows the use of an mmapped buffer to receive packets from the
kernel. This should cut down the number of host syscalls that need to be made
to receive packets when the underlying fd is a socket of the AF_PACKET type.
PiperOrigin-RevId: 233834998
Change-Id: I8060025c6ced206986e94cc46b8f382b81bfa47f
|
|
Nothing reads them and they can simply get stale.
Generated with:
$ sed -i "s/licenses(\(.*\)).*/licenses(\1)/" **/BUILD
PiperOrigin-RevId: 231818945
Change-Id: Ibc3f9838546b7e94f13f217060d31f4ada9d4bf0
|
|
This should reduce the number of syscalls required to process packets
significantly and improve throughputs.
PiperOrigin-RevId: 231366886
Change-Id: I8b38077262bf9c53176bc4a94b530188d3d7c0ca
|
|
...to (remote, local), reflecting the (correct) names in the implementation of
DeliverNetworkPacket (see tcpip/stack/nic.go).
Also trim the names in DeliverNetworkPacket and elsewhere to avoid stuttering;
since the type is tcpip.LinkAddress, there's no need to include "LinkAddr" in
the parameter names.
Note that every callsite passes arguments in the order (src, dst).
PiperOrigin-RevId: 221514396
Change-Id: I3637454ad0d6e62a19e4dcbc2a16493798bd0f09
|
|
This change also adds extensive testing to the p9 package via mocks. The sanity
checks and type checks are moved from the gofer into the core package, where
they can be more easily validated.
PiperOrigin-RevId: 218296768
Change-Id: I4fc3c326e7bf1e0e140a454cbacbcc6fd617ab55
|
|
PiperOrigin-RevId: 217951017
Change-Id: Ie08bf6987f98467d07457bcf35b5f1ff6e43c035
|
|
This allows a NetworkDispatcher to implement transparent bridging,
assuming all implementations of LinkEndpoint.WritePacket call eth.Encode
with header.EthernetFields.SrcAddr set to the passed
Route.LocalLinkAddress, if it is provided.
PiperOrigin-RevId: 213686651
Change-Id: I446a4ac070970202f0724ef796ff1056ae4dd72a
|
|
It is the same as buffer.Prependable.View.
PiperOrigin-RevId: 213064166
Change-Id: Ib33b8a2c4da864209d9a0be0a1c113be10b520d3
|
|
PiperOrigin-RevId: 213053370
Change-Id: I60ea89572b4fca53fd126c870fcbde74fcf52562
|
|
PiperOrigin-RevId: 212757571
Change-Id: I04200df9e45c21eb64951cd2802532fa84afcb1a
|
|
Makes it possible to avoid copying or allocating in cases where DeliverNetworkPacket (rx)
needs to turn around and call WritePacket (tx) with its VectorisedView.
Also removes the restriction on having VectorisedViews with multiple views in the write path.
PiperOrigin-RevId: 211728717
Change-Id: Ie03a65ecb4e28bd15ebdb9c69f05eced18fdfcff
|
|
PiperOrigin-RevId: 209819644
Change-Id: I329d054bf8f4999e7db0dcd95b13f7793c65d4e2
|
|
Add option to redirect packet back to netstack if it's destined to itself.
This fixes the problem where connecting to the local NIC address would
not work, e.g.:
echo bar | nc -l -p 8080 &
echo foo | nc 192.168.0.2 8080
PiperOrigin-RevId: 207995083
Change-Id: I17adc2a04df48bfea711011a5df206326a1fb8ef
|