Re: llvm toolchain packages

[Brook Milligan <brook%nmsu.edu@localhost>]

> On Jan 30, 2019, at 5:12 PM, Adam <adam%viamedium.pl@localhost> wrote:
> 
>> With two exceptions, I have successfully compiled what I believe to be all of the llvm toolchain packages on my Mac.
>> 
>> One exception is lang/compiler-rt; for that I needed the patch the PLIST.Darwin file as below.  Is that ok to commit or is there some reason this is incorrect?
>> 
>> The second exception is devel/lldb, which had the following error during compilation:
>> 
>> [ 36%] Building CXX object source/Plugins/ExpressionParser/Clang/CMakeFiles/lldbPluginExpressionParserClang.dir/ClangHost.cpp.o
>> /Volumes/NetBSD-boost/NetBSD/pkgsrc/pkgsrc-cvs/devel/lldb/work/lldb-7.0.1.src/source/Plugins/ExpressionParser/Clang/ClangHost.cpp:78:17: error: no matching function for call to 'next'
>> auto parent = std::next(rev_it);
>>               ^~~~~~~~~
>> /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/../include/c++/v1/iterator:521:25: note: candidate template ignored: disabled by 'enable_if' [with _ForwardIter = llvm::sys::path::reverse_iterator]
>>    typename enable_if<__is_forward_iterator<_ForwardIter>::value>::type* = 0)
>>                       ^
>> 
>> The requirement for std::next() is that the first argument be a LegacyForwardIterator (until c++17) or a LegacyInputIterator (since c++17).  In this case, it is a reverse iterator as indicated in the error. Clearly, the type traits in use claim that a reverse iterator does not meet the requirement, so the overload for std::next() is disabled because of SFINAE.
>> 
>> So, what should be done to make this compile on a Mac?  The compiler in use is AppleClang 8.0.0.8000042.
>> 
>> Thanks for your help.
>> 
>> Cheers,
>> Brook
> 
> I think I got different results for "Apple LLVM version 10.0.1 (clang-1001.0.37.7)".
> Let me try again…

Please do and report back.

Did you mean different results on both counts or just one?  I'm guessing you meant the compile error above, but did you also mean the PLIST issue as well?

Cheers,
Brook