https://sourceware.org/bugzilla/show_bug.cgi?id=25732
--- Comment #8 from Rainer Orth <ro at CeBiTec dot Uni-Bielefeld.DE> --- > --- Comment #6 from H.J. Lu <hjl.tools at gmail dot com> --- > I stopped checking Solaris cross target since the result isn't clean. > If Solaris cross target test becomes clean, I will test Solaris cross I doubt I will be able to do much more than the occasional bugfix in binutils. In particular, the bfd and ld code bases remain a complete mystery to me. > target again. BTW, you can disable ld for Solaris target by default > so that ld test won't run by default. No point in that: gld is by far good enough on Solaris to bootstrap and test gcc. It's certainly good to have it for comparisons when investigating issues with the Solaris ld. Looking at your list of failing ld tests FAIL: Build pr20995-2.so FAIL: pr20995-2 FAIL: ld-ifunc/ifunc-23a-x86 FAIL: ld-ifunc/ifunc-24a-x86 FAIL: ld-ifunc/ifunc-25a-x86 the last three are highly dubious: Solaris ld.so.1 doesn't support ifunc and never will, so I'm unsure if those tests can produce useful results. Certainly not if they are runtime tests. That leaves us with the two pr20995-2 ones... -- You are receiving this mail because: You are on the CC list for the bug.