I've got XCode 12.5 up and running on macOS 11.3 (M1 chip) in Github Actions as a self-hosted runner.
While my macOS workflow is able to find gl.h in the github-hosted machine just fine, my own self-hosted one reports this:
CMake Error at /Users/administrator/actions-runner/_work/Mudlet/Qt/5.14.2/clang_64/lib/cmake/Qt5Gui/Qt5GuiConfigExtras.cmake:9 (message):
Error: Failed to find "gl.h" in
"/System/Library/Frameworks/OpenGL.framework/Headers;/System/Library/Frameworks/AGL.framework/Headers".
Call Stack (most recent call first):
/Users/administrator/actions-runner/_work/Mudlet/Qt/5.14.2/clang_64/lib/cmake/Qt5Gui/Qt5GuiConfig.cmake:202 (include)
3rdparty/vcpkg/scripts/buildsystems/vcpkg.cmake:555 (_find_package)
/Users/administrator/actions-runner/_work/Mudlet/Qt/5.14.2/clang_64/lib/cmake/Qt5Widgets/Qt5WidgetsConfig.cmake:94 (find_package)
3rdparty/vcpkg/scripts/buildsystems/vcpkg.cmake:555 (_find_package)
/Users/administrator/actions-runner/_work/Mudlet/Qt/5.14.2/clang_64/lib/cmake/Qt5UiTools/Qt5UiToolsConfig.cmake:219 (find_package)
3rdparty/vcpkg/scripts/buildsystems/vcpkg.cmake:555 (_find_package)
3rdparty/edbee-lib/edbee-lib/CMakeLists.txt:177 (FIND_PACKAGE)
The code responsible is this line:
FIND_PACKAGE(Qt5UiTools REQUIRED)
What could be the reason that gl.h is not locatable?
I got this problem when I updated to MacOS Big Sur (11.6.2) when trying to make a qt app.
This was the error I got:
CMake Error at /usr/local/Qt-5.5.1/lib/cmake/Qt5Gui/Qt5GuiConfigExtras.cmake:9 (message):
Failed to find "gl.h" in
"/System/Library/Frameworks/OpenGL.framework/Headers;/System/Library/Frameworks/AGL.framework/Headers".
Call Stack (most recent call first):
/usr/local/Qt-5.5.1/lib/cmake/Qt5Gui/Qt5GuiConfig.cmake:139 (include)
My google-fu found this from 2019 [https://github.com/ziglang/zig/issues/2208]
https://github.com/ziglang/zig/issues/2208 :
The new path to the framework is:
/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk/System/Library/Frameworks
I looked, and gl.h was found in several places, including OpenGL.framework.
As a work-around, I changed the library path in /usr/local/Qt-5.5.1/lib/cmake/Qt5Gui/Qt5GuiConfigExtras.cmake:9 from:
set(_GL_INCDIRS "/System/Library/Frameworks/OpenGL.framework/Headers" "/System/Library/Frameworks/AGL.framework/Headers")
to:
set(_GL_INCDIRS "/System/Library/Frameworks/OpenGL.framework/Headers" "/System/Library/Frameworks/AGL.framework/Headers" "/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk/System/Library/Frameworks")
running on macOS 11.3 (M1 chip) in Github Actions as a self-hosted runner.
Try again with the latest version of self-hosted runners, since (Aug. 2022):
GitHub Actions: Self-hosted runners now support Apple M1 hardware
Actions runner support for Apple silicon hardware, such as the M1 chip, is now generally available.
This provides teams with the capability to run self-hosted macOS workflows in a macOS ARM64 runtime.
Now the Actions runner supports M1 and the ARM64 runtime meaning developers can run it on their own M1 or M2 hardware.
Based on initial testing, there are currently two issues to be aware of:
macOS ARM64 does not support node12.
Therefore, the runner will automatically use node16 to execute any javascript Action written for node12.
All actions provided by GitHub are compatible with the runner except for a known issue with setup-python. The fix for that can be tracked here.
For additional information on how to set up a self-hosted macOS ARM64 runner, please refer to our documentation.
If you have any feedback or questions for Actions self-hosted Apple silicon support, you can submit an issue in the runner repository.
Related
I am trying to run Corda nodes on macOS El Capitan with JDK 1.8.0_60 by following the instructions here: https://docs.corda.net/getting-set-up.html.
