The original motivation behind the sign/complete split was to avoid
needlessly hashing the input on each pass through the state machine, but
we're payload-based now and, in all cases, the payload is either cheap
to compute or readily available. (Even the hashing worry was probably
unnecessary.)
Tweak ssl_private_key_{sign,decrypt} to automatically call
ssl_private_key_complete as needed and take advantage of this in the
handshake state machines:
- TLS 1.3 signing now computes the payload each pass. The payload is
small and we're already allocating a comparable-sized buffer each
iteration to hold the signature. This shouldn't be a big deal.
- TLS 1.2 decryption code still needs two states due to reading the
message (fixed in new state machine style), but otherwise it just
performs cheap idempotent tasks again. The PSK code is reshuffled to
guarantee the callback is not called twice (though this was impossible
anyway because we don't support RSA_PSK).
- TLS 1.2 CertificateVerify signing is easy as the transcript is readily
available. The buffer is released very slightly later, but it
shouldn't matter.
- TLS 1.2 ServerKeyExchange signing required some reshuffling.
Assembling the ServerKeyExchange parameters is moved to the previous
state. The signing payload has some randoms prepended. This is cheap
enough, but a nuisance in C. Pre-prepend the randoms in
hs->server_params.
With this change, we are *nearly* rid of the A/B => same function
pattern.
BUG=128
Change-Id: Iec4fe0be7cfc88a6de027ba2760fae70794ea810
Reviewed-on: https://boringssl-review.googlesource.com/17265
Commit-Queue: David Benjamin <davidben@google.com>
Commit-Queue: Steven Valdez <svaldez@google.com>
Reviewed-by: Steven Valdez <svaldez@google.com>
We've got an asynchronous ServerKeyExchange state in the middle that
complicates things a bit, but this is still a little tighter.
BUG=128
Change-Id: I4ee2e3b85e677c9555d2fbddd387c12d41ab2b54
Reviewed-on: https://boringssl-review.googlesource.com/17250
Commit-Queue: Steven Valdez <svaldez@google.com>
Reviewed-by: Steven Valdez <svaldez@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
We can take advantage of our flight-by-flight model.
BUG=128
Change-Id: If27a5b6d88055da71199ef672d9c71969925aca9
Reviewed-on: https://boringssl-review.googlesource.com/17249
Reviewed-by: Steven Valdez <svaldez@google.com>
Commit-Queue: Steven Valdez <svaldez@google.com>
BUG=76
Change-Id: If58a73da38e46549fd55f84a9104e2dfebfda43f
Reviewed-on: https://boringssl-review.googlesource.com/14164
Reviewed-by: Steven Valdez <svaldez@google.com>
Reviewed-by: David Benjamin <davidben@google.com>
Commit-Queue: Steven Valdez <svaldez@google.com>
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
I doubt this matters, but this seems a little odd. In particular, this
avoids info_callback seeing the SSL_ST_OK once we stop switching
hs->state back and forth.
BUG=177
Change-Id: Ied39c0e94c242af9d5d0f26795d6e0f2f0b12406
Reviewed-on: https://boringssl-review.googlesource.com/13827
Reviewed-by: Steven Valdez <svaldez@google.com>
Reviewed-by: David Benjamin <davidben@google.com>
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Change-Id: If97da565155292d5f0de5c6a8b0fd8508398768a
Reviewed-on: https://boringssl-review.googlesource.com/13564
Commit-Queue: David Benjamin <davidben@google.com>
Reviewed-by: Adam Langley <agl@google.com>
The version negotiation logic was a little bizarrely wedged in the
middle of the state machine. (We don't support server renegotiation, so
have_version is always false here.)
BUG=128
Change-Id: I9448dce374004b92e8bd5172c36a4e0eea51619c
Reviewed-on: https://boringssl-review.googlesource.com/13561
Commit-Queue: David Benjamin <davidben@google.com>
Reviewed-by: Adam Langley <agl@google.com>
The TLS 1.2 state machine now looks actually much closer to the TLS 1.3
one on the write side. Although the write states still have a BIO-style
return, they don't actually send anything anymore. Only the BIO flush
state does. Reads are still integrated into the states themselves
though, so I haven't made it match TLS 1.3 yet.
