kc3-lang/libevent

Branch :

Log

Author Commit Date CI Message
9a9b92ed 2020-03-01 16:01:12 Add EVENT_BASE_FLAG_EPOLL_DISALLOW_TIMERFD flag (fixes: #958) By default we are using CLOCK_MONOTONIC_COARSE, but if EVENT_BASE_FLAG_PRECISE_TIMER isset, then CLOCK_MONOTONIC will be used, however this will also enable timerfd, while this is not always what someone wants, hence add a flag to control this (by default the old behavior is preserved, set new flag to change it).
4e5a41ca 2020-03-01 15:47:40 test-time: do not use deprecated API - event_init() -> event_base_new() - event_set() -> event_new() - check return value of event_base_dispatch() - use EXIT_SUCCESS/EXIT_FAILURE
a11edbfa 2020-03-01 14:54:36 test-time: enable debug mode if EVENT_DEBUG_LOGGING_ALL env set
5a5fe67e 2020-03-01 13:52:32 Fix typo in thread.h (s/event/evthread)
114b3836 2020-02-29 17:47:47 increase segment refcnt only if evbuffer_add_file_segment() succeeds
4da9f87c 2020-02-27 16:59:45 evdns: fix a crash when evdns_base with waiting requests is freed Fix undefined behaviour and application crash that might take place in some rare cases after calling evdns_base_free when there are requests in the waiting queue. Current cleanup procedure in evdns_base_free_and_unlock function includes 2 steps: 1. Finish all inflight requests. 2. Finish all waiting requests. During the first step we iterate over each list in req_heads structure and finish all requests in these lists. With current logic finishing an inflight request (function request_finished) removes it from the inflight requests container and forces a wating connection to be sent (by calling evdns_requests_pump_waiting_queue). When these new requests are sent it is possible that they will be inserted to the list in req_heads that we've already cleaned. So in some cases container of the inflight requests is not empty after this procedure and some requests are not finished and deleted. When timeouts for these requests expire evdns_request_timeout_callback is called but corresponding evdns_base has been already deleted which causes undefined behaviour and possible applicaton crash. It is interesting to note that in old versions of libevent such situation was not possible. This bug was introduced by the commit 14f84bbdc77d90b1d936076661443cdbf516c593. Before this commit nameservers were deleted before finishing the requests. Therefore it was not possible that requests from the waiting queue be sent while we finish the inflight requests.
968bbd5c 2020-02-04 16:05:02 event_base_once: fix potential null pointer threat supposing if base is null, EVBASE_ACQUIRE_LOCK and EVBASE_RELEASE_LOCK would get a coredump, so we add a guard for protection. Signed-off-by: chenguolong <cgl.chenguolong@huawei.com>
5fbe6313 2020-02-11 14:33:15 test-ratelim: add missing free
08e7d513 2020-01-28 14:31:39 misspelling of output in bufferevent_struct.h
0b6f29ac 2020-01-27 01:56:58 Merge branch 'github-actions-v2' (#951) * github-actions-v2: travis: disable doxygen and coveralls, in favor of github actions github actions: test and coverage github actions: doxygen Ignore truthy in yamllint (for github-actions) test: mark common_timeout as retriable cmake: set rpath for libraries on linux test-export: compatible with all versions of visual studio coverage: 'lcov --remove' need full path
19a55164 2020-01-27 01:53:27 travis: disable doxygen and coveralls, in favor of github actions
84bb2c18 2020-01-21 19:51:48 github actions: test and coverage
d151968f 2020-01-21 10:14:11 github actions: doxygen
cf379ecc 2020-01-26 21:30:15 Ignore truthy in yamllint (for github-actions)
ce8be238 2020-01-20 21:17:27 cmake: set rpath for libraries on linux
9adc9f14 2020-01-20 21:15:26 test-export: compatible with all versions of visual studio
1a99f9b2 2020-01-14 15:54:14 coverage: 'lcov --remove' need full path
f6d7992b 2020-01-23 21:34:18 test: mark common_timeout as retriable Refs: https://github.com/libevent/libevent/pull/951#issuecomment-576711224
2c473056 2020-01-21 20:10:15 Do not use shared global structures on CYGWIN Fixes: #950
ec775a96 2020-01-21 22:02:34 Add vcpkg installation instructions
ca2b72c5 2020-01-14 21:45:01 test: move thread into realtime class even on EVENT__DISABLE_THREAD_SUPPORT
d0adbc05 2020-01-14 10:20:12 test: fix compilation without thread support (EVENT__DISABLE_THREAD_SUPPORT=ON)
8dcccf35 2020-01-14 02:36:54 travis-ci: do not allow failures under osx
30fe1250 2020-01-14 02:14:16 test: fix bufferevent/bufferevent_connect_fail_eventcb* under osx/freebsd For OSX the socket should be closed, otherwise the "connection refused" will not be triggered. And freebsd can return error from the connect().
