Re: final (?) issue with darwin-x86

[Jonas Hahnfeld]

Am Mittwoch, den 08.04.2020, 20:13 +0200 schrieb Jonas Hahnfeld:
> Phil has hit another problem when building for darwin-x86:
> 
> Making flower/out/library.a 
> ar: `u' modifier ignored since `D' is the default (see `U')
> i686-apple-darwin8-ranlib: archive: out/library.a truncated or malformed 
> (archive header of next member extends past the end of the file)
> ar: out/library.a: Malformed archive
> .../gub/target/darwin-x86/root/usr/cross/bin/i686-apple-darwin8-ld: truncated 
> or malformed archive: ../flower/./out/library.a at offset 565244 (member 
> extends past the end of the file, can't load from it)
> collect2: error: ld returned 1 exit status
> make[1]: *** [out/lilypond] Error 1
> make: *** [all] Error 2
> 
> I looked into this and I think this happens if running 'ar' on the
> result of 'i686-apple-darwin8-ranlib'. I think the following should
> solve the issue:
> http://git.savannah.gnu.org/cgit/lilypond.git/commit/?h=dev/hahnjo/release-unstable
> 
> I'm currently running that branch through GUB, let's see what happens.

$ bin/gub --platform=linux-64 
darwin-x86::'git://git.sv.gnu.org/lilypond.git?branch=dev/hahnjo/release-unstable'
succeeded and I had already verified that it works in my usual (native)
build setup. Can't test the other platforms myself because building all
the dependencies would take another few hours.

Phil, would you want to give it another try today? The commit is still
in a separate branch, so I'll need to push into release/unstable first.

Jonas

signature.asc
Description: This is a digitally signed message part