BUG=72
Change-Id: I7708162efca13cd335723efa5080718a5f2808ab
Reviewed-on: https://boringssl-review.googlesource.com/13228
Reviewed-by: Adam Langley <agl@google.com>
This avoids needing a extra state around client certificates to avoid
calling the callbacks twice. This does, however, come with a behavior
change: configuring both callbacks won't work. No consumer does this.
(Except bssl_shim which needed slight tweaks.)
Change-Id: Ia5426ed2620e40eecdcf352216c4a46764e31a9a
Reviewed-on: https://boringssl-review.googlesource.com/12690
Reviewed-by: Adam Langley <agl@google.com>
This simplifies a little code around EMS and PSK KE modes, but requires
tweaking the SNI code.
The extensions that are more tightly integrated with the handshake are
still processed inline for now. It does, however, require an extra state
in 1.2 so the asynchronous session callback does not cause extensions to
be processed twice. Tweak a test enforce this.
This and a follow-up to move cert_cb before resumption are done in
preparation for resolving the cipher suite before resumption and only
resuming on match.
Note this has caller-visible effects:
- The legacy SNI callback happens before resumption.
- The ALPN callback happens before resumption.
- Custom extension ClientHello parsing callbacks also cannot depend on
resumption state.
- The DoS protection callback now runs after all the extension callbacks
as it is documented to be called after the resumption decision.
BUG=116
Change-Id: I1281a3b61789b95c370314aaed4f04c1babbc65f
Reviewed-on: https://boringssl-review.googlesource.com/11845
Reviewed-by: Adam Langley <agl@google.com>
Commit-Queue: Adam Langley <agl@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
If cert_cb runs asynchronously, we end up repeating a large part of very
stateful ClientHello processing. This seems to be mostly fine and there
are few users of server-side cert_cb (it's a new API in 1.0.2), but it's
a little scary.
This is also visible to external consumers because some callbacks get
called multiple times. We especially should try to avoid that as there
is no guarantee that these callbacks are idempotent and give the same
answer each time.
Change-Id: I212b2325eae2cfca0fb423dace101e466c5e5d4e
Reviewed-on: https://boringssl-review.googlesource.com/10224
Commit-Queue: David Benjamin <davidben@google.com>
Commit-Queue: Adam Langley <agl@google.com>
Reviewed-by: Adam Langley <agl@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
Share a bit more of it between TLS 1.2 and 1.3.
Change-Id: I43c9dbf785a3d33db1793cffb0fdbd3af075cc89
Reviewed-on: https://boringssl-review.googlesource.com/8849
Reviewed-by: David Benjamin <davidben@google.com>
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
This adds the machinery for doing TLS 1.3 1RTT.
Change-Id: I736921ffe9dc6f6e64a08a836df6bb166d20f504
Reviewed-on: https://boringssl-review.googlesource.com/8720
Reviewed-by: David Benjamin <davidben@google.com>
Commit-Queue: David Benjamin <davidben@google.com>
CQ-Verified: CQ bot account: commit-bot@chromium.org <commit-bot@chromium.org>
V2ClientHello is going to be ugly wherever we do it, but this hides it
behind the transport method table. It removes a place where the
handshake state machine reaches into ssl3_get_message's internal state.
ssl3_get_message will now silently translate V2ClientHellos into true
ClientHellos and manage the handshake hash appropriately.
Now the only accesses of init_buf from the handshake state machines are
to create and destroy the buffer.
Change-Id: I81467a038f6ac472a465eec7486a443fe50a98e1
Reviewed-on: https://boringssl-review.googlesource.com/8641
Reviewed-by: Adam Langley <agl@google.com>
To match the Go side. That message will never be used for anything else,
so there's not much need to give it such a long name.
Change-Id: I3396c9d513d02d873e59cd8e81ee64005c5c706c
Reviewed-on: https://boringssl-review.googlesource.com/8620
Reviewed-by: Steven Valdez <svaldez@google.com>
Reviewed-by: David Benjamin <davidben@google.com>
They're not necessary.