10504fca 2020-01-14 00:38:06 test: fix dst thread in move_pthread_to_realtime_scheduling_class (osx) Fixes the following tests on osx: - del_wait - no_events Refs: #940
34d51e1b 2020-01-14 00:27:21 test: fix compilation under win32 (rearrange thread_setup() code)
391003e9 2020-01-14 00:41:48 test: use THREAD_* wrappers over pthread* in del_notify
a6f81aa4 2020-01-13 23:36:40 Merge branch 'osx-clock' Moves the thread into real-time scheduling class, as recommended in [1], it fixes the separate test provided by @ygj6 [2] everywhere (github actions, travis-ci, appveyor) under osx. [1]: https://developer.apple.com/library/archive/technotes/tn2169/_index.html [2]: https://github.com/azat-archive/osx-timers/commit/dde1a6e4d04506d0e0fb193ebb9f49ae25873be6 Although even after this changes the following time-related tests failed on travis-ci: - no_events - del_wait But anyway I guess #940 can be closed, since this fixes the issue in common. * osx-clock: test: Use THREAD_* wrappers in del_notify/del_wait test: move threads created with THREAD_START() to realtime scheduling class too test: put thread into real time scheduling class on osx for better latencies Closes: #940
4b72024b 2020-01-13 22:24:54 test: Use THREAD_* wrappers in del_notify/del_wait
e6285eed 2020-01-13 22:24:54 test: move threads created with THREAD_START() to realtime scheduling class too
b1e46c32 2020-01-13 00:33:39 test: put thread into real time scheduling class on osx for better latencies
603d7937 2020-01-13 22:27:32 Merge branch 'ci-improvements' * ci-improvements: travis-ci: measure build/tests time travis-ci: drop travis_wait (does not work with sub processes) cmake: do not print used method (EVENT_SHOW_METHOD) while running tests cmake: run regress test quietly like autotools (makes CI logs cleaner)
b30717eb 2020-01-13 22:24:54 travis-ci: measure build/tests time
fbe6f7d9 2020-01-13 22:24:54 travis-ci: drop travis_wait (does not work with sub processes) Refs: https://github.com/travis-ci/travis-ci/issues/8526
f8a6f127 2020-01-13 22:24:54 cmake: do not print used method (EVENT_SHOW_METHOD) while running tests autotools don't, plus this will make CI logs cleaner
fa33819e 2020-01-13 22:24:54 cmake: run regress test quietly like autotools (makes CI logs cleaner)
02905413 2016-02-09 18:01:00 Add callback support for error pages The existing error pages are very basic and don't allow for multi-lingual support or for conformity with other pages in a web site. The aim of the callback functionality is to allow custom error pages to be supported for calls to evhttp_send_error() by both calling applications and Libevent itself. A backward-incompatible change has been made to the title of error pages sent by evhttp_send_error(). The original version of the function used the reason argument as part of the title. That might have unforeseen side-effects if it contains HTML tags. Therefore the title has been changed to always use the standard status text. An example of the error callback can be found in this [version](https://github.com/libevent/libevent/files/123607/http-server.zip) of the 'http-server' sample. It will output error pages with very bright backgrounds, the error code using a very large font size and the reason. Closes: #323 (cherr-picked from PR)
462f2e97 2020-01-12 15:34:51 Merge branch 'http-connect' After this patchset http-connect works with pproxy [1]: $ pproxy -l http://:8000/ -vvv & $ http-connect //127.1:8000 http://kernel.org:80/ [1]: https://pypi.org/project/pproxy/ * http-connect: http-connect: do not check connection on GET cb http-connect: set Host header (for CONNECT and GET) (like curl) http-connect: cleanup and helpers http: do not close connection for CONNECT http: do not assume body for CONNECT Fixes: #946
9c91fe14 2020-01-12 15:31:50 http-connect: do not check connection on GET cb
415ddee3 2020-01-12 15:24:35 http-connect: set Host header (for CONNECT and GET) (like curl)
e54d5fc5 2020-01-12 15:14:24 http-connect: cleanup and helpers
12ad0c8f 2020-01-12 15:29:48 http: do not close connection for CONNECT
