aboutsummaryrefslogtreecommitdiffstats
path: root/ext/openssl
Commit message (Collapse)AuthorAgeFilesLines
* Fix RDoc markuptopic/fix-rdoc-markupKazuki Yamaguchi2017-02-2422-381/+390
| | | | | Ruby core uses _str_ for emphasizing argument names and +str+ for codes. Match with the rule for better rendering.
* Revert r57690 except for read_nonblocknobu2017-02-241-4/+2
| | | | | | https://github.com/ruby/ruby/pull/1527#issuecomment-281867551 git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@57694 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* [DOC] mark up literalsnobu2017-02-241-4/+6
| | | | git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@57693 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* [DOC] keyword argument _exception_nobu2017-02-241-2/+2
| | | | git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@57692 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* [DOC] {read,write}_nonblock with exception: falsenobu2017-02-241-2/+2
| | | | | | | | | | Update docs to reflect EOF behavior change of read_nonblock and write_nonblock when using `exception: false`. [Fix GH-1527] Author: Russell Davis <russell-stripe@users.noreply.github.com> git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@57690 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* Merge branch 'topic/ssl-fixup-ex_data-handling'Kazuki Yamaguchi2017-02-191-27/+26
|\ | | | | | | | | | | * topic/ssl-fixup-ex_data-handling: ssl: assume SSL/SSL_CTX always have a valid reference to the Ruby object ssl: do not confuse different ex_data index registries
| * ssl: assume SSL/SSL_CTX always have a valid reference to the Ruby objecttopic/ssl-fixup-ex_data-handlingKazuki Yamaguchi2017-02-191-20/+5
| | | | | | | | It is impossible that they don't.
| * ssl: do not confuse different ex_data index registriesKazuki Yamaguchi2017-02-191-8/+22
| | | | | | | | | | | | | | Register necessary indices for these two independently. Similar to r55074 (76e933a5a517, "openssl: register ex_data index for X509_STORE{_CTX,} respectively", 2016-05-19), we can't blindly use the same ex_data index number for SSL and SSL_CTX.
* | ssl: remove OpenSSL::ExtConfig::TLS_DH_anon_WITH_AES_256_GCM_SHA384topic/ssl-remove-TLS_DH_anon_WITH_AES_256_GCM_SHA384Kazuki Yamaguchi2017-02-191-6/+0
|/ | | | | | | | | | | | | | The constant was initially introduced just to skip test cases that do not work with old versions without AES-GCM cipher suites support (< 1.0.1). However, the value of the constant has been always `false' because the macro TLS_DH_anon_WITH_AES_256_GCM_SHA384 does not exist in any version of OpenSSL. We no longer use it as of commit c9d1659f4027 ("test/utils: remove use_anon_cipher option from SSLTestCase#start_server", 2016-09-06). Let's just remove the constant.
* Merge branch 'maint'Kazuki Yamaguchi2017-01-311-1/+1
|\ | | | | | | | | | | | | | | | | | | * maint: Ruby/OpenSSL 2.0.3 .travis.yml: test with Ruby 2.4 ruby-openssl-docker: update versions of Ruby and OpenSSL x509: fix OpenSSL::X509::Name#eql? test/envutil: fix assert_raise_with_message buffering: fix typo in doc
| * Ruby/OpenSSL 2.0.3v2.0.3Kazuki Yamaguchi2017-01-311-1/+1
| |
| * x509: fix OpenSSL::X509::Name#eql?Kazuki Yamaguchi2017-01-281-1/+1
| | | | | | | | | | | | | | | | Commit 34e7fe34ee32 ("Use rb_obj_class() instead of CLASS_OF()", 2016-09-08) incorrectly inverted the result. Fix it, and add a test case for this. Fixes: 34e7fe34ee32 ("Use rb_obj_class() instead of CLASS_OF()")
* | Merge branch 'topic/ssl-certificate-verify-error-desc'Kazuki Yamaguchi2017-01-262-8/+25
|\ \ | | | | | | | | | | | | | | | * topic/ssl-certificate-verify-error-desc: ssl: show reason of 'certificate verify error' in exception message Make exceptions with the same format regardless of OpenSSL.debug
| * | ssl: show reason of 'certificate verify error' in exception messagetopic/ssl-certificate-verify-error-descKazuki Yamaguchi2017-01-241-0/+20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The 'certificate verify error' is one of the most common errors that can be raised by OpenSSL::SSL::SSLSocket#connect. The certificate verification may fail due to many different issues such as misconfigured trusted certificate store or inaccurate system clock. Unfortunately, since the detail is not put to the queue and is only accessible through OpenSSL::SSL::SSLSocket#verify_result, it is sometimes hard to figure out the real reason. Let's include a human readable reason message in the exception message. Like this: require "socket" require "openssl" ctx = OpenSSL::SSL::SSLContext.new ctx.set_params(cert_store: OpenSSL::X509::Store.new) ssl = OpenSSL::SSL::SSLSocket.new(Socket.tcp("www.ruby-lang.org", 443), ctx) ssl.connect #=> -:7:in `connect': SSL_connect returned=1 errno=0 state=error: certificate verify failed (unable to get local issuer certificate) (OpenSSL::SSL::SSLError) from -:7:in `<main>'
| * | Make exceptions with the same format regardless of OpenSSL.debugKazuki Yamaguchi2017-01-241-8/+5
| | | | | | | | | | | | | | | As the current behavior is useless. If OpenSSL.debug is set to true, errors put to the error queue will be printed to stderr anyway.
