serverAuthenticate only bounded real failures via MaxAuthTries.
PartialSuccessError responses and the publickey isQuery short-circuit
both kept the loop running without incrementing authFailures, so a
client could keep the server processing SSH_MSG_USERAUTH_REQUEST
messages indefinitely on a single connection.
Add an unconditional cap, maxAuthServerAttempts = 128, on the total
number of userauth requests handled per connection. The counter is
incremented at the top of the loop before method dispatch, so every
method and every isQuery / partial-success path counts. When the cap
is exceeded the server sends SSH_MSG_DISCONNECT with reason 2 ("too
many authentication attempts"), mirroring the MaxAuthTries handling.
The bound is well below OpenSSH's hard cap of 1024 but above any
realistic multi-step auth flow.
Change-Id: I56779fc55cd00ddfd32ec938f8de3a49be0145dc
Reviewed-on: https://go-review.googlesource.com/c/crypto/+/781903
Reviewed-by: Junyang Shao <shaojunyang@google.com>
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: golang-scoped@luci-project-accounts.iam.gserviceaccount.com <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Make NewClient automatically pipeline concurrent requests over its
connection when the supplied io.ReadWriter also implements io.Closer.
In this mode the client writes requests to the wire as soon as the
write path is available and dispatches responses back to callers in
FIFO order via a dedicated reader goroutine, instead of fully
serializing each call. Up to 32 requests may be in flight on a single
connection.
This lets an agent that load-balances signing across multiple backend
devices keep several of them busy concurrently without forcing callers
to maintain their own connection pool. Aggregate throughput scales
roughly linearly with the number of backend devices up to the
in-flight cap; the protocol's in-order response requirement means
slow requests still delay subsequent ones on the same connection,
which is unchanged.
The pipelined path requires io.Closer because, on a Write error, the
background reader goroutine must be unblocked by closing the
underlying connection; otherwise it would remain parked forever
waiting for a response that will never arrive, leaking the goroutine
and desynchronising the FIFO routing of responses for subsequent
successful writes. When the supplied transport does not implement
io.Closer, NewClient falls back to the previous fully-serialized
behavior: a single in-flight call at a time, with no background
goroutine.
Fixesgolang/go#78473
Change-Id: Icf14e5e8ca897506d68fb32c14fc72c774cd97b2
Reviewed-on: https://go-review.googlesource.com/c/crypto/+/768483
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
LUCI-TryBot-Result: golang-scoped@luci-project-accounts.iam.gserviceaccount.com <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Add ClientAuthCallback, a hook invoked before each authentication
attempt that allows the client to dynamically select an auth method
based on server capabilities, partial successes, or previous failures.
clientAuthenticate tracks partial successes in a slice and passes
them to AuthCallback via ClientAuthContext.PartialSuccessMethods,
along with the list of failed methods and currently allowed methods.
When AuthCallback returns a non-nil AuthMethod, it is used for the
next attempt, bypassing the static findNext selection. When it
returns (nil, nil), the static config.Auth selection proceeds as
before. If AuthCallback returns (nil, error) the handshake aborts
immediately with that error.
To bound resource use when AuthCallback keeps supplying methods, the
client caps the total number of authentication attempts (failures and
partial successes combined) at 64; exceeding the cap aborts the
handshake with an error.
Fixesgolang/go#76146
Change-Id: I0d02bea7b9dd724e95e5d9d49d85306666c0df7a
Reviewed-on: https://go-review.googlesource.com/c/crypto/+/717140
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
LUCI-TryBot-Result: golang-scoped@luci-project-accounts.iam.gserviceaccount.com <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Previously, channel.handlePacket sent channelRequestSuccess and
channelRequestFailure messages to ch.msg unconditionally via the default
arm of its type switch. Because ch.msg is a bounded buffer (chanSize),
a peer that sends a burst of unsolicited channel request responses for
an open, idle channel fills the buffer and blocks the mux read loop on
the next send. That stalls all packet processing on the connection,
and because readLoop then backs up on t.incoming, closing the
underlying net.Conn does not unblock either goroutine: user code
observes Close() returning promptly while Wait() hangs and the mux,
readLoop, and kexLoop goroutines leak permanently.