I am getting the following error:
CAPSULE EXCEPTION: Could not parse version: -9.0.1 while processing
system property java.home:
/Library/Java/JavaVirtualMachines/jdk1.8.0_60.jdk/Contents/Home/jre
java.lang.IllegalArgumentException: Could not parse version: -9.0.1
How can I fix this?
Although its older thread, i encountered it today and stumbled upon this page and thought to share my findings.
Actually Oracle changed the numbering style from java9 (like 9.0.1) and it seems you have it installed. You will have to point to jdk 1.8.x so make it work until capsule is updated to work with java9.
Running Corda nodes requires a JDK version greater than JDK 8u121. If you upgrade your JDK, this error will disappear.
I m having 2 version of VS2017 and a mac with latest macOS.
I transfer my Xamarin forms App file from VS2017 enterprise to a machine(more than 3 years old but running win10 pro) with VS2017 Community version. When I debug ( with selecting Android project and VS-android-23 X86 tablet as emulator ) this App in VS2017 Community version , I have no compilation error but have these message:
Found conflicts between different versions of the same dependent assembly
that could not be resolved.
These reference conflicts are listed in the build log when log verbosity is set to detailed.
I could not understand this message. I have no problem running the same app in VS2017 Enterprize ver but I have this error message in VS2017 Community ver.
Update
Click View > Output --> this will show the error
1>Starting deploy VisualStudio_android-23_x86_tablet ...
1>Starting emulator VisualStudio_android-23_x86_tablet ...
1>C:\Users\xxxxx\AppData\Local\Android\ANDROI~1\tools\emulator.EXE -partition-size 512 -no-boot-anim -avd VisualStudio_android-23_x86_tablet -prop monodroid.avdname=VisualStudio_android-23_x86_tablet
1>emulator: ERROR: x86 emulation currently requires hardware acceleration!
1>Please ensure Intel HAXM is properly installed and usable.
1>CPU acceleration status: Please disable Hyper-V before using the Android Emulator. Start a command prompt as Administrator, run 'bcdedit /set hypervisorlaunchtype off', reboot.
1>Emulator VisualStudio_android-23_x86_tablet cannot be started.
Please help.
I have updated my gnat to the latest version from AdaCore and trying to rebuild the latest AWS (Ada Web Server) also downloaded from AdaCore. I got some errors when building:
aws-services-web_block-context.ads:106:21: argument of pragma "SUPPRESS" is not valid check name
aws-services-web_block-context.ads:106:21: argument of pragma "SUPPRESS" is not valid check name
compilation of aws-services-web_block-context.adb failed
compilation of aws-services-web_block-registry.adb failed
gprbuild: *** compilation phase failed
make: *** [build-native] Error 4
What could have been wrong? Or, have I overlook something?
Thanks.
Adrian
The check that is invalid is Tampering_Check, which is to do with the resource-intensive ARM-specified checks on improper use of Containers (basically, is one part of the code trying to change the structure of a Container that another is already using?)
I don’t know why you’ve had this problem: both GNAT GPL 2016 and FSF GCC 6.1.0 are OK with the pragma. Are you sure you’re using the latest compiler?
One other thing to watch out for is that the master branch at Github may rely on compiler features not available in compiler releases available to unsupported customers; if using GNAT GPL 2016, maybe the gpl-2016 branch would be safer, if not up to date.
After upgrade XCode from 4.2 to 4.3 I've got below problem. I can't build it.
Do you guys have idea how can I resolve this problem? even some clue would be helpful to me.
ld: warning: ignoring file /Applications/Xcode.app/Contents/Developer/Library/Frameworks/SenTestingKit.framework/SenTestingKit, missing required architecture armv7 in file
Undefined symbols for architecture armv7:
"_STComposeString", referenced from:
-[NSMutableURLRequest_ParametersTest verifyDefaultParametersForRequest:] in NSMutableURLRequest+ParametersTest.o
-[NSMutableURLRequest_ParametersTest verifyModifiedParametersForRequest:] in NSMutableURLRequest+ParametersTest.o
-[NSMutableURLRequest_ParametersTest verifyEmptyParametersForRequest:] in NSMutableURLRequest+ParametersTest.o
-[NSString_URLEncodingTest testURLEncodedString] in NSString+URLEncodingTest.o
-[NSURL_BaseTest testURLStringWithoutQuery] in NSURL+BaseTest.o
-[OAHMAC_SHA1SignatureProviderTest testSignClearText] in OAHMAC_SHA1SignatureProviderTest.o
-[OAMutableURLRequestTest testGenerateNonce] in OAMutableURLRequestTest.o
...
