On 14 July 2016 at 15:27, Aaron Lindsay <alind...@codeaurora.org> wrote: > On Jul 14 14:23, Peter Maydell wrote: >> On 14 July 2016 at 14:15, Paolo Bonzini <pbonz...@redhat.com> wrote: >> > On 14/07/2016 15:13, Aaron Lindsay wrote: >> >> I'm configuring with: >> >> # ./configure \ >> >> --static \ >> >> --disable-gtk \ >> >> --target-list=aarch64-softmmu >> >> > Hmm, it's possible that we have to disable ifunc together with --static. >> >> I'm still tempted to say we should just forbid building the softmmu >> binaries with --static, unless somebody has a serious use case for it... > > FWIW, we do find it convenient to be able to compile one binary capable > of running on multiple systems with different (and incompatible) library > versions.
The difficulty here is that the system emulators may call functions in glibc which do dynamic resolution anyway, and so "require the shared libraries from the glibc version used for linking", to quote the linker warning. For the user-only binaries we can claim we don't make those function calls, but for softmmu we do. thanks -- PMM
