https://gcc.gnu.org/bugzilla/show_bug.cgi?id=124266
--- Comment #11 from Wilco <wilco at gcc dot gnu.org> --- (In reply to Richard Sandiford from comment #10) > (In reply to Wilco from comment #9) > > Eg. GCC13 generates: https://godbolt.org/z/qexq61E47 > > > > You can clearly see that the cfi notes are totally wrong even without the > > new ldp pass... > That's the intended behaviour though. The ABI is that the extra register > state is only preserved by normal returns, not by exceptional returns. The > .eh_frame information therefore only describes the base PCS set. Look at the notes, they don't even match up. So even if it didn't correctly implement asynchronous unwind (which we turn on by default), it is still totally wrong.
