Sorry, i'm not sure which script you're referring to!

> On Feb 15, 2019, at 1:44 PM, Gregory Casamento <greg.casame...@gmail.com> 
> wrote:
> 
> Are you talking about how the script configures things?  
> 
> On Fri, Feb 15, 2019 at 12:08 Jordan Schidlowsky <jor...@noodlecake.com 
> <mailto:jor...@noodlecake.com>> wrote:
> We will be shipping a large game in a couple weeks on android that is built 
> on gnustep.   We've been testing very extensively over the past 3-4 months 
> and things look quite stable (at least on armv7-a).  The game heavily uses 
> ARC, blocks, libdispatch....  Lemme just say the configuration of 
> gnustep-base, build, and project setup, were........ not ideal.
> 
>> On Feb 15, 2019, at 11:04 AM, Gregory Casamento <greg.casame...@gmail.com 
>> <mailto:greg.casame...@gmail.com>> wrote:
>> 
>> The challenge at this point is to get a working app.  
>> 
>> On Wed, Feb 13, 2019 at 14:47 Gregory Casamento <greg.casame...@gmail.com 
>> <mailto:greg.casame...@gmail.com>> wrote:
>> 
>> Ivan pointed out that I was using the wrong ninja script.  This solved my 
>> problem.   I am now working on getting base to build properly.
>> 
>> GC
>> 
>> On Wed, Feb 13, 2019 at 12:21 PM Ivan Vučica <i...@vucica.net 
>> <mailto:i...@vucica.net>> wrote:
>> FTR thank you for that, and for spotting -androideabi16!
>> 
>> On Wed, Feb 13, 2019 at 5:20 PM Jordan Schidlowsky <jor...@noodlecake.com 
>> <mailto:jor...@noodlecake.com>> wrote:
>> Agreed, just sharing whats working for us via gradle config.
>> 
>>> On Feb 13, 2019, at 11:17 AM, Ivan Vučica <i...@vucica.net 
>>> <mailto:i...@vucica.net>> wrote:
>>> 
>>> Aye; but we should not depend on using Gradle as the end state.
>>> 
>>> On Wed, Feb 13, 2019 at 5:09 PM Jordan Schidlowsky <jor...@noodlecake.com 
>>> <mailto:jor...@noodlecake.com>> wrote:
>>> Ya you're right,  It should be picked up.  For what it's worth, here's our 
>>> gradle config, (which ends up passing the correct flags to cmake for 
>>> building gnustep):
>>> defaultConfig {
>>>     applicationId "com.noodlecake.ssg4"
>>>     minSdkVersion 21
>>>     targetSdkVersion 28
>>>     versionCode 28
>>>     versionName "1.0.0"
>>>     testInstrumentationRunner 
>>> "android.support.test.runner.AndroidJUnitRunner"
>>>     externalNativeBuild {
>>>         cmake {
>>>             cppFlags "-std=c++14 -frtti -fexceptions 
>>> -fconstant-string-class=NSConstantString"
>>>             cFlags "-DANDROID -fconstant-string-class=NSConstantString"
>>>             arguments "-DCMAKE_VERBOSE_MAKEFILE=ON",
>>>                     "-DANDROID_STL=c++_static",
>>>                     "-DANDROID_DISABLE_FORMAT_STRING_CHECKS=TRUE"
>>>         }
>>>     }
>>>     ndk {
>>>         // Specifies the ABI configurations of your native
>>>         // libraries Gradle should build and package with your APK.
>>>         //abiFilters 'x86', 'x86_64', 'armeabi', 'armeabi-v7a',
>>>         //        'arm64-v8a'
>>>         abiFilters 'armeabi-v7a'
>>>     }
>>> }
>>> 
>>> 
>>> 
>>>> On Feb 13, 2019, at 11:03 AM, Ivan Vučica <i...@vucica.net 
>>>> <mailto:i...@vucica.net>> wrote:
>>>> 
>>>> Reading through the toolchain file 
>>>> <https://mailtrack.io/trace/link/faab2ea5ccbf2b25df57825eb4e8d933befc94aa?url=https%3A%2F%2Fandroid.googlesource.com%2Fplatform%2Fndk%2F%2B%2Fmaster%2Fbuild%2Fcmake%2Fandroid.toolchain.cmake&userId=2790543&signature=ae1079c361f3f494>,
>>>>  ANDROID_PLATFORM will be calculated from ANDROID_NATIVE_API_LEVEL. So 
>>>> that should not be the issue.
