summaryrefslogtreecommitdiff
path: root/drivers/edac/ghes_edac.c
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/edac/ghes_edac.c')
-rw-r--r--drivers/edac/ghes_edac.c771
1 files changed, 404 insertions, 367 deletions
diff --git a/drivers/edac/ghes_edac.c b/drivers/edac/ghes_edac.c
index bb534670ec02..d80c88818691 100644
--- a/drivers/edac/ghes_edac.c
+++ b/drivers/edac/ghes_edac.c
@@ -1,12 +1,10 @@
+// SPDX-License-Identifier: GPL-2.0-only
/*
* GHES/EDAC Linux driver
*
- * This file may be distributed under the terms of the GNU General Public
- * License version 2.
+ * Copyright (c) 2013 by Mauro Carvalho Chehab
*
- * Copyright (c) 2013 by Mauro Carvalho Chehab <mchehab@redhat.com>
- *
- * Red Hat Inc. http://www.redhat.com
+ * Red Hat Inc. https://www.redhat.com
*/
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
@@ -14,26 +12,52 @@
#include <acpi/ghes.h>
#include <linux/edac.h>
#include <linux/dmi.h>
-#include "edac_core.h"
+#include "edac_module.h"
#include <ras/ras_event.h>
+#include <linux/notifier.h>
+#include <linux/string.h>
-#define GHES_EDAC_REVISION " Ver: 1.0.0"
+#define OTHER_DETAIL_LEN 400
-struct ghes_edac_pvt {
- struct list_head list;
- struct ghes *ghes;
+struct ghes_pvt {
struct mem_ctl_info *mci;
/* Buffers for the error handling routine */
- char detail_location[240];
- char other_detail[160];
+ char other_detail[OTHER_DETAIL_LEN];
char msg[80];
};
-static LIST_HEAD(ghes_reglist);
-static DEFINE_MUTEX(ghes_edac_lock);
-static int ghes_edac_mc_num;
+static refcount_t ghes_refcount = REFCOUNT_INIT(0);
+
+/*
+ * Access to ghes_pvt must be protected by ghes_lock. The spinlock
+ * also provides the necessary (implicit) memory barrier for the SMP
+ * case to make the pointer visible on another CPU.
+ */
+static struct ghes_pvt *ghes_pvt;
+
+/*
+ * This driver's representation of the system hardware, as collected
+ * from DMI.
+ */
+static struct ghes_hw_desc {
+ int num_dimms;
+ struct dimm_info *dimms;
+} ghes_hw;
+
+/* GHES registration mutex */
+static DEFINE_MUTEX(ghes_reg_mutex);
+
+/*
+ * Sync with other, potentially concurrent callers of
+ * ghes_edac_report_mem_error(). We don't know what the
+ * "inventive" firmware would do.
+ */
+static DEFINE_SPINLOCK(ghes_lock);
+
+static bool system_scanned;
+static struct list_head *ghes_devs;
/* Memory Device - Type 17 of SMBIOS spec */
struct memdev_dmi_entry {
@@ -61,155 +85,221 @@ struct memdev_dmi_entry {
u16 conf_mem_clk_speed;
} __attribute__((__packed__));
-struct ghes_edac_dimm_fill {
- struct mem_ctl_info *mci;
- unsigned count;
-};
+static struct dimm_info *find_dimm_by_handle(struct mem_ctl_info *mci, u16 handle)
+{
+ struct dimm_info *dimm;
-char *memory_type[] = {
- [MEM_EMPTY] = "EMPTY",
- [MEM_RESERVED] = "RESERVED",
- [MEM_UNKNOWN] = "UNKNOWN",
- [MEM_FPM] = "FPM",
- [MEM_EDO] = "EDO",
- [MEM_BEDO] = "BEDO",
- [MEM_SDR] = "SDR",
- [MEM_RDR] = "RDR",
- [MEM_DDR] = "DDR",
- [MEM_RDDR] = "RDDR",
- [MEM_RMBS] = "RMBS",
- [MEM_DDR2] = "DDR2",
- [MEM_FB_DDR2] = "FB_DDR2",
- [MEM_RDDR2] = "RDDR2",
- [MEM_XDR] = "XDR",
- [MEM_DDR3] = "DDR3",
- [MEM_RDDR3] = "RDDR3",
-};
+ mci_for_each_dimm(mci, dimm) {
+ if (dimm->smbios_handle == handle)
+ return dimm;
+ }
+
+ return NULL;
+}
-static void ghes_edac_count_dimms(const struct dmi_header *dh, void *arg)
+static void dimm_setup_label(struct dimm_info *dimm, u16 handle)
{
- int *num_dimm = arg;
+ const char *bank = NULL, *device = NULL;
+
+ dmi_memdev_name(handle, &bank, &device);
- if (dh->type == DMI_ENTRY_MEM_DEVICE)
- (*num_dimm)++;
+ /*
+ * Set to a NULL string when both bank and device are zero. In this case,
+ * the label assigned by default will be preserved.
