Created attachment 711735 [details]
Firefox build.log without fix
I was failing to build Firefox for a while thinking it will get sorted out with an update. But it did not. While reading the Clang Gentoo Wiki , I learned about building packages with clang and they suggest the following env update in a per-package basis:
LDFLAGS="-fuse-ld=lld -rtlib=compiler-rt -unwindlib=libunwind"
After I did this for Firefox, it could compile properly.
Please find attached the build.log of Firefox without the fix above (so emerged as-is without any change), called `firefox-build-fail.7z`.
Created attachment 711738 [details]
Created attachment 711744 [details]
Forgot the put the link for the Clang wiki 
Created attachment 711780 [details]
(please don't use 7z for logs, re-attaching your log using xz for convenience)
Noted! I will use `xz` to compress logs from now on.
9:15.60 ld.lld: error: undefined symbol: _Unwind_Resume
Are you using default-* flags on sys-devel/clang by any chances? "Normally" setting anything clang-related isn't needed for firefox[clang]
`emerge -pvq sys-devel/clang` output might help
Probably(?) a duplicate of bug #791430
(In reply to Adel KARA SLIMANE from comment #5)
> Noted! I will use `xz` to compress logs from now on.
Well, only when logs are over 1MB though, leave uncompressed otherwise.
> Are you using default-* flags on sys-devel/clang by any chances? "Normally" setting anything clang-related isn't needed for firefox[clang]
Yes I do have `default-compiler-rt default-lld` USE flags enabled. Now that you bring it up this must be the culprit !
[ebuild R ] sys-devel/clang-12.0.0-r1 USE="default-compiler-rt default-lld static-analyzer xml -debug -default-libcxx -doc -test" ABI_X86="(64) -32 (-x32)" LLVM_TARGETS="AMDGPU BPF NVPTX (X86) -AArch64 -ARC -ARM -AVR (-CSKY) -Hexagon -Lanai -MSP430 -Mips -PowerPC -RISCV -Sparc -SystemZ -VE -WebAssembly -XCore" PYTHON_SINGLE_TARGET="python3_8 -python3_7 -python3_9"
> Probably(?) a duplicate of bug #791430
It really seems like it! And I seem to have more information on the matter, then. I will post there my fix and see if it fixes his issue.
> Well, only when logs are over 1MB though, leave uncompressed otherwise.
*** Bug 792675 has been marked as a duplicate of this bug. ***