1
0
mirror of https://github.com/moparisthebest/curl synced 2024-12-22 16:18:48 -05:00
Commit Graph

548 Commits

Author SHA1 Message Date
Thomas Glanzmann
4f8b17743d HTTPS Proxy: Implement CURLOPT_PROXY_PINNEDPUBLICKEY 2016-11-25 10:49:38 +01:00
Daniel Stenberg
6832c1d4b2 checksrc: move open braces to comply with function declaration style 2016-11-24 23:58:22 +01:00
Daniel Stenberg
8657c268e1 checksrc: white space edits to comply to stricter checksrc 2016-11-24 23:58:22 +01:00
Daniel Stenberg
dbadaebfc4 checksrc: code style: use 'char *name' style 2016-11-24 23:58:22 +01:00
Daniel Stenberg
74ffa040a4 openssl: remove dead code
Coverity CID 1394666
2016-11-24 23:41:45 +01:00
Okhin Vasilij
c6da05a5ec HTTPS-proxy: fixed mbedtls and polishing 2016-11-24 23:41:45 +01:00
Daniel Stenberg
49765cd75c darwinssl: adopted to the HTTPS proxy changes
It builds and runs all test cases. No adaptations for actual HTTPS proxy
support has been made.
2016-11-24 23:41:45 +01:00
Daniel Stenberg
8b4352658a gtls: fix indent to silence compiler warning
vtls/gtls.c: In function ‘Curl_gtls_data_pending’:
vtls/gtls.c:1429:3: error: this ‘if’ clause does not guard... [-Werror=misleading-indentation]
   if(conn->proxy_ssl[connindex].session &&
      ^~
      vtls/gtls.c:1433:5: note: ...this statement, but the latter is misleadingly indented as if it is guarded by the ‘if’
           return res;
2016-11-24 23:41:45 +01:00
Thomas Glanzmann
8cb872df10 mbedtls: Fix compile errors 2016-11-24 23:41:45 +01:00
Alex Rousskov
cb4e2be7c6 proxy: Support HTTPS proxy and SOCKS+HTTP(s)
* HTTPS proxies:

An HTTPS proxy receives all transactions over an SSL/TLS connection.
Once a secure connection with the proxy is established, the user agent
uses the proxy as usual, including sending CONNECT requests to instruct
the proxy to establish a [usually secure] TCP tunnel with an origin
server. HTTPS proxies protect nearly all aspects of user-proxy
communications as opposed to HTTP proxies that receive all requests
(including CONNECT requests) in vulnerable clear text.

With HTTPS proxies, it is possible to have two concurrent _nested_
SSL/TLS sessions: the "outer" one between the user agent and the proxy
and the "inner" one between the user agent and the origin server
(through the proxy). This change adds supports for such nested sessions
as well.

A secure connection with a proxy requires its own set of the usual SSL
options (their actual descriptions differ and need polishing, see TODO):

  --proxy-cacert FILE        CA certificate to verify peer against
  --proxy-capath DIR         CA directory to verify peer against
  --proxy-cert CERT[:PASSWD] Client certificate file and password
  --proxy-cert-type TYPE     Certificate file type (DER/PEM/ENG)
  --proxy-ciphers LIST       SSL ciphers to use
  --proxy-crlfile FILE       Get a CRL list in PEM format from the file
  --proxy-insecure           Allow connections to proxies with bad certs
  --proxy-key KEY            Private key file name
  --proxy-key-type TYPE      Private key file type (DER/PEM/ENG)
  --proxy-pass PASS          Pass phrase for the private key
  --proxy-ssl-allow-beast    Allow security flaw to improve interop
  --proxy-sslv2              Use SSLv2
  --proxy-sslv3              Use SSLv3
  --proxy-tlsv1              Use TLSv1
  --proxy-tlsuser USER       TLS username
  --proxy-tlspassword STRING TLS password
  --proxy-tlsauthtype STRING TLS authentication type (default SRP)

All --proxy-foo options are independent from their --foo counterparts,
except --proxy-crlfile which defaults to --crlfile and --proxy-capath
which defaults to --capath.

Curl now also supports %{proxy_ssl_verify_result} --write-out variable,
similar to the existing %{ssl_verify_result} variable.

Supported backends: OpenSSL, GnuTLS, and NSS.

