Use trace interface to elaborate all H/W error related
information.
Signed-off-by: Chen, Gong <gong.chen@xxxxxxxxxxxxxxx>
---
drivers/acpi/Kconfig | 7 ++-
drivers/acpi/Makefile | 4 ++
drivers/acpi/acpi_extlog.c | 28 +++++++++++-
drivers/acpi/apei/cper.c | 13 ++++--
drivers/acpi/debug_extlog.h | 16 +++++++
drivers/acpi/extlog_trace.c | 105 ++++++++++++++++++++++++++++++++++++++++++++
drivers/acpi/extlog_trace.h | 77 ++++++++++++++++++++++++++++++++
include/linux/cper.h | 2 +
8 files changed, 246 insertions(+), 6 deletions(-)
create mode 100644 drivers/acpi/debug_extlog.h
create mode 100644 drivers/acpi/extlog_trace.c
create mode 100644 drivers/acpi/extlog_trace.h
diff --git a/drivers/acpi/Kconfig b/drivers/acpi/Kconfig
index 1465fa8..9ea343e 100644
--- a/drivers/acpi/Kconfig
+++ b/drivers/acpi/Kconfig
@@ -372,12 +372,17 @@ config ACPI_BGRT
source "drivers/acpi/apei/Kconfig"
+config EXTLOG_TRACE
+ def_bool n
+
config ACPI_EXTLOG
tristate "Extended Error Log support"
depends on X86 && X86_MCE
+ select EXTLOG_TRACE
default n
help
This driver adds support for decoding extended errors from hardware.
- which allows the operating system to obtain data from trace.
+ which allows the operating system to obtain data from trace. It will
+ appear under /sys/kernel/debug/tracing/ras/ .
endif # ACPI
diff --git a/drivers/acpi/Makefile b/drivers/acpi/Makefile
index bce34af..a6e41b7 100644
--- a/drivers/acpi/Makefile
+++ b/drivers/acpi/Makefile
@@ -83,4 +83,8 @@ obj-$(CONFIG_ACPI_PROCESSOR_AGGREGATOR) += acpi_pad.o
obj-$(CONFIG_ACPI_APEI) += apei/
+# extended log support
+acpi-$(CONFIG_EXTLOG_TRACE) += extlog_trace.o
+CFLAGS_extlog_trace.o := -I$(src)
+
obj-$(CONFIG_ACPI_EXTLOG) += acpi_extlog.o
diff --git a/drivers/acpi/acpi_extlog.c b/drivers/acpi/acpi_extlog.c
index 3e3e286..ca51eb0 100644
--- a/drivers/acpi/acpi_extlog.c
+++ b/drivers/acpi/acpi_extlog.c
@@ -26,6 +26,7 @@
#include <asm/mce.h>
#include "apei/apei-internal.h"
+#include "debug_extlog.h"
#define EXT_ELOG_ENTRY_MASK 0xfffffffffffff /* elog entry address mask */
@@ -55,6 +56,8 @@ struct extlog_l1_head {
static u8 extlog_dsm_uuid[] = "663E35AF-CC10-41A4-88EA-5470AF055295";
+static const uuid_le invalid_uuid = NULL_UUID_LE;
+
/* L1 table related physical address */
static u64 elog_base;
static size_t elog_size;
@@ -143,7 +146,12 @@ static int print_extlog_rcd(const char *pfx,
static int extlog_print(const char *pfx, int cpu, int bank)
{
- struct acpi_generic_status *estatus;
+ struct acpi_generic_status *estatus, *tmp;
+ struct acpi_generic_data *gdata;
+ const uuid_le *fru_id = &invalid_uuid;
+ char *fru_text = "";
+ uuid_le *sec_type;
+ static u64 err_count;
int rc;
estatus = extlog_elog_entry_check(cpu, bank);
@@ -154,7 +162,23 @@ static int extlog_print(const char *pfx, int cpu, int bank)
/* clear record status to enable BIOS to update it again */
estatus->block_status = 0;
