On Tue, May 05, 2015 at 11:59:50AM +0200, Thomas Huth wrote: > On Tue, 5 May 2015 11:30:49 +0200 > mreza...@redhat.com wrote: > > > From: Miroslav Rezanina <mreza...@redhat.com> > > > > Disabling CONFIG_PARALLEL cause build failure as commit 07dc788 factored > > out initialization to parallel_hds_isa_init that is not build. > > > > Make calling parallel_hds_isa_init depending on CONFIG_PARALLEL so it can > > be correctly disabled. > > > > Signed-off-by: Miroslav Rezanina <mreza...@redhat.com> > > --- > > hw/i386/pc.c | 2 ++ > > hw/mips/mips_fulong2e.c | 2 ++ > > hw/mips/mips_malta.c | 2 ++ > > hw/sparc64/sun4u.c | 2 ++ > > 4 files changed, 8 insertions(+) > > > > diff --git a/hw/i386/pc.c b/hw/i386/pc.c > > index a8e6be1..560464e 100644 > > --- a/hw/i386/pc.c > > +++ b/hw/i386/pc.c > > @@ -1465,7 +1465,9 @@ void pc_basic_device_init(ISABus *isa_bus, qemu_irq > > *gsi, > > } > > > > serial_hds_isa_init(isa_bus, MAX_SERIAL_PORTS); > > +#ifdef CONFIG_PARALLEL > > parallel_hds_isa_init(isa_bus, MAX_PARALLEL_PORTS); > > +#endif > > Not sure, but is this pre-processor macro really defined if > CONFIG_PARALLEL has been set in the makefile? I've hit some similar > problem in the past and I had to discover that only the CONFIG_* > options from config-host.mak get translated into #defines, all the > others don't get translated. I might be wrong, but just to be sure, > could you please double-check that CONFIG_PARALLEL is #defined if it's > enabled in the .mak file? >
Yeah, you're right..I switch the behavior so thought of CONFIG_* from config-host.mak to be the one to not translated. And ifdef masked the issue. This patch is not working and has to be redone. > If not: Where does the build break exactly? Does it fail for all three > types, i386, mips and sun? > I can't test mips/sun build but for i386 it fails on linking where parallel_hds_isa_init does not exists. Mirek > Thomas >