* A SOCKS proxy + HTTP/HTTPS proxy combination:

If both --socks* and --proxy options are given, Curl first connects to
the SOCKS proxy and then connects (through SOCKS) to the HTTP or HTTPS
proxy.

TODO: Update documentation for the new APIs and --proxy-* options.
Look for "Added in 7.XXX" marks.
2016-11-24 23:41:44 +01:00
Patrick Monnerat
945f60e8a7 Limit ASN.1 structure sizes to 256K. Prevent some allocation size overflows.
See CRL-01-006.
2016-11-24 14:28:39 +01:00
Marcel Raad
c0ae2dbb86
vtls: fix build warnings
Fix warnings about conversions from long to time_t in openssl.c and
schannel.c.

Follow-up to de4de4e3c7
2016-11-19 14:09:03 +01:00
David Schweikert
7c9b9add6f darwinssl: fix SSL client certificate not found on MacOS Sierra
Reviewed-by: Nick Zitzmann

Closes #1105
2016-11-15 08:41:32 +01:00
Daniel Stenberg
f682156a4f Curl_rand: fixed and moved to rand.c
Now Curl_rand() is made to fail if it cannot get the necessary random
level.

Changed the proto of Curl_rand() slightly to provide a number of ints at
once.

Moved out from vtls, since it isn't a TLS function and vtls provides
Curl_ssl_random() for this to use.

Discussion: https://curl.haxx.se/mail/lib-2016-11/0119.html
2016-11-14 08:23:52 +01:00
Daniel Stenberg
ddefc056b6 openssl: make sure to fail in the unlikely event that PRNG seeding fails 2016-11-11 14:16:31 +01:00
Daniel Stenberg
942c952db6 openssl: avoid unnecessary seeding if already done
1.1.0+ does more of this by itself so we can avoid extra processing this
way.
2016-11-11 13:54:16 +01:00
Daniel Stenberg
f9de7c405f openssl: RAND_status always exists in OpenSSL >= 0.9.7
and remove RAND_screen from configure since nothing is using that
function
2016-11-11 13:44:09 +01:00
Daniel Stenberg
ea80a2dcfc openssl: raise the max_version to 1.3 if asked for
Now I've managed to negotiate TLS 1.3 with https://enabled.tls13.com/ when
using boringssl.
2016-11-10 11:16:25 +01:00
Jay Satiro
f43b7b6cb6 vtls: Fail on unrecognized param for CURLOPT_SSLVERSION
- Fix GnuTLS code for CURL_SSLVERSION_TLSv1_2 that broke when the
TLS 1.3 support was added in 6ad3add.

- Homogenize across code for all backends the error message when TLS 1.3
is not available to "<backend>: TLS 1.3 is not yet supported".

- Return an error when a user-specified ssl version is unrecognized.

---

Prior to this change our code for some of the backends used the
'default' label in the switch statement (ie ver unrecognized) for
ssl.version and treated it the same as CURL_SSLVERSION_DEFAULT.

Bug: https://curl.haxx.se/mail/lib-2016-11/0048.html
Reported-by: Kamil Dudka
2016-11-09 22:08:49 -05:00
Kamil Dudka
5e6c04f913 nss: silence warning 'SSL_NEXT_PROTO_EARLY_VALUE not handled in switch'
... with nss-3.26.0 and newer

Reported-by: Daniel Stenberg
2016-11-07 15:40:32 +01:00
Daniel Stenberg
9e8b0a283f openssl: initial TLS 1.3 adaptions
BoringSSL supports TLSv1.3 already, but these changes don't seem to be anough
to get it working.
2016-11-07 14:38:59 +01:00
Kamil Dudka
6ad3add606 vtls: support TLS 1.3 via CURL_SSLVERSION_TLSv1_3
Fully implemented with the NSS backend only for now.