- rc = print_extlog_rcd(pfx, (struct acpi_generic_status *)elog_buf, cpu);
+ tmp = (struct acpi_generic_status *)elog_buf;
+ gdata = (struct acpi_generic_data *)(tmp + 1);
+ rc = print_extlog_rcd(pfx, tmp, cpu);
+
+ /* trace extended error log */
+ err_count++;
+ if (gdata->validation_bits & CPER_SEC_VALID_FRU_ID)
+ fru_id = (uuid_le *)gdata->fru_id;
+ if (gdata->validation_bits & CPER_SEC_VALID_FRU_TEXT)
+ fru_text = gdata->fru_text;
+ sec_type = (uuid_le *)gdata->section_type;
+ if (!uuid_le_cmp(*sec_type, CPER_SEC_PLATFORM_MEM)) {
+ struct cper_sec_mem_err *mem_err = (void *)(gdata + 1);
+ if (gdata->error_data_length >= sizeof(*mem_err))
+ trace_mem_error(fru_id, fru_text, err_count,
+ gdata->error_severity, mem_err);
+ }
return rc;
}
diff --git a/drivers/acpi/apei/cper.c b/drivers/acpi/apei/cper.c
index 567410e..0b4cfad 100644
--- a/drivers/acpi/apei/cper.c
+++ b/drivers/acpi/apei/cper.c
@@ -56,11 +56,12 @@ static const char *cper_severity_strs[] = {
"info",
};
-static const char *cper_severity_str(unsigned int severity)
+const char *cper_severity_str(unsigned int severity)
{
return severity < ARRAY_SIZE(cper_severity_strs) ?
cper_severity_strs[severity] : "unknown";
}
+EXPORT_SYMBOL_GPL(cper_severity_str);
/*
* cper_print_bits - print strings for set bits
@@ -195,6 +196,13 @@ static const char *cper_mem_err_type_strs[] = {
"Physical Memory Map-out event",
};
+const char *cper_mem_err_type_str(unsigned int etype)
+{
+ return etype < ARRAY_SIZE(cper_mem_err_type_strs) ?
+ cper_mem_err_type_strs[etype] : "unknown";
+}
+EXPORT_SYMBOL_GPL(cper_mem_err_type_str);
+
static void cper_print_mem(const char *pfx, const struct cper_sec_mem_err *mem)
{
if (mem->validation_bits & CPER_MEM_VALID_ERROR_STATUS)
@@ -232,8 +240,7 @@ static void cper_print_mem(const char *pfx, const struct cper_sec_mem_err *mem)
if (mem->validation_bits & CPER_MEM_VALID_ERROR_TYPE) {
u8 etype = mem->error_type;
printk("%s""error_type: %d, %s\n", pfx, etype,
- etype < ARRAY_SIZE(cper_mem_err_type_strs) ?
- cper_mem_err_type_strs[etype] : "unknown");
+ cper_mem_err_type_str(etype));
}
if (mem->validation_bits & CPER_MEM_VALID_MODULE_HANDLE) {
const char *bank = NULL, *device = NULL;
diff --git a/drivers/acpi/debug_extlog.h b/drivers/acpi/debug_extlog.h
new file mode 100644
index 0000000..67bb2c5
--- /dev/null
+++ b/drivers/acpi/debug_extlog.h
@@ -0,0 +1,16 @@
+#ifndef __DEBUG_EXTLOG_H
+#define __DEBUG_EXTLOG_H
+
+#include <linux/cper.h>
+
+#ifdef CONFIG_EXTLOG_TRACE
+extern void trace_mem_error(const uuid_le *fru_id, char *fru_text,
+ u64 err_count, u32 severity, struct cper_sec_mem_err *mem);
+#else
+void trace_mem_error(const uuid_le *fru_id, char *fru_text,
+ u64 err_count, u32 severity, struct cper_sec_mem_err *mem)
+{
+}
+#endif
+
+#endif
diff --git a/drivers/acpi/extlog_trace.c b/drivers/acpi/extlog_trace.c
new file mode 100644
index 0000000..2b2824c