This change mirrors the fix for the mux-level SendRequest path: a
sentRequestPending atomic gate is set while a SendRequest with
WantReply is in flight, handlePacket drops responses when the gate is
closed, and uses a non-blocking send otherwise. SendRequest drains
any spurious response that slipped through before discarding it, so
the caller always observes the reply to its own request.
This aligns with OpenSSH, which silently ignores channel confirm
messages that do not match a pending request.
Fixesgolang/go#79564
Fixes CVE-2026-39830
Change-Id: I15e2add4bf7876bb0c6f921f8b57203d97e83f47
Reviewed-on: https://go-review.googlesource.com/c/crypto/+/781664
LUCI-TryBot-Result: golang-scoped@luci-project-accounts.iam.gserviceaccount.com <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Neal Patel <nealpatel@google.com>
Reviewed-by: Neal Patel <nealpatel@google.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
The internal 'min' helper function in channel.go incorrectly cast the
input data length (int) to uint32 before comparing it with the
maximum packet size. On 64-bit systems, if the data length is a
multiple of 2^32 (approx. 4GB), this cast results in 0.
Consequently, the function returns 0, causing the WriteExtended loop
to spin indefinitely because it attempts to reserve 0 bytes while
the remaining data length is still positive.
This change renames the helper to 'minPayloadSize' to avoid confusion
with the Go 1.21 built-in 'min' and updates the logic to use int64
for comparisons, preventing truncation and the resulting infinite loop.
This issue was found during a security audit by NCC Group Cryptography
Services, sponsored by Teleport.
Fixesgolang/go#79567
Fixes CVE-2026-39834
Change-Id: Id5bf81d9f06c7042452acffe1c76580ff878665e
Reviewed-on: https://go-review.googlesource.com/c/crypto/+/781663
Reviewed-by: Neal Patel <nealpatel@google.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
LUCI-TryBot-Result: golang-scoped@luci-project-accounts.iam.gserviceaccount.com <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Previously the library did not verify the "User Presence" (UP) bit in
signatures generated by FIDO/U2F security keys
(sk-ecdsa-sha2-nistp256@openssh.com and sk-ssh-ed25519@openssh.com).
This allowed signatures without physical interaction to be accepted
if the underlying hardware produced them, deviating from the default
secure behavior expected by the FIDO standards and OpenSSH.
skECDSAPublicKey.Verify and skEd25519PublicKey.Verify now enforce
the user-presence bit (0x01, constant flagUserPresence) by default.
Signatures whose flags byte has UP clear fail with the sentinel
errSKMissingUserPresence.
The server public-key authentication path honors the OpenSSH
"no-touch-required" extension as an opt-out. noTouchAllowed reports
true when the extension is present either in the Permissions
returned by PublicKeyCallback (authorized_keys-level opt-out) or in
the certificate's own Extensions (CA-level opt-out); in that case
skKeyWithoutUP is used to derive a clone of the SK public key (and,
for certificates, a clone of the wrapping Certificate whose inner
Key is the cloned SK key) whose Verify accepts UP-clear signatures.
The originals are never mutated, so a per-session opt-out cannot
leak across authentication attempts or connections. Matching
OpenSSH, the opt-out is read only from Extensions, never from
CriticalOptions.
skKeyWithoutUP is iterative and unwraps at most one level of
*Certificate: the SSH cert format forbids Certificate.Key from being
another Certificate (parseCert rejects it) but callers can still
construct such a value directly in Go, so a recursive descent would
be driven to unbounded depth by malformed or cyclic input. Any such
pathological *Certificate is returned unchanged.
CertChecker.CheckCert applies skKeyWithoutUP unconditionally to the
certificate's CA key before verifying the CA signature, matching
OpenSSH, which calls sshkey_verify with detailsp==NULL in
sshkey.c:cert_parse and never extracts or enforces UP/UV flags on
CA signatures. The UP bit on a CA signature reflects the CA
operator's presence at cert-issuance time, which has no bearing on
whether the user being authenticated is present now, so enforcing it
here would only break interop with certificates issued by
non-interactive SK CAs without a corresponding security benefit. The
skKeyWithoutUP call is a no-op for non-SK CA keys (the common case).