Change-Id: Ifeb3fae73a8b22f88019e6ef9f9ba5e64ed3cfab
Reviewed-on: https://boringssl-review.googlesource.com/8543
Reviewed-by: David Benjamin <davidben@google.com>
Match the actual name of the type.
Change-Id: I0ad27196ee2876ce0690d13068fa95f68b05b0da
Reviewed-on: https://boringssl-review.googlesource.com/8187
Reviewed-by: David Benjamin <davidben@google.com>
This keeps the naming convention in line with the actual spec.
Change-Id: I34673f78dbc29c1659b4da0e49677ebe9b79636b
Reviewed-on: https://boringssl-review.googlesource.com/8090
Reviewed-by: David Benjamin <davidben@google.com>
This renames the Channel ID EncryptedExtensions message to allow for
compatibility with TLS 1.3 EncryptedExtensions.
Change-Id: I5b67d00d548518045554becb1b7213fba86731f2
Reviewed-on: https://boringssl-review.googlesource.com/8040
Reviewed-by: Adam Langley <agl@google.com>
They're completely unused now. The handshake message reassembly logic should
not depend on the state machine. This should partially free it up (ugly as it
is) to be shared with a future TLS 1.3 implementation while, in parallel, it
and the layers below, get reworked. This also cuts down on the number of states
significantly.
Partially because I expect we'd want to get ssl_hash_message_t out of there
too. Having it in common code is fine, but it needs to be in the (supposed to
be) protocol-agnostic handshake state machine, not the protocol-specific
handshake message layer.
Change-Id: I12f9dc57bf433ceead0591106ab165d352ef6ee4
Reviewed-on: https://boringssl-review.googlesource.com/7949
Reviewed-by: Adam Langley <agl@google.com>
This dates to SSLeay 0.8.0 (or earlier). The use counter sees virtually
no hits.
Change-Id: Iff4c8899d5cb0ba4afca113c66d15f1d980ffe41
Reviewed-on: https://boringssl-review.googlesource.com/6558
Reviewed-by: Adam Langley <agl@google.com>
Like tls1.h, ssl3.h is now just a bundle of protocol constants.
Hopefully we can opaquify this struct in due time, but for now it's
still public.
Change-Id: I68366eb233702e149c92e21297f70f8a4a45f060
Reviewed-on: https://boringssl-review.googlesource.com/6300
Reviewed-by: Adam Langley <alangley@gmail.com>
This dates all the way to SSLeay 0.9.0b. At this point the
application/handshake interleave logic in ssl3_read_bytes was already
present:
((
(s->state & SSL_ST_CONNECT) &&
(s->state >= SSL3_ST_CW_CLNT_HELLO_A) &&
(s->state <= SSL3_ST_CR_SRVR_HELLO_A)
) || (
(s->state & SSL_ST_ACCEPT) &&
(s->state <= SSL3_ST_SW_HELLO_REQ_A) &&
(s->state >= SSL3_ST_SR_CLNT_HELLO_A)
)
The comment is attached to SSL3_ST_SR_CLNT_HELLO_A, so I suspect this is
what it was about. This logic is gone now, so let's remove that scary
warning.
Change-Id: I45f13b53b79e35d80e6074b0942600434deb0684
Reviewed-on: https://boringssl-review.googlesource.com/6299
Reviewed-by: Adam Langley <alangley@gmail.com>
Applications may require the stapled OCSP response in order to verify
the certificate within the verification callback.
Change-Id: I8002e527f90c3ce7b6a66e3203c0a68371aac5ec
Reviewed-on: https://boringssl-review.googlesource.com/5730
Reviewed-by: David Benjamin <davidben@chromium.org>
Reviewed-by: Adam Langley <agl@google.com>
The handshake state machine is still rather messy (we should switch to CBB,
split the key exchanges apart, and also pull reading and writing out), but this
version makes it more obvious to the compiler that |p| and |sig_len| are
initialized. The old logic created a synchronous-only state which, if enterred
directly, resulted in some variables being uninitialized.