* | | Merge branch 'topic/test-memory-leak'Kazuki Yamaguchi2017-01-171-9/+100
|\ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * topic/test-memory-leak: Enable OSSL_MDEBUG on CI builds Add OpenSSL.print_mem_leaks test: prepare test PKey instances on demand test: let OpenSSL::TestCase include OpenSSL::TestUtils Don't define main() when built with --enable-debug
| * | | Add OpenSSL.print_mem_leaksKazuki Yamaguchi2017-01-171-0/+102
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add a wrapper method for CRYPTO_mem_leaks_fp(stderr). Calling the method at the end of programs helps debugging memory leak bugs in Ruby/OpenSSL. This is defined only when --enable-debug option is given when building Ruby/OpenSSL, and the OpenSSL version is capable. The test suite recognizes 'OSSL_MDEBUG' environment variable. Set to '1' to enable the memory leak checker. This would prevent creating another memory leak problem at least on the success paths. Note that this may print some false-positives with OpenSSL <= 1.0.2. It was once introduced by f0754f0b2f33 ("test: add test/mdebug extension", 2016-08-06) as a separate native extension, but reverted by 4c1ca7669180 ("Remove test/mdebug", 2016-08-26) because it didn't work on Windows. Let's re-introduce as part of openssl.so.
| * | | Don't define main() when built with --enable-debugKazuki Yamaguchi2017-01-161-11/+0
| | | | | | | | | | | | | | | | It is unnecessary as we have a test suite that does the job.
* | | | Merge branch 'maint'Kazuki Yamaguchi2017-01-172-1/+7
|\ \ \ \ | | |_|/ | |/| | | | | | | | | | | | | | | | | | * maint: appveyor.yml: update OpenSSL version to 1.0.2j Fix build with static OpenSSL libraries on Windows Fix for ASN1::Constructive 'each' implementation
| * | | Merge branch 'topic/windows-static-linking-without-pkg-config' into maintKazuki Yamaguchi2017-01-051-0/+6
| |\ \ \ | | | | | | | | | | | | | | | | | | | | * topic/windows-static-linking-without-pkg-config: Fix build with static OpenSSL libraries on Windows
| | * | | Fix build with static OpenSSL libraries on Windowstopic/windows-static-linking-without-pkg-configKazuki Yamaguchi2016-12-301-0/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | OpenSSL <= 1.0.2 requires gdi32 for RAND_screen(). OpenSSL >= 1.1.0 no longer has RAND_screen() but it now requires crypt32. If pkg-config is usable, they are automatically linked, but if it is not, configuring Ruby/OpenSSL fails. Fixes: https://bugs.ruby-lang.org/issues/13080
| * | | | Fix for ASN1::Constructive 'each' implementationCorey Bonnell2016-12-291-1/+1
| |/ / /
* | | | Merge branch 'topic/bn-updates'Kazuki Yamaguchi2016-12-261-0/+50
|\ \ \ \ | |_|/ / |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Tests for OpenSSL::BN are re-written. OpenSSL::BN now implements unary+ operator, unary- operator and negative? method. * topic/bn-updates: bn: implement OpenSSL::BN#negative? bn: implement unary {plus,minus} operators for OpenSSL::BN bn: refine tests
| * | | bn: implement OpenSSL::BN#negative?topic/bn-updatesKazuki Yamaguchi2016-12-231-0/+16
| | | | | | | | | | | | | | | | | | | | Numeric class implemented #negative? and #positive? in Ruby 2.3. Let's follow that.
| * | | bn: implement unary {plus,minus} operators for OpenSSL::BNKazuki Yamaguchi2016-12-231-0/+34
| | |/ | |/| | | | | | | | | | For consistency with Numeric. Not sure why they aren't currently; maybe they were simply forgotten.
