Taken from my implementation: https://bitbucket.org/ede/sha3
Performance gain from using less memory and more registers.
benchmark old ns/op new ns/op delta
BenchmarkPermutationFunction 1484 1118 -24.66%
BenchmarkBulkKeccak512 374993 295178 -21.28%
BenchmarkBulkKeccak256 215496 172335 -20.03%
benchmark old MB/s new MB/s speedup
BenchmarkPermutationFunction 134.76 178.80 1.33x
BenchmarkBulkKeccak512 43.69 55.51 1.27x
BenchmarkBulkKeccak256 76.03 95.07 1.25x
R=jcb, agl
CC=golang-dev, nigeltao
https://golang.org/cl/8088044
time); delete some comments that didn't add much and were incorrect
anyway (the test specification was tc, not t).
R=jcb
CC=agl, golang-dev
https://golang.org/cl/7665045
When running the ssh tests several times (e.g. with -cpu 1,2), the
second run would fail because testing globals had been altered. This
change avoids altering the globals since the default worked anyway.
Fixesgolang/go#4715.
R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/7903045
Added a pure Go implementation of SHA3 (Keccak) which implements the hash.Hash interface.
A test file is included with performance benchmarks and standard test vectors.
R=agl, nigeltao
CC=golang-dev
https://golang.org/cl/7760044
The session test previously had a one second timeout for the output of
stty and this was leading to flakiness. This change removes the timeout
since go test has a generic timeout mechanism.
Additionally, the test was looking for "-echo" in the output to test
the value of the echo flag. However, there are also typically "echoe",
"echok" and "echonl" flags, and "-echo" could be a prefix of any of
time. Thus we now also match a trailing space.
R=golang-dev, rsc, extraterrestrial.neighbour
CC=golang-dev
https://golang.org/cl/7579043
terminal contains a number of utility functions that are currently only
implemented for Linux. Darwin uses different named constants for
getting and setting the terminal state so this change splits them off
as constants and defines them for each arch.
R=golang-dev, minux.ma
CC=golang-dev
https://golang.org/cl/7286043
Work on uint32 slices instead of bytes.
Replace usage of Salsa20/8 from salsa package with
the specialized version.
benchmark old ns/op new ns/op delta
BenchmarkKey 266430525 126657130 -52.46%
R=agl
CC=golang-dev
https://golang.org/cl/7139050
This is a followup to issue 6927044. I *really* don't want to
break streaming for large encrypted data packets.
Removing the automatic re-parsing. OpaqueReader can be
used for recovering useful information from mangled/unsupported packets.
R=agl
CC=golang-dev
https://golang.org/cl/6944056
Simplify MarshalAuthorizedKey by using the algoName func.
Make the algoName func be very specific about supported key types in openssh certs.
Generalize some of the commentary that previously mentioned specific key types.
R=agl, dave
CC=golang-dev
https://golang.org/cl/6938067
These are improvements I've made as necessary to develop Hockeypuck,
an OpenPGP keyserver (https://launchpad.net/hockeypuck).
The max armor line length was increased to 96 because some keyservers (SKS)
will output armor with lines greater than 64 (the prior max).
I've exposed packet.ReadEntity to support stream-parsing, useful for
large SKS dump files.
ReadKeyRing attempts to recover in the event of malformed data.
Because many packet types are not yet supported, I added OpaquePacket to
capture unsupported types for offline storage and later reparsing.
R=agl
CC=golang-dev
https://golang.org/cl/6927044
Export key and certificate algorithm names.
Switch from string literals over to using the constants for any key/cert algorithm references.
Make URL references visible in the godoc web display.
Standardize url reference names with surrounding [].
R=dave, agl, jonathan.mark.pittman
CC=golang-dev
https://golang.org/cl/6944047
This allows easy import/export of public keys in the format
expected by OpenSSH for authorized_keys files, as well as
allowing comparisons with ServerConfig's PublickeyCallback.
Fixesgolang/go#3908.
R=agl, dave, golang-dev, bradfitz
CC=agl, golang-dev
https://golang.org/cl/6855107
The go.crypto repo was missed when this function was removed from the
ciphers in the standard library.
R=agl
CC=golang-dev
https://golang.org/cl/6904062
t caches the value of z squared, but wasn't getting correctly
initialised when the value of z was reset.
R=golang-dev
CC=golang-dev
https://golang.org/cl/6850091
user.Current() currently requires cgo - if an error is returned
attempt to get the username from the environment.
R=golang-dev, minux.ma, bradfitz, dave
CC=golang-dev
https://golang.org/cl/6819113
Fixesgolang/go#4285.
Adding a new cipher that is supported by the remote end, but not supported by our client causes that cipher to be considered a valid candidate. This fails later in setupKeys when there is no cipherModes configuration.
In summary, unsupported ciphers cannot be willed into existence by adding them to the client config. This change enforces this.
R=golang-dev, agl
CC=golang-dev
https://golang.org/cl/6780047
It's useful to expose this bit of state because, although callers can
keep track of it themselves, it's already in the Conversation so it
seems wasteful to make them do so.
R=golang-dev, ioerror
CC=golang-dev
https://golang.org/cl/6724056
Also fix salt expansion code to fill the properly sized array,
since the following code never uses more than 4 uints.
Also remove unused expandedKey array.
R=golang-dev, agl
CC=golang-dev
https://golang.org/cl/6622050
This proposal is an attempt to improve the state of functional testing in the ssh package. The previous functional tests required the user to give away some personal details, like their password and private key to run the tests, and so were probably not run as frequently as they should.
R=agl, gustav.paul, kardianos, fullung
CC=golang-dev
https://golang.org/cl/6601043
This change alters the assembly to use FSUBD instructions such that
6l will actually emit the correct FSUBRD instructions and enables
the assembly code.
R=golang-dev, r
CC=golang-dev
https://golang.org/cl/6514044
The check for a sensible packet length was removed a while ago
when the window size and channel packet size checks were moved
into channel.go. While the RFC suggests that any packet of size
less than uint32 -1 is valid, most implmentations limit the size
to a smaller value. OpenSSH chose 256kb, so that sounds like a
sensible default.
R=agl, huin, kardianos
CC=golang-dev
https://golang.org/cl/6490098
Fixesgolang/go#3810.
Fixes chanWriter Write after close behaviour bug.
Fixes serverChan writePacket after close bug.
Addresses final comments by agl on 6405064, plus various cleanups.
R=agl, kardianos, gustav.paul, fullung
CC=golang-dev
https://golang.org/cl/6479056
Fixesgolang/go#3989.
Tested for several hours on an 8 core ec2 instance with
random GOMAXPROC values.
Also, rolls server_test.go into session_test using the
existing dial() framework.
R=fullung, agl, kardianos
CC=golang-dev
https://golang.org/cl/6475063
Use a handler that does not attempt to send a status message
as the failing test closes the connection abruptly.
Also, check the err response on all shell.ReadLine operations.
R=agl, minux.ma, kardianos
CC=golang-dev
https://golang.org/cl/6487043
Fixesgolang/go#3810.
This change introduces an atomic boolean to guard the close
of the clientChan. Previously the client code was quite
lax with the ordering of the close messages and could allow
window adjustment or EOF messages to leak after Close had
been signaled.
Consolidating the changes to the serverChan will be handled
in a following CL.
R=agl, kardianos, gustav.paul
CC=golang-dev
https://golang.org/cl/6405064