Commit Graph

327 Commits

Author SHA1 Message Date
Rasmus Lerdorf
8d5f83dde5 Another openssl test that is dependent on the openssl version. The output has
changed in more recent versions. Synch with newer output and consider changing
the test to only pick out the more stable fields instead of all of them.
2012-02-05 10:08:16 +00:00
Rasmus Lerdorf
e4fb44c8b6 Need EXPECTF here, of course 2012-02-05 09:52:41 +00:00
Rasmus Lerdorf
f3a7ba75cf Getting different hashes here. But this test isn't testing the hashes,
it is just making sure we actually get a hash and don't crash.
2012-02-05 09:50:14 +00:00
Rasmus Lerdorf
a06e8ca56d According to the reports on qa this test is failing the same way for everyone.
See: http://qa.php.net/reports/viewreports.php?version=5.3.10&test=%2Fext%2Fopenssl%2Ftests%2Fbug28382.phpt
I'm not sure if this is due to a change in the openssl library or in the extension, so perhaps the test
itself needs to change, but for now synch it with the new output and watch for failures.
2012-02-05 09:32:20 +00:00
Scott MacVicar
398c6e6d11 MFH r322485
Fix possible attack in SSL sockets with SSL 3.0 / TLS 1.0.
CVE-2011-3389
2012-01-26 05:15:57 +00:00
Felipe Pena
4e19825281 - Year++ 2012-01-01 13:15:04 +00:00
Stanislav Malyshev
2f15c1d717 fix uninitialized var that may lead to crash 2011-12-26 02:16:37 +00:00
Scott MacVicar
c36926ba4c Add tests for OpenSSL crash. 2011-12-19 03:09:05 +00:00
Scott MacVicar
beda5efd41 Fix segfault in older versions of OpenSSL (before 0.9.8i) 2011-12-18 05:14:32 +00:00
Scott MacVicar
287e1917cf Make sure that we set the strong crypto result to false as well as returning false. 2011-12-07 20:50:33 +00:00
Mateusz Kocielski
a9482367f8 - Fixed NULL pointer dereference in stream_socket_enable_crypto, case when
ssl_handle of session_stream is not initialized.
2011-11-12 10:36:55 +00:00
Pierre Joye
2f3adeb083 - Revert r313616 (When we have a blocking SSL socket, respect the timeout
option, scottmac)

# This caused bug #55283 and #55848, we should investigate a proper solution without
# breaking anything.
2011-10-05 05:20:51 +00:00
Gustavo André dos Santos Lopes
4de9123b85 - ext/openssl/tests/bug36732.phpt more portable. 2011-09-14 10:55:46 +00:00
Gustavo André dos Santos Lopes
428ef23067 - Fixed test ext/openssl/tests/004.phpt.
- Made ext/openssl/tests/bug55646.phpt more reproducible by giving it a custom
  openssl.cnf.
2011-09-14 09:56:59 +00:00
Gustavo André dos Santos Lopes
b6aabaae59 - Test for bug #55646. 2011-09-12 20:05:07 +00:00
Gustavo André dos Santos Lopes
1d5028be3d - Fixed bug #55646: textual data is returned in UTF-8, but is input in
another encoding. 5.4 only as this implies a BC break.
2011-09-12 17:23:10 +00:00
Hannes Magnusson
7e986b2a46 Fix the path to the .pem 2011-09-08 09:27:47 +00:00
Felipe Pena
4b30846b50 - Make usage of new PHP_FE_END macro 2011-07-25 11:35:02 +00:00
Scott MacVicar
ebbb2b1df1 When we have a blocking SSL socket, respect the timeout option.
reading from SSL sockets could block indefinitely due to the lack
of timeout
2011-07-23 01:29:44 +00:00
Ryan Biesemeyer
530311f4ba removing openssl test for bug #55169 per Scott MacVicar's request; duplicate coverage of opensssl_random_pseudo_bytes.phpt 2011-07-20 21:25:39 +00:00
Ryan Biesemeyer
386abb9f5a update test for bug #55169, fix skipif 2011-07-20 21:13:03 +00:00
Ryan Biesemeyer
08cab10dc9 update test for bug #55169, don't require the extension in INI 2011-07-20 20:57:17 +00:00
Ryan Biesemeyer
cd6f4bcf8d tests for bug #55169 (mcrypt and openssl) 2011-07-20 18:59:05 +00:00
Pierre Joye
b03817afb5 - re apply the rng change specific to windows, long term it should be a std function but as this function was badly introduced in the 1st place, we have to fix the bad things here instead, pls do not revert again, bad idea. 2011-07-19 23:01:41 +00:00
Scott MacVicar
591e0e446f Revert change to use a special Windows version of openssl_random_pseudo_bytes().
Lets discuss this on internals first. We're advertising something from the OpenSSL library 
and then subverting it with another Windows OS call.