Change-Id: Ia3ac9397d523fe299c50a95dc82a9b26304cea96
Reviewed-on: https://boringssl-review.googlesource.com/5765
Reviewed-by: Adam Langley <agl@google.com>
Gets another field out of the SSL_SESSION.
Change-Id: I9a27255533f8e43e152808427466ec1306cfcc60
Reviewed-on: https://boringssl-review.googlesource.com/5756
Reviewed-by: Adam Langley <agl@google.com>
This begins decoupling the transport from the SSL state machine. The buffering
logic is hidden behind an opaque API. Fields like ssl->packet and
ssl->packet_length are gone.
ssl3_get_record and dtls1_get_record now call low-level tls_open_record and
dtls_open_record functions that unpack a single record independent of who owns
the buffer. Both may be called in-place. This removes ssl->rstate which was
redundant with the buffer length.
Future work will push the buffer up the stack until it is above the handshake.
Then we can expose SSL_open and SSL_seal APIs which act like *_open_record but
return a slightly larger enum due to other events being possible. Likewise the
handshake state machine will be detached from its buffer. The existing
SSL_read, SSL_write, etc., APIs will be implemented on top of SSL_open, etc.,
combined with ssl_read_buffer_* and ssl_write_buffer_*. (Which is why
ssl_read_buffer_extend still tries to abstract between TLS's and DTLS's fairly
different needs.)
The new buffering logic does not support read-ahead (removed previously) since
it lacks a memmove on ssl_read_buffer_discard for TLS, but this could be added
if desired. The old buffering logic wasn't quite right anyway; it tried to
avoid the memmove in some cases and could get stuck too far into the buffer and
not accept records. (The only time the memmove is optional is in DTLS or if
enough of the record header is available to know that the entire next record
would fit in the buffer.)
The new logic also now actually decrypts the ciphertext in-place again, rather
than almost in-place when there's an explicit nonce/IV. (That accidentally
switched in https://boringssl-review.googlesource.com/#/c/4792/; see
3d59e04bce96474099ba76786a2337e99ae14505.)
BUG=468889
Change-Id: I403c1626253c46897f47c7ae93aeab1064b767b2
Reviewed-on: https://boringssl-review.googlesource.com/5715
Reviewed-by: Adam Langley <agl@google.com>
This is a simpler implementation than OpenSSL's, lacking responder IDs
and request extensions support. This mirrors the client implementation
already present.
Change-Id: I54592b60e0a708bfb003d491c9250401403c9e69
Reviewed-on: https://boringssl-review.googlesource.com/5700
Reviewed-by: Adam Langley <agl@google.com>
I'm not sure why I made a separate one. (Not quite how the V2ClientHello
code will look in the buffer-free API yet. Probably the future
refactored SSL_HANDSHAKE gadget will need separate entry points to
consume a handshake message or V2ClientHello and the driver deals with
framing.)
This also means that ssl3_setup_read_buffer is never called external to
ssl3_read_n.
BUG=468889
Change-Id: I872f1188270968bf53ee9d0488a761c772a11e9e
Reviewed-on: https://boringssl-review.googlesource.com/5713
Reviewed-by: Adam Langley <agl@google.com>
Rather than support arbitrarily many handshake hashes in the general
case (which the PRF logic assumes is capped at two), special-case the
MD5/SHA1 two-hash combination and otherwise maintain a single rolling
hash.
Change-Id: Ide9475565b158f6839bb10b8b22f324f89399f92
Reviewed-on: https://boringssl-review.googlesource.com/5618
Reviewed-by: Adam Langley <agl@google.com>
A memory BIO is internally a BUF_MEM anyway. There's no need to bring
BIO_write into the mix. BUF_MEM is size_t clean.
Change-Id: I4ec6e4d22c72696bf47c95861771013483f75cab
Reviewed-on: https://boringssl-review.googlesource.com/5616
Reviewed-by: Adam Langley <agl@google.com>
The split was only needed for buffering records. Likewise, the extra
seq_num field is now unnecessary.
