Enables HTTP/2 on any servers used with the autocert listener
by setting "h2" in NextProtos of the listener *tls.Config.
Also adds a warning to the listener documentation that it
enables HTTP/2.
Fixesgolang/go#20572
Change-Id: If7c0f5722f0b1781789219fc4e84da3f19a89ab7
Reviewed-on: https://go-review.googlesource.com/45630
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
These largely follow the pattern laid out in the secretbox example
file. Hopefully they should demonstrate the best ways to use the `box`
primitives.
Change-Id: I99be5f886e6ffe5f859268d6caa282e263eeacd0
Reviewed-on: https://go-review.googlesource.com/44736
Reviewed-by: Adam Langley <agl@golang.org>
Although a 128-bit digest provides little collision resistance it can be
used as a MAC, as found in the WireGuard protocol.
Change-Id: I06ee528e5711035da8f1b254c223a6fda8890abf
Reviewed-on: https://go-review.googlesource.com/39370
Reviewed-by: Adam Langley <agl@golang.org>
ParseResponseForCert would previously complain about an invalid hash OID
if the response contained no SingleResponse matching the certificate
provided by the caller.
Fixesgolang/go#19540
Change-Id: I0354c4048707a788ed3d184cc88b4f13f65544ba
Reviewed-on: https://go-review.googlesource.com/44005
Reviewed-by: Adam Langley <agl@golang.org>
This change does two things:
1. Fix a length checking bug in the Decrypt function.
2. Use binary.LittleEndian for byte conversions.
Fixesgolang/go#19881
Change-Id: I9d33b92f2bd7e6ca8f69308388f1e8a5c6df2be8
Reviewed-on: https://go-review.googlesource.com/39954
Reviewed-by: Adam Langley <agl@golang.org>
SSH host certificates are expected to contain hostnames only,
not "host:port" format.
This change allows Go clients to connect to OpenSSH servers that
use host certificates.
Note, this change will break any clients that use ssh.NewClientConn()
with an `addr` that is not in `host:port` format (they will see a
"missing port in address" error).
Fixes bug 20273.
Change-Id: I5a306c6b7b419a737e1f0f9c5ca8c585e21a45a4
Reviewed-on: https://go-review.googlesource.com/43475
Reviewed-by: Han-Wen Nienhuys <hanwen@google.com>
Run-TryBot: Han-Wen Nienhuys <hanwen@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
The bcrypt implementation must append a zero byte to the user provided key
to be compatible to C implementations.
This will change the user provided key if the slice has enough capacity to
hold the extra zero byte.
This change always allocates a new slice for the C-compatible key.
Fixesgolang/go#20425
Change-Id: I8dc4e840c29711daabdabe58d83643cc0103cedd
Reviewed-on: https://go-review.googlesource.com/43715
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Dialing the 0.0.0.0 address (as returned by net.Addr().String() for a
net.Listen("tcp", ":1") address) is not yet guaranteed to work. It's
currently OS-dependent.
For some reason it works on Go 1.8+, but it hasn't yet been defined to
work reliably.
Fix the tests for now (since we need to support older Go releases),
even if this might work in the future.
Updates golang/go#18806
Change-Id: I2f0476b1d4f2673ab64ffedfa733f2d92fceb6ff
Reviewed-on: https://go-review.googlesource.com/42496
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Han-Wen Nienhuys <hanwen@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Han-Wen Nienhuys <hanwen@google.com>
This change makes the Manager try creating a certificate
again, after a previously unsuccessful attempt.
The implementation is based on a timer, to prevent hitting
an ACME CA with too high QPS when under a heavy load.
The timer is hardcoded to 1 minute.
Fixesgolang/go#17740.
Change-Id: I46a49201cf423be3360633a89209d7b2bccc1d76
Reviewed-on: https://go-review.googlesource.com/41694
Run-TryBot: Alex Vaghin <ddos@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
This is a breaking change.
This adds a new hostkey callback which takes the hostname field
restrictions into account when validating host certificates.
Prior to this, a known_hosts file with the following entry
@cert-authority *.example.com ssh-rsa <example.com public key>
would, when passed to knownhosts.New() generate an ssh.HostKeyCallback
that would accept all host certificates signed by the example.com public
key, no matter what host the client was connecting to.
After this change, that known_hosts entry can only be used to validate
host certificates presented when connecting to hosts under *.example.com
This also renames IsAuthority to IsUserAuthority to make its intended
purpose more clear.
Change-Id: I7188a53fdd40a8c0bc21983105317b3498f567bb
Reviewed-on: https://go-review.googlesource.com/41751
Reviewed-by: Han-Wen Nienhuys <hanwen@google.com>
Run-TryBot: Han-Wen Nienhuys <hanwen@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
A cached cert data may be corrupted or simply contain an expired
certificate, which results in GetCertificate returning an error.
This change makes the Manager ignore those invalid and expired
cache entries, treating them as nonexistent.
Fixesgolang/go#20035.
Change-Id: I5345291ecb1aab1cf19671cf0a383135c7102038
Reviewed-on: https://go-review.googlesource.com/41690
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Without this, autocert will panic with an unhelpful nil pointer inside
the acme client.
Reorganized the test suite's ACME server stub creation, as I needed to
be able to stand up a test where GetCertificate was expected to fail.