+ */
+ snprintf(dimm->label, sizeof(dimm->label), "%s%s%s",
+ (bank && *bank) ? bank : "",
+ (bank && *bank && device && *device) ? " " : "",
+ (device && *device) ? device : "");
}
-static void ghes_edac_dmidecode(const struct dmi_header *dh, void *arg)
+static void assign_dmi_dimm_info(struct dimm_info *dimm, struct memdev_dmi_entry *entry)
{
- struct ghes_edac_dimm_fill *dimm_fill = arg;
- struct mem_ctl_info *mci = dimm_fill->mci;
-
- if (dh->type == DMI_ENTRY_MEM_DEVICE) {
- struct memdev_dmi_entry *entry = (struct memdev_dmi_entry *)dh;
- struct dimm_info *dimm = EDAC_DIMM_PTR(mci->layers, mci->dimms,
- mci->n_layers,
- dimm_fill->count, 0, 0);
-
- if (entry->size == 0xffff) {
- pr_info("Can't get DIMM%i size\n",
- dimm_fill->count);
- dimm->nr_pages = MiB_TO_PAGES(32);/* Unknown */
- } else if (entry->size == 0x7fff) {
- dimm->nr_pages = MiB_TO_PAGES(entry->extended_size);
- } else {
- if (entry->size & 1 << 15)
- dimm->nr_pages = MiB_TO_PAGES((entry->size &
- 0x7fff) << 10);
- else
- dimm->nr_pages = MiB_TO_PAGES(entry->size);
- }
+ u16 rdr_mask = BIT(7) | BIT(13);
- switch (entry->memory_type) {
- case 0x12:
- if (entry->type_detail & 1 << 13)
- dimm->mtype = MEM_RDDR;
- else
- dimm->mtype = MEM_DDR;
- break;
- case 0x13:
- if (entry->type_detail & 1 << 13)
- dimm->mtype = MEM_RDDR2;
- else
- dimm->mtype = MEM_DDR2;
- break;
- case 0x14:
- dimm->mtype = MEM_FB_DDR2;
- break;
- case 0x18:
- if (entry->type_detail & 1 << 13)
- dimm->mtype = MEM_RDDR3;
- else
- dimm->mtype = MEM_DDR3;
- break;
- default:
- if (entry->type_detail & 1 << 6)
- dimm->mtype = MEM_RMBS;
- else if ((entry->type_detail & ((1 << 7) | (1 << 13)))
- == ((1 << 7) | (1 << 13)))
- dimm->mtype = MEM_RDR;
- else if (entry->type_detail & 1 << 7)
- dimm->mtype = MEM_SDR;
- else if (entry->type_detail & 1 << 9)
- dimm->mtype = MEM_EDO;
- else
- dimm->mtype = MEM_UNKNOWN;
- }
+ if (entry->size == 0xffff) {
+ pr_info("Can't get DIMM%i size\n", dimm->idx);
+ dimm->nr_pages = MiB_TO_PAGES(32);/* Unknown */
+ } else if (entry->size == 0x7fff) {
+ dimm->nr_pages = MiB_TO_PAGES(entry->extended_size);
+ } else {
+ if (entry->size & BIT(15))
+ dimm->nr_pages = MiB_TO_PAGES((entry->size & 0x7fff) << 10);
+ else
+ dimm->nr_pages = MiB_TO_PAGES(entry->size);
+ }
- /*
- * Actually, we can only detect if the memory has bits for
- * checksum or not
- */
- if (entry->total_width == entry->data_width)
- dimm->edac_mode = EDAC_NONE;
+ switch (entry->memory_type) {
+ case 0x12:
+ if (entry->type_detail & BIT(13))
+ dimm->mtype = MEM_RDDR;
else
- dimm->edac_mode = EDAC_SECDED;
+ dimm->mtype = MEM_DDR;
+ break;
+ case 0x13:
+ if (entry->type_detail & BIT(13))
+ dimm->mtype = MEM_RDDR2;
+ else
+ dimm->mtype = MEM_DDR2;
+ break;
+ case 0x14:
+ dimm->mtype = MEM_FB_DDR2;
+ break;
+ case 0x18:
+ if (entry->type_detail & BIT(12))
+ dimm->mtype = MEM_NVDIMM;
+ else if (entry->type_detail & BIT(13))
+ dimm->mtype = MEM_RDDR3;
+ else
+ dimm->mtype = MEM_DDR3;