This change breaks backward compatibility for clients or keys that
generate user-authentication signatures without the User Presence
flag set. Previously those signatures were accepted by the server.
They will now be rejected with "ssh: signature missing required user
presence flag" unless the "no-touch-required" extension is
explicitly granted to the session by the server callbacks, or
carried by the user certificate.
This issue was found during a security audit by NCC Group
Cryptography Services, sponsored by Teleport.
Fixesgolang/go#79566
Fixes CVE-2026-39831
Change-Id: I74b6de3bb6a2d7a0f34d7fa36bbbbf06f0b3fc6b
Reviewed-on: https://go-review.googlesource.com/c/crypto/+/781662
Reviewed-by: Neal Patel <nealpatel@google.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
LUCI-TryBot-Result: golang-scoped@luci-project-accounts.iam.gserviceaccount.com <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
The parseDSA function previously accepted DSA keys with arbitrary values
for the sub-prime Q and did not validate that group elements G and Y
were within the modulus P.
Malicious actors could provide a key with a massively large Q (e.g.,
millions of bits), leading to excessive CPU consumption during signature
verification.
This change restricts the sub-prime Q to exactly 160 bits, as required
by FIPS 186-2, and ensures that G and Y are strictly less than P.
This issue was found during a security audit by NCC Group Cryptography
Services, sponsored by Teleport.
Fixesgolang/go#79565
Fixes CVE-2026-39829
Change-Id: I526118d94684076088d0625178844f64c1303ec8
Reviewed-on: https://go-review.googlesource.com/c/crypto/+/781661
Reviewed-by: Roland Shoemaker <roland@golang.org>
LUCI-TryBot-Result: golang-scoped@luci-project-accounts.iam.gserviceaccount.com <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Neal Patel <nealpatel@google.com>
Previously, the RSA key parser accepted keys with arbitrary modulus
sizes. Processing keys with extremely large moduli (e.g., > 8192 bits)
can consume excessive CPU resources during verification, potentially
leading to a Denial of Service (DoS).
This change introduces a limit of 8192 bits for the RSA modulus in
parseRSA, rejecting keys that exceed this size in line with the limit
enforced by crypto/tls.
This issue was found during a security audit by NCC Group Cryptography
Services, sponsored by Teleport.
Fixesgolang/go#79565
Fixes CVE-2026-39829
Change-Id: Ibdddad1859a4d9db5c9f052d06c82f29bfc2e5e5
Reviewed-on: https://go-review.googlesource.com/c/crypto/+/781641
Reviewed-by: Neal Patel <nealpatel@google.com>
LUCI-TryBot-Result: golang-scoped@luci-project-accounts.iam.gserviceaccount.com <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Previously, the mux implementation handled global request responses by
blocking until the response could be sent to the globalResponses channel.
Since this channel has a buffer size of 1, unsolicited responses from a
server (or responses arriving after a timeout) would fill the buffer.
Subsequent unsolicited responses would block handleGlobalPacket, stalling
the entire connection's read loop and causing a denial of service.
This change modifies handleGlobalPacket to use a non-blocking send. If
no goroutine is waiting for a response (or the buffer is full), the
message is dropped. This aligns with OpenSSH behavior, which ignores
unexpected global responses.
Additionally, SendRequest now drains the globalResponses channel after
acquiring the mutex but before sending the request. This ensures that
any stale responses or "spam" buffered just before the lock was acquired
are discarded, preventing race conditions where a legitimate request
might otherwise consume an unrelated response.
This issue was found during a security audit by NCC Group Cryptography
Services, sponsored by Teleport.
Fixesgolang/go#79564
Fixes CVE-2026-39830
Change-Id: Ia0c46355203d557eadcd432c10b87c8a044e1089
Reviewed-on: https://go-review.googlesource.com/c/crypto/+/781640
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Neal Patel <nealpatel@google.com>
LUCI-TryBot-Result: golang-scoped@luci-project-accounts.iam.gserviceaccount.com <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
In serverAuthenticate, the permissions variable is reset to nil at the
beginning of the authentication loop. If an authentication callback
returns a PartialSuccessError along with non-nil Permissions, those
permissions are currently silently discarded before the next
authentication step.