This also fixes a bug where dtls1_process_record will push an error on
the queue if the decrypted record is too large, which dtls1_get_record
will ignore but fail to clear, leaving garbage on the error queue. The
error is now treated as fatal; the reason DTLS silently drops invalid
packets is worrying about ease of DoS, but after SSL_AEAD_CTX_open, the
packet has been authenticated. (Unless it's the null cipher, but that's
during the handshake and the handshake is already DoS-able by breaking
handshake reassembly state.)
The function is still rather a mess. Later changes will clean this up.
BUG=468889
Change-Id: I96a54afe0755d43c34456f76e77fc4ee52ad01e3
Reviewed-on: https://boringssl-review.googlesource.com/5557
Reviewed-by: Adam Langley <agl@google.com>
The only point format that we ever support is uncompressed, which the
RFC says implementations MUST support. The TLS 1.3 and Curve25519
forecast is that point format negotiation is gone. Each curve has just
one point format and it's labeled, for historial reasons, as
"uncompressed".
Change-Id: I8ffc8556bed1127cf288d2a29671abe3c9b3c585
Reviewed-on: https://boringssl-review.googlesource.com/5542
Reviewed-by: Adam Langley <agl@google.com>
The RSA key exchange needs decryption and is still unsupported.
Change-Id: I8c13b74e25a5424356afbe6e97b5f700a56de41f
Reviewed-on: https://boringssl-review.googlesource.com/5467
Reviewed-by: Adam Langley <agl@google.com>
This change mirrors upstream's custom extension API because we have some
internal users that depend on it.
Change-Id: I408e442de0a55df7b05c872c953ff048cd406513
Reviewed-on: https://boringssl-review.googlesource.com/5471
Reviewed-by: Adam Langley <agl@google.com>
This also removes support for the “old” Channel ID extension.
Change-Id: I1168efb9365c274db6b9d7e32013336e4404ff54
Reviewed-on: https://boringssl-review.googlesource.com/5462
Reviewed-by: Adam Langley <agl@google.com>
This is certainly far from exhaustive, but get rid of these.
Change-Id: Ie96925bcd452873ed8399b68e1e71d63e5a0929b
Reviewed-on: https://boringssl-review.googlesource.com/5357
Reviewed-by: Adam Langley <agl@google.com>
The SSL23_ST_foo macros are only used in ssl_stat.c.
However, these states are never set and can be removed.
Move the two remaining SSLv2 client hello record macros to ssl3.h
Change-Id: I76055405a9050cf873b4d1cbc689e54dd3490b8a
Reviewed-on: https://boringssl-review.googlesource.com/4160
Reviewed-by: Adam Langley <agl@google.com>
Rather than four massive functions that handle every extension,
organise the code by extension with four smaller functions for each.
Change-Id: I876b31dacb05aca9884ed3ae7c48462e6ffe3b49
Reviewed-on: https://boringssl-review.googlesource.com/5142
Reviewed-by: Adam Langley <agl@google.com>
This adds a new API, SSL_set_private_key_method, which allows the consumer to
customize private key operations. For simplicity, it is incompatible with the
multiple slots feature (which will hopefully go away) but does not, for now,
break it.
The new method is only routed up for the client for now. The server will
require a decrypt hook as well for the plain RSA key exchange.
BUG=347404
Change-Id: I35d69095c29134c34c2af88c613ad557d6957614
Reviewed-on: https://boringssl-review.googlesource.com/5049
Reviewed-by: Adam Langley <agl@google.com>
We shouldn't have protocol constraints that are sensitive to whether
data is returned synchronously or not.
Per https://boringssl-review.googlesource.com/#/c/4112/, the original
limitation was to avoid OpenSSL ABI changes. This is no longer a
concern.
Add tests for the sync and async case. Send the empty records in two
batches to ensure the count is reset correctly.
Change-Id: I3fee839438527e71adb83d437879bb0d49ca5c07
Reviewed-on: https://boringssl-review.googlesource.com/5040
Reviewed-by: Adam Langley <agl@google.com>
When the peer or caller requests a renegotiation, OpenSSL doesn't
renegotiate immediately. It sets a flag to begin a renegotiation as soon
as record-layer read and write buffers are clear. One reason is that
OpenSSL's record layer cannot write a handshake record while an
application data record is being written. The buffer consistency checks
around partial writes will break.
