re: We finally have the answer!
@vaporeon_ i thought i remembered git being reasonable enough to compile, but i wasn’t sure
re: We finally have the answer!
@aescling I think that ultimately, the reason for the problem was switching to GCC 11, but I'm not sure. Switching back to GCC 14 on the PowerPC computer didn't help...
Meson (used to build Git) complained that it couldn't find a working Ninja. If I tried ninja --version
, it'd return an error GLIBCXX_3.4.29 not found
...
I figured that re-compiling Ninja with GCC 11 might fix the problem, but that didn't work on the PowerPC system, it kept throwing some weird Python error (something like proc.communicate()[0]
returned an empty string and then an int()
conversion failed) during the configuration steps, and there were no relevant search results for it...
And the cross-compiling environment was compiling an LLVM at that point for other reasons, and I didn't want to interrupt it...
I had to wait that out, and then I cross-compiled Ninja in the cross-compiling enviroment (after also switching it go GCC 11), that didn't throw any weird errors, and then re-installing the Ninja binpkg on my PowerPC system fixed the problem!
re: We finally have the answer!
@vaporeon_ anyway i’m curious what was broken and how you fixed it