aboutsummaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
...
* pthreads.h: avoid undefined behaviorTheo Buehler2024-03-031-3/+8
| | | | | | | You can't pass a function pointer through a void pointer. So wrap the pthread callback in a struct. Fixes #966
* zap stray whitespaceTheo Buehler2024-03-031-1/+1
|
* Update .github/workflows/release_by_tag.ymlIlya Shipitsin2024-03-031-1/+2
| | | Co-authored-by: Joshua Sing <joshua@hypera.dev>
* CI: automatically create release, upload windows buildsIlya Shipitsin2024-03-031-0/+66
|
* windows: minor compat header fixesViktor Szakats2024-03-033-2/+2
| | | | | | - posix_win.c: use `snprintf` as-is with _MSC_VER >= 1900 - stdio.h: include socket header before windows.h - pthread.h: delete exec permission from source file
* Add configure check for strcasecmpTheo Buehler2024-03-031-1/+2
|
* cmake: auto-detect `strtonum`Viktor Szakats2024-03-031-0/+5
| | | | | | | | | | | | | | Notice that just like in autotools, this detection also doesn't take into account the targeted OS version. Meaning it detects `strtonum` even if targeting e.g. macOS older than release v11 Big Sur (which introduced this funcitions), if the SDK declares it. Wrong detection will either cause a binary broken on older macOS and/or trigger compiler warnings. Ref: https://github.com/libressl/portable/issues/928#issuecomment-1850178282 Ref: https://github.com/libressl/portable/issues/928#issuecomment-1850276298 Prerequisite: https://github.com/libressl/portable/issues/928#issuecomment-1850356408
* delete broken linkViktor Szakats2024-03-031-2/+1
|
* upgrade links in comments to HTTPSViktor Szakats2024-03-033-4/+4
|
* check-release.sh: use HTTPS when downloading sourcesViktor Szakats2024-03-031-1/+1
|
* windows: always upload build artifacts, even in case of failureIlya Shipitsin2024-03-031-0/+1
|
* Update solaris vm to latest versionneil2024-03-031-2/+2
| | | Update solaris vm to latest version
* there has to be some trigger, even a manual oneBrent Cook2024-03-032-0/+6
|
* remove Solaris build statusBrent Cook2024-03-031-1/+0
|
* disable periodic builds of broken CI workflowsBrent Cook2024-03-032-10/+0
| | | | | | | | Neither of these has built in over a month. Solaris looks to be an upstream timeout issue (seems the hack it uses may not work anymore). Not sure if cifuzz is really needed either, since we're still being tested through clusterfuzz actively.
* remove additional endbr64 patch, add tests for macos-11Brent Cook2024-03-032-1/+21
|
* Replace rust-openssl patch with an ed scriptTheo Buehler2024-03-032-13/+9
| | | | | It is annoying to update this patch all the time. An easy ed script does the trick just as well.
* Fix rust-openssl regressTheo Buehler2024-03-031-2/+1
| | | | | Change around the order slightly. It seems more logical to apply the patch first, then export the env.
* ci: add --verbose flag to rust-openssl cargo testJoshua Sing2024-03-031-1/+1
|
* ci: tidy up rust-openssl libressl build stepJoshua Sing2024-03-031-5/+0
|
* ci: improve rust-openssl workflowJoshua Sing2024-03-032-33/+40
|
* readme: fix CI workflow badge imagesJoshua Sing2024-03-031-5/+3
| | | | | Update the workflow badge images to use the new workflow names introduced in: #942, #941, #934
* Land #953, Fix build on mipsBrent Cook2023-12-171-0/+2
|
* Land #941, improve Linux github workflowsBrent Cook2023-11-085-87/+62
|\
| * ci: shrink and improve Linux workflowsJoshua Sing2023-11-095-87/+62
| | | | | | | | | | Add a new `linux` workflow that replaces cross_test, linux_ci, linux_ci_asan, and linux_ci_asan_noasm.
