Re: sparc64 vs. gcc 4.8
Masao Uebayashi <[email protected]>
| Newsgroups | gmane.os.netbsd.ports.sparc64 |
|---|---|
| Message-ID | <CADbF7efZJwZoCHQ_fNXVkL85jtCFPuoAM2XSbDQ2DTVGvYZjqg@mail.gmail.com> |
Compiler switch is such a big change. This not only annoys conservative engineers who closely check generated code, but also affects "forthcoming" binary patches. Clear design policy with respect to release engineering will be needed later. On Sun, Oct 5, 2014 at 8:17 AM, Erik Fair <[email protected]> wrote: > > On Oct 4, 2014, at 00:44, Martin Husemann <[email protected]> wrote: > >> On Fri, Oct 03, 2014 at 01:54:05PM -0700, Erik Fair wrote: >>> This suggests a need for "elfcmp" or "objcmp" or equivalent, to >>> compare like ELF sections (text, data, bss, symbols) in object files >>> against each other, while excluding sections that would always be >>> different (e.g. notes containing time stamps, compiler versions). >> >> This does not make a lot of sense for such different compiler versions >> involved. > > So you’re saying that the code generation has changed so much between those versions that few if any object modules will have the same instructions in the same order, despite identical source input? > > Erik <[email protected]> >