>>>> 
>>>> I'd start by pepper-spraying message() calls in toolchain .cmake file, and 
>>>> in other .cmake files as necessary, too.
>>>> 
>>>> On Wed, Feb 13, 2019 at 5:01 PM Jordan Schidlowsky <jor...@noodlecake.com 
>>>> <mailto:jor...@noodlecake.com>> wrote:
>>>> Ah, i didn't see that shell script...  I think 
>>>> -DANDROID_PLATFORM=android-21 (or 23) is what you want there.
>>>> 
>>>> 
>>>>> On Feb 13, 2019, at 10:53 AM, Ivan Vučica <i...@vucica.net 
>>>>> <mailto:i...@vucica.net>> wrote:
>>>>> 
>>>>> Actually -- it doesn't explain why this is happening, as the shell script 
>>>>> (which I failed to notice Gregory attached) matches what I sent him. It's 
>>>>> using -DANDROID_NATIVE_API_LEVEL=23 which should make it use 
>>>>> ...-androideabi23.
>>>>> 
>>>>> Strange.
>>>>> 
>>>>> On Wed, Feb 13, 2019 at 4:47 PM Ivan Vučica <i...@vucica.net 
>>>>> <mailto:i...@vucica.net>> wrote:
>>>>> Ah, that doesn't match what I sent out and makes me feel better ;-)
>>>>> 
>>>>> On Wed, Feb 13, 2019 at 4:41 PM Jordan Schidlowsky <jor...@noodlecake.com 
>>>>> <mailto:jor...@noodlecake.com>> wrote:
>>>>> I think this line in his output indicates he's building for API 16:
>>>>> 
>>>>>  --target=armv7-none-linux-androideabi16 
>>>>> 
>>>>>> On Feb 13, 2019, at 10:07 AM, Ivan Vučica <i...@vucica.net 
>>>>>> <mailto:i...@vucica.net>> wrote:
>>>>>> 
>>>>>> Since Greg mentioned me: 
>>>>>> 
>>>>>> Instructions/commands I came up with and that I sent over to Gregory 
>>>>>> should attempt using API level 23. Totally arbitrarily picked. Use of 
>>>>>> pre-21 API is not happening so not an issue.
>>>>>> 
>>>>>> Needless to say, building works for me. I don’t have a self-contained 
>>>>>> script to share, but it’s super simplistic and what Greg described (incl 
>>>>>> using GUI to install NDK) is what I did.
>>>>>> 
>>>>>> I’m only sure it builds, not that it works, as I am yet to try running 
>>>>>> the code; I don’t have a build script ready for producing an APK (the 
>>>>>> old approach from 2013 and 2014 is a mess and needs to be reworked).
>>>>>> 
>>>>>> On Wed 13 Feb 2019 at 15:34 Jordan Schidlowsky <jor...@noodlecake.com 
>>>>>> <mailto:jor...@noodlecake.com>> wrote:
>>>>>> I've got some patches but they are pretty ugly and I want to clean them 
>>>>>> up properly before submitting...
>>>>>> 
>>>>>> 
>>>>>>> On Feb 13, 2019, at 8:41 AM, Jordan Schidlowsky <jor...@noodlecake.com 
>>>>>>> <mailto:jor...@noodlecake.com>> wrote:
>>>>>>> 
>>>>>>> An NDK app can chose to bundle in a (c++_shared) library in your app, 
>>>>>>> or you can link with a static (c++_static) standard library.   This is 
>>>>>>> actually dependant on what setting you chose for this in your gradle 
>>>>>>> build file, as gradle will pass that along to cmake which will link it 
>>>>>>> in.  
>>>>>>> 
>>>>>>> That cxx runtime test doesn't quite work correctly using an android 
>>>>>>> toolchain.  But if you want to configure your ndk app using c++_static 
>>>>>>> you can remove that test section from CMakeLists.txt and add in 
>>>>>>> manually below:
>>>>>>> 
>>>>>>> set(CXXRT_IS_STDLIB true)
>>>>>>> target_link_libraries(objc c++_static stdc++)
>>>>>>> 
>>>>>>> I will also note, that I am still thinking about a way to run that test 
>>>>>>> suite while cross compiling...     