* / | Remove support for OpenSSL 0.9.8 and 1.0.0topic/drop-openssl-098-and-100Kazuki Yamaguchi2016-12-2215-287/+32
|/ / | | | | | | | | | | | | | | | | | | They are no longer receiving security updates from the OpenSSL development team since 2015-12. We have kept basic compatibility until now because RHEL 5 still uses an (heavily modified) OpenSSL 0.9.8e. The RHEL 5 will reach EOL on 2017-03, thus it is now safe to assume nobody is still using such old versions of OpenSSL.
* / Start preparing for 2.1.0Kazuki Yamaguchi2016-12-211-1/+1
|/
* Ruby/OpenSSL 2.0.2v2.0.2Kazuki Yamaguchi2016-12-211-1/+1
|
* pkey: allow instantiating OpenSSL::PKey::PKey with unsupported key typeKazuki Yamaguchi2016-12-211-5/+10
| | | | | | | | | | Fix 'unsupported key type' error if OpenSSL::SSL::SSLSocket#tmp_key is called when X25519 is used for key exchange. EVP_PKEY may have a key type that we don't have have a dedicated subclass. Let's allow instantiating OpenSSL::PKey::PKey with such an EVP_PKEY, although the resulting instance is not so useful because it can't be exported at the moment.
* ssl: use SSL_SESSION_get_protocol_version()Kazuki Yamaguchi2016-12-213-5/+9
| | | | | | | | Restore the old behavior of OpenSSL::SSL::Session#==. SSL_SESSION_get_protocol_version() was missing in OpenSSL master at the time r55287 (cad3226a06a1, "openssl: adapt to OpenSSL 1.1.0 opaque structs", 2016-06-05).
* Rename functions in openssl_missing.cKazuki Yamaguchi2016-12-212-20/+29
| | | | | To avoid symbol conflict that would occur if two versions of OpenSSL are loaded at the same time.
* ssl: check for SSL_CTX_clear_options()Kazuki Yamaguchi2016-12-182-0/+5
| | | | | | | SSL_CTX_clear_options() first appeared in OpenSSL 0.9.8m. Add alternative macro definition for ancient versions of OpenSSL. http://blade.nagaokaut.ac.jp/cgi-bin/scat.rb/ruby/ruby-core/78693
* Ruby/OpenSSL 2.0.1v2.0.1Kazuki Yamaguchi2016-12-101-1/+1
|
* Merge changes from Ruby trunk r56927..r56953Kazuki Yamaguchi2016-12-101-1/+1
|\ | | | | | | | | | | | | | | | | * ruby-trunk r56927..r56953: (3 commits) (r56953) openssl: import fixes from upstream (r56948) ossl.c: cast (r56946) openssl: import v2.0.0 Sync-with-trunk: r56953
| * openssl: import fixes from upstreamrhe2016-12-102-0/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Import the following two commits from upstream: commit 72126d6c8b88abd69c3565fc3bbbd5ed1e401611 Author: Kazuki Yamaguchi <k@rhe.jp> Date: Thu Dec 1 22:27:03 2016 +0900 pkey: check existence of EVP_PKEY_get0() EVP_PKEY_get0() did not exist in early OpenSSL 0.9.8 series. So define ourselves if needed. commit 94a1c4e0c5705ad1e9a4ca08cacaa6cba8b1e6f5 Author: Kazuki Yamaguchi <k@rhe.jp> Date: Thu Dec 1 22:13:22 2016 +0900 test/test_cipher: fix test with OpenSSL 1.0.1 before 1.0.1d Set the authentication tag before the AAD when decrypting. Before OpenSSL commit 96f7fafa2431 ("Don't require tag before ciphertext in AESGCM mode", 2012-10-16, at OpenSSL_1_0_1-stable branch, included in OpenSSL 1.0.1d), the authentication tag must be set before any calls of EVP_CipherUpdate(). They should fix build on CentOS 5 and Ubuntu 12.04 respectively. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56953 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
| * ossl.c: castnobu2016-12-101-1/+1
| | | | | | | | | | | | | | * ext/openssl/ossl.c (ossl_pem_passwd_cb): cast to int. it's safe because len does not exceed int max_len. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56948 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* | Merge pull request #88 from yogo1212/generalizedtime_formatKazuki Yamaguchi2016-12-101-3/+9
|\ \ | | | | | | asn1: handle GENERALIZEDTIME without seconds
| * | asn1: more output on error 'bad GENERALIZEDTIME'Leon M. George2016-12-091-1/+2
| | |
| * | asn1: handle GENERALIZEDTIME without secondsLeon M. George2016-12-091-2/+7
| |/
* | Merge pull request #89 from koic/fix_typo_in_ossl_engineKazuki Yamaguchi2016-12-091-1/+1
|\ \ | | | | | | Fix a typo in ossl_engine.c
| * | Fix a typo in ossl_engine.cKoichi ITO2016-12-091-1/+1
| |/
* | ssl: prevent encoded NPN advertised protocol list from being GCedKazuki Yamaguchi2016-12-051-1/+4
| | | | | | | | | | | | | | | | | | SSLContext#setup encodes the protocol list set in @npn_protocols into a String. The String is passed to SSL_CTX_set_next_protos_advertised_cb() and OpenSSL invokes the callback function with the String. However since Ruby's GC can't find the reference to the String from the inside of OpenSSL, it can be free'd before the callback is invoked. So store the String in an instance variable to prevent this.
