Build results from 2020-12-27#15353
Conversation
PASSING, REMOVE FROM FAIL LIST: capnproto:x64-windows (.\scripts\ci.baseline.txt) PASSING, REMOVE FROM FAIL LIST: capnproto:x64-windows-static (.\scripts\ci.baseline.txt) PASSING, REMOVE FROM FAIL LIST: capnproto:x86-windows (.\scripts\ci.baseline.txt) These were fixed by the update to VS2019 version 16.8. Other capnproto skips moved into "supports" because: capnp.lib(serialize-packed.c++.obj) : error LNK2019: unresolved external symbol __popcnt referenced in function "int __cdecl kj::popCount(unsigned int)" (?popCount@kj@@yahi@Z) i.e. the port wants popcnt to exist and it doesn't, so this isn't an artifact of our CI.
… Acquire using HTTPS.
``` Starting package 759/1414: leaf:x86-windows Building package leaf[core]:x86-windows... -- Downloading https://github.com/zajo/leaf/archive/0.2.2.tar.gz... -- Extracting source D:/downloads/zajo-leaf-0.2.2.tar.gz -- Using source at D:/buildtrees/leaf/src/0.2.2-26417cde9d.clean -- Configuring x86-windows -- Building x86-windows-dbg -- Building x86-windows-rel -- Installing: D:/packages/leaf_x86-windows/share/leaf/copyright -- Performing post-build validation -- Performing post-build validation done Uploaded binaries to 1 HTTP remotes. Building package leaf[core]:x86-windows... done Installing package leaf[core]:x86-windows... The following files are already installed in D:/installed/x86-windows and are in conflict with leaf:x86-windows Installed by boost-leaf:x86-windows include/boost/leaf/capture.hpp include/boost/leaf/common.hpp include/boost/leaf/context.hpp include/boost/leaf/detail/all.hpp include/boost/leaf/detail/demangle.hpp include/boost/leaf/detail/function_traits.hpp include/boost/leaf/detail/mp11.hpp include/boost/leaf/detail/optional.hpp include/boost/leaf/detail/print.hpp include/boost/leaf/error.hpp include/boost/leaf/exception.hpp include/boost/leaf/result.hpp Elapsed time for package leaf:x86-windows: 2.931 s ``` Caused by the update to Boost 1.75.0. Because this port is touching include/boost, I'm interpreting that as "the author intends users to use it as a Boost library", and have deleted this port in favor of the Boost version.
|
@BillyONeal That's strange. Python 3.9 should build correctly with VS 2015 or higher. I'll fire up a VS 2017 virtual machine and look soon. |
|
@Hoikas to clarify, when I try to build this locally it works fine, it's only broken in CI; I suspect it's some hidden conflict with another port. But I didn't see where |
ports/leaf/CONTROL
Outdated
| Source: leaf | ||
| Version: 0.2.2 | ||
| Description: Lightweight error augmentation framework No newline at end of file | ||
| Version: removed |
There was a problem hiding this comment.
With versioning, this would need to be 0.2.2#1 to enable graceful constraint transition.
There was a problem hiding this comment.
Hmmm... doesn't that make it a lie though?
There was a problem hiding this comment.