+ break;
+ case 0x1a:
+ if (entry->type_detail & BIT(12))
+ dimm->mtype = MEM_NVDIMM;
+ else if (entry->type_detail & BIT(13))
+ dimm->mtype = MEM_RDDR4;
+ else
+ dimm->mtype = MEM_DDR4;
+ break;
+ default:
+ if (entry->type_detail & BIT(6))
+ dimm->mtype = MEM_RMBS;
+ else if ((entry->type_detail & rdr_mask) == rdr_mask)
+ dimm->mtype = MEM_RDR;
+ else if (entry->type_detail & BIT(7))
+ dimm->mtype = MEM_SDR;
+ else if (entry->type_detail & BIT(9))
+ dimm->mtype = MEM_EDO;
+ else
+ dimm->mtype = MEM_UNKNOWN;
+ }
- dimm->dtype = DEV_UNKNOWN;
- dimm->grain = 128; /* Likely, worse case */
-
- /*
- * FIXME: It shouldn't be hard to also fill the DIMM labels
- */
-
- if (dimm->nr_pages) {
- edac_dbg(1, "DIMM%i: %s size = %d MB%s\n",
- dimm_fill->count, memory_type[dimm->mtype],
- PAGES_TO_MiB(dimm->nr_pages),
- (dimm->edac_mode != EDAC_NONE) ? "(ECC)" : "");
- edac_dbg(2, "\ttype %d, detail 0x%02x, width %d(total %d)\n",
- entry->memory_type, entry->type_detail,
- entry->total_width, entry->data_width);
+ /*
+ * Actually, we can only detect if the memory has bits for
+ * checksum or not
+ */
+ if (entry->total_width == entry->data_width)
+ dimm->edac_mode = EDAC_NONE;
+ else
+ dimm->edac_mode = EDAC_SECDED;
+
+ dimm->dtype = DEV_UNKNOWN;
+ dimm->grain = 128; /* Likely, worse case */
+
+ dimm_setup_label(dimm, entry->handle);
+
+ if (dimm->nr_pages) {
+ edac_dbg(1, "DIMM%i: %s size = %d MB%s\n",
+ dimm->idx, edac_mem_types[dimm->mtype],
+ PAGES_TO_MiB(dimm->nr_pages),
+ (dimm->edac_mode != EDAC_NONE) ? "(ECC)" : "");
+ edac_dbg(2, "\ttype %d, detail 0x%02x, width %d(total %d)\n",
+ entry->memory_type, entry->type_detail,
+ entry->total_width, entry->data_width);
+ }
+
+ dimm->smbios_handle = entry->handle;
+}
+
+static void enumerate_dimms(const struct dmi_header *dh, void *arg)
+{
+ struct memdev_dmi_entry *entry = (struct memdev_dmi_entry *)dh;
+ struct ghes_hw_desc *hw = (struct ghes_hw_desc *)arg;
+ struct dimm_info *d;
+
+ if (dh->type != DMI_ENTRY_MEM_DEVICE)
+ return;
+
+ /* Enlarge the array with additional 16 */
+ if (!hw->num_dimms || !(hw->num_dimms % 16)) {
+ struct dimm_info *new;
+
+ new = krealloc_array(hw->dimms, hw->num_dimms + 16,
+ sizeof(struct dimm_info), GFP_KERNEL);
+ if (!new) {
+ WARN_ON_ONCE(1);
+ return;
}
- dimm_fill->count++;
+ hw->dimms = new;
}
+
+ d = &hw->dimms[hw->num_dimms];
+ d->idx = hw->num_dimms;
+
+ assign_dmi_dimm_info(d, entry);
+
+ hw->num_dimms++;
+}
+
+static void ghes_scan_system(void)
+{
+ if (system_scanned)
+ return;
+
+ dmi_walk(enumerate_dimms, &ghes_hw);
+
+ system_scanned = true;
}
-void ghes_edac_report_mem_error(struct ghes *ghes, int sev,
- struct cper_sec_mem_err *mem_err)
+static int print_mem_error_other_detail(const struct cper_sec_mem_err *mem, char *msg,
+ const char *location, unsigned int len)
{
- enum hw_event_mc_err_type type;
+ u32 n;
+
+ if (!msg)
+ return 0;
+
+ n = 0;
+ len -= 1;
+
+ n += scnprintf(msg + n, len - n, "APEI location: %s ", location);
+
+ if (!(mem->validation_bits & CPER_MEM_VALID_ERROR_STATUS))
+ goto out;
+
+ n += scnprintf(msg + n, len - n, "status(0x%016llx): ", mem->error_status);
+ n += scnprintf(msg + n, len - n, "%s ", cper_mem_err_status_str(mem->error_status));
+
+out:
+ msg[n] = '\0';
+
+ return n;
+}
+
+static int ghes_edac_report_mem_error(struct notifier_block *nb,
+ unsigned long val, void *data)
+{
+ struct cper_sec_mem_err *mem_err = (struct cper_sec_mem_err *)data;
+ struct cper_mem_err_compact cmem;
struct edac_raw_error_desc *e;
struct mem_ctl_info *mci;
- struct ghes_edac_pvt *pvt = NULL;
+ unsigned long sev = val;
+ struct ghes_pvt *pvt;
+ unsigned long flags;
char *p;
- u8 grain_bits;
- list_for_each_entry(pvt, &ghes_reglist, list) {
- if (ghes == pvt->ghes)
- break;
- }
- if (!pvt) {
- pr_err("Internal error: Can't find EDAC structure\n");
- return;
- }
+ /*
+ * We can do the locking below because GHES defers error processing
+ * from NMI to IRQ context. Whenever that changes, we'd at least
+ * know.
+ */
+ if (WARN_ON_ONCE(in_nmi()))
+ return NOTIFY_OK;
+
+ spin_lock_irqsave(&ghes_lock, flags);
+
+ pvt = ghes_pvt;
+ if (!pvt)
+ goto unlock;
+
mci = pvt->mci;
e = &mci->error_desc;
/* Cleans the error report buffer */
memset(e, 0, sizeof (*e));
e->error_count = 1;
- strcpy(e->label, "unknown label");
+ e->grain = 1;
e->msg = pvt->msg;
e->other_detail = pvt->other_detail;
e->top_layer = -1;
@@ -220,17 +310,17 @@ void ghes_edac_report_mem_error(struct ghes *ghes, int sev,
switch (sev) {
case GHES_SEV_CORRECTED:
- type = HW_EVENT_ERR_CORRECTED;
+ e->type = HW_EVENT_ERR_CORRECTED;
break;
case GHES_SEV_RECOVERABLE:
- type = HW_EVENT_ERR_UNCORRECTED;
+ e->type = HW_EVENT_ERR_UNCORRECTED;
break;
case GHES_SEV_PANIC:
- type = HW_EVENT_ERR_FATAL;
+ e->type = HW_EVENT_ERR_FATAL;
break;
default:
case GHES_SEV_NO:
- type = HW_EVENT_ERR_INFO;
+ e->type = HW_EVENT_ERR_INFO;
}
edac_dbg(1, "error validation_bits: 0x%08llx\n",
@@ -238,266 +328,147 @@ void ghes_edac_report_mem_error(struct ghes *ghes, int sev,
/* Error type, mapped on e->msg */
if (mem_err->validation_bits & CPER_MEM_VALID_ERROR_TYPE) {
+ u8 etype = mem_err->error_type;
+
p = pvt->msg;
- switch (mem_err->error_type) {
- case 0:
- p += sprintf(p, "Unknown");
- break;
- case 1:
- p += sprintf(p, "No error");
- break;
- case 2:
- p += sprintf(p, "Single-bit ECC");
- break;
- case 3:
- p += sprintf(p, "Multi-bit ECC");
- break;
- case 4:
- p += sprintf(p, "Single-symbol ChipKill ECC");
- break;
- case 5:
- p += sprintf(p, "Multi-symbol ChipKill ECC");
- break;
- case 6:
- p += sprintf(p, "Master abort");
- break;
- case 7:
- p += sprintf(p, "Target abort");
- break;
- case 8:
- p += sprintf(p, "Parity Error");
- break;
- case 9:
- p += sprintf(p, "Watchdog timeout");
- break;
- case 10:
- p += sprintf(p, "Invalid address");
- break;
- case 11:
- p += sprintf(p, "Mirror Broken");
- break;
- case 12:
- p += sprintf(p, "Memory Sparing");
