On 17/06/2020 13.31, Gerd Hoffmann wrote: > On Wed, Jun 17, 2020 at 10:40:24AM +0200, Thomas Huth wrote: >> On 17/06/2020 09.11, Gerd Hoffmann wrote: >>> DSDT change: isa device order changes in case MI1 (ipmi) is present. >>> >>> Signed-off-by: Gerd Hoffmann <kra...@redhat.com> >>> Reviewed-by: Igor Mammedov <imamm...@redhat.com> >>> --- >>> hw/block/fdc.c | 83 ++++++++++++++++++++++++++++++++ >>> hw/i386/acpi-build.c | 83 -------------------------------- >>> stubs/cmos.c | 7 +++ >>> stubs/Makefile.objs | 1 + >>> tests/data/acpi/pc/DSDT.ipmikcs | Bin 5086 -> 5086 bytes >>> 5 files changed, 91 insertions(+), 83 deletions(-) >>> create mode 100644 stubs/cmos.c >> >> Hmm, not all targets that use CONFIG_FDC also have ACPI ... would it be >> possible to move the ACPI-related code into a new file, say fdc-acpi.c, >> instead and only compile that if both, CONFIG_FDC and CONFIG_ACPI are >> enabled? > > Possible sure, but is that really worth it? It isn't that much > (possibly dead) code, and we have stubs for the aml_*() functions so it > builds just fine with CONFIG_ACPI=n ...
Ok, fine for me, too. It's really not that much code, indeed. Thomas
