[tip:ras/core] x86/mce/AMD, EDAC/mce_amd: Define and use tables for known SMCA IP types

From: tip-bot for Yazen Ghannam
Date: Tue Sep 13 2016 - 09:31:24 EST


Commit-ID: 5896820e0aa32572ad03b30563c539655b6c6375
Gitweb: http://git.kernel.org/tip/5896820e0aa32572ad03b30563c539655b6c6375
Author: Yazen Ghannam <Yazen.Ghannam@xxxxxxx>
AuthorDate: Mon, 12 Sep 2016 09:59:34 +0200
Committer: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
CommitDate: Tue, 13 Sep 2016 15:23:10 +0200

x86/mce/AMD, EDAC/mce_amd: Define and use tables for known SMCA IP types

Scalable MCA defines a number of IP types. An MCA bank on an SMCA
system is defined as one of these IP types. A bank's type is uniquely
identified by the combination of the HWID and MCATYPE values read from
its MCA_IPID register.

Add the required tables in order to be able to lookup error descriptions
based on a bank's type and the error's extended error code.

[ bp: Align comments, simplify a bit. ]

Signed-off-by: Yazen Ghannam <Yazen.Ghannam@xxxxxxx>
Signed-off-by: Borislav Petkov <bp@xxxxxxx>
Link: http://lkml.kernel.org/r/1472741832-1690-1-git-send-email-Yazen.Ghannam@xxxxxxx
Signed-off-by: Thomas Gleixner <tglx@xxxxxxxxxxxxx>

---
arch/x86/include/asm/mce.h | 61 +++++------
arch/x86/kernel/cpu/mcheck/mce_amd.c | 104 ++++++++++++++-----
drivers/edac/mce_amd.c | 194 +++++++----------------------------
3 files changed, 147 insertions(+), 212 deletions(-)

diff --git a/arch/x86/include/asm/mce.h b/arch/x86/include/asm/mce.h
index 21bc5a3..9bd7ff5 100644
--- a/arch/x86/include/asm/mce.h
+++ b/arch/x86/include/asm/mce.h
@@ -337,44 +337,47 @@ extern void apei_mce_report_mem_error(int corrected,
* Scalable MCA.
*/
#ifdef CONFIG_X86_MCE_AMD
-enum amd_ip_types {
- SMCA_F17H_CORE = 0, /* Core errors */
- SMCA_DF, /* Data Fabric */
- SMCA_UMC, /* Unified Memory Controller */
- SMCA_PB, /* Parameter Block */
- SMCA_PSP, /* Platform Security Processor */
- SMCA_SMU, /* System Management Unit */
- N_AMD_IP_TYPES
-};
-
-struct amd_hwid {
- const char *name;
- unsigned int hwid;
-};

-extern struct amd_hwid amd_hwids[N_AMD_IP_TYPES];
-
-enum amd_core_mca_blocks {
+/* These may be used by multiple smca_hwid_mcatypes */
+enum smca_bank_types {
SMCA_LS = 0, /* Load Store */
SMCA_IF, /* Instruction Fetch */
- SMCA_L2_CACHE, /* L2 cache */
- SMCA_DE, /* Decoder unit */
- RES, /* Reserved */
- SMCA_EX, /* Execution unit */
+ SMCA_L2_CACHE, /* L2 Cache */
+ SMCA_DE, /* Decoder Unit */
+ SMCA_EX, /* Execution Unit */
SMCA_FP, /* Floating Point */
- SMCA_L3_CACHE, /* L3 cache */
- N_CORE_MCA_BLOCKS
+ SMCA_L3_CACHE, /* L3 Cache */
+ SMCA_CS, /* Coherent Slave */
+ SMCA_PIE, /* Power, Interrupts, etc. */
+ SMCA_UMC, /* Unified Memory Controller */
+ SMCA_PB, /* Parameter Block */
+ SMCA_PSP, /* Platform Security Processor */
+ SMCA_SMU, /* System Management Unit */
+ N_SMCA_BANK_TYPES
+};
+
+struct smca_bank_name {
+ const char *name; /* Short name for sysfs */
+ const char *long_name; /* Long name for pretty-printing */
};

