On Fri, Mar 17, 2023 at 2:00 PM Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> wrote: > > On Fri, Mar 17, 2023 at 1:51 PM Linus Torvalds > <torvalds@xxxxxxxxxxxxxxxxxxxx> wrote: > > > > but I would *really* like clang to be fixed to not silently generate > > code that does insane things and would be basically impossible to > > debug if it ever triggers. > > Side note: the key word here is "silently". > > If clang notices that it generates crazy code, a warning at build-time > would be preferable to the "oh, we noticed the crazy code generation > because we do sanity checking that just happened to catch it". That's fair. I have something hacked up locally that can spot the fallthough from m5mols_set_fmt() as objtool did. With some polish, we can likely ship that as a compiler warning. Then we can have these checks regardless of objtool arch support. First I need to teach LLVM that __stack_chk_fail is noreturn, though I've only verified that thus far in glibc, musl, and bionic; I still need to check that's the case for the BSDs' libcs. -- Thanks, ~Nick Desaulniers