Reviewed-by: Ray Satiro
2016-11-07 11:52:07 +01:00
Kamil Dudka
5d45ced7a4 nss: map CURL_SSLVERSION_DEFAULT to NSS default
... but make sure we use at least TLSv1.0 according to libcurl API

Reported-by: Cure53
Reviewed-by: Ray Satiro
2016-11-07 11:51:33 +01:00
Daniel Stenberg
27302abb94 s/cURL/curl
We're mostly saying just "curl" in lower case these days so here's a big
cleanup to adapt to this reality. A few instances are left as the
project could still formally be considered called cURL.
2016-11-07 10:36:23 +01:00
Daniel Stenberg
9ea3a6e150 mbedtls: fix build with mbedtls versions < 2.4.0
Regression added in 62a8095e71

Reported-by: Tony Kelman

Discussed in #1087
2016-11-04 11:14:15 +01:00
Kamil Dudka
1071680946 gskit, nss: do not include strequal.h
follow-up to 811a693b80
2016-10-31 17:36:22 +01:00
Dan Fandrich
9363f1a37f strcasecompare: replaced remaining rawstr.h with strcase.h
This is a followup to commit 811a693b
2016-10-31 12:57:58 +01:00
Daniel Stenberg
811a693b80 strcasecompare: all case insensitive string compares ignore locale now
We had some confusions on when each function was used. We should not act
differently on different locales anyway.
2016-10-31 08:46:35 +01:00
Daniel Stenberg
502acba2af strcasecompare: is the new name for strequal()
... to make it less likely that we forget that the function actually
does case insentive compares. Also replaced several invokes of the
function with a plain strcmp when case sensitivity is not an issue (like
comparing with "-").
2016-10-31 08:46:35 +01:00
Daniel Stenberg
62a8095e71 mbedtls: stop using deprecated include file
Reported-by: wyattoday
Fixes #1087
2016-10-26 23:38:04 +02:00
Martin Frodl
eb84412b33 nss: fix tight loop in non-blocking TLS handhsake over proxy
... in case the handshake completes before entering
CURLM_STATE_PROTOCONNECT

Bug: https://bugzilla.redhat.com/1388162
2016-10-25 11:56:37 +02:00
Daniel Stenberg
88753c1e89 polarssl: indented code, removed unused variables 2016-10-18 13:27:51 +02:00
Daniel Stenberg
8748d4787f polarssl: reduce #ifdef madness with a macro 2016-10-18 13:17:05 +02:00
Daniel Stenberg
1e3161a20d polarssl: fix unaligned SSL session-id lock 2016-10-18 13:14:11 +02:00
Daniel Stenberg
ace237ae4e Curl_polarsslthreadlock_thread_setup: clear array at init
... since if it fails to init the entire array and then tries to clean
it up, it would attempt to work on an uninitialized pointer.
2016-10-18 13:05:43 +02:00
Daniel Stenberg
8a6e89a9eb select: switch to macros in uppercase
Curl_select_ready() was the former API that was replaced with
Curl_select_check() a while back and the former arg setup was provided
with a define (in order to leave existing code unmodified).

Now we instead offer SOCKET_READABLE and SOCKET_WRITABLE for the most
common shortcuts where only one socket is checked. They're also more
visibly macros.
2016-10-18 11:05:45 +02:00
Daniel Stenberg
1671d84b38 vtls: only re-use session-ids using the same scheme
To make it harder to do cross-protocol mistakes
2016-10-13 11:24:16 +02:00
Kamil Dudka
d1f1c857ad nss: add chacha20-poly1305 cipher suites if supported by NSS 2016-09-22 10:20:12 +02:00
Kamil Dudka
049aa92546 nss: add cipher suites using SHA384 if supported by NSS 2016-09-22 10:17:26 +02:00
Kamil Dudka
7591220270 nss: fix typo in ecdhe_rsa_null cipher suite string
As it seems to be a rarely used cipher suite (for securely established
but _unencrypted_ connections), I believe it is fine not to provide an
alias for the misspelled variant.
2016-09-22 10:17:26 +02:00
Daniel Gustafsson
79607eec51 darwinssl: Fix typo in comment
Closes https://github.com/curl/curl/pull/1028
2016-09-19 17:06:00 -04:00
Bernard Spil
01c8db47aa libressl: fix version output
LibreSSL defines `OPENSSL_VERSION_NUMBER` as `0x20000000L` for all
versions returning `LibreSSL/2.0.0` for any LibreSSL version.

This change provides a local OpenSSL_version_num function replacement
returning LIBRESSL_VERSION_NUMBER instead.

Closes #1029
2016-09-19 22:47:35 +02:00
Daniel Stenberg
a2b360f890 openssl: don't call ERR_remote_thread_state on >= 1.1.0
Follow-up fix to d9321562
2016-09-19 14:32:59 +02:00
Daniel Stenberg
de71e68000 openssl: don’t call CRYTPO_cleanup_all_ex_data
The OpenSSL function CRYTPO_cleanup_all_ex_data() cannot be called
multiple times without crashing - and other libs might call it! We
basically cannot call it without risking a crash. The function is a
no-op since OpenSSL 1.1.0.