What are the implications of this? Should we make this available in ext/standard/ instead?
2011-07-19 22:29:55 +00:00
Scott MacVicar
64a0c79a07 openssl_encrypt() / openssl_decrypt() were flawed and truncated the key to the default size for the case of a variable key length cipher.
The result is a key of 448 bits being passed to the blowfish algorithm would be truncated to 128 bit.

Also fixed an error in the zend_parse_parameters() having an invalid character being used.
2011-07-19 22:15:56 +00:00
Pierre Joye
5fb2570742 - use php_win32_get_random_bytes instead of over slow and partially wrong openssl's version 2011-07-10 14:59:33 +00:00
Pierre Joye
104cd28e95 - did I not kill that already? (do not use rand_screen, pointless on server and not TS) 2011-06-16 13:06:41 +00:00
Felipe Pena
15f5dd5cb3 - Fixed bug #55028 (// is abad comment) 2011-06-10 22:48:36 +00:00
Gustavo André dos Santos Lopes
c27079d9e0 - Fixed bug #54992: Stream not closed and error not returned when SSL CN_match
fails.
2011-06-08 00:23:02 +00:00
Felipe Pena
32b5f8a1a3 - Added new parameter parsing option (p - for valid path (string without null byte in the middle))
# The tests will be fixed in the next commits
2011-06-06 21:28:16 +00:00
Scott MacVicar
30f19bd524 Allow management of your own padding in openssl_encrypt/decrypt.
For using mcrypt / openssl interchangeabley managing your own padding is the only solution.
2011-05-20 18:56:13 +00:00
Martin Jansen
0c8438462c The project calls itself OpenSSL and not openSSL, so let's keep it
that way in our code as well.
2011-04-25 16:50:30 +00:00
Rasmus Lerdorf
380c3e5127 SSLV2 patch cleanup 2011-04-24 23:27:48 +00:00
Rasmus Lerdorf
f1806e67e6 Support for openssl without SSLv2 supprot compiled in. Distros are starting to
remove support now and this wasn't compiling anymore on my Debian dev box.
2011-04-24 20:47:22 +00:00
Raphael Geissert
a286fa3523 Add the libraries we actually test and need for LDAP and OpenSSL, fixes bug #53339
Patch by Clint Byrum
2011-03-18 18:47:09 +00:00
Pierre Joye
32c4a3a061 - fix bug #54061, memory leak in openssl_decrypt 2011-02-21 12:50:26 +00:00
Pierre Joye
f922e31129 - fix bug #54061, memory leak in openssl_decrypt 2011-02-21 12:47:38 +00:00
Pierre Joye
b526a6ef44 - fix bug #54060, memory leak in openssl_encrypt 2011-02-21 10:07:31 +00:00
Felipe Pena
0203cc3d44 - Year++ 2011-01-01 02:17:06 +00:00
Gustavo André dos Santos Lopes
063393f29b - Fixed bug #53592 (stream_socket_enable_crypto() busy-waits in client mode).
- Fixed stream_socket_enable_crypto() not honoring the socket timeout in
  server mode.
2010-12-23 01:44:54 +00:00
Antony Dovgal
3722811395 make sure the stream context is present before looking for any options
and fix segfault
2010-12-04 21:54:20 +00:00
Adam Harvey
18ec6dae2c Implemented FR #53447 (Cannot disable SessionTicket extension for servers that
do not support it).

I haven't written a test due to the need for such a test to have a HTTPS server
available which mishandles SessionTicket requests; it's likely that server
administrators will gradually fix this either intentionally or through OpenSSL
upgrades. That said, if there's a great clamoring for a test, I'll work one up.
2010-12-03 09:34:35 +00:00
Pierre Joye
3fd615cc8d - not TS and useless on server, also not required anymore with the supported windows versions 2010-12-02 11:37:43 +00:00
Felipe Pena
6e8c2ba690 - Fixed bug #53136 (Invalid read on openssl_csr_new()) 2010-11-12 23:34:03 +00:00
Adam Harvey
86944b47a6 Fix vim marker folds. 2010-10-13 09:23:39 +00:00
Ilia Alshanetsky
412d151681 Fixed extrenous warning inside openssl_encrypt() for cases where iv not provided, but algo does not require an iv 2010-10-07 12:32:00 +00:00
Felipe Pena
2d8a4ea299 - Fixed bug #52947 (segfault when ssl stream option capture_peer_cert_chain used) 2010-09-29 01:25:35 +00:00
Pierre Joye
aa0ed267a2 - use TSRMLS_*C instead of TSRMLS_FETCH in zend_list_insert 2010-09-16 09:13:19 +00:00
Felipe Pena
80926568f1 - Fixed bug #52183 (Reflectionfunction reports invalid number of arguments for function aliases) 2010-06-26 16:03:39 +00:00