static linking to .dylib in xcode 4.5 error - dll

I'm trying to use FMOD, I'm recreating one of the example projects. When the program loads, it says
dyld: Library not loaded: ./libfmodex.dylib
Referenced from: (...project filepath...)
Reason: image not found
I'm linking by going to build phases of the target, expanding link with binaries, clicking the + sign, clicking other, then navigating to the libfmodex.dylib. It then automatically fills out the library search path.
The program builds and runs but stops instantly with the above error message. Is there some kind of compiler flag that I'm missing?

I found out elsewhere that it needs a build script added (target->build phases->add build phase->run script)
cp "/absolute/filepath/of/your/libfmodex.dylib" .
cp -f "$TARGET_BUILD_DIR/$PRODUCT_NAME" .
cp "/absolute/filepath/of/your/libfmodex.dylib" "$TARGET_BUILD_DIR"

Related

RNCryptor Dyld Error Message: Library not loaded

So, I've tired everything to get this to work after I've archived the application, but it keeps failing to open with the error message:
Dyld Error Message:
Library not loaded: /Library/Frameworks/RNCryptor.framework/Versions/A/RNCryptor
Referenced from: /Users/USER/Desktop/Project.app/Contents/MacOS/Project
Reason: image not found
I've added the Framework to the Copy Files Build Phase, I've changed the Runpath Search Paths to #loader_path/../Frameworks and the Framework Search Path to #executable_path/../Frameworks and it still doesn't want to open.
It opens fine in Xcode itself, but as soon as it's archived, it complains and doesn't run.
Any help is much appreciated :)
Ok, so I was able to fix it.
So the application was actually looking for the Framework in /Library/Frameworks/ on my Mac, I originally had thought it was looking for a reference in my actual .app file.
To fix it I opened up Terminal, cd' into the folder containing a copy of RNCryptor.framework and used this command
install_name_tool -id #executable_path/../Frameworks/RNCryptor.framework/RNCryptor RNCryptor.framework/RNCryptor
Then removed the old framework from the Xcode Project and re-inserted it using the Framework I just used from the directory I cd' into, and added the new Framework to the Copy Files Build Phase.
And voila, it worked :)

Xcode build failure "Undefined symbols for architecture x86_64"

