[PATCH v6 4/4] x86/mce: Add support for Extended Physical Address MCA changes

From: Yazen Ghannam
Date: Tue Dec 06 2022 - 12:37:31 EST


From: Smita Koralahalli <Smita.KoralahalliChannabasappa@xxxxxxx>

Newer AMD CPUs support more physical address bits.

That is the MCA_ADDR registers on Scalable MCA systems contain the
ErrorAddr in bits [56:0] instead of [55:0]. Hence the existing LSB field
from bits [61:56] in MCA_ADDR must be moved around to accommodate the
larger ErrorAddr size.

MCA_CONFIG[McaLsbInStatusSupported] indicates this change. If set, the
LSB field will be found in MCA_STATUS rather than MCA_ADDR.

Each logical CPU has unique MCA bank in hardware and is not shared with
other logical CPUs. Additionally on SMCA systems, each feature bit may be
different for each bank within same logical CPU.

Check for MCA_CONFIG[McaLsbInStatusSupported] for each MCA bank and for
each CPU.

Additionally, all MCA banks do not support maximum ErrorAddr bits in
MCA_ADDR. Some banks might support fewer bits but the remaining bits are
marked as reserved.

[Yazen: Rebased and fixed up formatting.]

Signed-off-by: Smita Koralahalli <Smita.KoralahalliChannabasappa@xxxxxxx>
Signed-off-by: Yazen Ghannam <yazen.ghannam@xxxxxxx>
---

Link:
https://lore.kernel.org/r/20220412154038.261750-3-Smita.KoralahalliChannabasappa@xxxxxxx

v2:
Declared lsb_in_status in existing mce_bank[] struct.
Moved struct mce_bank[] declaration from core.c -> internal.h
v3:
Rebased on the latest tip tree. No functional changes.
v4:
No change.
v5:
Extend comment for smca_extract_err_addr if AddrLsb is found in
MCA_STATUS registers.
v6:
Rebase and fix up formatting.

arch/x86/kernel/cpu/mce/amd.c | 2 ++
arch/x86/kernel/cpu/mce/core.c | 8 +-------
arch/x86/kernel/cpu/mce/internal.h | 31 +++++++++++++++++++++++++++++-
3 files changed, 33 insertions(+), 8 deletions(-)

diff --git a/arch/x86/kernel/cpu/mce/amd.c b/arch/x86/kernel/cpu/mce/amd.c
index 85977ca07825..d4ec9b3481b8 100644
--- a/arch/x86/kernel/cpu/mce/amd.c
+++ b/arch/x86/kernel/cpu/mce/amd.c
@@ -306,6 +306,8 @@ static void smca_configure(unsigned int bank, unsigned int cpu)
if ((low & BIT(5)) && !((high >> 5) & 0x3))
high |= BIT(5);

+ this_cpu_ptr(mce_banks_array)[bank].lsb_in_status = !!(low & BIT(8));
+
wrmsr(smca_config, low, high);
}

diff --git a/arch/x86/kernel/cpu/mce/core.c b/arch/x86/kernel/cpu/mce/core.c
index 757cc46298d3..8b67e0284564 100644
--- a/arch/x86/kernel/cpu/mce/core.c
+++ b/arch/x86/kernel/cpu/mce/core.c
@@ -67,13 +67,7 @@ DEFINE_PER_CPU(unsigned, mce_exception_count);

DEFINE_PER_CPU_READ_MOSTLY(unsigned int, mce_num_banks);

-struct mce_bank {
- u64 ctl; /* subevents to enable */
-
- __u64 init : 1, /* initialise bank? */
- __reserved_1 : 63;
-};
-static DEFINE_PER_CPU_READ_MOSTLY(struct mce_bank[MAX_NR_BANKS], mce_banks_array);
+DEFINE_PER_CPU_READ_MOSTLY(struct mce_bank[MAX_NR_BANKS], mce_banks_array);

#define ATTR_LEN 16
/* One object for each MCE bank, shared by all CPUs */
diff --git a/arch/x86/kernel/cpu/mce/internal.h b/arch/x86/kernel/cpu/mce/internal.h
index 6dcb94fe0f65..867bcf9ee424 100644
--- a/arch/x86/kernel/cpu/mce/internal.h
+++ b/arch/x86/kernel/cpu/mce/internal.h
@@ -177,6 +177,24 @@ struct mce_vendor_flags {

extern struct mce_vendor_flags mce_flags;

+struct mce_bank {
+ /* subevents to enable */
+ u64 ctl;
+
+ /* initialise bank? */
+ __u64 init : 1,
+
+ /*
+ * (AMD) MCA_CONFIG[McaLsbInStatusSupported]: This bit indicates
+ * the LSB field is found in MCA_STATUS, when set.
+ */
+ lsb_in_status : 1,
+
+ __reserved_1 : 62;
+};
+
+DECLARE_PER_CPU_READ_MOSTLY(struct mce_bank[MAX_NR_BANKS], mce_banks_array);
+
enum mca_msr {
MCA_CTL,
MCA_STATUS,
@@ -190,7 +208,10 @@ extern bool filter_mce(struct mce *m);
#ifdef CONFIG_X86_MCE_AMD
extern bool amd_filter_mce(struct mce *m);

-/* Extract [55:<lsb>] where lsb is the LS-*valid* bit of the address bits. */
+/*
+ * If MCA_CONFIG[McaLsbInStatusSupported] is set, extract ErrAddr in bits
+ * [56:0], else in bits [55:0] of MCA_ADDR.
+ */
static __always_inline void smca_extract_err_addr(struct mce *m)
{
u8 lsb;
@@ -198,6 +219,14 @@ static __always_inline void smca_extract_err_addr(struct mce *m)
if (!mce_flags.smca)
return;

+ if (this_cpu_ptr(mce_banks_array)[m->bank].lsb_in_status) {
+ lsb = (m->status >> 24) & 0x3f;
+
+ m->addr &= GENMASK_ULL(56, lsb);
+
+ return;
+ }
+
lsb = (m->addr >> 56) & 0x3f;

m->addr &= GENMASK_ULL(55, lsb);
--
2.34.1