https://gcc.gnu.org/bugzilla/show_bug.cgi?id=116035
--- Comment #7 from Sam James <sjames at gcc dot gnu.org> --- (In reply to Patrick O'Neill from comment #5) > (In reply to Sam James from comment #3) > > I assume this is a '[14/15 regression]', or does < 14 not support these > > instructions and you're using it just to say what's affected? If the latter, > > it's counter to the normal conventions and maybe better to use "Known to > > work" / "Known to fail". > > > > (This is a general question as I've noticed this with a lot of riscv bugs > > and I'm not sure.) > > Typically I use [14/15 Regression] if I'm sure it's a regression (ex. the > feature exists in prior versions and worked there). [14/15] is for cases > like this where the isa extension wasn't implemented before so it's not > technically a regression. How about [14/15 only] for those? > I like to err on the side of [14/15] if I'm not > sure if it's a regression. > If it's preferred I can remove the [14/15 <Regression>] from my titles and > start exclusively using known to work/known to fail or just use known to > work/fail in addition to the title. I don't want to make life difficult for you/your workflow, my only real concern is people copying it and misunderstanding the notation and/or bugs being mislabelled. Would the "only" compromise work for you?