An Xcode beginner's question:
It is my first experience with Xcode 4.6.3.
I am trying to write a very simple console program, that searches for paired BT devices and prints them to an NSLog.
It builds with the following error:
Undefined symbols for architecture x86_64:
"_OBJC_CLASS_$_IOBluetoothDevice", referenced from:
objc-class-ref in main.o
ld: symbol(s) not found for architecture x86_64
clang: error: linker command failed with exit code 1 (use -v to see invocation)
I searched like crazy. The common problem should be a reference to a file, of which only the header files are imported and no implementation (*.m-file) is found by the linker. The IOBluetooth library is however, a standard Framework like the Foundation Framework.
What am I missing in my above statement?
I also have tried building it for a 32-bit machine (build fails again). It is clearly a linker error, however I have no idea, to what it relates, except that there is an issue with finding the implementation for IOBluetoothDevice, on both x86 and x64 architecture, while the header files are from a standard included Framework, called IOBluetooth?
For your information my main code "main.m" being:
#import <Foundation/Foundation.h>
#import <IOBluetooth/objc/IOBluetoothDevice.h> // Note the import for bluetooth
#import <IOBluetooth/objc/IOBluetoothDeviceInquiry.h> // Note the import for bluetooth
int main(int argc, const char * argv[])
{
#autoreleasepool {
IOBluetoothDevice *currentDevice;
NSArray *devices = [ IOBluetoothDevice pairedDevices];
for (id currentDevice in devices){
NSLog(#"%i : %#",[ currentDevice classOfDevice ], [ currentDevice name ]);
}
}
return 0;
}
Thanks for any help or pointers to the right direction.
It looks like you are missing including the IOBluetooth.framework in your project. You can add it by:
Clicking on your project in the upper left of the left pane (the blue icon).
In the middle pane, click on the Build Phases tab.
Under "Link Binary With Libraries", click on the plus button.
Find the IOBluetooth.framework from the list and hit Add.
This will make sure that the IOBluetooth.framework definitions are found by the linker. You can see that the framework is a member of your target by clicking on the framework in the left pane and seeing the framework's target membership in the right pane (note I've moved the framework under the Frameworks group for organization purposes):
UPD
Apple requires to use arm64 architecture. Do not use x32 libraries in your project
So the answer below is not correct anymore!
Old answer
The new Xcode 5.1 sets the architecture armv7,armv7s,and arm64 as default.
And sometimes the error "build failure “Undefined symbols for architecture x86_64”" may be caused by this. Because, some libs (not Apple's) were compiled for x32 originally and doesn't support x64.
So what you need, is to change the "Architectures" for your project target like this
NB. If you're using Cocoapods - you should do the same for "Pods" target.
Undefined symbols for architecture x86_64: "_OBJC_CLASS_$_xxx",
referenced from:
objc-class-ref in yyy.o
This generally means, you are calling "xxx" (it may be a framework or class) from the class "yyy". The compiler can not locate the "xxx" so this error occurs.
You need to add the missing files(in this case "xxx") by right click on your project folder in navigator window and tap on "Add files to "YourProjectName"" option.
A popup window will open your project files in Finder. There, you can see the missing files and just add them to your project. Don't forget to check the "Copy items if needed" box. Good luck!!
I have also seen this error on Xcode 7.2 when the derived data becomes corrupted (in my case I interrupted a build and suspect that was the root cause).
So if the other solutions (notably Chris's and BraveS's which I suspect are more likely) do not fit your problem try deleting derived data (Select: Window / Projects / Derived Data -> Delete) and re-building.
(Added for reference by others - I know the original question has been answered correctly).
Under Xcode 9.0b5 you may encounter this because Xcode 9.0b5 has a bug in it where when you add source code, it does not honor the target settings. You must go in and set each file's target manually afterwords:
In my Case , it was not a library, it was some classes ..
Undefined symbols for architecture x86_64:
"_OBJC_CLASS_$_ClassNmae", referenced from: objc-class-ref in
SomeClassName" . . .
d: symbol(s) not found for architecture x86_64
clang: error: linker command failed with exit code 1 (use -v to see
invocation)
Solution
I had several targets in Xcode with several schemas ( Production , Dev etc ) .. some of my newly added implementation ( Class.m ) were missing in
Xcode->Targets->Build Phases->Compile Sources
So I had to add them manually.
then I could compile & build successfully.
I also encountered the same problem , the above methods will not work . I accidentally deleted the files in the following directory on it .
Or
~/Library/Developer/Xcode/DerivedData/
I tried just about everything here but my problem turned out to be the remnants of a previous cocoapods build. What worked for me was:
rm -Rf Pods; pod install
Delete Derived Data (Window/Projects... select your target. click Delete Button)
Rebuild
I have faced this issue many times. This usually comes when you delete your build folder.
The easy solution is to de-integrate and install the pod files again.
pod deintegrate
pod install
When updating to Xcode 7.1, you might see this type of error, and it can't be resolved by any of the above answers. One of the symptoms in my case was that the app runs on the device not in the simulator. You'll probably see a huge number of errors related to pretty much all of the frameworks you're using.
The fix is actually quite simple. You just need to delete an entry from the "Framework Search Paths" setting, found in your TARGETS > Build Settings > Search Paths section (make sure the "All" tab is selected)
If you see another entry here (besides $(inherited)) for your main target(s) or your test target, just delete the faulty path from all targets and rebuild.
I have found this can also occur if you drag a folder with Objective-C files into your project. If that folder appears blue I think it indicates its not properly linked. You can verify this (if you use version control) because whenever you add new files the pbxproj file should update with links to those new files. However you may find that after you added a folder that the pbxproj file did not change (and hence there is a linking error). So you will get auto-complete working and it will find the classes you imported, but when it goes to actually build the image it fails with this error code.
The solution is to not add the folder but rather add the files. Do this and you should see the pbxproj file update and it should fix this error.
This also assumes you've done what was suggested above and have properly linked all the right frameworks.
I know it's an old question but today got the same error and non of the above solutions worked.
Have fixed it however by setting option:
Project -> Architecture -> Build Active Architecture Only
to Yes
and project compiles and builds properly
I had the same error, because instead of deleting a file I simply removed references to it. Locating the file in Finder and removing it helped.
In my case, I built a custom framework with Deployment target set to 9.1, but the Deployment target of my app was lower, which supports 8.1. Minimize the custom framework Deployment target solved my problem.
If you're getting this error when trying to link to a C file, first double check the function names for typos. Next double check that you are not trying to call a C function from a C++ / Objective-C++ environment without using the extern C {} construct. I was tearing my hair out because I had a class that was in a .mm file which was trying to call C functions. It doesn't work because in C++, the symbols are mangled. You can actually see the concrete symbols generated using the nm tool. Terminal to the path of the .o files, and run nm -g on the file that is calling the symbol and the one that should have the symbol, and you should see if they match up or not, which can provide clues for the error.
nm -g file.o
You can inspect the C++ symbols demangled with this:
nm -gC file.o
I got it solved by adding "-lc++" in Other Linker Flags in Build Settings.
In my case problem was compiled framework architecture.
I'm running Xcode 11 and using Swift 5.1
I had 3 target like:
MyApp
MyAppTests
MyAppFrameWork
I was tried to run tests but MyAppFrameWork product was compiled for Generic iOS Devices and the Test target needed an arm x86-64, So I rebuilt Framework for iOS Simulators and test cases successfuly start running.
This is also happend with apple M1 chip.
Here is my solution just check Open using Rosetta
Steps:
Go to application > right click xcode > get info > check Open using Rosetta
Restart the system.
I am late to the party but thought of sharing one more scenario where this could happen.
I was working on a framework and was distributing it over cocoapods.
The framework had both objective c and swift classes and protocols and it was building successfully.
While using pod in another framework or project it was giving this error as I forgot to include .m files in podspec.
Please include .swtift,.h and .m files in your podspec sources as below:
s.source_files = "Projectname/Projectname/**/*.{swift,h,m}"
I hope it saves someone else's time.
in my case I had to add
target 'SomeTargetTests' do
inherit! :search_paths
end
to my podfile and then delete the /Pods directory and run `pod install`
(Xcode 10.1)
This might help somebody. It took me days to finally figure it out. I am working in OBJ-C and I went to:
Project -> Build Phases -> Compile sources and added the new VC.m file I had just added.
I am working with legacy code and I am usually a swifty, new to OBJ-C so I didn't even think to import my .m files into a sources library.
EDIT:
Ran into this problem a second time and it was something else. This answer saved me after 5 hours of debugging. Tried all of the options on this thread and more. https://stackoverflow.com/a/13625967/7842175 Please give him credit if this helps you, but basically you might need to set your file to its target in file inspector.
All in all, this is a very vague Error code that could be caused for a lot of reasons, so keep on trying different options.
What helped me was adding s.static_framework = true to my /podspec in the project that was throwing the error.
For me, this started to happen after merge conflict.
I tried to clean and remove build folder, but none of it helped. This issue kept happening regardless. Then I relinked the reference by deleting the groups that was problematic and re-added to the project and it worked.
Could also be an #include <windows.h> in the .c file that you're trying to compile.
Sometime, I forget to copy library from Release-universal and mistakenly copy from Release-iphoneos. Usually Release-iphoneos contains .a file which has been pruned for X86. and so it gives the error.
in my case, removing selection of target membership and then select again fix the issue.
Check William Cerniuk answer with the attachment photo.
Undefined symbols for architecture x86_64
I have run on this issue when used CocoaPods with some dependency which did not have specific version, that is why after pod update it downloaded the latest version which included some breaking changes
Upgrade dependencies and code for using them
Set specific version of pod
Remove Derived Data[About] folder
In my case I was getting this error: Undefined symbols for architecture x86_64: "_OBJC_CLASS _ $ _ RCTImageLoader"
And I was able to fix it by adding the following line in my Podfile file:
pod 'React-RCTImage', :path => '../node_modules/react-native/Libraries/Image'
Reference
It happens when you are using architecture arm6, arm7 in your current project but any 3rd party framework which you are trying to consume in your project is built over x86_64.
For e.g If you are using Cocoa Pods in your project, then you need to add following script to make sure all 3rd party frameworks ensure arm6, arm7. i.e
Sample podfile with script to be add at end
target 'XYZ_ProjectTarget' do
# Comment the next line if you don't want to use dynamic frameworks
use_frameworks!
# Pods for XYZ_ProjectTarget
pod 'pod_name'
target 'XYZ_TargetTests' do
inherit! :search_paths
# Pods for testing
end
target 'XYZ_TargetUITests' do
# Pods for testing
end
end
post_install do |installer_representation|
installer_representation.project.targets.each do |target|
target.build_configurations.each do |config|
config.build_settings['ARCHS'] = 'armv7 armv7s'
end
end
end

