dmd
dmd copied to clipboard
CI: Switch to macOS-11 as 10.15 will be removed Aug 30th
We've just been hit by a brownout.
Thanks for your pull request, @Geod24!
Bugzilla references
Your PR doesn't reference any Bugzilla issue.
If your PR contains non-trivial changes, please reference a Bugzilla issue or create a manual changelog.
Testing this PR locally
If you don't have a local development environment setup, you can use Digger to test this PR:
dub run digger -- build "master + dmd#14333"
We have another brownout, what needs to be done to move this forward?
We have another brownout, what needs to be done to move this forward?
Was a bit puzzled by the issue, just found out that we need to change Phobos first :upside_down_face: https://github.com/dlang/phobos/pull/8527
Dependant PR merged
ping
# All hosts are 64 bits but let's be explicit
[3](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:3)
./dmd/compiler/src/build.d -j2 MODEL=64
[4](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:4)
make -C dmd/druntime -f posix.mak -j2 MODEL=64
[5](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:5)
make -C phobos -f posix.mak -j2 MODEL=64
[6](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:6)
# Both version can live side by side (they end up in a different directory)
[7](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:7)
# However, since clang does not provide a multilib package, only test 32 bits with g++
[8](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:8)
if [ clang == "g++" ]; then
[9](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:9)
./dmd/compiler/src/build.d install -j2 MODEL=32
[10](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:10)
make -C dmd/druntime -f posix.mak install -j2 MODEL=32
[11](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:11)
make -C phobos -f posix.mak install -j2 MODEL=32
[12](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:12)
fi
[13](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:13)
shell: /bin/bash -e {0}
[14](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:14)
env:
[15](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:15)
DC: dmd
[16](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:16)
LD_LIBRARY_PATH: /Users/runner/hostedtoolcache/dc/dmd-2.100.0/x64/dmd2/linux/lib64
[17](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:17)
CC: /Users/runner/work/dmd/dmd/clang+llvm-11.0.0-x86_64-apple-darwin/bin/clang-wrapper
[18](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:18)
CXX: /Users/runner/work/dmd/dmd/clang+llvm-11.0.0-x86_64-apple-darwin/bin/clang++-wrapper
[19](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:20)
Warning: Syncing file access because of OSX!
[20](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:21)
Warning: Syncing file access because of OSX!
[21](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:22)
ld: library not found for -lpthread
[22](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:23)
clang-11: error: linker command failed with exit code 1 (use -v to see invocation)
[23](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:24)
Error: linker exited with status 1
[24](https://github.com/dlang/dmd/runs/7817610235?check_suite_focus=true#step:16:25)
Error: Process completed with exit code 1.
@Geod24 all passing now. :-)
@ibuclaw : Thanks ! I rebased on master, trimmed down the commit, and went with a slightly different approach: Instead of setting SDKROOT, I just changed the key of the cache to trigger a cache miss (if you look, it was already using -1 as a suffix for this reason).
Checking if this pass, but if it does, we probably don't need the wrapper scripts (we can just set SDKROOT, which I didn't knew about).
@ibuclaw : Thanks ! I rebased on master, trimmed down the commit, and went with a slightly different approach: Instead of setting
SDKROOT, I just changed the key of the cache to trigger a cache miss (if you look, it was already using-1as a suffix for this reason). Checking if this pass, but if it does, we probably don't need the wrapper scripts (we can just set SDKROOT, which I didn't knew about).
Some primer information about SDKROOT/sysroot/syslibroot: https://lists.apple.com/archives/xcode-users/2005/Dec/msg00524.html
I read this comment before settling on SDKROOT: https://trac.macports.org/ticket/57612#comment:18
Reverted back to your diff.