- break;
- case 13:
- p += sprintf(p, "Scrub corrected error");
- break;
- case 14:
- p += sprintf(p, "Scrub uncorrected error");
- break;
- case 15:
- p += sprintf(p, "Physical Memory Map-out event");
- break;
- default:
- p += sprintf(p, "reserved error (%d)",
- mem_err->error_type);
- }
+ p += snprintf(p, sizeof(pvt->msg), "%s", cper_mem_err_type_str(etype));
} else {
- strcpy(pvt->msg, "unknown error");
+ strscpy(pvt->msg, "unknown error");
}
/* Error address */
- if (mem_err->validation_bits & CPER_MEM_VALID_PHYSICAL_ADDRESS) {
- e->page_frame_number = mem_err->physical_addr >> PAGE_SHIFT;
- e->offset_in_page = mem_err->physical_addr & ~PAGE_MASK;
+ if (mem_err->validation_bits & CPER_MEM_VALID_PA) {
+ e->page_frame_number = PHYS_PFN(mem_err->physical_addr);
+ e->offset_in_page = offset_in_page(mem_err->physical_addr);
}
/* Error grain */
- if (mem_err->validation_bits & CPER_MEM_VALID_PHYSICAL_ADDRESS_MASK) {
- e->grain = ~(mem_err->physical_addr_mask & ~PAGE_MASK);
- }
+ if (mem_err->validation_bits & CPER_MEM_VALID_PA_MASK)
+ e->grain = ~mem_err->physical_addr_mask + 1;
/* Memory error location, mapped on e->location */
p = e->location;
- if (mem_err->validation_bits & CPER_MEM_VALID_NODE)
- p += sprintf(p, "node:%d ", mem_err->node);
- if (mem_err->validation_bits & CPER_MEM_VALID_CARD)
- p += sprintf(p, "card:%d ", mem_err->card);
- if (mem_err->validation_bits & CPER_MEM_VALID_MODULE)
- p += sprintf(p, "module:%d ", mem_err->module);
- if (mem_err->validation_bits & CPER_MEM_VALID_BANK)
- p += sprintf(p, "bank:%d ", mem_err->bank);
- if (mem_err->validation_bits & CPER_MEM_VALID_ROW)
- p += sprintf(p, "row:%d ", mem_err->row);
- if (mem_err->validation_bits & CPER_MEM_VALID_COLUMN)
- p += sprintf(p, "col:%d ", mem_err->column);
- if (mem_err->validation_bits & CPER_MEM_VALID_BIT_POSITION)
- p += sprintf(p, "bit_pos:%d ", mem_err->bit_pos);
+ cper_mem_err_pack(mem_err, &cmem);
+ p += cper_mem_err_location(&cmem, p);
+
+ if (mem_err->validation_bits & CPER_MEM_VALID_MODULE_HANDLE) {
+ struct dimm_info *dimm;
+
+ p += cper_dimm_err_location(&cmem, p);
+ dimm = find_dimm_by_handle(mci, mem_err->mem_dev_handle);
+ if (dimm) {
+ e->top_layer = dimm->idx;
+ strscpy(e->label, dimm->label);
+ }
+ }
if (p > e->location)
*(p - 1) = '\0';
+ if (!*e->label)
+ strscpy(e->label, "unknown memory");
+
/* All other fields are mapped on e->other_detail */
p = pvt->other_detail;
- if (mem_err->validation_bits & CPER_MEM_VALID_ERROR_STATUS) {
- u64 status = mem_err->error_status;
-
- p += sprintf(p, "status(0x%016llx): ", (long long)status);
- switch ((status >> 8) & 0xff) {
- case 1:
- p += sprintf(p, "Error detected internal to the component ");
- break;
- case 16:
- p += sprintf(p, "Error detected in the bus ");
- break;
- case 4:
- p += sprintf(p, "Storage error in DRAM memory ");
- break;
- case 5:
- p += sprintf(p, "Storage error in TLB ");
- break;
- case 6:
- p += sprintf(p, "Storage error in cache ");
