On Nov 24, 1:02pm, [email protected] ("William D. Jones") wrote: -- Subject: Re: NetBSD/i386 7.0 hangs booting 486
| Was this a regression from the problems that I had in February? Or did | gcc/llvm make changes to their code that assume CPUID exists | unconditionally? I seem to recall the problem last time was i386-specific | code in the NetBSD source tree proper unconditionally trying to call CPUID. | I should've been updating my kernel more frequently :/, but alas my time | management isn't great. Yes, it was the same problem, and your report helped fixing it :-) | LLVM doesn't support anything lower than the Pentium, so I'm unsure if | they'll accept the patch :(. I know, but they still have code that supports i386 and they follow the same API's as gcc... So they are 1/2 broken: http://nxr.netbsd.org/xref/src/external/bsd/llvm/dist/clang/lib/Headers/cpuid.h#184 The check for level is missing here: http://nxr.netbsd.org/xref/src/external/bsd/llvm/dist/clang/lib/Headers/cpuid.h#177 And the bug report for gcc: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68491 christos