This change returns an error if a callback returns both a PartialSuccessError
and non-nil Permissions, preventing API misuse where the user might
erroneously expect those permissions to be preserved or merged into the
final session permissions.
This issue was found during a security audit by NCC Group Cryptography
Services, sponsored by Teleport.
Fixesgolang/go#79562
Fixes CVE-2026-39828
Change-Id: I632c9e46e2b5e8804ef88081063a3612a2462f9f
Reviewed-on: https://go-review.googlesource.com/c/crypto/+/781621
Reviewed-by: Neal Patel <nealpatel@google.com>
LUCI-TryBot-Result: golang-scoped@luci-project-accounts.iam.gserviceaccount.com <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Gopher Robot <gobot@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
When a server rejects an incoming channel request via
NewChannel.Reject, the channel is left in the multiplexer's
channel list. Because the channel is never explicitly removed or
closed, its internal buffers and sync primitives remain allocated
for the lifetime of the SSH connection.
A malicious client could exploit this behavior by repeatedly
requesting to open channels that are destined to be rejected,
causing unbounded memory growth and potentially leading to a
Denial of Service (DoS) via resource exhaustion.
This change fixes the leak by calling ch.mux.chanList.remove
within the Reject method, removing the channel from the list and allowing the
garbage collector to reclaim the associated memory immediately.
Fixesgolang/go#35127
Fixes CVE-2026-3982
Change-Id: Iaa177f5dfd151812dd404e528a4a1c77527a0e29
Reviewed-on: https://go-review.googlesource.com/c/crypto/+/781320
LUCI-TryBot-Result: golang-scoped@luci-project-accounts.iam.gserviceaccount.com <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Nicholas Husin <nsh@golang.org>
Reviewed-by: Nicholas Husin <husin@google.com>
The in-memory keyring supports the "lifetime" constraint but does not
implement the "confirm" constraint. Previously, keyring.Add silently
ignored ConfirmBeforeUse: the key was stored, advertised through List,
and used for signing without any interactive confirmation, potentially
misleading callers into believing this security measure was enforced.
Return an error when ConfirmBeforeUse is set instead of silently
downgrading the caller's security expectations. Implementing real
confirm-before-use in an in-memory library keyring is infeasible (there
is no UI or confirmation callback), so failing closed is the correct
behavior; adding actual confirm support would require an API addition
and is out of scope.
This is a deliberate behavior change: keyring.Add previously accepted
and ignored ConfirmBeforeUse and now returns an error. This change also
updates the keyring doc comments to document the supported constraints.
This issue was found during a security audit by NCC Group Cryptography
Services, sponsored by Teleport.
Fixes CVE-2026-39833
Updates golang/go#47533Fixesgolang/go#79436
Change-Id: I1b3a286f0c1e4a4e08ac37109f7e491692ca90ae
Reviewed-on: https://go-review.googlesource.com/c/crypto/+/778642
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Neal Patel <nealpatel@google.com>
Reviewed-by: Neal Patel <neal@golang.org>
Auto-Submit: Neal Patel <nealpatel@google.com>
LUCI-TryBot-Result: golang-scoped@luci-project-accounts.iam.gserviceaccount.com <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
The client Add method only serialized the lifetime and confirm
constraints and silently dropped AddedKey.ConstraintExtensions before
sending the SSH_AGENTC_ADD_IDENTITY request. As a result the remote
agent always received the key with no extension constraints, regardless
of what the caller requested.
Applications that add a key believing custom constraint extensions
(such as restrict-destination-v00@openssh.com) would be enforced
instead loaded a completely unrestricted key into the agent. For
example, an administrator forwarding their agent into an untrusted jump
host and trying to limit the forwarded key with restrict-destination
never had that restriction reach the agent: any user or compromised
process on that host could make the agent sign arbitrary challenges.
Serialize each entry in key.ConstraintExtensions as an
agentConstrainExtension constraint so the constraints reach the agent,
and add a round-trip regression test that verifies the extensions
survive client serialization and server parsing.