@vicroms Maybe we want some kind of way to say "this one is replaced with that one" in version-space
Co-authored-by: ras0219 <533828+ras0219@users.noreply.github.com>
|
LGTM. |
* [capnproto] Remove from baseline. PASSING, REMOVE FROM FAIL LIST: capnproto:x64-windows (.\scripts\ci.baseline.txt) PASSING, REMOVE FROM FAIL LIST: capnproto:x64-windows-static (.\scripts\ci.baseline.txt) PASSING, REMOVE FROM FAIL LIST: capnproto:x86-windows (.\scripts\ci.baseline.txt) These were fixed by the update to VS2019 version 16.8. Other capnproto skips moved into "supports" because: capnp.lib(serialize-packed.c++.obj) : error LNK2019: unresolved external symbol __popcnt referenced in function "int __cdecl kj::popCount(unsigned int)" (?popCount@kj@@yahi@Z) i.e. the port wants popcnt to exist and it doesn't, so this isn't an artifact of our CI. * Remove "passing: remove from fail list" entries. * [atk, atkmm, gdk-pixbuf, glibmm, gtkmm, libsigcpp, libxmlpp, pangomm] Acquire using HTTPS. * [chardir] Update osx download. * [leaf] Remove leaf in favor of boost-leaf. ``` Starting package 759/1414: leaf:x86-windows Building package leaf[core]:x86-windows... -- Downloading https://github.com/zajo/leaf/archive/0.2.2.tar.gz... -- Extracting source D:/downloads/zajo-leaf-0.2.2.tar.gz -- Using source at D:/buildtrees/leaf/src/0.2.2-26417cde9d.clean -- Configuring x86-windows -- Building x86-windows-dbg -- Building x86-windows-rel -- Installing: D:/packages/leaf_x86-windows/share/leaf/copyright -- Performing post-build validation -- Performing post-build validation done Uploaded binaries to 1 HTTP remotes. Building package leaf[core]:x86-windows... done Installing package leaf[core]:x86-windows... The following files are already installed in D:/installed/x86-windows and are in conflict with leaf:x86-windows Installed by boost-leaf:x86-windows include/boost/leaf/capture.hpp include/boost/leaf/common.hpp include/boost/leaf/context.hpp include/boost/leaf/detail/all.hpp include/boost/leaf/detail/demangle.hpp include/boost/leaf/detail/function_traits.hpp include/boost/leaf/detail/mp11.hpp include/boost/leaf/detail/optional.hpp include/boost/leaf/detail/print.hpp include/boost/leaf/error.hpp include/boost/leaf/exception.hpp include/boost/leaf/result.hpp Elapsed time for package leaf:x86-windows: 2.931 s ``` Caused by the update to Boost 1.75.0. Because this port is touching include/boost, I'm interpreting that as "the author intends users to use it as a Boost library", and have deleted this port in favor of the Boost version. * [magic-get] Remove magic-get in favor of boost-pfr. * Update ports/chartdir/portfile.cmake Co-authored-by: ras0219 <533828+ras0219@users.noreply.github.com> * Restore versions. * format-manifest Co-authored-by: ras0219 <533828+ras0219@users.noreply.github.com>
|
@Hoikas OK I got the python3 regression to repro consistently locally by running CI on my local box overrnight; will keep you posted |
|
Great work! I was at a loss on this one. |
Build results from 2020-12-27 https://dev.azure.com/vcpkg/public/_build/results?buildId=47034
PASSING, REMOVE FROM FAIL LIST: capnproto:x64-windows (.\scripts\ci.baseline.txt)
PASSING, REMOVE FROM FAIL LIST: capnproto:x64-windows-static (.\scripts\ci.baseline.txt)
PASSING, REMOVE FROM FAIL LIST: capnproto:x86-windows (.\scripts\ci.baseline.txt)
These were fixed by the update to VS2019 version 16.8. Other capnproto skips moved into "supports" because:
capnp.lib(serialize-packed.c++.obj) : error LNK2019: unresolved external symbol __popcnt referenced in function "int __cdecl kj::popCount(unsigned int)" (?popCount@kj@@yahi@Z)
i.e. the port wants popcnt to exist and it doesn't, so this isn't an artifact of our CI.
PASSING, REMOVE FROM FAIL LIST: libepoxy:arm64-windows (.\scripts\ci.baseline.txt)
PASSING, REMOVE FROM FAIL LIST: libsamplerate:arm-uwp (.\scripts\ci.baseline.txt)
PASSING, REMOVE FROM FAIL LIST: libsamplerate:arm64-windows (.\scripts\ci.baseline.txt)
PASSING, REMOVE FROM FAIL LIST: nanogui:arm64-windows (.\scripts\ci.baseline.txt)
Not sure why these got fixed.