998e6834 2020-01-12 13:43:18 http: do not assume body for CONNECT
f0e79baf 2020-01-08 20:37:16 Added uninstall target check to cmakelists
08981f8d 2020-01-05 19:02:22 Fix compilation without OPENSSL_API_COMPAT Use the following for openssl 1.1+: - X509_getm_notBefore over X509_get_notBefore - X509_getm_notAfter over X509_get_notAfter - use OPENSSL_VERSION_NUMBER over SSLeay() - add missing headers Refs: openssl/openssl@0b7347effee5
6412f34f 2019-12-27 17:53:28 evutil_time: Implements usleep() using wait funtion on Windows
1edb6f61 2019-12-29 20:52:17 Initialize variable to 0 replace use memset function in sample/hello-world.c
ea12428d 2019-12-10 09:25:39 appveyor: switch to Previous Visual Studio 2019 (VS issues after update) After update [1] of the Visual Studio 2019 image cmake unable to find path to the c compiler [2]: cmake : CMake Error at CMakeLists.txt:47 (project): At line:1 char:1 + cmake -G 'Visual Studio 16 2019' -A x64 .. + ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + CategoryInfo : NotSpecified: (CMake Error at ...t:47 (project)::String) [], RemoteException + FullyQualifiedErrorId : NativeCommandError The CMAKE_C_COMPILER: C:/Program Files (x86)/Microsoft Visual Studio/2019/Community/VC/Tools/MSVC/14.23.28105/bin/Hostx64/x64/cl.exe is not a full path to an existing compiler tool. [1]: https://github.com/appveyor/ci/issues/3231 https://www.appveyor.com/updates/2019/12/09/ [2]: https://ci.appveyor.com/project/libevent/libevent/builds/29431286/job/a68h7dn9rcride9g
f76456b0 2019-12-04 17:56:54 Add support for priority inheritance Add support for posix mutex priority inheritance. This is important to avoid priority inversion in systems running with threads with different priorities. Signed-off-by: Andre Azevedo <andre.azevedo@gmail.com>
8f13c170 2019-11-30 15:48:36 cmake: set CMAKE_{RUNTIME,LIBRARY,ARCHIVE}_OUTPUT_DIRECTORY they are not defined This will allow overriding them in parent cmake rules, i.e. if libevent is used via add_subdirectory(). Closes: #931
b9f442e2 2019-11-30 15:53:52 cmake: use CMAKE_LIBRARY_OUTPUT_DIRECTORY for the final shared library symlink Fixes: 669a53f3 ("cmake: set library names to be the same as with autotools")
aff23122 2019-11-17 23:22:53 appveyor: fix EVENT_CMAKE_OPTIONS expansion (for multiple arguments) Fixes: 63f73ee2 ("CI: integrate testing for components export")
8a348699 2019-11-17 18:13:51 test-ratelim: calculate timers bias (for slow CPUs) to avoid false-positive This can be/should be done for regression tests too. Refs: https://ci.appveyor.com/project/libevent/libevent/builds/28916689/job/kg621aa194a0qbym Refs: https://github.com/libevent/libevent/pull/917#issuecomment-553811834 v2: EVENT_BASE_FLAG_PRECISE_TIMER
026590f9 2019-11-17 15:41:22 mailmap: add name/email aliases for yuangongji (name and email) Before: $ git log --format='%aE %aN' --author=yuangongji | sort | uniq -c 30 82787816@qq.com yuangongji 1 82787816@qq.com yuangongji (A) After: $ git log --format='%aE %aN' --author=yuangongji | sort | uniq -c 34 yuangongji@foxmail.com yuangongji
8be8ac46 2019-11-17 02:45:54 Merge #929 -- cmake package improvements Example: find_package(Libevent 2.2.0 REQUIRED COMPONENTS core) add_executable(test test.c) target_link_libraries(test ${LIBEVENT_LIBRARIES}) * upstream/pr/929: CI: integrate testing for components export cmake: test for find_package() cmake: improve package config file
02dc7a1b 2019-10-22 22:29:12 appveyor: fix openssl version mismatch warning yuangongji: "Many warnings appear when building and running with Visual Studio 2019 in Appveyor: WARN C:\projects\libevent\test\regress_ssl.c:210: Version mismatch for openssl: compiled with 1000214f but running with 1000212f Simply add the openssl binary to the "PATH" environment variable to fix it. I wrote a simple demo to reproduce it: https://github.com/ygj6/verify I see there are dozens of openssl libraries in the system of appveyor: https://ci.appveyor.com/project/ygj6/verify/builds/28290688 If you do not specify which openssl to use, the system cannot find the correct library."