* | bn: keep reference to temporary OpenSSL::BN object created by GetBNPtr()Kazuki Yamaguchi2016-12-043-19/+31
| | | | | | | | | | | | | | | | | | GetBNPtr() accepts both OpenSSL::BN and Ruby integers. In the latter case, it creates a temporary OpenSSL::BN internally. The OpenSSL::BN object immediately disappears from the stack and is not protected from GC. Fixes: https://github.com/ruby/openssl/issues/87
* | pkey: check existence of EVP_PKEY_get0()Kazuki Yamaguchi2016-12-012-0/+5
|/ | | | | EVP_PKEY_get0() did not exist in early OpenSSL 0.9.8 series. So define ourselves if needed.
* ssl: follow-up for the workaround for OpenSSL 1.1.0c's SSL_read()Kazuki Yamaguchi2016-11-301-7/+7
| | | | | | | | Remove the comment added by commit 072d53ecf984 ("ssl: workaround for new behavior of SSL_read() in OpenSSL >= 1.1.0c"). The breaking change in OpenSSL 1.1.0c has been reverted in the 1.1.0 branch. However, for the sake of safety, ensure that we never call rb_sys_fail() with errno == 0. So there is no change in the actual code.
* ssl: make OpenSSL::SSL::SSLContext#freeze an alias of #setuptopic/ssl-make-sslctx-freeze-alias-of-setupKazuki Yamaguchi2016-11-291-0/+1
| | | | | | | | | | | | | | | | | | | | SSLSocket#setup uses the frozen state as "SSL_CTX is already set up". If an user manually freeze the context, it misunderstands as if #setup is already called, leading to unexpected behaviors because parameters the user set won't be actually set to the underlying SSL_CTX and thus ignored. Ideally, #setup should go and be replaced with setters. But we don't do this now because it is not that simple: some of them would produce new ordering issues, e.g. 'ca_file' property which loads a file into SSL_CTX::cert_store and 'cert_store' which replaces SSL_CTX::cert_store would conflict. Fixing this properly would require deprecating 'ca_file' first. So, let's take the second best way: make it "just work" instead of break silently. Fixes: https://github.com/ruby/openssl/issues/85
* ssl: fix possible exception from non-protected codeKazuki Yamaguchi2016-11-281-42/+66
| | | | | | | | | | | rb_ary_new_from_args() is called from non-protected callback function which will be directly called from OpenSSL. It may raise NoMemoryError and may corrupt the internal state of SSL object. So, avoid creating Array here and pass raw values to the protected function instead. The same change has been applied to ALPN/NPN selection callbacks in 3a926047a729 ("ssl: catch exceptions raised in ALPN/NPN callbacks", 2016-08-30).
* ssl: inline ossl_ssl_shutdown() into ossl_ssl_stop()Kazuki Yamaguchi2016-11-281-20/+15
| | | | | | | | We call SSL_shutdown() four times at most meaninglessly. Since the underlying socket is in non-blocking mode, if the first call failed because the underlying socket is not write/readable, the subsequent calls would just fail with the same error. Just call once, and give up if it fails.
* engine: undef alloc functionKazuki Yamaguchi2016-11-161-18/+1
| | | | | | This prevents users from allocating OpenSSL::Engine instance using OpenSSL::Engine.allocate. Undef'ing alloc function also allows us to remove explicit undef of OpenSSL::Engine.new and #initialize_copy.
* asn1: fix possible segfault in OpenSSL::ASN1::Constructive#eachKazuki Yamaguchi2016-11-161-2/+6
| | | | | Don't blindy assume that the value which can be modified from Ruby code is always an Array, and just call its #each method.
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-06 04:40:27 +0000