New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Build failures on macOS #107
Comments
Looks like the zlib failure shows up in an unrelated project: bazelbuild/bazel#17956 There are some proposed fixes in the issue's comments, including a three-line patch and changing the flags used to build zlib, not sure of the best approach here. The My hunch is that some macOS version bump also bumped clang's version and caused some previously benign issues to raise warnings. |
I forgot to note that on the I've been able to fix the |
Another idea that I've seen mentioned is changing clang's build args to ignore this warning. I'm new to C++ and CMake but can dig into the feasibility of doing that since it seems preferable over the patch approach. While we could make that work for |
A few updates on this:
|
@bkirz I'm not sure if downgrading the requirement would be the correct fix / don't have time to dig into this myself, but: Try C_STANDARD instead of CXX_STANDARD and it might do it - C and C++ standards are different things, and zlib is a mostly C library. For the warnings - It's entirely valid to not treat warnings as errors for 3rd party dependencies. That might just be a top level '-Werror=all' or similar for the itgmania code, that's then inherited through the CMake file hierarchy. And for zlib - I'd guess it's a missing include or something, maybe fixed in a newer zlib? You're on the right track though as the two fixes are to update zlib itself (and risk breaking other platforms), or tweak the compile flags to be happy. |
Thanks for the tips @geefr! I tried swapping to
Not what I was hoping for, but at least we can rule out that option now. Seems changing the args is the next best option since updates to zlib and tomcrypt don't yet exist that fix these issues. I'm not sure how to set arbitrary warning config with CMake and will dig into that next. I'm running into a weird issue with calling
Strangely, this persists if I set CMAKE_OSX_ARCHITECTURES as an env arg. I suspect I'll have to sit down and Actually Learn how to use CMake to get around this. |
See itgmania#107 for error details.
See itgmania#107 for error details.
Turns out I just needed to swap out the |
I'm seeing this on Linux too with a recent clang, the fix (edit: *workaround) probably shouldn't be limited to mac |
I'm seeing the same warning building with clang on Linux.
I'm seeing the same warning building with clang on Linux.
* beta: When sorting course songs by GRADEBEST or GRADEWORST, check if GAMESTATE has a master player before calling SongUtil::SortSongPointerArrayByGrades(). Added missing GRADEBEST and GRADEWORST cases for CourseWriterCRS Automatically create Songs folder on Memory Cards Bump actions/checkout from 3 to 4 pref to bass simplify lighting MergeFromOtherHSL() now 'Removes All But One Of Each Name' per preference nudge enchantment noteskin's first two frames 1px Fix groove radar value calculation Make tomcrypt workaround for itgmania#107 apply everywhere # Conflicts: # src/Course.cpp # src/CourseWriterCRS.cpp
I missed zlib in itgmania#150.
I'm running into errors building some dependencies on macOS 13.3.1. Here's a full build log if I run
./Utils/build-release-macos.sh
: https://gist.github.com/bkirz/02660a7e26d0dcfbfe2128c7189fa84aSpecifically, I'm seeing two classes of errors:
mbedtls
, I get the failure/Users/bkirz/dev/itgmania/extern/mbedtls/library/bignum.c:1395:29: error: variable 't' set but not used [-Werror,-Wunused-but-set-variable] mbedtls_mpi_uint c = 0, t = 0; ^
zlib
, I get multiple failures of the formerror: call to undeclared function 'lseek'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration]
.I did some googling but didn't see sny obvious solutions for either problem. Would bumping the version of either library help? Is that even feasible? I'm not sure how up-to-date SM5's dependencies are kept.
Test Environment
The text was updated successfully, but these errors were encountered: