ACPI, APEI, Generic Hardware Error Source memory error support
Generic Hardware Error Source provides a way to report platform
hardware errors (such as that from chipset). It works in so called
"Firmware First" mode, that is, hardware errors are reported to
firmware firstly, then reported to Linux by firmware. This way, some
non-standard hardware error registers or non-standard hardware link
can be checked by firmware to produce more valuable hardware error
information for Linux.
Now, only SCI notification type and memory errors are supported. More
notification type and hardware error type will be added later. These
memory errors are reported to user space through /dev/mcelog via
faking a corrected Machine Check, so that the error memory page can be
offlined by /sbin/mcelog if the error count for one page is beyond the
threshold.
On some machines, Machine Check can not report physical address for
some corrected memory errors, but GHES can do that. So this simplified
GHES is implemented firstly.
Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
2010-05-18 14:35:20 +08:00
|
|
|
/*
|
|
|
|
* Bridge between MCE and APEI
|
|
|
|
*
|
|
|
|
* On some machine, corrected memory errors are reported via APEI
|
|
|
|
* generic hardware error source (GHES) instead of corrected Machine
|
|
|
|
* Check. These corrected memory errors can be reported to user space
|
|
|
|
* through /dev/mcelog via faking a corrected Machine Check, so that
|
|
|
|
* the error memory page can be offlined by /sbin/mcelog if the error
|
|
|
|
* count for one page is beyond the threshold.
|
|
|
|
*
|
ACPI, APEI, Use ERST for persistent storage of MCE
Traditionally, fatal MCE will cause Linux print error log to console
then reboot. Because MCE registers will preserve their content after
warm reboot, the hardware error can be logged to disk or network after
reboot. But system may fail to warm reboot, then you may lose the
hardware error log. ERST can help here. Through saving the hardware
error log into flash via ERST before go panic, the hardware error log
can be gotten from the flash after system boot successful again.
The fatal MCE processing procedure with ERST involved is as follow:
- Hardware detect error, MCE raised
- MCE read MCE registers, check error severity (fatal), prepare error record
- Write MCE error record into flash via ERST
- Go panic, then trigger system reboot
- System reboot, /sbin/mcelog run, it reads /dev/mcelog to check flash
for error record of previous boot via ERST, and output and clear
them if available
- /sbin/mcelog logs error records into disk or network
ERST only accepts CPER record format, but there is no pre-defined CPER
section can accommodate all information in struct mce, so a customized
section type is defined to hold struct mce inside a CPER record as an
error section.
Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
2010-05-18 14:35:22 +08:00
|
|
|
* For fatal MCE, save MCE record into persistent storage via ERST, so
|
|
|
|
* that the MCE record can be logged after reboot via ERST.
|
|
|
|
*
|
ACPI, APEI, Generic Hardware Error Source memory error support
Generic Hardware Error Source provides a way to report platform
hardware errors (such as that from chipset). It works in so called
"Firmware First" mode, that is, hardware errors are reported to
firmware firstly, then reported to Linux by firmware. This way, some
non-standard hardware error registers or non-standard hardware link
can be checked by firmware to produce more valuable hardware error
information for Linux.
Now, only SCI notification type and memory errors are supported. More
notification type and hardware error type will be added later. These
memory errors are reported to user space through /dev/mcelog via
faking a corrected Machine Check, so that the error memory page can be
offlined by /sbin/mcelog if the error count for one page is beyond the
threshold.
On some machines, Machine Check can not report physical address for
some corrected memory errors, but GHES can do that. So this simplified
GHES is implemented firstly.
Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
2010-05-18 14:35:20 +08:00
|
|
|
* Copyright 2010 Intel Corp.
|
|
|
|
* Author: Huang Ying <ying.huang@intel.com>
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or
|
|
|
|
* modify it under the terms of the GNU General Public License version
|
|
|
|
* 2 as published by the Free Software Foundation.
|
|
|
|
*
|
|
|
|
* This program is distributed in the hope that it will be useful,
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
* GNU General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU General Public License
|
|
|
|
* along with this program; if not, write to the Free Software
|
|
|
|
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/kernel.h>
|
|
|
|
#include <linux/acpi.h>
|
|
|
|
#include <linux/cper.h>
|
|
|
|
#include <acpi/apei.h>
|
|
|
|
#include <asm/mce.h>
|
|
|
|
|
|
|
|
#include "mce-internal.h"
|
|
|
|
|
|
|
|
void apei_mce_report_mem_error(int corrected, struct cper_sec_mem_err *mem_err)
|
|
|
|
{
|
|
|
|
struct mce m;
|
|
|
|
|
|
|
|
/* Only corrected MC is reported */
|
|
|
|
if (!corrected)
|
|
|
|
return;
|
|
|
|
|
|
|
|
mce_setup(&m);
|
|
|
|
m.bank = 1;
|
|
|
|
/* Fake a memory read corrected error with unknown channel */
|
|
|
|
m.status = MCI_STATUS_VAL | MCI_STATUS_EN | MCI_STATUS_ADDRV | 0x9f;
|
|
|
|
m.addr = mem_err->physical_addr;
|
|
|
|
mce_log(&m);
|
|
|
|
mce_notify_irq();
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(apei_mce_report_mem_error);
|
ACPI, APEI, Use ERST for persistent storage of MCE
Traditionally, fatal MCE will cause Linux print error log to console
then reboot. Because MCE registers will preserve their content after
warm reboot, the hardware error can be logged to disk or network after
reboot. But system may fail to warm reboot, then you may lose the
hardware error log. ERST can help here. Through saving the hardware
error log into flash via ERST before go panic, the hardware error log
can be gotten from the flash after system boot successful again.
The fatal MCE processing procedure with ERST involved is as follow:
- Hardware detect error, MCE raised
- MCE read MCE registers, check error severity (fatal), prepare error record
- Write MCE error record into flash via ERST
- Go panic, then trigger system reboot
- System reboot, /sbin/mcelog run, it reads /dev/mcelog to check flash
for error record of previous boot via ERST, and output and clear
them if available
- /sbin/mcelog logs error records into disk or network
ERST only accepts CPER record format, but there is no pre-defined CPER
section can accommodate all information in struct mce, so a customized
section type is defined to hold struct mce inside a CPER record as an
error section.
Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
2010-05-18 14:35:22 +08:00
|
|
|
|
|
|
|
#define CPER_CREATOR_MCE \
|
|
|
|
UUID_LE(0x75a574e3, 0x5052, 0x4b29, 0x8a, 0x8e, 0xbe, 0x2c, \
|
|
|
|
0x64, 0x90, 0xb8, 0x9d)
|
|
|
|
#define CPER_SECTION_TYPE_MCE \
|
|
|
|
UUID_LE(0xfe08ffbe, 0x95e4, 0x4be7, 0xbc, 0x73, 0x40, 0x96, \
|
|
|
|
0x04, 0x4a, 0x38, 0xfc)
|
|
|
|
|
|
|
|
/*
|
|
|
|
* CPER specification (in UEFI specification 2.3 appendix N) requires
|
|
|
|
* byte-packed.
|
|
|
|
*/
|
|
|
|
struct cper_mce_record {
|
|
|
|
struct cper_record_header hdr;
|
|
|
|
struct cper_section_descriptor sec_hdr;
|
|
|
|
struct mce mce;
|
|
|
|
} __packed;
|
|
|
|
|
|
|
|
int apei_write_mce(struct mce *m)
|
|
|
|
{
|
|
|
|
struct cper_mce_record rcd;
|
|
|
|
|
|
|
|
memset(&rcd, 0, sizeof(rcd));
|
|
|
|
memcpy(rcd.hdr.signature, CPER_SIG_RECORD, CPER_SIG_SIZE);
|
|
|
|
rcd.hdr.revision = CPER_RECORD_REV;
|
|
|
|
rcd.hdr.signature_end = CPER_SIG_END;
|
|
|
|
rcd.hdr.section_count = 1;
|
2010-08-02 15:48:23 +08:00
|
|
|
rcd.hdr.error_severity = CPER_SEV_FATAL;
|
ACPI, APEI, Use ERST for persistent storage of MCE
Traditionally, fatal MCE will cause Linux print error log to console
then reboot. Because MCE registers will preserve their content after
warm reboot, the hardware error can be logged to disk or network after
reboot. But system may fail to warm reboot, then you may lose the
hardware error log. ERST can help here. Through saving the hardware
error log into flash via ERST before go panic, the hardware error log
can be gotten from the flash after system boot successful again.
The fatal MCE processing procedure with ERST involved is as follow:
- Hardware detect error, MCE raised
- MCE read MCE registers, check error severity (fatal), prepare error record
- Write MCE error record into flash via ERST
- Go panic, then trigger system reboot
- System reboot, /sbin/mcelog run, it reads /dev/mcelog to check flash
for error record of previous boot via ERST, and output and clear
them if available
- /sbin/mcelog logs error records into disk or network
ERST only accepts CPER record format, but there is no pre-defined CPER
section can accommodate all information in struct mce, so a customized
section type is defined to hold struct mce inside a CPER record as an
error section.
Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
2010-05-18 14:35:22 +08:00
|
|
|
/* timestamp, platform_id, partition_id are all invalid */
|
|
|
|
rcd.hdr.validation_bits = 0;
|
|
|
|
rcd.hdr.record_length = sizeof(rcd);
|
|
|
|
rcd.hdr.creator_id = CPER_CREATOR_MCE;
|
|
|
|
rcd.hdr.notification_type = CPER_NOTIFY_MCE;
|
|
|
|
rcd.hdr.record_id = cper_next_record_id();
|
|
|
|
rcd.hdr.flags = CPER_HW_ERROR_FLAGS_PREVERR;
|
|
|
|
|
|
|
|
rcd.sec_hdr.section_offset = (void *)&rcd.mce - (void *)&rcd;
|
|
|
|
rcd.sec_hdr.section_length = sizeof(rcd.mce);
|
|
|
|
rcd.sec_hdr.revision = CPER_SEC_REV;
|
|
|
|
/* fru_id and fru_text is invalid */
|
|
|
|
rcd.sec_hdr.validation_bits = 0;
|
|
|
|
rcd.sec_hdr.flags = CPER_SEC_PRIMARY;
|
|
|
|
rcd.sec_hdr.section_type = CPER_SECTION_TYPE_MCE;
|
2010-08-02 15:48:23 +08:00
|
|
|
rcd.sec_hdr.section_severity = CPER_SEV_FATAL;
|
ACPI, APEI, Use ERST for persistent storage of MCE
Traditionally, fatal MCE will cause Linux print error log to console
then reboot. Because MCE registers will preserve their content after
warm reboot, the hardware error can be logged to disk or network after
reboot. But system may fail to warm reboot, then you may lose the
hardware error log. ERST can help here. Through saving the hardware
error log into flash via ERST before go panic, the hardware error log
can be gotten from the flash after system boot successful again.
The fatal MCE processing procedure with ERST involved is as follow:
- Hardware detect error, MCE raised
- MCE read MCE registers, check error severity (fatal), prepare error record
- Write MCE error record into flash via ERST
- Go panic, then trigger system reboot
- System reboot, /sbin/mcelog run, it reads /dev/mcelog to check flash
for error record of previous boot via ERST, and output and clear
them if available
- /sbin/mcelog logs error records into disk or network
ERST only accepts CPER record format, but there is no pre-defined CPER
section can accommodate all information in struct mce, so a customized
section type is defined to hold struct mce inside a CPER record as an
error section.
Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
2010-05-18 14:35:22 +08:00
|
|
|
|
|
|
|
memcpy(&rcd.mce, m, sizeof(*m));
|
|
|
|
|
|
|
|
return erst_write(&rcd.hdr);
|
|
|
|
}
|
|
|
|
|
|
|
|
ssize_t apei_read_mce(struct mce *m, u64 *record_id)
|
|
|
|
{
|
|
|
|
struct cper_mce_record rcd;
|
|
|
|
ssize_t len;
|
|
|
|
|
|
|
|
len = erst_read_next(&rcd.hdr, sizeof(rcd));
|
|
|
|
if (len <= 0)
|
|
|
|
return len;
|
|
|
|
/* Can not skip other records in storage via ERST unless clear them */
|
|
|
|
else if (len != sizeof(rcd) ||
|
|
|
|
uuid_le_cmp(rcd.hdr.creator_id, CPER_CREATOR_MCE)) {
|
|
|
|
if (printk_ratelimit())
|
|
|
|
pr_warning(
|
|
|
|
"MCE-APEI: Can not skip the unknown record in ERST");
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
|
|
|
memcpy(m, &rcd.mce, sizeof(*m));
|
|
|
|
*record_id = rcd.hdr.record_id;
|
|
|
|
|
|
|
|
return sizeof(*m);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Check whether there is record in ERST */
|
|
|
|
int apei_check_mce(void)
|
|
|
|
{
|
|
|
|
return erst_get_record_count();
|
|
|
|
}
|
|
|
|
|
|
|
|
int apei_clear_mce(u64 record_id)
|
|
|
|
{
|
|
|
|
return erst_clear(record_id);
|
|
|
|
}
|