REGRESSION: atk:x86-windows. If expected, add atk:x86-windows=fail to .\scripts\ci.baseline.txt
REGRESSION: dcmtk:x86-windows. If expected, add dcmtk:x86-windows=fail to .\scripts\ci.baseline.txt.
REGRESSION: libsigcpp:x86-windows. If expected, add libsigcpp:x86-windows=fail to .\scripts\ci.baseline.txt.
These all had various timeouts and TLS errors connecting ftp.gnome.org; I updated all ports targeting this server to use HTTPs.
REGRESSION: chartdir:x64-osx. If expected, add chartdir:x64-osx=fail to .\scripts\ci.baseline.txt.
Upstream sources broke:
REGRESSION: leaf:arm-uwp. If expected, add leaf:arm-uwp=fail to .\scripts\ci.baseline.txt.
REGRESSION: leaf:arm64-windows. If expected, add leaf:arm64-windows=fail to .\scripts\ci.baseline.txt.
REGRESSION: leaf:x64-linux. If expected, add leaf:x64-linux=fail to .\scripts\ci.baseline.txt.
REGRESSION: leaf:x64-osx. If expected, add leaf:x64-osx=fail to .\scripts\ci.baseline.txt.
REGRESSION: leaf:x64-uwp. If expected, add leaf:x64-uwp=fail to .\scripts\ci.baseline.txt
REGRESSION: leaf:x64-windows-static. If expected, add leaf:x64-windows-static=fail to .\scripts\ci.baseline.txt.
REGRESSION: leaf:x64-windows. If expected, add leaf:x64-windows=fail to .\scripts\ci.baseline.txt
REGRESSION: leaf:x86-windows. If expected, add leaf:x86-windows=fail to .\scripts\ci.baseline.txt.
Caused by the update to Boost 1.75.0. Because this port is touching include/boost, I'm interpreting that as "the author intends users to use it as a Boost library", and have deleted this port in favor of the Boost version.
REGRESSION: magic-get:arm-uwp. If expected, add magic-get:arm-uwp=fail to .\scripts\ci.baseline.txt.
REGRESSION: magic-get:arm64-windows. If expected, add magic-get:arm64-windows=fail to .\scripts\ci.baseline.txt.
REGRESSION: magic-get:x64-linux. If expected, add magic-get:x64-linux=fail to .\scripts\ci.baseline.txt.
REGRESSION: magic-get:x64-osx. If expected, add magic-get:x64-osx=fail to .\scripts\ci.baseline.txt.
REGRESSION: magic-get:x64-uwp. If expected, add magic-get:x64-uwp=fail to .\scripts\ci.baseline.txt.
REGRESSION: magic-get:x64-windows-static. If expected, add magic-get:x64-windows-static=fail to .\scripts\ci.baseline.txt
REGRESSION: magic-get:x64-windows. If expected, add magic-get:x64-windows=fail to .\scripts\ci.baseline.txt.
REGRESSION: magic-get:x86-windows. If expected, add magic-get:x86-windows=fail to .\scripts\ci.baseline.txt.
This library got accepted as boost-pfr. Changing to use the boost port under the same rationale as leaf.
REGRESSION: python3:x64-windows-static. If expected, add python3:x64-windows-static=fail to .\scripts\ci.baseline.txt.
REGRESSION: python3:x64-windows. If expected, add python3:x64-windows=fail to .\scripts\ci.baseline.txt.
REGRESSION: python3:x86-windows. If expected, add python3:x86-windows=fail to .\scripts\ci.baseline.txt.
Probably caused by #14891
These files are C++ files getting built as C. mpdecimal.h seems to be the root of it but it has:
so I'm not sure exactly what is up right now....
@Hoikas any ideas?