How to run objective-c console application (with a library) outside of xcode?

I have a simple console application intended to exercise a library from a larger project. The console application is one target within an XCode project that also contains the library ( plus some other targets I don't care about right now). The console target references the library by listing it in the Build Phases "Target Dependencies" and "Link Binary with Libraries".
I can run the console program just fine from within XCode by clicking the "Run" button and examining the console output under "All Output".
However I get a load error if I try to run outside of XCode. I select the console application under "Products" in the project navigator, and then I locate the "Full Path" in the "Identity" section of the information panel. When I copy this path to a terminal window and run it, I get an error like:
dyld: Library not loaded:
#executable_path/../Frameworks/[LIBRARY].framework/Versions/A/[LIBRARY]
Referenced from: [THE PATH THAT I TYPED]
Reason: image not found
What should I do differently to get my console application to execute from the console?
(Note, in the "Copy Files" phase, I have tried leaving the "Subpath" blank and setting it to "#executable_path/../Frameworks" per this answer, which was about a cocoa app, but neither way has helped.)
As the commenters above explain, dyld is the dynamic loader and it is looking for my library at the path it printed. (When it says #executable_path, it means the location of the console application I'm actually trying to run.)
Looking at the compiled output from XCode, I see a folder hierarchy like this:
Debug/
MyProgram
Frameworks/[LIBRARY].framework/Versions/A/[LIBRARY]
For whatever reason, this structure does not match what the loader wants -- it is looking for the library at ../Frameworks rather than ./Frameworks.
Rather than messing with the environment variables used by the loader or trying to tweak the program itself, I just copied these files a new folder with the structure specified in the error message:
foo/
bin/MyProgram
Frameworks/[LIBRARY].framework/Versions/A/[LIBRARY]
I can now run foo/bin/MyProgram from the terminal without any problem.

