On 12 November 2013 14:57, Paolo Bonzini <pbonz...@redhat.com> wrote: > Il 12/11/2013 15:09, Gleb Natapov ha scritto: >> On Tue, Nov 12, 2013 at 02:57:49PM +0100, Paolo Bonzini wrote: >>> Il 12/11/2013 14:23, Gleb Natapov ha scritto: >>>>> If -O0 does not do that, let's move debug builds to -O1. >>>> >>>> Why not enable dce with -fdce? >>> >>> First, because clang doesn't have fine-tuned optimization options (at >>> least I couldn't find them and -fdce doesn't work). >> >> -O1 then for clang. > > We can even test in configure for the exact optimizations we want, in > fact. But I think -O1 doesn't sacrifice debuggability that much:
I'm afraid I still don't see why you'd want to sacrifice it at all, when the alternative is "provide a three line stub function in a file we already have all the build machinery to compile in the config where it's needed". I just don't see why you'd worry about the fact that there's no longer a compile error if you try to call this obviously kvm specific function in a non-kvm-enabled code path, when we already have large numbers of kvm-specific functions that have stubs (and when in general, eg QOM APIs, we seem to be entirely happy to have things be runtime errors rather than compile time). To me the benefit seems entirely on the side of "have standard compliant code" rather than "attempt to find various odd workarounds for the fact that some compilers will correctly complain about our code". -- PMM