Not calling this function only risks a small memory leak with OpenSSL <
1.1.0.

Bug: https://curl.haxx.se/mail/lib-2016-09/0045.html
Reported-by: Todd Short
2016-09-19 11:33:14 +02:00
Nick Zitzmann
227ee9d84d darwinssl: disable RC4 cipher-suite support
RC4 was a nice alternative to CBC back in the days of BEAST, but it's insecure and obsolete now.
2016-09-18 15:01:03 -05:00
Daniel Stenberg
d93215621f openssl: fix per-thread memory leak usiong 1.0.1 or 1.0.2
OpenSSL 1.0.1 and 1.0.2 build an error queue that is stored per-thread
so we need to clean it when easy handles are freed, in case the thread
will be killed in which the easy handle was used. All OpenSSL code in
libcurl should extract the error in association with the error already
so clearing this queue here should be harmless at worst.

Fixes #964
2016-09-14 14:35:44 +02:00
Daniel Stenberg
83ef21e5e9 openssl: fix bad memory free (regression)
... by partially reverting f975f06033. The allocation could be made by
OpenSSL so the free must be made with OPENSSL_free() to avoid problems.

Reported-by: Harold Stuart
Fixes #1005
2016-09-09 23:35:10 +02:00
Gaurav Malhotra
8e176a7c79 openssl: fix CURLINFO_SSL_VERIFYRESULT
CURLINFO_SSL_VERIFYRESULT does not get the certificate verification
result when SSL_connect fails because of a certificate verification
error.

This fix saves the result of SSL_get_verify_result so that it is
returned by CURLINFO_SSL_VERIFYRESULT.

Closes https://github.com/curl/curl/pull/995
2016-09-06 16:17:37 -04:00
Daniel Gustafsson
022dbdb8ac darwinssl: test for errSecSuccess in PKCS12 import rather than noErr (#993)
While noErr and errSecSuccess are defined as the same value, the API
documentation states that SecPKCS12Import() returns errSecSuccess if
there were no errors in importing. Ensure that a future change of the
defined value doesn't break (however unlikely) and be consistent with
the API docs.
2016-09-06 10:37:31 +02:00
Marcel Raad
4af389353a openssl: Fix compilation with OPENSSL_API_COMPAT=0x10100000L
With OPENSSL_API_COMPAT=0x10100000L (OpenSSL 1.1 API), the cleanup
functions are unavailable (they're no-ops anyway in OpenSSL 1.1). The
replacements for SSL_load_error_strings, SSLeay_add_ssl_algorithms, and
OpenSSL_add_all_algorithms are called automatically [1][2]. SSLeay() is
now called OpenSSL_version_num().

[1]: https://www.openssl.org/docs/man1.1.0/ssl/OPENSSL_init_ssl.html
[2]: https://www.openssl.org/docs/man1.1.0/crypto/OPENSSL_init_crypto.html

Closes #992
2016-09-05 22:51:12 +02:00
Jay Satiro
895168bfd3 schannel: Disable ALPN for Wine since it is causing problems
- Disable ALPN on Wine.

- Don't pass input secbuffer when ALPN is disabled.

When ALPN support was added a change was made to pass an input secbuffer
to initialize the context. When ALPN is enabled the buffer contains the
ALPN information, and when it's disabled the buffer is empty. In either
case this input buffer caused problems with Wine and connections would
not complete.

Bug: https://github.com/curl/curl/issues/983
Reported-by: Christian Fillion
2016-08-26 15:35:16 -04:00
Peter Wang
3a5d5de9ef nss: work around race condition in PK11_FindSlotByName()
Serialise the call to PK11_FindSlotByName() to avoid spurious errors in
a multi-threaded environment. The underlying cause is a race condition
in nssSlot_IsTokenPresent().