Is there any way to change the dynamic library search path in Xcode

Below is my scenario,
In my Application i had to make use of libopus library , i downloaded and install, compile --> install procedure is normal as its for any other open source library,
I linked libopus.a with my application, the way i did is , by default it will get installed in /usr/local/lib, so i drag from there and add it to my application,
Worked fine and no error on my machine,
On Another machine, i was expecting it to be run smoothly as i included this library statically, but its throwing error as
dyld: Library not loaded: /usr/local/lib/libopus.0.dylib
so i concluded, libopus.a somehow including libopus.0.dylib also dynamically,
Now i am able ot add a copy phase in my build setting , so it will get copied in ../Framework folder
if i do otool -L libpus.a then it shows following result
otool -L /usr/local/lib/libopus.a
Archive : /usr/local/lib/libopus.a
/usr/local/lib/libopus.a(bands.o):
/usr/local/lib/libopus.a(celt.o):
/usr/local/lib/libopus.a(cwrs.o):
/usr/local/lib/libopus.a(entcode.o):
/usr/local/lib/libopus.a(entdec.o):
/usr/local/lib/libopus.a(entenc.o):
/usr/local/lib/libopus.a(repacketizer.o):
It doesn't show as its depend upon the dylib library
Now my Question is
How to tell Application to look into this path first
I tried following option,
install_name_tool but it seems it will work on other machine , so the user need to run this script NOT DEVELOPER,
trying to set the some option in the xcode to set the RUNTIME Search path to locate that particular dylib but not getting succeed so far
install_name_tool is run by the developer during the build process, not by the user.
If you're building the library, you should use libtool(1) with the option -install_name #rpath; otherwise, you can use install_name_tool(1) with -id #rpath to do the same thing on the dylib. Then, when you're building your application, set the "Runpath search paths" to the path where you will install the library.
Apple has some good documentation on this in their Mach-O Programming Topics and Dynamic Library Programming Topics.

Xcode 4 adding dylib

I am trying to create and then add the dylib to a project.
I created it by using the "Cocoa-Library" template and setting the type to "Dynamic" (not sure if it should be dynamic or static...). Then I created a simple obj-c class called Test and wrote a method in it that prints out something to console.
I compiled and used the generated .dylib file and put it in another project. Now whenever I try to use it, I get this message on runtime"
dyld: Library not loaded: /usr/local/lib/TESTLib.dylib
Referenced from: /Users/***/Library/Developer/Xcode/DerivedData/TestingDYLIB-axmoocnxbwznoyerfogosumqufxc/Build/Products/Debug/TestingDYLIB.app/Contents/MacOS/TestingDYLIB
Reason: image not found
I also created a Copy File phase and set the destination to "Frameworks". I still get the same error.
What am I doing wrong?
The issue is not where Xcode is looking for the library at compile time, which is what Simon Whitaker's answer addresses.
The issue is that the application which uses the dylib cannot find it at runtime. When an application is built that uses a dynamic library, it copies the install_name of the dylib into the output binary.
Creating a copy files phase and copying the dylib to the Frameworks subdirectory of the app's bundle is the right thing do do.
However, you need to do an additional step. You need to compile the dynamic library with an install_name appropriate for a bundle app. By default, a dynamic library is created with an install_name of /usr/local/lib. The app can't find your library there because it doesn't exist. Even if you put the library there, your users certainly won't have it, so that would be the wrong solution.
The right solution is bundling the library with the app. To set the install name for the dynamic library, go into the dynamic library project and set the "Dynamic Library Install Name" option to: #executable_path/../Frameworks/libmydynamiclibrary.dylib
Xcode is looking in /usr/local/lib/ for the library. If the library is in another location, add that location to your Library Search Paths:
Select project file in Xcode 4
Select the target, then click the Build Settings tab
Make sure All is selected in the filter bar (not Basic)
Scroll down to the Search Paths section and you'll find Library Search Paths in there