None of these cases are relevant for the client auth hack. We already
require that renego come in at a quiescent part of the application
protocol by forbidding handshake/app_data interleave.
The new behavior is now: when a HelloRequest comes in, if the record
layer is not idle, the renegotiation is rejected as if
SSL_set_reject_peer_renegotiations were set. Otherwise we immediately
begin the new handshake. The server may not send any application data
between HelloRequest and completing the handshake. The HelloRequest may
not be consumed if an SSL_write is pending.
Note this does require that Chromium's HTTP stack not attempt to read
the HTTP response until the request has been written, but the
renegotiation logic already assumes it. Were Chromium to drive the
SSL_read state machine early and the server, say, sent a HelloRequest
after reading the request headers but before we've sent the whole POST
body, the SSL state machine may racily enter renegotiate early, block
writing the POST body on the new handshake, which would break Chromium's
ERR_SSL_CLIENT_AUTH_CERT_NEEDED plumbing.
BUG=429450
Change-Id: I6278240c3bceb5d2e1a2195bdb62dd9e0f4df718
Reviewed-on: https://boringssl-review.googlesource.com/4825
Reviewed-by: Adam Langley <agl@google.com>
tls1_enc is now SSL_AEAD_CTX_{open,seal}. This starts tidying up a bit
of the record-layer logic. This removes rr->input, as encrypting and
decrypting records no longer refers to various globals. It also removes
wrec altogether. SSL3_RECORD is now only used to maintain state about
the current incoming record. Outgoing records go straight to the write
buffer.
This also removes the outgoing alignment memcpy and simply calls
SSL_AEAD_CTX_seal with the parameters as appropriate. From bssl speed
tests, this seems to be faster on non-ARM and a bit of a wash on ARM.
Later it may be worth recasting these open/seal functions to write into
a CBB (tweaked so it can be malloc-averse), but for now they take an
out/out_len/max_out trio like their EVP_AEAD counterparts.
BUG=468889
Change-Id: Ie9266a818cc053f695d35ef611fd74c5d4def6c3
Reviewed-on: https://boringssl-review.googlesource.com/4792
Reviewed-by: Adam Langley <agl@google.com>
There's multiple different versions of this check, between
s->s3->have_version (only works at some points), s->new_session (really
weird and not actually right), s->renegotiate (fails on the server
because it's always 2 after ClientHello), and s->s3->tmp.finish_md_len
(super confusing). Add an explicit bit with clear meaning. We'll prune
some of the others later; notably s->renegotiate can go away when
initiating renegotiation is removed.
This also tidies up the extensions to be consistent about whether
they're allowed during renego:
- ALPN failed to condition when accepting from the server, so even
if the client didn't advertise, the server could.
- SCTs now *are* allowed during renego. I think forbidding it was a
stray copy-paste. It wasn't consistently enforced in both ClientHello
and ServerHello, so the server could still supply it. Moreover, SCTs
are part of the certificate, so we should accept it wherever we accept
certificates, otherwise that session's state becomes incomplete. This
matches OCSP stapling. (NB: Chrome will never insert a session created
on renego into the session cache and won't accept a certificate
change, so this is moot anyway.)
Change-Id: Ic9bd1ebe2a2dbe75930ed0213bf3c8ed8170e251
Reviewed-on: https://boringssl-review.googlesource.com/4730
Reviewed-by: Adam Langley <agl@google.com>
The only difference is SSL_clear_num_renegotiations which is never
called.
Change-Id: Id661c71e89d34d834349ad1f1a296e332606e6cc
Reviewed-on: https://boringssl-review.googlesource.com/4564
Reviewed-by: Adam Langley <agl@google.com>
Compression is gone, so don't allow for compression overhead. With that fixed,
the second rr->length check in ssl3_get_record matches the length computation
which sizes the read buffer. The first is wrong and doesn't account for the
alignment padding. Move the second to the first.
Change-Id: I3f4f05de9fdf5c645ff24493bbfdf303dcc1aa90
Reviewed-on: https://boringssl-review.googlesource.com/4236
Reviewed-by: Adam Langley <agl@google.com>