>>>>>>> 
>>>>>>> 
>>>>>>>> On Feb 13, 2019, at 6:41 AM, Gregory Casamento 
>>>>>>>> <greg.casame...@gmail.com <mailto:greg.casame...@gmail.com>> wrote:
>>>>>>>> 
>>>>>>>> 
>>>>>>>> A little more context...  my build environment is a MacPro 2010 
>>>>>>>> running the latest version of Mojave.  I have downloaded and installed 
>>>>>>>> the latest of Android studio and installed the latest SDK and NDK 
>>>>>>>> using the menu under tools and the SDK manager.  The version of the 
>>>>>>>> NDK I'm using is 19.0.5232133.   I am utterly stumped as to why this 
>>>>>>>> is not working.  Also, it seems as though Ivan's installation of this 
>>>>>>>> is working which seems to indicate that this is a configuration issue.
>>>>>>>> 
>>>>>>>> Any input would be appreciated.
>>>>>>>> 
>>>>>>>> On Wed, Feb 13, 2019 at 7:09 AM Gregory Casamento 
>>>>>>>> <greg.casame...@gmail.com <mailto:greg.casame...@gmail.com>> wrote:
>>>>>>>> 
>>>>>>>> ### Build libobjc2
>>>>>>>> -- The ASM compiler identification is Clang
>>>>>>>> -- Found assembler: 
>>>>>>>> /Users/heron/Library/Android/sdk/ndk-bundle/toolchains/llvm/prebuilt/darwin-x86_64/bin/clang
>>>>>>>> -- Check for working C compiler: 
>>>>>>>> /Users/heron/Library/Android/sdk/ndk-bundle/toolchains/llvm/prebuilt/darwin-x86_64/bin/clang
>>>>>>>> -- Check for working C compiler: 
>>>>>>>> /Users/heron/Library/Android/sdk/ndk-bundle/toolchains/llvm/prebuilt/darwin-x86_64/bin/clang
>>>>>>>>  -- works
>>>>>>>> -- Detecting C compiler ABI info
>>>>>>>> -- Detecting C compiler ABI info - done
>>>>>>>> -- Detecting C compile features
>>>>>>>> -- Detecting C compile features - done
>>>>>>>> -- Check for working CXX compiler: 
>>>>>>>> /Users/heron/Library/Android/sdk/ndk-bundle/toolchains/llvm/prebuilt/darwin-x86_64/bin/clang++
>>>>>>>> -- Check for working CXX compiler: 
>>>>>>>> /Users/heron/Library/Android/sdk/ndk-bundle/toolchains/llvm/prebuilt/darwin-x86_64/bin/clang++
>>>>>>>>  -- works
>>>>>>>> -- Detecting CXX compiler ABI info
>>>>>>>> -- Detecting CXX compiler ABI info - done
>>>>>>>> -- Detecting CXX compile features
>>>>>>>> -- Detecting CXX compile features - done
>>>>>>>> -- Testing C++ interop
>>>>>>>> -- Testing C++ standard library
>>>>>>>> -- No useable C++ runtime found
>>>>>>>> -- Looking for pthread.h
>>>>>>>> -- Looking for pthread.h - found
>>>>>>>> -- Looking for pthread_create
>>>>>>>> -- Looking for pthread_create - found
>>>>>>>> -- Found Threads: TRUE  
>>>>>>>> -- GNUstep install type set to NONE