Bug: https://bugzilla.mozilla.org/1297397

Closes #985
2016-08-26 14:42:22 +02:00
Kamil Dudka
7700fcba64 nss: refuse previously loaded certificate from file
... when we are not asked to use a certificate from file
2016-08-26 13:30:27 +02:00
Daniel Stenberg
39ba8dd09b TLS: random file/egd doesn't have to match for conn reuse 2016-08-26 10:01:19 +02:00
Daniel Stenberg
9cb851e371 openssl: make build with 1.1.0 again
synced with OpenSSL git master commit cc06906707
2016-08-25 12:27:31 +02:00
Daniel Stenberg
667fcb04a6 nss: make the global variables static 2016-08-17 10:50:06 +02:00
Daniel Stenberg
f975f06033 openssl: use regular malloc instead of OPENSSL_malloc
This allows for better memmory debugging and torture tests.
2016-08-16 10:58:33 +02:00
Jay Satiro
b6fcdc32eb openssl: accept subjectAltName iPAddress if no dNSName match
Undo change introduced in d4643d6 which caused iPAddress match to be
ignored if dNSName was present but did not match.

Also, if iPAddress is present but does not match, and dNSName is not
present, fail as no-match. Prior to this change in such a case the CN
would be checked for a match.

Bug: https://github.com/curl/curl/issues/959
Reported-by: wmsch@users.noreply.github.com
2016-08-13 02:14:46 -04:00
Thomas Glanzmann
7b4bf37a44 mbedtls: set debug threshold to 4 (verbose) when MBEDTLS_DEBUG is defined
In order to make MBEDTLS_DEBUG work, the debug threshold must be unequal
to 0.  This patch also adds a comment how mbedtls must be compiled in
order to make debugging work, and explains the possible debug levels.
2016-08-05 00:23:17 +02:00
Daniel Stenberg
11ec5ad435 TLS: only reuse connections with the same client cert
CVE-2016-5420
Bug: https://curl.haxx.se/docs/adv_20160803B.html
2016-08-03 00:34:27 +02:00
Daniel Stenberg
247d890da8 TLS: switch off SSL session id when client cert is used
CVE-2016-5419
Bug: https://curl.haxx.se/docs/adv_20160803A.html
Reported-by: Bru Rom
Contributions-by: Eric Rescorla and Ray Satiro
2016-08-03 00:34:27 +02:00
Thomas Glanzmann
aa9f536a18 mbedtls: Fix debug function name
This patch is necessary so that curl compiles if MBEDTLS_DEBUG is
defined.

Bug: https://curl.haxx.se/mail/lib-2016-08/0001.html
2016-08-01 13:16:42 -04:00
Daniel Stenberg
434f8d0389 internals: rename the SessionHandle struct to Curl_easy 2016-06-22 10:28:41 +02:00
Jay Satiro
04b4ee5498 vtls: Only call add/getsession if session id is enabled
Prior to this change we called Curl_ssl_getsessionid and
Curl_ssl_addsessionid regardless of whether session ID reusing was
enabled. According to comments that is in case session ID reuse was
disabled but then later enabled.

The old way was not intuitive and probably not something users expected.
When a user disables session ID caching I'd guess they don't expect the
session ID to be cached anyway in case the caching is later enabled.
2016-06-22 02:33:29 -04:00
Daniel Stenberg
bb4e7921e7 openssl: use more 'const' to fix build warnings with 1.1.0 branch 2016-06-19 23:21:54 +02:00
Daniel Stenberg
d4643d6e79 openssl: fix cert check with non-DNS name fields present
Regression introduced in 5f5b62635 (released in 7.48.0)

Reported-by: Fabian Ruff
Fixes #875
2016-06-16 10:33:15 +02:00
Dan Fandrich
b1839f6ed8 axtls: Use Curl_wait_ms instead of the less-portable usleep 2016-06-16 08:44:08 +02:00
Dan Fandrich
52c5e9488c axtls: Fixed compile after compile 31c521b0 2016-06-16 08:29:10 +02:00
Steve Holme
34855feeb4 schannel: Disable ALPN on Windows < 8.1
Calling QueryContextAttributes with SECPKG_ATTR_APPLICATION_PROTOCOL
fails on Windows < 8.1 so we need to disable ALPN on these OS versions.

