On Wed, 20 Dec 2023, Suma Hegde wrote:
> AMD EPYC family 0x1A and Model 0x0-0xF are having different
> mailbox message ID offset compared to previous
> platforms. In case of ACPI based BIOS, this information will be read
> from ACPI table, for non ACPI BIOS, this needs to be #defined.
non-ACPI
> Signed-off-by: Suma Hegde <suma.hegde@xxxxxxx>
> Signed-off-by: Naveen Krishna Chatradhi <nchatrad@xxxxxxx>
> Reviewed-by: Hans de Goede <hdegoede@xxxxxxxxxx>
>
> Changes since v1:
> Add "Reviewed-by: Hans de Goede <hdegoede@xxxxxxxxxx>"
>
> ---
> drivers/platform/x86/amd/hsmp.c | 21 ++++++++++++++++++++-
> 1 file changed, 20 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/platform/x86/amd/hsmp.c b/drivers/platform/x86/amd/hsmp.c
> index 726ba73e23dd..95afea56d333 100644
> --- a/drivers/platform/x86/amd/hsmp.c
> +++ b/drivers/platform/x86/amd/hsmp.c
> @@ -44,6 +44,7 @@
> */
> #define SMN_HSMP_BASE 0x3B00000
> #define SMN_HSMP_MSG_ID 0x0010534
> +#define SMN_HSMP_MSG_ID_F1A_M0H 0x0010934
> #define SMN_HSMP_MSG_RESP 0x0010980
> #define SMN_HSMP_MSG_DATA 0x00109E0
>
> @@ -712,6 +713,15 @@ static int hsmp_cache_proto_ver(u16 sock_ind)
> return ret;
> }
>
> +static inline bool is_f1a_m0h(void)
> +{
> + if (boot_cpu_data.x86 == 0x1A &&
> + (boot_cpu_data.x86_model >= 0x00 && boot_cpu_data.x86_model <= 0x0F))
> + return true;
> +
> + return false;
> +}
> +
> static int initialize_platdev(struct device *dev)
> {
> int ret, i;
> @@ -723,7 +733,16 @@ static int initialize_platdev(struct device *dev)
> plat_dev.sock[i].sock_ind = i;
> plat_dev.sock[i].dev = dev;
> plat_dev.sock[i].mbinfo.base_addr = SMN_HSMP_BASE;
> - plat_dev.sock[i].mbinfo.msg_id_off = SMN_HSMP_MSG_ID;
> +
> + /*
> + * This is a tranisitional change from non ACPI to ACPI, only
> + * family 0x1A, model 0x00 platform is supported for both ACPI and non ACPI.
non-ACPI x2
> + */
> + if (is_f1a_m0h())
> + plat_dev.sock[i].mbinfo.msg_id_off = SMN_HSMP_MSG_ID_F1A_M0H;
> + else
> + plat_dev.sock[i].mbinfo.msg_id_off = SMN_HSMP_MSG_ID;
You could have made all these lines shorter if you'd have added an local
variable for sock = &plat_dev.sock[i] (in the earlier patch where this was
introduced).
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@xxxxxxxxxxxxxxx>
--
i.