>>>>>>>> -- Performing Test CXA_ALLOCATE_EXCEPTION_NOEXCEPT_COMPILES
>>>>>>>> -- Performing Test CXA_ALLOCATE_EXCEPTION_NOEXCEPT_COMPILES - Success
>>>>>>>> -- Configuring done
>>>>>>>> -- Generating done
>>>>>>>> -- Build files have been written to: 
>>>>>>>> /Users/heron/Development/Algoriddim/gnustep-toolchain/gnustep-android/gnustep/libobjc2/build
>>>>>>>> /Users/heron/Development/Algoriddim/gnustep-toolchain/gnustep-android/gnustep/libobjc2/build/CMake
>>>>>>>> [1/1] Linking CXX executable test_cxx_runtime
>>>>>>>> FAILED: test_cxx_runtime 
>>>>>>>> : && 
>>>>>>>> /Users/heron/Library/Android/sdk/ndk-bundle/toolchains/llvm/prebuilt/darwin-x86_64/bin/clang++
>>>>>>>>  --target=armv7-none-linux-androideabi16 
>>>>>>>> --gcc-toolchain=/Users/heron/Library/Android/sdk/ndk-bundle/toolchains/llvm/prebuilt/darwin-x86_64
>>>>>>>>   --sysroot 
>>>>>>>> /Users/heron/Library/Android/sdk/ndk-bundle/toolchains/llvm/prebuilt/darwin-x86_64/sysroot
>>>>>>>>  -g -DANDROID -fdata-sections -ffunction-sections -funwind-tables 
>>>>>>>> -fstack-protector-strong -no-canonical-prefixes -mfpu=vfpv3-d16 
>>>>>>>> -fno-addrsig -mthumb -Wa,--noexecstack -Wformat 
>>>>>>>> -Werror=format-security -stdlib=libc++  -Wl,--exclude-libs,libgcc.a 
>>>>>>>> -Wl,--exclude-libs,libatomic.a -static-libstdc++ -Wl,--build-id 
>>>>>>>> -Wl,--warn-shared-textrel  -Wl,--exclude-libs,libunwind.a 
>>>>>>>> -Wl,--no-undefined -Qunused-arguments -Wl,-z,noexecstack -Wl,-z,relro 
>>>>>>>> -Wl,-z,now -Wl,--gc-sections 
>>>>>>>> CMakeFiles/test_cxx_runtime.dir/typeinfo_test.cc.o  -o 
>>>>>>>> test_cxx_runtime  -latomic -lm && :
>>>>>>>> CMakeFiles/test_cxx_runtime.dir/typeinfo_test.cc.o:typeinfo_test.cc 
>>>>>>>> <https://mailtrack.io/trace/link/60886d21dd395f248f150bf2063af90bce602b5b?url=http%3A%2F%2Ftypeinfo_test.cc%2F&userId=2790543&signature=a72ed8c78f119f8b>:vtable
>>>>>>>>  for type_info2: error: undefined reference to 
>>>>>>>> 'std::type_info::__do_upcast(__cxxabiv1::__class_type_info const*, 
>>>>>>>> void**) const'
>>>>>>>> clang++: error: linker command failed with exit code 1 (use -v to see 
>>>>>>>> invocation)
>>>>>>>> ninja: build stopped: subcommand failed.
>>>>>>>> ### LIBOBJC2 BUILD FAILED!!!
>>>>>>>> 
>>>>>>>> I am running into the above error when trying to build for Android.  I 
>>>>>>>> have the latest ndk and sdk installed with Android studio.  Does 
>>>>>>>> anyone have any idea offhand why C++ isn't finding it's runtime?   I 
>>>>>>>> will attach the build script I am using to generate the ninja 
>>>>>>>> script....