"_OBJC_CLASS_$_SenTestCase", referenced from:
_OBJC_CLASS_$_NSMutableURLRequest_ParametersTest in NSMutableURLRequest+ParametersTest.o
_OBJC_CLASS_$_NSString_URLEncodingTest in NSString+URLEncodingTest.o
_OBJC_CLASS_$_NSURL_BaseTest in NSURL+BaseTest.o
_OBJC_CLASS_$_OAHMAC_SHA1SignatureProviderTest in OAHMAC_SHA1SignatureProviderTest.o
_OBJC_CLASS_$_OAMutableURLRequestTest in OAMutableURLRequestTest.o
_OBJC_CLASS_$_OAPlaintextSignatureProviderTest in OAPlaintextSignatureProviderTest.o
_OBJC_CLASS_$_OARequestParameterTest in OARequestParameterTest.o
...
"_OBJC_METACLASS_$_SenTestCase", referenced from:
_OBJC_METACLASS_$_NSMutableURLRequest_ParametersTest in NSMutableURLRequest+ParametersTest.o
_OBJC_METACLASS_$_NSString_URLEncodingTest in NSString+URLEncodingTest.o
_OBJC_METACLASS_$_NSURL_BaseTest in NSURL+BaseTest.o
_OBJC_METACLASS_$_OAHMAC_SHA1SignatureProviderTest in OAHMAC_SHA1SignatureProviderTest.o
_OBJC_METACLASS_$_OAMutableURLRequestTest in OAMutableURLRequestTest.o
_OBJC_METACLASS_$_OAPlaintextSignatureProviderTest in OAPlaintextSignatureProviderTest.o
_OBJC_METACLASS_$_OARequestParameterTest in OARequestParameterTest.o
...
ld: symbol(s) not found for architecture armv7
clang: error: linker command failed with exit code 1 (use -v to see invocation)
I was getting the same linking error for the Simulator after upgrading from Xcode 4.2 -> 4.3.1. I noticed that new projects in 4.3.1 do work with Unit Tests, so just compared the project differences.
What fixed it for me was:-
Select Project in project navigator
Select Test Target
Select Build Settings
Go down to Framework Search Paths and single-click it
Delete whatever is there
Enter $(SDKROOT)/Developer/Library/Frameworks $(DEVELOPER_LIBRARY_DIR)/Frameworks
If you had any custom frameworks, add those back
Actually, Unit Tests run fine on the Device, too.
Looks like this was either an accidental (or on purpose) omission on Apple's part.
If (and only if) you are a registered Apple Developer, you can log into the Apple Developer Forums and if you copy this linked search query into your browser address bar, you will see other people have had the same problem.
For now, it looks like you can only use SenTestingKit on the simulator.
I had a similar issue with Xcode 4.4.
My solution path:
Removed SetTestingKit.framework from the main target's (the system under test) "link binary with libraries", leaving it in the test target as is. Not sure why it ended up there in the first place.
I believe this bug is now fixed in Xcode 4.4.1
Using xcode 4.6 and trying to add facebook SDK 3.1 I had this problem because I didn't select to add the files to my test target.
You should check both targets when you copy the framework.
I just upgraded to 4.3.1 There is a problem with SenTestingKit framework. I removed the test source code from the project, it builds ok. And it can run on 5.1 device. But I got _dyld_dyld_start error when I ran it on simulator. I had to remove SenTestingKit from the project to make it work on simulator.
I'm trying to install a linux driver for my wireless network card (D-Link DWL-G510) on my Red Hat Linux 7.1 machine with a 2.4.37.9 kernel. I downloaded the serial monkey driver from the sourceforge site and was able to successfully compile the module. However, whenever I do a "make install", the make script executes a "depmod -a" command which then complains with the message: "Unresolved symbols in /lib/modules/2.4.37.9/extra/rt73.o".
I then executed a "depmod -e" command to show unresolved symbols and it indicates the following information:
request_firmware_Rsmp_38ce5074
release_firmware_Rsmp_33934162
I did a grep on the above information and it showed no source files making reference to it. I searched for it on google and it returned no results. Can anyone help?
I found out the reason - it was missing the firmware_class module. I had to rebuild kernel modules with the experimental feature CONFIG_FW_LOADER as a module which is found under Library routines\Hotplug firmware loading support. After loading the firmware_class module, the rt73 module loaded successfully also.