Change-Id: Ie5e19c6e7766b4578c9b3c16789d7b27bd3be163
Reviewed-on: https://go-review.googlesource.com/40951
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Consider the following code:
var p *int
a := []int{0, 1, 2, 3}
for _, i := range a {
if i == 1 {
p = &i
}
}
fmt.Println(*p) // Prints 3
This prints 3 because the variable i is the exact same variable across
all iterations of the loop. When the address is taken for some specific
iteration, the user's intent is to capture the value of i at that
given loop, but instead the value of i in the last loop is what remains.
A bug this sort occurs in the check logic since the address of the
knownKey is taken, but is changed upon subsequent iterations of the
loop (which happens when there are multiple lines).
Change-Id: Ic626778cdcde3968dcff4fa5e7206274957dcb04
Reviewed-on: https://go-review.googlesource.com/40937
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
After receiving a badNonce error, the call can be safely retried. Nonce
errors can happen unexpectedly based on an unknown expiration date or
server-side changes. Rather than force the caller handle these errors,
retryPostJWS will keep retrying until success or a different error.
According to the spec, the error returned should be
"urn:ietf:params:acme:error:badNonce", but the error that Let's Encrypt
returns is "urn:acme:error:badNonce" so we just check the suffix.
Fixesgolang/go#19703
Change-Id: Id15012dff91e51d28ed8bc54f13a6212186cb7df
Reviewed-on: https://go-review.googlesource.com/40130
Run-TryBot: Alex Vaghin <ddos@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Alex Vaghin <ddos@google.com>
Fixes a nil pointer dereference that slipped through buildbots because
it was introduced by the last two commits.
Change-Id: Ib269e910956cd8b3b46e217b03fde1b61572260a
Reviewed-on: https://go-review.googlesource.com/40530
Reviewed-by: Han-Wen Nienhuys <hanwen@google.com>
Run-TryBot: Han-Wen Nienhuys <hanwen@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
The normal handshake kicks off with a waitSession(), which guarantees
that we never attempt to send data before the first kex is completed,
but ensuring readPacketsLeft > 0 and writePacketsLeft > 0 helps
understand that thresholds can never cause spurious rekeying at the
start of a connection.
Change-Id: If5bcafcda0c7d16fd21f22c664101ac5f5b487d7
Reviewed-on: https://go-review.googlesource.com/38696
Reviewed-by: Adam Langley <agl@golang.org>
Run-TryBot: Adam Langley <agl@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Simplify the constant swap function.
On amd64: Replace the CMOVQEQ scheme with SSE2 code similar to the non-amd64 code.
On non-amd64: Avoid unnecessary loop iterations.
The result is less and slightly faster code.
name old time/op new time/op delta
ScalarBaseMult-4 653µs ± 0% 636µs ± 0% ~ (p=0.100 n=3+3)
name old time/op new time/op delta
ConstantSwap-4 10.4ns ± 1% 6.2ns ± 0% -39.86% (p=0.029 n=4+4)
On an i7-65000U
Change-Id: Ia5eea92e0b3eabb6c291d25229aa582b51278552
Reviewed-on: https://go-review.googlesource.com/39693
Reviewed-by: Adam Langley <agl@golang.org>
Run-TryBot: Adam Langley <agl@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
This change breaks backwards compatibility.
MaxAuthTries specifies the maximum number of authentication attempts
permitted per connection. If set to a negative number, the server will
allow unlimited authentication attempts. MaxAuthTries defaults to 6 if
not specified, which is a backwards incompatible change. On exceeding
maximum authentication attempts, the server will send a disconnect
message to the client.
This configuration property mirrors a similar property in sshd_config
and prevents bad actors from continuously trying authentication.
Change-Id: Ic77d2c29ee2fd2ae5c764becf7df91d29d03131b
Reviewed-on: https://go-review.googlesource.com/35230
Reviewed-by: Han-Wen Nienhuys <hanwen@google.com>
Run-TryBot: Han-Wen Nienhuys <hanwen@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
The ctxhttp package used to be big and gross before net/http supported
contexts natively. Nowadays it barely does anything. Stop using it,
because it just pulls in the old context package anyway. (We can't
really clean up the ctxhttp package until Go 1.9)
Change-Id: I48b11f2f483783a32cbaa75e244301148a304c08
Reviewed-on: https://go-review.googlesource.com/40110
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Alex Vaghin <ddos@google.com>
A certificate must have the Server Auth Extended Key Usage to be used
for TLS, and an ECDSA certificate must have the Digital Signature Key
Usage to be used at all (you can't encrypt to an ECDSA key).
crypto/tls ignores (E)KUs when serving certificates, and most browsers
do as well, so it works, but OpenSSL would refuse to serve these
certificates, and clients would be allowed to reject them.
Change-Id: I699e58e613f01077e6b67fdb9e789d46e1672112
Reviewed-on: https://go-review.googlesource.com/39913
Run-TryBot: Alex Vaghin <ddos@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Alex Vaghin <ddos@google.com>
Previously, the public key authentication for clients would send an
enquiry to the remote for every key specified before attempting to
authenticate with the server.
Now, we immediately try to authenticate once a valid key is found.
This results in exchanging fewer packets if the valid key is near the
top of the list. If all keys fail, then the number of packets exchanged
by the client and server is unaffected.
For OpenSSH daemon, an enquiry into the validity of a key without
authentication is still recorded as an authentication attempt, so any
clients with more than MaxAuthTries public keys would not be able to
authenticate using the previous implementation. This change will allow
clients to succeed authentication if the successful key is at the start
of the list of keys.
Change-Id: I8ea42caf40c0864752218c3f6934e86b12f5b81a
Reviewed-on: https://go-review.googlesource.com/38890
Reviewed-by: Adam Langley <agl@golang.org>