forked from Minki/linux
3f8dec1162
Platforms with large BERT table data can trigger soft lockup errors while attempting to print the entire BERT table data to the console at boot: watchdog: BUG: soft lockup - CPU#160 stuck for 23s! [swapper/0:1] Observed on Ampere Altra systems with a single BERT record of ~250KB. The original bert driver appears to have assumed relatively small table data. Since it is impractical to reassemble large table data from interwoven console messages, and the table data is available in /sys/firmware/acpi/tables/data/BERT limit the size for tables printed to the console to 1024 (for no reason other than it seemed like a good place to kick off the discussion, would appreciate feedback from existing users in terms of what size would maintain their current usage model). Alternatively, we could make printing a CONFIG option, use the bert_disable boot arg (or something similar), or use a debug log level. However, all those solutions require extra steps or change the existing behavior for small table data. Limiting the size preserves existing behavior on existing platforms with small table data, and eliminates the soft lockups for platforms with large table data, while still making it available. Signed-off-by: Darren Hart <darren@os.amperecomputing.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
156 lines
4.0 KiB
C
156 lines
4.0 KiB
C
// SPDX-License-Identifier: GPL-2.0-only
|
|
/*
|
|
* APEI Boot Error Record Table (BERT) support
|
|
*
|
|
* Copyright 2011 Intel Corp.
|
|
* Author: Huang Ying <ying.huang@intel.com>
|
|
*
|
|
* Under normal circumstances, when a hardware error occurs, the error
|
|
* handler receives control and processes the error. This gives OSPM a
|
|
* chance to process the error condition, report it, and optionally attempt
|
|
* recovery. In some cases, the system is unable to process an error.
|
|
* For example, system firmware or a management controller may choose to
|
|
* reset the system or the system might experience an uncontrolled crash
|
|
* or reset.The boot error source is used to report unhandled errors that
|
|
* occurred in a previous boot. This mechanism is described in the BERT
|
|
* table.
|
|
*
|
|
* For more information about BERT, please refer to ACPI Specification
|
|
* version 4.0, section 17.3.1
|
|
*/
|
|
|
|
#include <linux/kernel.h>
|
|
#include <linux/module.h>
|
|
#include <linux/init.h>
|
|
#include <linux/acpi.h>
|
|
#include <linux/io.h>
|
|
|
|
#include "apei-internal.h"
|
|
|
|
#undef pr_fmt
|
|
#define pr_fmt(fmt) "BERT: " fmt
|
|
#define ACPI_BERT_PRINT_MAX_LEN 1024
|
|
|
|
static int bert_disable;
|
|
|
|
static void __init bert_print_all(struct acpi_bert_region *region,
|
|
unsigned int region_len)
|
|
{
|
|
struct acpi_hest_generic_status *estatus =
|
|
(struct acpi_hest_generic_status *)region;
|
|
int remain = region_len;
|
|
u32 estatus_len;
|
|
|
|
while (remain >= sizeof(struct acpi_bert_region)) {
|
|
estatus_len = cper_estatus_len(estatus);
|
|
if (remain < estatus_len) {
|
|
pr_err(FW_BUG "Truncated status block (length: %u).\n",
|
|
estatus_len);
|
|
return;
|
|
}
|
|
|
|
/* No more error records. */
|
|
if (!estatus->block_status)
|
|
return;
|
|
|
|
if (cper_estatus_check(estatus)) {
|
|
pr_err(FW_BUG "Invalid error record.\n");
|
|
return;
|
|
}
|
|
|
|
pr_info_once("Error records from previous boot:\n");
|
|
if (region_len < ACPI_BERT_PRINT_MAX_LEN)
|
|
cper_estatus_print(KERN_INFO HW_ERR, estatus);
|
|
else
|
|
pr_info_once("Max print length exceeded, table data is available at:\n"
|
|
"/sys/firmware/acpi/tables/data/BERT");
|
|
|
|
/*
|
|
* Because the boot error source is "one-time polled" type,
|
|
* clear Block Status of current Generic Error Status Block,
|
|
* once it's printed.
|
|
*/
|
|
estatus->block_status = 0;
|
|
|
|
estatus = (void *)estatus + estatus_len;
|
|
remain -= estatus_len;
|
|
}
|
|
}
|
|
|
|
static int __init setup_bert_disable(char *str)
|
|
{
|
|
bert_disable = 1;
|
|
|
|
return 1;
|
|
}
|
|
__setup("bert_disable", setup_bert_disable);
|
|
|
|
static int __init bert_check_table(struct acpi_table_bert *bert_tab)
|
|
{
|
|
if (bert_tab->header.length < sizeof(struct acpi_table_bert) ||
|
|
bert_tab->region_length < sizeof(struct acpi_bert_region))
|
|
return -EINVAL;
|
|
|
|
return 0;
|
|
}
|
|
|
|
static int __init bert_init(void)
|
|
{
|
|
struct apei_resources bert_resources;
|
|
struct acpi_bert_region *boot_error_region;
|
|
struct acpi_table_bert *bert_tab;
|
|
unsigned int region_len;
|
|
acpi_status status;
|
|
int rc = 0;
|
|
|
|
if (acpi_disabled)
|
|
return 0;
|
|
|
|
if (bert_disable) {
|
|
pr_info("Boot Error Record Table support is disabled.\n");
|
|
return 0;
|
|
}
|
|
|
|
status = acpi_get_table(ACPI_SIG_BERT, 0, (struct acpi_table_header **)&bert_tab);
|
|
if (status == AE_NOT_FOUND)
|
|
return 0;
|
|
|
|
if (ACPI_FAILURE(status)) {
|
|
pr_err("get table failed, %s.\n", acpi_format_exception(status));
|
|
return -EINVAL;
|
|
}
|
|
|
|
rc = bert_check_table(bert_tab);
|
|
if (rc) {
|
|
pr_err(FW_BUG "table invalid.\n");
|
|
goto out_put_bert_tab;
|
|
}
|
|
|
|
region_len = bert_tab->region_length;
|
|
apei_resources_init(&bert_resources);
|
|
rc = apei_resources_add(&bert_resources, bert_tab->address,
|
|
region_len, true);
|
|
if (rc)
|
|
goto out_put_bert_tab;
|
|
rc = apei_resources_request(&bert_resources, "APEI BERT");
|
|
if (rc)
|
|
goto out_fini;
|
|
boot_error_region = ioremap_cache(bert_tab->address, region_len);
|
|
if (boot_error_region) {
|
|
bert_print_all(boot_error_region, region_len);
|
|
iounmap(boot_error_region);
|
|
} else {
|
|
rc = -ENOMEM;
|
|
}
|
|
|
|
apei_resources_release(&bert_resources);
|
|
out_fini:
|
|
apei_resources_fini(&bert_resources);
|
|
out_put_bert_tab:
|
|
acpi_put_table((struct acpi_table_header *)bert_tab);
|
|
|
|
return rc;
|
|
}
|
|
|
|
late_initcall(bert_init);
|