--- /dev/null
+++ b/drivers/acpi/extlog_trace.c
@@ -0,0 +1,105 @@
+#include <linux/export.h>
+#include <linux/dmi.h>
+#include "debug_extlog.h"
+
+#define CREATE_TRACE_POINTS
+#include "extlog_trace.h"
+
+static char mem_location[LOC_LEN];
+static char dimm_location[LOC_LEN];
+
+static void mem_err_location(struct cper_sec_mem_err *mem)
+{
+ char *p;
+ u32 n = 0;
+
+ memset(mem_location, 0, LOC_LEN);
+ p = mem_location;
+ if (mem->validation_bits & CPER_MEM_VALID_NODE)
+ n += sprintf(p + n, " node: %d", mem->node);
+ if (n >= LOC_LEN)
+ goto end;
+ if (mem->validation_bits & CPER_MEM_VALID_CARD)
+ n += sprintf(p + n, " card: %d", mem->card);
+ if (n >= LOC_LEN)
+ goto end;
+ if (mem->validation_bits & CPER_MEM_VALID_MODULE)
+ n += sprintf(p + n, " module: %d", mem->module);
+ if (n >= LOC_LEN)
+ goto end;
+ if (mem->validation_bits & CPER_MEM_VALID_RANK_NUMBER)
+ n += sprintf(p + n, " rank: %d", mem->rank);
+ if (n >= LOC_LEN)
+ goto end;
+ if (mem->validation_bits & CPER_MEM_VALID_BANK)
+ n += sprintf(p + n, " bank: %d", mem->bank);
+ if (n >= LOC_LEN)
+ goto end;
+ if (mem->validation_bits & CPER_MEM_VALID_DEVICE)
+ n += sprintf(p + n, " device: %d", mem->device);
+ if (n >= LOC_LEN)
+ goto end;
+ if (mem->validation_bits & CPER_MEM_VALID_ROW)
+ n += sprintf(p + n, " row: %d", mem->row);
+ if (n >= LOC_LEN)
+ goto end;
+ if (mem->validation_bits & CPER_MEM_VALID_COLUMN)
+ n += sprintf(p + n, " column: %d", mem->column);
+ if (n >= LOC_LEN)
+ goto end;
+ if (mem->validation_bits & CPER_MEM_VALID_BIT_POSITION)
+ n += sprintf(p + n, " bit_position: %d", mem->bit_pos);
+ if (n >= LOC_LEN)
+ goto end;
+ if (mem->validation_bits & CPER_MEM_VALID_REQUESTOR_ID)
+ n += sprintf(p + n, " requestor_id: 0x%016llx",
+ mem->requestor_id);
+ if (n >= LOC_LEN)
+ goto end;
+ if (mem->validation_bits & CPER_MEM_VALID_RESPONDER_ID)
+ n += sprintf(p + n, " responder_id: 0x%016llx",
+ mem->responder_id);
+ if (n >= LOC_LEN)
+ goto end;
+ if (mem->validation_bits & CPER_MEM_VALID_TARGET_ID)
+ n += sprintf(p + n, " target_id: 0x%016llx", mem->target_id);
+end:
+ return;
+}
+
+static void dimm_err_location(struct cper_sec_mem_err *mem)
+{
+ memset(dimm_location, 0, LOC_LEN);
+ if (mem->validation_bits & CPER_MEM_VALID_MODULE_HANDLE) {
+ const char *bank = NULL, *device = NULL;
+ dmi_memdev_name(mem->mem_dev_handle, &bank, &device);
+ if (bank != NULL && device != NULL)
+ snprintf(dimm_location, LOC_LEN - 1,
+ "%s %s", bank, device);
+ else
+ snprintf(dimm_location, LOC_LEN - 1,
+ "DMI handle: 0x%.4x", mem->mem_dev_handle);
+ }
+}
+
+void trace_mem_error(const uuid_le *fru_id, char *fru_text,
+ u64 err_count, u32 severity, struct cper_sec_mem_err *mem)
+{
+ u32 etype = ~0U;
+ u64 phy_addr = 0;
+
+ if (mem->validation_bits & CPER_MEM_VALID_ERROR_TYPE)
+ etype = mem->error_type;
+ if (mem->validation_bits & CPER_MEM_VALID_PHYSICAL_ADDRESS) {