>>>>>>>> 
>>>>>>>> GC
>>>>>>>> 
>>>>>>>> -- 
>>>>>>>> Gregory Casamento
>>>>>>>> GNUstep Lead Developer / OLC, Principal Consultant
>>>>>>>> http://www.gnustep.org 
>>>>>>>> <https://mailtrack.io/trace/link/5539c79bf8ea11ab855df69cf27c47504df4029a?url=http%3A%2F%2Fwww.gnustep.org&userId=2790543&signature=a3bacb38978af1d3>
>>>>>>>>  - http://heronsperch.blogspot.com 
>>>>>>>> <https://mailtrack.io/trace/link/0e7d2849e32e6a5545dee4db716d416d8a769d54?url=http%3A%2F%2Fheronsperch.blogspot.com&userId=2790543&signature=794ec8837a56aa7c>
>>>>>>>> http://ind.ie/phoenix/ 
>>>>>>>> <https://mailtrack.io/trace/link/2957e53708e301d78eebf58dc506d80bbe462935?url=http%3A%2F%2Find.ie%2Fphoenix%2F&userId=2790543&signature=4c11c7cb082d4c41>
>>>>>>>> 
>>>>>>>> -- 
>>>>>>>> Gregory Casamento
>>>>>>>> GNUstep Lead Developer / OLC, Principal Consultant
>>>>>>>> http://www.gnustep.org 
>>>>>>>> <https://mailtrack.io/trace/link/395c4e14c1327f2eb68a925849b3ef4405ef1c3c?url=http%3A%2F%2Fwww.gnustep.org&userId=2790543&signature=97bd0bcba9a39e96>
>>>>>>>>  - http://heronsperch.blogspot.com 
>>>>>>>> <https://mailtrack.io/trace/link/df75d951421ec7ec49a7f3777a60a26416e028ad?url=http%3A%2F%2Fheronsperch.blogspot.com&userId=2790543&signature=9831445b2546de15>
>>>>>>>> http://ind.ie/phoenix/ 
>>>>>>>> <https://mailtrack.io/trace/link/aae8e7cc987e35195aa951657c98b8697f0f1d6e?url=http%3A%2F%2Find.ie%2Fphoenix%2F&userId=2790543&signature=124dae0a65061977>_______________________________________________
>>>>>>>> Discuss-gnustep mailing list
>>>>>>>> Discuss-gnustep@gnu.org <mailto:Discuss-gnustep@gnu.org>
>>>>>>>> https://lists.gnu.org/mailman/listinfo/discuss-gnustep 
>>>>>>>> <https://mailtrack.io/trace/link/f80f650ce2e69eb21b34eff589f9ace84a85b5ba?url=https%3A%2F%2Flists.gnu.org%2Fmailman%2Flistinfo%2Fdiscuss-gnustep&userId=2790543&signature=fd9e027ef774428f>
>>>>>> 
>>>>>> _______________________________________________
>>>>>> Gnustep-dev mailing list
>>>>>> gnustep-...@gnu.org <mailto:gnustep-...@gnu.org>
>>>>>> https://lists.gnu.org/mailman/listinfo/gnustep-dev 
>>>>>> <https://mailtrack.io/trace/link/12db2ea57bbeafe9fb5617a92f224d44ca31af35?url=https%3A%2F%2Flists.gnu.org%2Fmailman%2Flistinfo%2Fgnustep-dev&userId=2790543&signature=69fbc0c26214836b>
>>>>>> -- 
>>>>>> Sent from Gmail Mobile
>>>>> 
>>>> 
>>> 
>> 
>> 
>> 
>> -- 
>> Gregory Casamento
>> GNUstep Lead Developer / OLC, Principal Consultant
>> http://www.gnustep.org 
>> <https://mailtrack.io/trace/link/f96c84969d3eb1649c6152bfd4e969a2d0bc04ca?url=http%3A%2F%2Fwww.gnustep.org&userId=2790543&signature=cf673d8f7c9de2ee>
>>  - http://heronsperch.blogspot.com 
>> <https://mailtrack.io/trace/link/6dcc6d47fb32cb89a517b10436520e8740ac82ac?url=http%3A%2F%2Fheronsperch.blogspot.com&userId=2790543&signature=6596886ce09bf75f>
>> http://ind.ie/phoenix/ 
>> <https://mailtrack.io/trace/link/29eac4e9399908a6ec029d8842d4a0c867a6b255?url=http%3A%2F%2Find.ie%2Fphoenix%2F&userId=2790543&signature=b50eb38ce5efa054>--
>>  
>> Gregory Casamento
>> GNUstep Lead Developer / OLC, Principal Consultant
>> http://www.gnustep.org <http://www.gnustep.org/> - 
>> http://heronsperch.blogspot.com <http://heronsperch.blogspot.com/>
>> http://ind.ie/phoenix/ <http://ind.ie/phoenix/>
> -- 
> Gregory Casamento
> GNUstep Lead Developer / OLC, Principal Consultant
> http://www.gnustep.org <http://www.gnustep.org/> - 
> http://heronsperch.blogspot.com <http://heronsperch.blogspot.com/>
> http://ind.ie/phoenix/ <http://ind.ie/phoenix/>
_______________________________________________
Discuss-gnustep mailing list
Discuss-gnustep@gnu.org
https://lists.gnu.org/mailman/listinfo/discuss-gnustep

Reply via email to