On 28 December 2017 at 14:58, Igor Mammedov <imammedo@xxxxxxxxxx> wrote: > On Thu, 28 Dec 2017 13:54:17 +0800 > Dongjiu Geng <gengdongjiu@xxxxxxxxxx> wrote: > >> Support this feature since version 2.10, disable it by >> default in the old version. > patch should go before acpi tables are actually added, > otherwise it might break bisectability. > >> >> Signed-off-by: Dongjiu Geng <gengdongjiu@xxxxxxxxxx> >> --- >> Address Shannon's comments to add platform version in [1]. >> >> [1]: https://lkml.org/lkml/2017/8/25/821 >> >> Signed-off-by: Dongjiu Geng <gengdongjiu@xxxxxxxxxx> >> --- >> hw/arm/virt-acpi-build.c | 14 +++++++++----- >> hw/arm/virt.c | 4 ++++ >> include/hw/arm/virt.h | 1 + >> 3 files changed, 14 insertions(+), 5 deletions(-) >> >> diff --git a/hw/arm/virt-acpi-build.c b/hw/arm/virt-acpi-build.c >> index 06c14b3..b6974ef 100644 >> --- a/hw/arm/virt-acpi-build.c >> +++ b/hw/arm/virt-acpi-build.c >> @@ -801,10 +801,11 @@ void virt_acpi_build(VirtMachineState *vms, AcpiBuildTables *tables) >> acpi_add_table(table_offsets, tables_blob); >> build_spcr(tables_blob, tables->linker, vms); >> >> - acpi_add_table(table_offsets, tables_blob); >> - build_hardware_error_table(tables->hardware_errors, tables->linker); >> - build_apei_ghes(tables_blob, tables->hardware_errors, tables->linker); >> - >> + if (!vmc->no_ras) { > > it's better to avoid no_foo, use something like > > vmc->has_ras Fields in VirtMachineClass for this kind of thing tend to end up having to be no_foo, because the default (false) must be the setting for the most up to date version of the virt board, because of the way the virt_machine_X_XX_options() functions chain together. So no_ras matches the sense used for all the existing bools in VirtMachineClass. (In contrast, bools in the VirtMachineState struct are the conventional sense, so there we have highmem/its/virt/etc.) thanks -- PMM