+ phy_addr = mem->physical_addr;
+ if (mem->validation_bits &
+ CPER_MEM_VALID_PHYSICAL_ADDRESS_MASK)
+ phy_addr &= mem->physical_addr_mask;
+ }
+ mem_err_location(mem);
+ dimm_err_location(mem);
+
+ trace_extlog_mem_event(etype, dimm_location, fru_id, fru_text,
+ err_count, severity, phy_addr, mem_location);
+}
+EXPORT_SYMBOL_GPL(trace_mem_error);
diff --git a/drivers/acpi/extlog_trace.h b/drivers/acpi/extlog_trace.h
new file mode 100644
index 0000000..21f0887
--- /dev/null
+++ b/drivers/acpi/extlog_trace.h
@@ -0,0 +1,77 @@
+#if !defined(_TRACE_EXTLOG_H) || defined(TRACE_HEADER_MULTI_READ)
+#define _TRACE_EXTLOG_H
+
+#include <linux/tracepoint.h>
+#include <linux/cper.h>
+
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM extlog
+
+/*
+ * MCE Extended Error Log Trace event
+ *
+ * These events are generated when hardware detects a corrected or
+ * uncorrected event.
+ *
+ */
+
+/* memory trace event */
+
+#define LOC_LEN 512
+#define MSG_LEN ((LOC_LEN) * 2)
+
+TRACE_EVENT(extlog_mem_event,
+ TP_PROTO(u32 etype,
+ char *dimm_loc,
+ const uuid_le *fru_id,
+ char *fru_text,
+ u64 error_count,
+ u32 severity,
+ u64 phy_addr,
+ char *mem_loc),
+
+ TP_ARGS(etype, dimm_loc, fru_id, fru_text, error_count, severity,
+ phy_addr, mem_loc),
+
+ TP_STRUCT__entry(
+ __field(u32, etype)
+ __dynamic_array(char, dimm_info, LOC_LEN)
+ __field(u64, error_count)
+ __field(u32, severity)
+ __dynamic_array(char, msg, MSG_LEN)
+ ),
+
+ TP_fast_assign(
+ __entry->error_count = error_count;
+ __entry->severity = severity;
+ __entry->etype = etype;
+ if (dimm_loc[0] != '\0')
+ snprintf(__get_dynamic_array(dimm_info), LOC_LEN - 1,
+ "on %s", dimm_loc);
+ else
+ __assign_str(dimm_info, "");
+ if (phy_addr != 0)
+ snprintf(__get_dynamic_array(msg), MSG_LEN - 1,
+ "(FRU: %pUl %.20s physical addr: 0x%016llx%s)",
+ fru_id, fru_text, phy_addr, mem_loc);
+ else
+ __assign_str(msg, "");
+ ),
+
+ TP_printk("%llu %s error%s:%s %s%s",
+ __entry->error_count,
+ cper_severity_str(__entry->severity),
+ __entry->error_count > 1 ? "s" : "",
+ cper_mem_err_type_str(__entry->etype),
+ __get_str(dimm_info),
+ __get_str(msg))
+);
+
+#endif /* _TRACE_EXTLOG_H */
+
+/* This part must be outside protection */
+#undef TRACE_INCLUDE_PATH
+#define TRACE_INCLUDE_PATH .
+#undef TRACE_INCLUDE_FILE
+#define TRACE_INCLUDE_FILE extlog_trace
+#include <trace/define_trace.h>
diff --git a/include/linux/cper.h b/include/linux/cper.h
index bd01c9a..c00eb55 100644
--- a/include/linux/cper.h
+++ b/include/linux/cper.h
@@ -395,6 +395,8 @@ struct cper_sec_pcie {
#pragma pack()
u64 cper_next_record_id(void);
+const char *cper_severity_str(unsigned int);
+const char *cper_mem_err_type_str(unsigned int);
void cper_print_bits(const char *prefix, unsigned int bits,
const char *strs[], unsigned int strs_size);
--
1.8.4.rc3
--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