This issue was found during a security audit by NCC Group Cryptography
Services, sponsored by Teleport.
Updates CVE-2026-39832
Updates golang/go#79435
Change-Id: I14c5583b106cbf0d282d2ba01e000e0f586f08c7
Reviewed-on: https://go-review.googlesource.com/c/crypto/+/778640
Reviewed-by: Neal Patel <neal@golang.org>
Reviewed-by: Neal Patel <nealpatel@google.com>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: golang-scoped@luci-project-accounts.iam.gserviceaccount.com <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Previously, pickSignatureAlgorithm constructed the list of candidate
algorithms by iterating over the static list returned by
algorithmsForKeyFormat. This caused the Signer's preference order
to be ignored in favor of the library's default internal order.
This change inverts the filtering logic to iterate over the signer's
supported algorithms first. This ensures that if a MultiAlgorithmSigner
explicitly prefers a specific algorithm (e.g., rsa-sha2-512 over
rsa-sha2-256), that preference is preserved and respected during the
handshake negotiation.
Fixesgolang/go#78248
Change-Id: I48a0aac720be7f973963342b82047ce32fc96699
Reviewed-on: https://go-review.googlesource.com/c/crypto/+/746020
Reviewed-by: Lonny Wong <lonnywang.cn@gmail.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Nicola Murino <nicola.murino@gmail.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Carlos Amedee <carlos@golang.org>
Previously, an attacker could specify an integer up to 0xFFFFFFFF
that would directly allocate memory despite the observability of
the rest of the payload. This change places a hard cap on the
amount of mechanisms that can be specified and encoded in the
payload. Additionally, it performs a small sanity check to deny
payloads whose stated size is contradictory to the observed payload.
Thank you to Jakub Ciolek for reporting this issue.
Fixes CVE-2025-58181
Fixesgolang/go#76363
Change-Id: I0307ab3e906a3f2ae763b5f9f0310f7073f84485
Reviewed-on: https://go-review.googlesource.com/c/crypto/+/721961
Auto-Submit: Roland Shoemaker <roland@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Skip the key exchange test when using the system's ssh CLI if the
required KEX algorithm (e.g., mlkem768x25519-sha256) is not supported.
This is determined by running ssh -Q kex and checking for the presence
of the target algorithm.
Prevents false test failures in CI environments with older or limited
SSH implementations.
Cq-Include-Trybots: luci.golang.try:x_crypto-gotip-darwin-amd64-longtest,x_crypto-gotip-linux-amd64-longtest,x_crypto-gotip-windows-amd64-longtest
Change-Id: I3fac703ec70559e18b30d5fff88274335a7c3952
Reviewed-on: https://go-review.googlesource.com/c/crypto/+/679195
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Carlos Amedee <carlos@golang.org>
We add this support for the following reasons:
- We are planning to expose recommended (secure) vs. supported (works,
not necessarily recommended) algorithms. The DHGEX kex is currently
only exposed as a client-side kex. To simplify the calling convention
for this follow-on, we expose the server side too.
- Some clients are quite inflexible with reference to kex algorithms
choice, for example they offer:
diffie-hellman-group-exchange-sha256, diffie-hellman-group-exchange-sha1,
diffie-hellman-group14-sha1, diffie-hellman-group1-sha1
therefore DHGEX helps interoperability.
We do not recommend the DHGEX kex as a whole:
- the negotiation requires an extra round trip
- the server must generate parameters (slow) or hardcode them, which
defeats the security benefit over traditional DH.
In this implementation we hardcode sending Oakley Group 14, Oakley
Group 15 or Oakley Group 16 based on the requested max size.
Users that are concerned with security of classical DH kex should
migrate to kex based on EC or Ed25519.
Fixesgolang/go#54743
Change-Id: I127822e90efc36821af4aca679931f40a2023021
Reviewed-on: https://go-review.googlesource.com/c/crypto/+/532415
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Nicola Murino <nicola.murino@gmail.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Han-Wen Nienhuys <hanwen@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Nicola Murino
Neal Patel
dingandCyrus Katrak
Neal Patel
Rob Picard
Michael Stapelberg
Santhanam
cuishuang
Filippo Valsorda