Inspiration provide by: Daniel Seither

Closes #848
Fixes #840
2016-06-06 20:53:30 +01:00
Steve Holme
332e8d6164 win32: Used centralised verify windows version function
Closes #845
2016-06-04 21:24:09 +01:00
Ivan Avdeev
31c521b047 vtls: fix ssl session cache race condition
Sessionid cache management is inseparable from managing individual
session lifetimes. E.g. for reference-counted sessions (like those in
SChannel and OpenSSL engines) every session addition and removal
should be accompanied with refcount increment and decrement
respectively. Failing to do so synchronously leads to a race condition
that causes symptoms like use-after-free and memory corruption.
This commit:
 - makes existing session cache locking explicit, thus allowing
   individual engines to manage lock's scope.
 - fixes OpenSSL and SChannel engines by putting refcount management
   inside this lock's scope in relevant places.
 - adds these explicit locking calls to other engines that use
   sessionid cache to accommodate for this change. Note, however,
   that it is unknown whether any of these engines could also have
   this race.

Bug: https://github.com/curl/curl/issues/815
Fixes #815
Closes #847
2016-06-01 09:40:55 +02:00
Andrew Kurushin
6cabd78531 schannel: add CURLOPT_CERTINFO support
Closes #822
2016-06-01 08:50:01 +02:00
Daniel Stenberg
142ee9fa15 openssl: rename the private SSL_strerror
... to make it not look like an OpenSSL function
2016-05-31 19:54:35 +02:00
Michael Kaufmann
7108e53fb5 openssl: Use correct buffer sizes for error messages
Closes #844
2016-05-31 19:52:45 +02:00
Renaud Lehoux
2072b4ae4f mbedtls: removed unused variables
Closes #838
2016-05-30 23:05:51 +02:00
Marcel Raad
c9b4e6e859 openssl: fix build with OPENSSL_NO_COMP
With OPENSSL_NO_COMP defined, there is no function
SSL_COMP_free_compression_methods

Closes #836
2016-05-30 15:31:14 +02:00
Daniel Stenberg
b15a17c702 mbedtls: fix includes so snprintf() works
Regression from the previous *printf() rearrangements, this file missed to
include the correct header to make sure snprintf() works universally.

Reported-by: Moti Avrahami
Bug: https://curl.haxx.se/mail/lib-2016-05/0196.html
2016-05-24 12:14:18 +02:00
Jay Satiro
3caaeffbe8 openssl: cleanup must free compression methods
- Free compression methods if OpenSSL 1.0.2 to avoid a memory leak.

Bug: https://github.com/curl/curl/issues/817
Reported-by: jveazey@users.noreply.github.com
2016-05-20 16:44:01 -04:00
Daniel Stenberg
48114a8634 openssl: ERR_remove_thread_state() is deprecated in latest 1.1.0
See OpenSSL commit 21e001747d4a
2016-05-19 11:39:59 +02:00
Marcel Raad
125827e60e schannel: fix compile break with MSVC XP toolset
For the Windows XP toolset of Visual C++ 2013/2015, the old Windows SDK
7.1 is used. In this case, _USING_V110_SDK71_ is defined.

Closes #812
2016-05-18 12:52:41 +02:00
Daniel Stenberg
6efd2fa529 mbedtls/polarssl: set "hostname" unconditionally
...as otherwise the TLS libs will skip the CN/SAN check and just allow
connection to any server. curl previously skipped this function when SNI
wasn't used or when connecting to an IP address specified host.

CVE-2016-3739

Bug: https://curl.haxx.se/docs/adv_20160518A.html
Reported-by: Moti Avrahami
2016-05-17 14:48:17 +02:00
Daniel Stenberg
675c30abc2 openssl: get_cert_chain: fix NULL dereference
CID 1361815: Explicit null dereferenced (FORWARD_NULL)
2016-05-17 09:34:33 +02:00
Daniel Stenberg
8132fe11b3 openssl: get_cert_chain: avoid NULL dereference
CID 1361811: Explicit null dereferenced (FORWARD_NULL)
2016-05-17 09:14:06 +02:00
Viktor Szakats
bf418d2df0 darwinssl.c: fix OS X codename typo in comment 2016-05-13 09:59:17 +02:00
Per Malmberg
3cf339901e darwinssl: fix certificate verification disable on OS X 10.8
The new way of disabling certificate verification doesn't work on
Mountain Lion (OS X 10.8) so we need to use the old way in that version
too. I've tested this solution on versions 10.7.5, 10.8, 10.9, 10.10.2
and 10.11.