-extern const char * const amd_core_mcablock_names[N_CORE_MCA_BLOCKS];
+extern struct smca_bank_name smca_bank_names[N_SMCA_BANK_TYPES];
+
+#define HWID_MCATYPE(hwid, mcatype) ((hwid << 16) | mcatype)

-enum amd_df_mca_blocks {
- SMCA_CS = 0, /* Coherent Slave */
- SMCA_PIE, /* Power management, Interrupts, etc */
- N_DF_BLOCKS
+struct smca_hwid_mcatype {
+ unsigned int bank_type; /* Use with smca_bank_types for easy indexing. */
+ u32 hwid_mcatype; /* (hwid,mcatype) tuple */
+ u32 xec_bitmap; /* Bitmap of valid ExtErrorCodes; current max is 21. */
};

-extern const char * const amd_df_mcablock_names[N_DF_BLOCKS];
+struct smca_bank_info {
+ struct smca_hwid_mcatype *type;
+ u32 type_instance;
+};
+
+extern struct smca_bank_info smca_banks[MAX_NR_BANKS];
+
#endif

#endif /* _ASM_X86_MCE_H */
diff --git a/arch/x86/kernel/cpu/mcheck/mce_amd.c b/arch/x86/kernel/cpu/mcheck/mce_amd.c
index 9da92fb..3b74b62 100644
--- a/arch/x86/kernel/cpu/mcheck/mce_amd.c
+++ b/arch/x86/kernel/cpu/mcheck/mce_amd.c
@@ -63,34 +63,55 @@ static const char * const th_names[] = {
"execution_unit",
};

-/* Define HWID to IP type mappings for Scalable MCA */
-struct amd_hwid amd_hwids[] = {
- [SMCA_F17H_CORE] = { "f17h_core", 0xB0 },
- [SMCA_DF] = { "data_fabric", 0x2E },
- [SMCA_UMC] = { "umc", 0x96 },
- [SMCA_PB] = { "param_block", 0x5 },
- [SMCA_PSP] = { "psp", 0xFF },
- [SMCA_SMU] = { "smu", 0x1 },
+struct smca_bank_name smca_bank_names[] = {
+ [SMCA_LS] = { "load_store", "Load Store Unit" },
+ [SMCA_IF] = { "insn_fetch", "Instruction Fetch Unit" },
+ [SMCA_L2_CACHE] = { "l2_cache", "L2 Cache" },
+ [SMCA_DE] = { "decode_unit", "Decode Unit" },
+ [SMCA_EX] = { "execution_unit", "Execution Unit" },
+ [SMCA_FP] = { "floating_point", "Floating Point Unit" },
+ [SMCA_L3_CACHE] = { "l3_cache", "L3 Cache" },
+ [SMCA_CS] = { "coherent_slave", "Coherent Slave" },
+ [SMCA_PIE] = { "pie", "Power, Interrupts, etc." },
+ [SMCA_UMC] = { "umc", "Unified Memory Controller" },
+ [SMCA_PB] = { "param_block", "Parameter Block" },
+ [SMCA_PSP] = { "psp", "Platform Security Processor" },
+ [SMCA_SMU] = { "smu", "System Management Unit" },
};
-EXPORT_SYMBOL_GPL(amd_hwids);
-
-const char * const amd_core_mcablock_names[] = {
- [SMCA_LS] = "load_store",
- [SMCA_IF] = "insn_fetch",
- [SMCA_L2_CACHE] = "l2_cache",
- [SMCA_DE] = "decode_unit",
- [RES] = "",
- [SMCA_EX] = "execution_unit",
- [SMCA_FP] = "floating_point",
- [SMCA_L3_CACHE] = "l3_cache",
-};
-EXPORT_SYMBOL_GPL(amd_core_mcablock_names);
+EXPORT_SYMBOL_GPL(smca_bank_names);
+
+static struct smca_hwid_mcatype smca_hwid_mcatypes[] = {
+ /* { bank_type, hwid_mcatype, xec_bitmap } */
+
+ /* ZN Core (HWID=0xB0) MCA types */
+ { SMCA_LS, HWID_MCATYPE(0xB0, 0x0), 0x1FFFEF },
+ { SMCA_IF, HWID_MCATYPE(0xB0, 0x1), 0x3FFF },
+ { SMCA_L2_CACHE, HWID_MCATYPE(0xB0, 0x2), 0xF },
+ { SMCA_DE, HWID_MCATYPE(0xB0, 0x3), 0x1FF },
+ /* HWID 0xB0 MCATYPE 0x4 is Reserved */
+ { SMCA_EX, HWID_MCATYPE(0xB0, 0x5), 0x7FF },
+ { SMCA_FP, HWID_MCATYPE(0xB0, 0x6), 0x7F },
+ { SMCA_L3_CACHE, HWID_MCATYPE(0xB0, 0x7), 0xFF },
+
+ /* Data Fabric MCA types */
+ { SMCA_CS, HWID_MCATYPE(0x2E, 0x0), 0x1FF },
+ { SMCA_PIE, HWID_MCATYPE(0x2E, 0x1), 0xF },
+
+ /* Unified Memory Controller MCA type */
+ { SMCA_UMC, HWID_MCATYPE(0x96, 0x0), 0x3F },
+
+ /* Parameter Block MCA type */
+ { SMCA_PB, HWID_MCATYPE(0x05, 0x0), 0x1 },