- break;
- case 7:
- p += sprintf(p, "Error in one or more functional units ");
- break;
- case 8:
- p += sprintf(p, "component failed self test ");
- break;
- case 9:
- p += sprintf(p, "Overflow or undervalue of internal queue ");
- break;
- case 17:
- p += sprintf(p, "Virtual address not found on IO-TLB or IO-PDIR ");
- break;
- case 18:
- p += sprintf(p, "Improper access error ");
- break;
- case 19:
- p += sprintf(p, "Access to a memory address which is not mapped to any component ");
- break;
- case 20:
- p += sprintf(p, "Loss of Lockstep ");
- break;
- case 21:
- p += sprintf(p, "Response not associated with a request ");
- break;
- case 22:
- p += sprintf(p, "Bus parity error - must also set the A, C, or D Bits ");
- break;
- case 23:
- p += sprintf(p, "Detection of a PATH_ERROR ");
- break;
- case 25:
- p += sprintf(p, "Bus operation timeout ");
- break;
- case 26:
- p += sprintf(p, "A read was issued to data that has been poisoned ");
- break;
- default:
- p += sprintf(p, "reserved ");
- break;
- }
- }
- if (mem_err->validation_bits & CPER_MEM_VALID_REQUESTOR_ID)
- p += sprintf(p, "requestorID: 0x%016llx ",
- (long long)mem_err->requestor_id);
- if (mem_err->validation_bits & CPER_MEM_VALID_RESPONDER_ID)
- p += sprintf(p, "responderID: 0x%016llx ",
- (long long)mem_err->responder_id);
- if (mem_err->validation_bits & CPER_MEM_VALID_TARGET_ID)
- p += sprintf(p, "targetID: 0x%016llx ",
- (long long)mem_err->responder_id);
+ p += print_mem_error_other_detail(mem_err, p, e->location, OTHER_DETAIL_LEN);
if (p > pvt->other_detail)
*(p - 1) = '\0';
- /* Generate the trace event */
- grain_bits = fls_long(e->grain);
- sprintf(pvt->detail_location, "APEI location: %s %s",
- e->location, e->other_detail);
- trace_mc_event(type, e->msg, e->label, e->error_count,
- mci->mc_idx, e->top_layer, e->mid_layer, e->low_layer,
- PAGES_TO_MiB(e->page_frame_number) | e->offset_in_page,
- grain_bits, e->syndrome, pvt->detail_location);
-
- /* Report the error via EDAC API */
- edac_raw_mc_handle_error(type, mci, e);
+ edac_raw_mc_handle_error(e);
+
+unlock:
+ spin_unlock_irqrestore(&ghes_lock, flags);
+
+ return NOTIFY_OK;
}
-EXPORT_SYMBOL_GPL(ghes_edac_report_mem_error);
-int ghes_edac_register(struct ghes *ghes, struct device *dev)
+static struct notifier_block ghes_edac_mem_err_nb = {
+ .notifier_call = ghes_edac_report_mem_error,
+ .priority = 0,
+};
+
+static int ghes_edac_register(struct device *dev)
{
bool fake = false;
- int rc, num_dimm = 0;
struct mem_ctl_info *mci;
+ struct ghes_pvt *pvt;
struct edac_mc_layer layers[1];
- struct ghes_edac_pvt *pvt;
- struct ghes_edac_dimm_fill dimm_fill;
+ unsigned long flags;
+ int rc = 0;
+
+ /* finish another registration/unregistration instance first */
+ mutex_lock(&ghes_reg_mutex);
- /* Get the number of DIMMs */
- dmi_walk(ghes_edac_count_dimms, &num_dimm);
+ /*
+ * We have only one logical memory controller to which all DIMMs belong.
