On 25/01/19 15:49, Thomas Huth wrote: > On 2019-01-25 11:06, Paolo Bonzini wrote: >> ARM does not have an ISA bus, this device should not be enabled. Kconfig >> catches the contradiction because ARM does not define CONFIG_ISA_BUS. > > That description is a little bit wrong, since ARM includes pci.mak, and > pci.mak enables CONFIG_ISA_BUS=y for the IDE code.
Right: ARM should not have an ISA bus, this device should not be enabled. Kconfig allows to clean up the dependencies and remove CONFIG_ISA_BUS=y from ARM, and then catches a contradiction between the hardcoded CONFIG_SERIAL_ISA=y and CONFIG_ISA_BUS=n. Paolo >> Signed-off-by: Paolo Bonzini <pbonz...@redhat.com> >> --- >> default-configs/arm-softmmu.mak | 1 - >> 1 file changed, 1 deletion(-) >> >> diff --git a/default-configs/arm-softmmu.mak >> b/default-configs/arm-softmmu.mak >> index 2420491..3c9351f 100644 >> --- a/default-configs/arm-softmmu.mak >> +++ b/default-configs/arm-softmmu.mak >> @@ -6,7 +6,6 @@ CONFIG_VGA=y >> CONFIG_NAND=y >> CONFIG_ECC=y >> CONFIG_SERIAL=y >> -CONFIG_SERIAL_ISA=y >> CONFIG_PTIMER=y >> CONFIG_SD=y >> CONFIG_MAX7310=y > > Anyway, looking through the code and the description of commit > 1401c322c8d431fa750cff8276 I think you're right that CONFIG_SERIAL_ISA > is not required for ARM, so if you change the commit message a little bit: > > Reviewed-by: Thomas Huth <th...@redhat.com> >