Closes #802
2016-05-12 21:37:38 +02:00
Kamil Dudka
ea06ad2e31 openssl: fix compile-time warning in Curl_ossl_check_cxn()
... introduced in curl-7_48_0-293-g2968c83:

Error: COMPILER_WARNING:
lib/vtls/openssl.c: scope_hint: In function ‘Curl_ossl_check_cxn’
lib/vtls/openssl.c:767:15: warning: conversion to ‘int’ from ‘ssize_t’
may alter its value [-Wconversion]
2016-05-12 08:36:21 +02:00
Jay Satiro
2968c83967 openssl: stricter connection check function
- In the case of recv error, limit returning 'connection still in place'
to EINPROGRESS, EAGAIN and EWOULDBLOCK.

This is an improvement on the parent commit which changed the openssl
connection check to use recv MSG_PEEK instead of SSL_peek.

Ref: https://github.com/curl/curl/commit/856baf5#comments
2016-05-11 21:21:15 -04:00
Anders Bakken
856baf5a46 TLS: SSL_peek is not a const operation
Calling SSL_peek can cause bytes to be read from the raw socket which in
turn can upset the select machinery that determines whether there's data
available on the socket.

Since Curl_ossl_check_cxn only tries to determine whether the socket is
alive and doesn't actually need to see the bytes SSL_peek seems like
the wrong function to call.

We're able to occasionally reproduce a connect timeout due to this
bug. What happens is that Curl doesn't know to call SSL_connect again
after the peek happens since data is buffered in the SSL buffer and thus
select won't fire for this socket.

Closes #795
2016-05-11 00:06:40 +02:00
Daniel Stenberg
f6767f5435 TLS: move the ALPN/NPN enable bits to the connection
Only protocols that actually have a protocol registered for ALPN and NPN
should try to get that negotiated in the TLS handshake. That is only
HTTPS (well, http/1.1 and http/2) right now. Previously ALPN and NPN
would wrongly be used in all handshakes if libcurl was built with it
enabled.

Reported-by: Jay Satiro

Fixes #789
2016-05-09 15:30:25 -04:00
Daniel Stenberg
283babfaf8 tls: make setting pinnedkey option fail if not supported
to make it obvious to users trying to use the feature with TLS backends
not supporting it.

Discussed in #781
Reported-by: Travis Burtrum
2016-05-01 17:05:38 +02:00
Jay Satiro
9f498de9a2 mbedtls: Fix session resume
This also fixes PolarSSL session resume.

Prior to this change the TLS session information wasn't properly
saved and restored for PolarSSL and mbedTLS.

Bug: https://curl.haxx.se/mail/lib-2016-01/0070.html
Reported-by: Thomas Glanzmann

Bug: https://curl.haxx.se/mail/lib-2016-04/0095.html
Reported-by: Moti Avrahami
2016-04-28 02:57:38 -04:00
Daniel Stenberg
ab691309ce openssl: avoid BN_print a NULL bignum
OpenSSL 1.1.0-pre seems to return NULL(?) for a whole lot of those
numbers so make sure the function handles this.

Reported-by: Linus Nordberg
2016-04-26 23:55:31 +02:00
Daniel Stenberg
05492fe790 mbedtls.c: silly spellfix of a comment 2016-04-26 00:28:40 +02:00
33623d7196 PolarSSL: Implement public key pinning 2016-04-24 12:27:39 +02:00
Daniel Stenberg
cfe16c22d7 openssl: builds with OpenSSL 1.1.0-pre5
The RSA, DSA and DH structs are now opaque and require use of new APIs

Fixes #763
2016-04-21 10:52:24 +02:00
Daniel Stenberg
cefe9f4f90 vtls.h: remove a space before semicolon
... that the new checksrc detected
2016-04-19 08:53:31 +02:00
Daniel Stenberg
b2d246cf6c darwinssl: removed commented out code 2016-04-19 08:53:31 +02:00
Michael Kaufmann
cd8d236245 news: CURLOPT_CONNECT_TO and --connect-to
Makes curl connect to the given host+port instead of the host+port found
in the URL.
2016-04-17 23:50:59 +02:00
Damien Vielpeau
83b39a4f4d mbedtls: fix MBEDTLS_DEBUG builds 2016-04-07 16:12:50 +02:00
Daniel Stenberg
c111178bd4 mbedtls: implement and provide *_data_pending()
... as otherwise we might get stuck thinking there's no more data to
handle.

Reported-by: Damien Vielpeau

Fixes #737
2016-04-07 16:10:10 +02:00