https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61577
--- Comment #101 from dave.anglin at bell dot net --- On 2019-07-25 8:46 a.m., cameron.heide at betasystems dot com wrote: > Looks like it's PHSS_34860, which is probably pretty old (it's also 11.23). In > case they're also relevant, it's using binutils 2.32 for the specified > assembler, and gcc-4.7.4 as the 'native' compiler, previously built per the > advice in PR64919. GMP, MPFR, MPC, and ISL were prebuilt with an even older > compiler (gcc-4.1.1) since some of the self-tests crashed when built with > 4.7.4. Can 4.7.4 rebuild itself? It not, it's probably somewhat broken. Need to figure out whether it is good enough to build stage1 at -O0. I think the early faults at -O2/-O1 should be looked at to see if they are problems in current code or problems with 4.7.4. Generally, we need preprocessed source and information about the fault (e.g., is it another pointer extend problem, etc). If the problem is a regression, then a regression hunt is in order. > > The patch set is pretty similar to yours, though a bit cruder (I just #if 0 > out > the nftw call) and I just have a minor one from PR67022 that you didn't, to > make DWARF2 the default. DWARF2 is the way to go. However, some work may be needed to provide unwind support.