28ba0a47 2019-11-17 02:37:59 appveyor: disable parallel tests execution Refs: https://github.com/libevent/libevent/pull/917#issuecomment-553784701
63f73ee2 2019-11-14 18:34:56 CI: integrate testing for components export
cd15112b 2019-11-14 18:28:31 cmake: test for find_package()
dc4be869 2019-11-07 18:26:47 cmake: improve package config file
1495f8b6 2019-11-06 21:41:38 Link with iphlpapi only on windows Fixes: 9fecb59a ("Parse IPv6 scope IDs.") Refs: #923
9fecb59a 2019-10-29 15:48:53 Parse IPv6 scope IDs.
4436287d 2019-10-31 09:18:58 Relax bufferevent_connect_hostname_emfile Do not do any assumptions on the error for the EMFILE from getaddrinfo(), expect just any error. Fixes: #924
53b3be4e 2019-10-26 22:31:18 autotools: fails build when need but can not find openssl
5fe83d74 2019-10-25 21:54:13 cmake: eliminate duplicate installation of public headers
1ba94bdf 2019-10-21 22:21:19 Merge pull request #915 from ygj6 -- evutil_socketpair win32 fixes * upstream/pr/915: appveyor: add vs2019 os to test some new features brought by Win10 test: add testcase for evutil_socketpair() evutil: make evutil_socketpair() have the same behavior on Windows with build number lower and higher than 17063
84affc18 2019-10-21 10:36:49 append to CMAKE_MODULE_PATH Don't override any -DCMAKE_MODULE_PATH= passed from CLI to eg. test custom Platform/ support.
38eb4846 2019-10-18 23:32:14 appveyor: add vs2019 os to test some new features brought by Win10
55d60c92 2019-10-18 21:11:37 test: add testcase for evutil_socketpair()
879d2496 2019-10-18 21:03:40 evutil: make evutil_socketpair() have the same behavior on Windows with build number lower and higher than 17063
b9b9f190 2019-10-16 01:11:51 Do not use sysctl.h on linux (it had been deprecated) It had been deprecated for a long time (AFAIK), but since glibc-2.29.9000-309-g744e829637 it produces a #warning
dda8968c 2019-10-12 18:45:52 evutil: implement socketpair with unix domain socket on Win10
11bdd41c 2019-10-14 23:22:55 travis-ci: add OPENSSL_1_1=yes for the openssl 1.1 to distinguish it in list
fb95772a 2019-10-14 23:20:23 travis-ci: use matrix.include over env.matrix/matrix.exclude for coveralls
f2bdd6ee 2019-10-14 22:32:01 travis-ci: build with clang only basic configurations I tried to exclude clang instead (for PRs), but matrix.exclude cannot have conditions (i..e matrix.exclude.if).
c1c8ae71 2019-10-14 12:20:45 Use matrix.include.if over matrix.exclude.if for doxygen And use if.repo over if.slug Fixes: aeb014cc ("Do not try to deploy documentaion for PR")
aeb014cc 2019-10-14 02:49:55 Do not try to deploy documentaion for PR Anyway access to security variables is forbidden for PRs. v2: use travis-ci conditions, to avoid running addons
b32a6443 2019-10-09 00:16:50 Fix all all yamllint warnings in travis/appveyor rules
c88c578a 2019-10-09 00:13:01 appveyor: do not allow any failures Yes we still have flacky tests, but anyway two main configuration can fail too (and they do fail from time to time) so let's not allow any failures and see how this will go. (Credits to @ygj6 via #910)
72d140c2 2019-10-08 22:01:01 appveyor: fix build script for compiling using mingw-w64
78b5bca6 2019-10-08 21:38:58 sample/https-client: link crypt32 explicitly when build with mingw-w64
c328999e 2019-10-07 22:10:14 Merge remote-tracking branch 'upstream/pr/908' * upstream/pr/908: Avoid transforming base C_FLAGS set deliberately
72e6eff0 2019-10-04 01:26:47 Fix compat with NetBSD >= 10 kevent::udata was switched from intptr_t to void*. Handle both cases with the GCC extension typeof().
