Fixes#18690
In some circumstances, it's possible that when using an external
database for the session cache, that pulling in an entry from that
cache to the internal cache will cause the newly added entry to
be deleted from the internal cache. This is likely to happen when
the internal cache is set to have a small size, and the newly added
entry's timeout places it at the end of the cache list.
This could be fixed by updating the timestamp of the session (via
`SSL_SESSION_set_time()` or `SSL_SESSION_set_timeout()`) before
adding to the cache. But that may not be desireable.
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18905)
As EVP_CIPHER_CTX_get_iv_length indicates failure with -1, this error
should be processed. Also the result of this function shouldn't be
assigned to an unsigned variable.
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18891)
If app data is received before a Finished message in DTLS then we buffer
it to return later. The function SSL_pending() is supposed to tell you
how much processed app data we have already buffered, and SSL_has_pending()
is supposed to tell you if we have any data buffered (whether processed or
not, and whether app data or not).
Neither SSL_pending() or SSL_has_pending() were taking account of this
DTLS specific app data buffer.
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18868)
This adds functions for encoding and decoding QUIC frames.
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18795)
Make the SSL object polymorphic based on whether this is
a traditional SSL connection, QUIC connection, or later
to be implemented a QUIC stream.
It requires adding if after every SSL_CONNECTION_FROM_SSL() call
which itself has to be added to almost every public SSL_ API call.
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18612)
This rather long function used to call BIO_get_ktls_send
mutliple times, although that result cannot change during
the execution of that function.
There was a similar unnecessary call to BIO_get_ktls_recv
in ssl3_get_record.
And while I'm already there, rewrite ssl3_write_bytes
to use BIO_get_ktls_send as a boolean (so using "!" instead
of "== 0").
Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18304)
Fix multiple places that could potentially segfault if memory
allocations fail. e.g. ssl_load_ciphers() could fail while calling
ssl_evp_md_fetch().
Found by #18355
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/18784)
As the potential failure of the BN_dup,
it should be better to check the return value
in order to guarantee the success.
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Todd Short <todd.short@me.com>
(Merged from https://github.com/openssl/openssl/pull/18699)
- If keymgmmt is not available, it's not an error but the error message
persists in stack
- when setting groups, it's worth saying which group is not available
Fixes#18585
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18591)
When TLS-1.3 is used and the server does not send any CA names
the ca_dn will be NULL. sk_X509_NAME_num() returns -1 on null
argument.
Reviewed-by: Todd Short <todd.short@me.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/17986)
This requires some code being pulled into the empty protocol
implementation so the state machinery works.
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18307)
Related: CVE-2022-29242
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18381)
Prior to the crash there is an out of memory error
in X509_verify_cert which makes the chain NULL or
empty. The error is ignored by ssl_add_cert_chain,
and ssl_security_cert_chain crashes due to the
unchecked null pointer.
This is reproducible with my error injection patch.
The test vector has been validated on the 1.1.1 branch
but the issue is of course identical in all branches.