+ */
+ if (refcount_inc_not_zero(&ghes_refcount))
+ goto unlock;
+
+ ghes_scan_system();
/* Check if we've got a bogus BIOS */
- if (num_dimm == 0) {
+ if (!ghes_hw.num_dimms) {
fake = true;
- num_dimm = 1;
+ ghes_hw.num_dimms = 1;
}
layers[0].type = EDAC_MC_LAYER_ALL_MEM;
- layers[0].size = num_dimm;
+ layers[0].size = ghes_hw.num_dimms;
layers[0].is_virt_csrow = true;
- /*
- * We need to serialize edac_mc_alloc() and edac_mc_add_mc(),
- * to avoid duplicated memory controller numbers
- */
- mutex_lock(&ghes_edac_lock);
- mci = edac_mc_alloc(ghes_edac_mc_num, ARRAY_SIZE(layers), layers,
- sizeof(*pvt));
+ mci = edac_mc_alloc(0, ARRAY_SIZE(layers), layers, sizeof(struct ghes_pvt));
if (!mci) {
pr_info("Can't allocate memory for EDAC data\n");
- mutex_unlock(&ghes_edac_lock);
- return -ENOMEM;
+ rc = -ENOMEM;
+ goto unlock;
}
- pvt = mci->pvt_info;
- memset(pvt, 0, sizeof(*pvt));
- list_add_tail(&pvt->list, &ghes_reglist);
- pvt->ghes = ghes;
- pvt->mci = mci;
- mci->pdev = dev;
+ pvt = mci->pvt_info;
+ pvt->mci = mci;
+ mci->pdev = dev;
mci->mtype_cap = MEM_FLAG_EMPTY;
mci->edac_ctl_cap = EDAC_FLAG_NONE;
mci->edac_cap = EDAC_FLAG_NONE;
mci->mod_name = "ghes_edac.c";
- mci->mod_ver = GHES_EDAC_REVISION;
mci->ctl_name = "ghes_edac";
mci->dev_name = "ghes";
- if (!ghes_edac_mc_num) {
- if (!fake) {
- pr_info("This EDAC driver relies on BIOS to enumerate memory and get error reports.\n");
- pr_info("Unfortunately, not all BIOSes reflect the memory layout correctly.\n");
- pr_info("So, the end result of using this driver varies from vendor to vendor.\n");
- pr_info("If you find incorrect reports, please contact your hardware vendor\n");
- pr_info("to correct its BIOS.\n");
- pr_info("This system has %d DIMM sockets.\n",
- num_dimm);
- } else {
- pr_info("This system has a very crappy BIOS: It doesn't even list the DIMMS.\n");
- pr_info("Its SMBIOS info is wrong. It is doubtful that the error report would\n");
- pr_info("work on such system. Use this driver with caution\n");
- }
+ if (fake) {
+ pr_info("This system has a very crappy BIOS: It doesn't even list the DIMMS.\n");
+ pr_info("Its SMBIOS info is wrong. It is doubtful that the error report would\n");
+ pr_info("work on such system. Use this driver with caution\n");
}
+ pr_info("This system has %d DIMM sockets.\n", ghes_hw.num_dimms);
+
if (!fake) {
- /*
- * Fill DIMM info from DMI for the memory controller #0
- *
- * Keep it in blank for the other memory controllers, as
- * there's no reliable way to properly credit each DIMM to
- * the memory controller, as different BIOSes fill the
- * DMI bank location fields on different ways
- */
- if (!ghes_edac_mc_num) {
- dimm_fill.count = 0;
- dimm_fill.mci = mci;
- dmi_walk(ghes_edac_dmidecode, &dimm_fill);
+ struct dimm_info *src, *dst;
+ int i = 0;
+
+ mci_for_each_dimm(mci, dst) {
+ src = &ghes_hw.dimms[i];
+
+ dst->idx = src->idx;
+ dst->smbios_handle = src->smbios_handle;
+ dst->nr_pages = src->nr_pages;
+ dst->mtype = src->mtype;
+ dst->edac_mode = src->edac_mode;
+ dst->dtype = src->dtype;
+ dst->grain = src->grain;
+
+ /*
+ * If no src->label, preserve default label assigned
+ * from EDAC core.