af4b07a5 2019-10-03 11:40:52 Avoid transforming base C_FLAGS set deliberately The CMAKE_C_FLAGS_DEBUG, CMAKE_C_FLAGS_RELEASE, CMAKE_C_FLAGS_MINSIZEREL and CMAKE_C_FLAGS_RELWITHDEBINFO options are correctly and deliberately toggled to use the libcmt (/MT) flag options in place of the usual msvcrt (/MD) options, but this isn't necessarily desired by the user. The default choice can be overriden with the EVENT__MSVC_STATIC_RUNTIME cmake option. However, the /MD flag that is the choice of CMake only enters into play for the four types of builds above. If the user introduces another CMAKE_BUILD_TYPE, the base CMAKE_C_FLAGS must not be manipulated, as that value (and the CMAKE_C_FLAGS_{custom} value) have been explicitly chosen by the user/developer deploying this library, and the mismatch between these flags in different dependencies results in link errors. The CMake build schema itself doesn't place an /MD flag in CMAKE_BUILD_TYPE so any /M compile option in that variable needs to be retained. Signed-off-by: William Rowe <wrowe@pivotal.io> Signed-off-by: Yechiel Kalmenson <ykalmenson@pivotal.io>
8d5c5650 2019-09-26 21:47:51 tinytest: support timeout on Windows
a977d696 2019-09-22 18:44:55 Merge branch 'upstream/pr/899' (evbuffer_freeze testcase enhancements) * upstream/pr/899: improve the description of parameter to evbuffer_read() regress_buffer: improve testcase for evbuffer_freeze()
5aae74d7 2019-09-21 23:37:38 improve the description of parameter to evbuffer_read()
6769f692 2019-09-19 22:19:58 regress_buffer: improve testcase for evbuffer_freeze()
4727150a 2019-09-21 23:30:48 evbuffer_add_file: fix freeing of segment in the error path if evbuffer_add_file_segment() fails it returns -1, so we should call evbuffer_file_segment_free() only on error, and this -1 not 0. Fixes: 6a81b1f5 ("Avoid double-free on error in evbuffer_add_file. Found by coverity.") Backport-to: 2.1
148d12ad 2019-09-19 15:24:51 evutil_time: detect and use _gmtime64_s()/_gmtime64()
0cd536b0 2019-09-18 23:12:59 http: rename bind_socket_ai() to create_bind_socket_nonblock()
1c78451f 2019-09-16 23:24:32 http: make sure the other fields in ext_method are not changed by the callback
572a5651 2019-09-15 21:45:26 https-client: load certificates from the system cert store on Windows
095c8ae1 2019-09-08 22:30:54 Build doxygen documentation via cmake (to fill variables) Refs: #782
05467445 2019-09-08 11:47:04 Deploy documentation to libevent-doc.github.io Based-on: https://github.com/ygj6/libevent/commit/4a86dcb4df0f9dc69f7722a1ba7567ef3b8447c8.patch (by @ygj6)
c3f35345 2019-09-05 23:31:19 Fix checking return value of the evdns_base_resolv_conf_parse() Reported-by: Maxim Gorbachyov <maxim.gorbachyov@gmail.com>
194a5d82 2019-09-04 00:46:01 cmake: fix getrandom() detection Fixes: 86f55b04 ("arc4random: replace sysctl() with getrandom (on linux)")
86f55b04 2019-09-03 00:34:35 arc4random: replace sysctl() with getrandom (on linux) Since sysctl() is deprecated for a long-long time, according to sysctl(2): Since Linux 2.6.24, uses of this system call result in warnings in the kernel log. Fixes: #890 Suggested-by: Pierce Lopez
13b8fc39 2019-08-30 10:21:07 Upgrade autoconf (after upgrading minimum required to 2.67) - AC_PROG_SED - AC_USE_SYSTEM_EXTENSIONS - AC_TRY_COMPILE -> AC_COMPILE_IFELSE - AC_TRY_RUN -> AC_RUN_IFELSE ... Also use: - AC_CONFIG_AUX_DIR Refs: #870 Fixes: 3f09e923 ("Change the minimum version of automake to 1.13 and autoconf to 2.67")
ddacaef5 2019-08-29 22:57:44 Revert "Warn if forked from the event loop during event_reinit()" Thinking about this more and realizing that this was a mistake, so should be reverted. In a nut shell I guess most of the apps calls event_reinit() from the loop (see [1] for example), and this should be totally fine (the bit with the signals [2] handled in event_reinit() gracefully) [1]: https://archives.seul.org/libevent/users/Aug-2019/msg00009.html [2]: https://github.com/libevent/libevent/pull/833#issuecomment-501834453 This reverts commit 497ef904d544ac51de43934549dbeccce8e6e8f8. Reported-by: mikulas@twibright.com Backport-to: 2.1