-const char * const amd_df_mcablock_names[] = {
- [SMCA_CS] = "coherent_slave",
- [SMCA_PIE] = "pie",
+ /* Platform Security Processor MCA type */
+ { SMCA_PSP, HWID_MCATYPE(0xFF, 0x0), 0x1 },
+
+ /* System Management Unit MCA type */
+ { SMCA_SMU, HWID_MCATYPE(0x01, 0x0), 0x1 },
};
-EXPORT_SYMBOL_GPL(amd_df_mcablock_names);
+
+struct smca_bank_info smca_banks[MAX_NR_BANKS];
+EXPORT_SYMBOL_GPL(smca_banks);

static DEFINE_PER_CPU(struct threshold_bank **, threshold_banks);
static DEFINE_PER_CPU(unsigned int, bank_map); /* see which banks are on */
@@ -108,6 +129,36 @@ void (*deferred_error_int_vector)(void) = default_deferred_error_interrupt;
* CPU Initialization
*/

+static void get_smca_bank_info(unsigned int bank)
+{
+ unsigned int i, hwid_mcatype, cpu = smp_processor_id();
+ struct smca_hwid_mcatype *type;
+ u32 high, instanceId;
+ u16 hwid, mcatype;
+
+ /* Collect bank_info using CPU 0 for now. */
+ if (cpu)
+ return;
+
+ if (rdmsr_safe_on_cpu(cpu, MSR_AMD64_SMCA_MCx_IPID(bank), &instanceId, &high)) {
+ pr_warn("Failed to read MCA_IPID for bank %d\n", bank);
+ return;
+ }
+
+ hwid = high & MCI_IPID_HWID;
+ mcatype = (high & MCI_IPID_MCATYPE) >> 16;
+ hwid_mcatype = HWID_MCATYPE(hwid, mcatype);
+
+ for (i = 0; i < ARRAY_SIZE(smca_hwid_mcatypes); i++) {
+ type = &smca_hwid_mcatypes[i];
+ if (hwid_mcatype == type->hwid_mcatype) {
+ smca_banks[bank].type = type;
+ smca_banks[bank].type_instance = instanceId;
+ break;
+ }
+ }
+}
+
struct thresh_restart {
struct threshold_block *b;
int reset;
@@ -425,6 +476,9 @@ void mce_amd_feature_init(struct cpuinfo_x86 *c)
int offset = -1;

for (bank = 0; bank < mca_cfg.banks; ++bank) {
+ if (mce_flags.smca)
+ get_smca_bank_info(bank);
+
for (block = 0; block < NR_BLOCKS; ++block) {
address = get_block_address(cpu, address, low, high, bank, block);
if (!address)
diff --git a/drivers/edac/mce_amd.c b/drivers/edac/mce_amd.c
index ea549a9..99b3bf3 100644
--- a/drivers/edac/mce_amd.c
+++ b/drivers/edac/mce_amd.c
@@ -283,6 +283,27 @@ static const char * const smca_smu_mce_desc[] = {
"SMU RAM ECC or parity error",
};