+ */
+ if (strlen(src->label))
+ memcpy(dst->label, src->label, sizeof(src->label));
+
+ i++;
}
+
} else {
- struct dimm_info *dimm = EDAC_DIMM_PTR(mci->layers, mci->dimms,
- mci->n_layers, 0, 0, 0);
+ struct dimm_info *dimm = edac_get_dimm(mci, 0, 0, 0);
dimm->nr_pages = 1;
dimm->grain = 128;
@@ -508,30 +479,96 @@ int ghes_edac_register(struct ghes *ghes, struct device *dev)
rc = edac_mc_add_mc(mci);
if (rc < 0) {
- pr_info("Can't register at EDAC core\n");
+ pr_info("Can't register with the EDAC core\n");
edac_mc_free(mci);
- mutex_unlock(&ghes_edac_lock);
+ rc = -ENODEV;
+ goto unlock;
+ }
+
+ spin_lock_irqsave(&ghes_lock, flags);
+ ghes_pvt = pvt;
+ spin_unlock_irqrestore(&ghes_lock, flags);
+
+ ghes_register_report_chain(&ghes_edac_mem_err_nb);
+
+ /* only set on success */
+ refcount_set(&ghes_refcount, 1);
+
+unlock:
+
+ /* Not needed anymore */
+ kfree(ghes_hw.dimms);
+ ghes_hw.dimms = NULL;
+
+ mutex_unlock(&ghes_reg_mutex);
+
+ return rc;
+}
+
+static void ghes_edac_unregister(struct ghes *ghes)
+{
+ struct mem_ctl_info *mci;
+ unsigned long flags;
+
+ mutex_lock(&ghes_reg_mutex);
+
+ system_scanned = false;
+ memset(&ghes_hw, 0, sizeof(struct ghes_hw_desc));
+
+ if (!refcount_dec_and_test(&ghes_refcount))
+ goto unlock;
+
+ /*
+ * Wait for the irq handler being finished.
+ */
+ spin_lock_irqsave(&ghes_lock, flags);
+ mci = ghes_pvt ? ghes_pvt->mci : NULL;
+ ghes_pvt = NULL;
+ spin_unlock_irqrestore(&ghes_lock, flags);
+
+ if (!mci)
+ goto unlock;
+
+ mci = edac_mc_del_mc(mci->pdev);
+ if (mci)
+ edac_mc_free(mci);
+
+ ghes_unregister_report_chain(&ghes_edac_mem_err_nb);
+
+unlock:
+ mutex_unlock(&ghes_reg_mutex);
+}
+
+static int __init ghes_edac_init(void)
+{
+ struct ghes *g, *g_tmp;
+
+ ghes_devs = ghes_get_devices();
+ if (!ghes_devs)
return -ENODEV;
+
+ if (list_empty(ghes_devs)) {
+ pr_info("GHES probing device list is empty\n");
+ return -ENODEV;
+ }
+
+ list_for_each_entry_safe(g, g_tmp, ghes_devs, elist) {
+ ghes_edac_register(g->dev);
}
- ghes_edac_mc_num++;
- mutex_unlock(&ghes_edac_lock);
return 0;
}
-EXPORT_SYMBOL_GPL(ghes_edac_register);
+module_init(ghes_edac_init);
-void ghes_edac_unregister(struct ghes *ghes)
+static void __exit ghes_edac_exit(void)
{
- struct mem_ctl_info *mci;
- struct ghes_edac_pvt *pvt, *tmp;
-
- list_for_each_entry_safe(pvt, tmp, &ghes_reglist, list) {
- if (ghes == pvt->ghes) {
- mci = pvt->mci;
- edac_mc_del_mc(mci->pdev);
- edac_mc_free(mci);
- list_del(&pvt->list);
- }
+ struct ghes *g, *g_tmp;
+
+ list_for_each_entry_safe(g, g_tmp, ghes_devs, elist) {
+ ghes_edac_unregister(g);
}
}
-EXPORT_SYMBOL_GPL(ghes_edac_unregister);
+module_exit(ghes_edac_exit);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Output ACPI APEI/GHES BIOS detected errors via EDAC");