$ ERROR_INJECT=1652848273 ../util/shlib_wrap.sh ./server-test ./corpora/server/47c8e933c4ec66fa3c309422283dfe0f31aafae8# ./corpora/server/47c8e933c4ec66fa3c309422283dfe0f31aafae8
#0 0x7f3a8f766eba in __sanitizer_print_stack_trace ../../../../gcc-trunk/libsanitizer/asan/asan_stack.cpp:87
#1 0x403ba4 in my_malloc fuzz/test-corpus.c:114
#2 0x7f3a8f39a430 in CRYPTO_zalloc crypto/mem.c:230
#3 0x7f3a8f46bd3b in sk_reserve crypto/stack/stack.c:180
#4 0x7f3a8f46bd3b in OPENSSL_sk_insert crypto/stack/stack.c:242
#5 0x7f3a8f4a4fd8 in sk_X509_push include/openssl/x509.h:99
#6 0x7f3a8f4a4fd8 in X509_verify_cert crypto/x509/x509_vfy.c:286
#7 0x7f3a8fed726e in ssl_add_cert_chain ssl/statem/statem_lib.c:959
#8 0x7f3a8fed726e in ssl3_output_cert_chain ssl/statem/statem_lib.c:1015
#9 0x7f3a8fee1c50 in tls_construct_server_certificate ssl/statem/statem_srvr.c:3812
#10 0x7f3a8feb8b0a in write_state_machine ssl/statem/statem.c:843
#11 0x7f3a8feb8b0a in state_machine ssl/statem/statem.c:443
#12 0x7f3a8fe84b3f in SSL_do_handshake ssl/ssl_lib.c:3718
#13 0x403202 in FuzzerTestOneInput fuzz/server.c:740
#14 0x40371b in testfile fuzz/test-corpus.c:182
#15 0x402856 in main fuzz/test-corpus.c:226
#16 0x7f3a8e859f44 in __libc_start_main (/lib/x86_64-linux-gnu/libc.so.6+0x21f44)
#17 0x402936 (/home/ed/OPC/openssl/fuzz/server-test+0x402936)
AddressSanitizer:DEADLYSIGNAL
=================================================================
==8400==ERROR: AddressSanitizer: SEGV on unknown address 0x000000000158 (pc 0x7f3a8f4d822f bp 0x7ffc39b76190 sp 0x7ffc39b760a0 T0)
==8400==The signal is caused by a READ memory access.
==8400==Hint: address points to the zero page.
#0 0x7f3a8f4d822f in x509v3_cache_extensions crypto/x509v3/v3_purp.c:386
#1 0x7f3a8f4d9d3a in X509_check_purpose crypto/x509v3/v3_purp.c:84
#2 0x7f3a8f4da02a in X509_get_extension_flags crypto/x509v3/v3_purp.c:921
#3 0x7f3a8feff7d2 in ssl_security_cert_sig ssl/t1_lib.c:2518
#4 0x7f3a8feff7d2 in ssl_security_cert ssl/t1_lib.c:2542
#5 0x7f3a8feffa03 in ssl_security_cert_chain ssl/t1_lib.c:2562
#6 0x7f3a8fed728d in ssl_add_cert_chain ssl/statem/statem_lib.c:963
#7 0x7f3a8fed728d in ssl3_output_cert_chain ssl/statem/statem_lib.c:1015
#8 0x7f3a8fee1c50 in tls_construct_server_certificate ssl/statem/statem_srvr.c:3812
#9 0x7f3a8feb8b0a in write_state_machine ssl/statem/statem.c:843
#10 0x7f3a8feb8b0a in state_machine ssl/statem/statem.c:443
#11 0x7f3a8fe84b3f in SSL_do_handshake ssl/ssl_lib.c:3718
#12 0x403202 in FuzzerTestOneInput fuzz/server.c:740
#13 0x40371b in testfile fuzz/test-corpus.c:182
#14 0x402856 in main fuzz/test-corpus.c:226
#15 0x7f3a8e859f44 in __libc_start_main (/lib/x86_64-linux-gnu/libc.so.6+0x21f44)
#16 0x402936 (/home/ed/OPC/openssl/fuzz/server-test+0x402936)
AddressSanitizer can not provide additional info.
SUMMARY: AddressSanitizer: SEGV crypto/x509v3/v3_purp.c:386 in x509v3_cache_extensions
==8400==ABORTING
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18376)
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18368)
Just extending the comment so coveriety knows as well.
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18326)
It seems the return value of EVP_PKEY_copy_parameters() in
ssl_set_cert_and_key(), and could lead to null pointer dereference in
EVP_PKEY_eq() function.
However those functions are complicated and this fix is suggested by
a static analyzer, so please advise.
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18071)
Even though the function is not part of the public api, it is not
entirely removed, in order to minimize the chance of breakage,
because it is exported from libcrypto. Instead, we keep a dummy
implementation.
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/17974)
rotated_mac is a 64-byte aligned buffer of size 64 and rotate_offset is secret.
Consider a weaker leakage model(CL) where only cacheline base address is leaked,
i.e address/32 for 32-byte cacheline(CL32).