* | Land #942, clean up macOS workflowBrent Cook2023-11-082-21/+31
|\ \
| * | ci: clean up macOS workflowJoshua Sing2023-11-082-21/+31
| |/
* | Land #940, avoid redefinition of OPENSSL_NO_ASMBrent Cook2023-11-082-2/+5
|\ \
| * | amd64/bn_arch.h: avoid redefinition of `OPENSSL_NO_ASM`Viktor Szakats2023-11-081-2/+4
| |/ | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes this warning when `OPENSSL_NO_ASM` is already set by the build system (seen with CMake): ``` In file included from ./libressl/crypto/bn/bn_mul.c:65: ./libressl/crypto/bn/arch/amd64/bn_arch.h:24:9: warning: 'OPENSSL_NO_ASM' macro redefined [-Wmacro-redefined] ^ <command line>:10:9: note: previous definition is here ^ ``` Reported in #910
| * Land #938, cmake: simplify if expressions, accept CPU values more consistentlyBrent Cook2023-11-071-19/+19
| |\
| * \ Land #937, remove unused C definesBrent Cook2023-11-071-1/+1
| |\ \
| * | | Update ChangeLogTheo Buehler2023-11-071-0/+1
| | | |
* | | | Land #938, cmake: simplify if expressions, accept CPU values more consistentlyBrent Cook2023-11-071-19/+19
|\ \ \ \ | | |_|/ | |/| |
| * | | always check `CMAKE_SYSTEM_PROCESSOR` with `MATCHES`Viktor Szakats2023-11-071-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | This makes it accept values consistently. Before this patch mingw-w64, Apple and SunOS did not accept a CPU if it had a suffix or prefix (e.g. a triplet), while other targets did.
| * | | simplify `MATCHES`/`STREQUAL` syntaxViktor Szakats2023-11-071-19/+19
| |/ /
* | | Land #937, remove unused C definesBrent Cook2023-11-073-31/+32
|\ \ \ | | |/ | |/|
| * | cmake: stop passing unused C macrosViktor Szakats2023-11-071-1/+1
| |/ | | | | | | | | | | | | | | - `-DCPPFLAGS`: probably a copy-paste typo from the initial CMake commit. - `-DNO_CRYPT`: `NO_CRYPT` is no longer used in the source and this macro is no longer set by autotools.
| * Land #936, fix arm64 builds to not trigger armv4 asmBrent Cook2023-11-071-1/+2
| |\
| * | Improve Android CI workflowJoshua Sing2023-11-072-30/+31
| | |
* | | Land #936, cmake: fix arm64 to not trigger armv4 ASMBrent Cook2023-11-071-1/+2
|\ \ \ | | |/ | |/|
| * | cmake: fix arm64 to not trigger armv4 ASMViktor Szakats2023-11-071-1/+2
| |/ | | | | | | | | Reported here: https://github.com/libressl/portable/pull/935#issuecomment-1798345787
| * Land #932, use existing crypto/ssl libs for tls testsBrent Cook2023-11-052-4/+3
| |\
| * \ Land #930, consistently use default CA cert location for WindowsBrent Cook2023-11-055-30/+16
| |\ \
* | \ \ Land #932, use existing crypto/ssl libs for tls testsBrent Cook2023-11-052-4/+3
|\ \ \ \ | | |_|/ | |/| |
| * | | slim `tls-static` library (used in tests)Viktor Szakats2023-11-042-4/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Instead of including a full copy of libcrypto and libssl in libtls-static, link existing libcrytpo-static and libssl-static to the test targets. This wasn't causing any issue, just unnecessarily duplicating a lot of objects.
* | | | Land #930, consistently use default CA cert location for WindowsBrent Cook2023-11-055-30/+16
|\ \ \ \ | | |_|/ | |/| |
| * | | also fix ocspcheck and nc, dedupe build logicViktor Szakats2023-11-055-34/+16
| | | |
| * | | cmake: fix default `TLS_DEFAULT_CA_FILE` value in libtls for WindowsViktor Szakats2023-11-041-1/+5
| |/ / | | | | | | | | | | | | | | | | | | Matching the logic used for `libcrypto` (and in autotools builds for both `libcrypto` and `libtls`), set the default `TLS_DEFAULT_CA_FILE` value to `C:/Windows/libressl/ssl/...`, replacing the current, potentialy world-writable filename.
* | | Land #931, dynamic link libtls with libssl/libcrypto in cmake buildsBrent Cook2023-11-054-13/+12
|\ \ \ | | |/ | |/|
| * | slim down `libtls`Viktor Szakats2023-11-041-3/+2
| |/ | | | | | | | | | | | | By dropping copies of `libcrypto` and `libssl` lib objects from it. This brings back its size to the normal ~100KB from ~4MB. autotools already worked like this.