+struct smca_mce_desc {
+ const char * const *descs;
+ unsigned int num_descs;
+};
+
+static struct smca_mce_desc smca_mce_descs[] = {
+ [SMCA_LS] = { smca_ls_mce_desc, ARRAY_SIZE(smca_ls_mce_desc) },
+ [SMCA_IF] = { smca_if_mce_desc, ARRAY_SIZE(smca_if_mce_desc) },
+ [SMCA_L2_CACHE] = { smca_l2_mce_desc, ARRAY_SIZE(smca_l2_mce_desc) },
+ [SMCA_DE] = { smca_de_mce_desc, ARRAY_SIZE(smca_de_mce_desc) },
+ [SMCA_EX] = { smca_ex_mce_desc, ARRAY_SIZE(smca_ex_mce_desc) },
+ [SMCA_FP] = { smca_fp_mce_desc, ARRAY_SIZE(smca_fp_mce_desc) },
+ [SMCA_L3_CACHE] = { smca_l3_mce_desc, ARRAY_SIZE(smca_l3_mce_desc) },
+ [SMCA_CS] = { smca_cs_mce_desc, ARRAY_SIZE(smca_cs_mce_desc) },
+ [SMCA_PIE] = { smca_pie_mce_desc, ARRAY_SIZE(smca_pie_mce_desc) },
+ [SMCA_UMC] = { smca_umc_mce_desc, ARRAY_SIZE(smca_umc_mce_desc) },
+ [SMCA_PB] = { smca_pb_mce_desc, ARRAY_SIZE(smca_pb_mce_desc) },
+ [SMCA_PSP] = { smca_psp_mce_desc, ARRAY_SIZE(smca_psp_mce_desc) },
+ [SMCA_SMU] = { smca_smu_mce_desc, ARRAY_SIZE(smca_smu_mce_desc) },
+};
+
static bool f12h_mc0_mce(u16 ec, u8 xec)
{
bool ret = false;
@@ -827,175 +848,32 @@ static void decode_mc6_mce(struct mce *m)
pr_emerg(HW_ERR "Corrupted MC6 MCE info?\n");
}

-static void decode_f17h_core_errors(const char *ip_name, u8 xec,
- unsigned int mca_type)
-{
- const char * const *error_desc_array;
- size_t len;
-
- pr_emerg(HW_ERR "%s Error: ", ip_name);
-
- switch (mca_type) {
- case SMCA_LS:
- error_desc_array = smca_ls_mce_desc;
- len = ARRAY_SIZE(smca_ls_mce_desc) - 1;
-
- if (xec == 0x4) {
- pr_cont("Unrecognized LS MCA error code.\n");
- return;
- }
- break;
-
- case SMCA_IF:
- error_desc_array = smca_if_mce_desc;
- len = ARRAY_SIZE(smca_if_mce_desc) - 1;
- break;
-
- case SMCA_L2_CACHE:
- error_desc_array = smca_l2_mce_desc;
- len = ARRAY_SIZE(smca_l2_mce_desc) - 1;
- break;
-
- case SMCA_DE:
- error_desc_array = smca_de_mce_desc;
- len = ARRAY_SIZE(smca_de_mce_desc) - 1;
- break;
-
- case SMCA_EX:
- error_desc_array = smca_ex_mce_desc;
- len = ARRAY_SIZE(smca_ex_mce_desc) - 1;
- break;
-
- case SMCA_FP:
- error_desc_array = smca_fp_mce_desc;
- len = ARRAY_SIZE(smca_fp_mce_desc) - 1;
- break;
-
- case SMCA_L3_CACHE:
- error_desc_array = smca_l3_mce_desc;
- len = ARRAY_SIZE(smca_l3_mce_desc) - 1;
- break;
-
- default:
- pr_cont("Corrupted MCA core error info.\n");
- return;
- }
-
- if (xec > len) {
- pr_cont("Unrecognized %s MCA bank error code.\n",
- amd_core_mcablock_names[mca_type]);
- return;
- }
-
- pr_cont("%s.\n", error_desc_array[xec]);
-}
-
-static void decode_df_errors(u8 xec, unsigned int mca_type)
-{
- const char * const *error_desc_array;
- size_t len;
-
- pr_emerg(HW_ERR "Data Fabric Error: ");
-
- switch (mca_type) {
- case SMCA_CS:
- error_desc_array = smca_cs_mce_desc;
- len = ARRAY_SIZE(smca_cs_mce_desc) - 1;
- break;
-
- case SMCA_PIE:
- error_desc_array = smca_pie_mce_desc;
- len = ARRAY_SIZE(smca_pie_mce_desc) - 1;
- break;
-
- default:
- pr_cont("Corrupted MCA Data Fabric info.\n");
- return;
- }
-
- if (xec > len) {
- pr_cont("Unrecognized %s MCA bank error code.\n",
- amd_df_mcablock_names[mca_type]);
- return;
- }
-
- pr_cont("%s.\n", error_desc_array[xec]);
-}
-
/* Decode errors according to Scalable MCA specification */
static void decode_smca_errors(struct mce *m)
{
- u32 addr = MSR_AMD64_SMCA_MCx_IPID(m->bank);
- unsigned int hwid, mca_type, i;
- u8 xec = XEC(m->status, xec_mask);
- const char * const *error_desc_array;
+ struct smca_hwid_mcatype *type;
+ unsigned int bank_type;
const char *ip_name;
- u32 low, high;
- size_t len;
+ u8 xec = XEC(m->status, xec_mask);