Previous code used to perform two loads
1. rotated_mac[rotate_offset ^ 32] and
2. rotated_mac[rotate_offset++]
which would leak 2q + 1, 2q for 0 <= rotate_offset < 32
and 2q, 2q + 1 for 32 <= rotate_offset < 64
The proposed fix performs load operations which will always leak 2q, 2q + 1 and
selects the appropriate value in constant-time.
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18033)
This is in line with the NEWS entry (erroneously) announcing such for 3.0.
Fixes#18194
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/18236)
CLA: trivial
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Todd Short <todd.short@me.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18220)
This fixes an internal error alert from the server and
an unexpected connection failure in the release version,
but a failed assertion and a server crash in the
debug version.
Reproduce this issue with a DTLS server/client like that:
./openssl s_server -dtls -mtu 1500
./openssl s_client -dtls -maxfraglen 512
In the debug version a crash happens in the Server now:
./openssl s_server -dtls -mtu 1500
Using default temp DH parameters
ACCEPT
ssl/statem/statem_dtls.c:269: OpenSSL internal error: Assertion failed: len == written
Aborted (core dumped)
While in the release version the handshake exceeds the
negotiated max fragment size, and fails because of this:
$ ./openssl s_server -dtls -mtu 1500
Using default temp DH parameters
ACCEPT
ERROR
4057152ADA7F0000:error:0A0000C2:SSL routines:do_dtls1_write:exceeds max fragment size:ssl/record/rec_layer_d1.c:826:
shutting down SSL
CONNECTION CLOSED
From the client's point of view the connection fails
with an Internal Error Alert:
$ ./openssl s_client -dtls -maxfraglen 512
Connecting to ::1
CONNECTED(00000003)
40B76343377F0000:error:0A000438:SSL routines:dtls1_read_bytes:tlsv1 alert internal error:ssl/record/rec_layer_d1.c:613:SSL alert number 80
and now the connection attempt fails unexpectedly.
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18093)
Currently we do not have any way to retrieve these values once set.
Fixes#18035.
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/18038)
This removes a guard condition that prevents KTLS being enabled for
receiving in TLS 1.3. Use the correct sequence number and BIO for
receive vs transmit offload.
Co-authored-by: John Baldwin <jhb@FreeBSD.org>
Signed-off-by: Daiki Ueno <dueno@redhat.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/17942)
- Don't unpad records, check the outer record type, or extract the
inner record type from TLS 1.3 records handled by the kernel. KTLS
performs all of these steps and returns the inner record type in the
TLS header.
- When checking the length of a received TLS 1.3 record don't allow
for the extra byte for the nested record type when KTLS is used.
- Pass a pointer to the record type in the TLS header to the
SSL3_RT_INNER_CONTENT_TYPE message callback. For KTLS, the old
pointer pointed to the last byte of payload rather than the record
type. For the non-KTLS case, the TLS header has been updated with
the inner type before this callback is invoked.
Co-authored-by: John Baldwin <jhb@FreeBSD.org>
Signed-off-by: Daiki Ueno <dueno@redhat.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/17942)
When KTLS receive is enabled, pending data may still be present due to
read ahead. This data must still be processed the same as records
received without KTLS. To ease readability (especially in
consideration of additional checks which will be added for TLS 1.3),
add a helper variable 'using_ktls' that is true when the KTLS receive
path is being used to receive a record.
Co-authored-by: Dmitry Podgorny <pasis.ua@gmail.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/17942)
KTLS implementations currently assume that the start of the in-kernel
socket buffer is aligned with the start of a TLS record for the
receive side. The socket option to enable KTLS specifies the TLS
sequence number of this initial record.
When read ahead is enabled, data can be pending in the SSL read buffer
after negotiating session keys. This pending data must be examined to
ensurs that the kernel's socket buffer does not contain a partial TLS
record as well as to determine the correct sequence number of the
first TLS record to be processed by the kernel.
In preparation for enabling receive kernel offload for TLS 1.3, move
the existing logic to handle read ahead from t1_enc.c into ktls.c and
invoke it from ktls_configure_crypto().
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/17942)
For TLSv1.3, limit ticket lifetime hint to 1 week per RFC8446
Fixes#17948
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/17952)
time_t is a 64 bits type on this platform.
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/17917)