- if (rdmsr_safe(addr, &low, &high)) {
- pr_emerg(HW_ERR "Invalid IP block specified.\n");
+ if (m->bank >= ARRAY_SIZE(smca_banks))
return;
- }
-
- hwid = high & MCI_IPID_HWID;
- mca_type = (high & MCI_IPID_MCATYPE) >> 16;

- pr_emerg(HW_ERR "MC%d IPID value: 0x%08x%08x\n", m->bank, high, low);
-
- /*
- * Based on hwid and mca_type values, decode errors from respective IPs.
- * Note: mca_type values make sense only in the context of an hwid.
- */
- for (i = 0; i < ARRAY_SIZE(amd_hwids); i++)
- if (amd_hwids[i].hwid == hwid)
- break;
-
- switch (i) {
- case SMCA_F17H_CORE:
- ip_name = (mca_type == SMCA_L3_CACHE) ?
- "L3 Cache" : "F17h Core";
- return decode_f17h_core_errors(ip_name, xec, mca_type);
- break;
-
- case SMCA_DF:
- return decode_df_errors(xec, mca_type);
- break;
-
- case SMCA_UMC:
- error_desc_array = smca_umc_mce_desc;
- len = ARRAY_SIZE(smca_umc_mce_desc) - 1;
- break;
-
- case SMCA_PB:
- error_desc_array = smca_pb_mce_desc;
- len = ARRAY_SIZE(smca_pb_mce_desc) - 1;
- break;
-
- case SMCA_PSP:
- error_desc_array = smca_psp_mce_desc;
- len = ARRAY_SIZE(smca_psp_mce_desc) - 1;
- break;
-
- case SMCA_SMU:
- error_desc_array = smca_smu_mce_desc;
- len = ARRAY_SIZE(smca_smu_mce_desc) - 1;
- break;
-
- default:
- pr_emerg(HW_ERR "HWID:%d does not match any existing IPs.\n", hwid);
+ type = smca_banks[m->bank].type;
+ if (!type)
return;
- }

- ip_name = amd_hwids[i].name;
- pr_emerg(HW_ERR "%s Error: ", ip_name);
+ bank_type = type->bank_type;
+ ip_name = smca_bank_names[bank_type].long_name;

- if (xec > len) {
- pr_cont("Unrecognized %s MCA bank error code.\n", ip_name);
- return;
- }
+ pr_emerg(HW_ERR "%s Extended Error Code: %d\n", ip_name, xec);

- pr_cont("%s.\n", error_desc_array[xec]);
+ /* Only print the decode of valid error codes */
+ if (xec < smca_mce_descs[bank_type].num_descs &&
+ (type->xec_bitmap & BIT_ULL(xec))) {
+ pr_emerg(HW_ERR "%s Error: ", ip_name);
+ pr_cont("%s.\n", smca_mce_descs[bank_type].descs[xec]);
+ }
}

static inline void amd_decode_err_code(u16 ec)