Merge tag 'edac_updates_for_5.9' of git://git.kernel.org/pub/scm/linux/kernel/git/ras/ras

Pull EDAC updates from Tony Luck:
 "Boris is on vacation and aske me to send you the EDAC changes"

* tag 'edac_updates_for_5.9' of git://git.kernel.org/pub/scm/linux/kernel/git/ras/ras:
  EDAC: Fix reference count leaks
  EDAC: Remove edac_get_dimm_by_index()
  EDAC/ghes: Scan the system once on driver init
  EDAC/ghes: Remove unused members of struct ghes_edac_pvt, rename it to ghes_pvt
  EDAC/ghes: Setup DIMM label from DMI and use it in error reports
  EDAC, {skx,i10nm}: Use CPU stepping macro to pass configurations
  EDAC/mc: Call edac_inc_ue_error() before panic
  EDAC, pnd2: Set MCE_PRIO_EDAC priority for pnd2_mce_dec notifier

7 files changed, 204 insertions, 141 deletions

diff --git a/drivers/edac/edac_device_sysfs.c b/drivers/edac/edac_device_sysfs.c
index 0e7ea3591b78..5e7593753799 100644
--- a/drivers/edac/edac_device_sysfs.c
+++ b/drivers/edac/edac_device_sysfs.c
@@ -275,6 +275,7 @@ int edac_device_register_sysfs_main_kobj(struct edac_device_ctl_info *edac_dev)
 
 	/* Error exit stack */
 err_kobj_reg:
+	kobject_put(&edac_dev->kobj);
 	module_put(edac_dev->owner);
 
 err_out:
diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
index 5813e931f2f0..01ff71f7b645 100644
--- a/drivers/edac/edac_mc.c
+++ b/drivers/edac/edac_mc.c
@@ -950,6 +950,8 @@ static void edac_ue_error(struct edac_raw_error_desc *e)
 			e->other_detail);
 	}
 
+	edac_inc_ue_error(e);
+
 	if (edac_mc_get_panic_on_ue()) {
 		panic("UE %s%son %s (%s page:0x%lx offset:0x%lx grain:%ld%s%s)\n",
 			e->msg,
@@ -959,8 +961,6 @@ static void edac_ue_error(struct edac_raw_error_desc *e)
 			*e->other_detail ? " - " : "",
 			e->other_detail);
 	}
-
-	edac_inc_ue_error(e);
 }
 
 static void edac_inc_csrow(struct edac_raw_error_desc *e, int row, int chan)
diff --git a/drivers/edac/edac_pci_sysfs.c b/drivers/edac/edac_pci_sysfs.c
index 72c9eb9fdffb..53042af7262e 100644
--- a/drivers/edac/edac_pci_sysfs.c
+++ b/drivers/edac/edac_pci_sysfs.c
@@ -386,7 +386,7 @@ static int edac_pci_main_kobj_setup(void)
 
 	/* Error unwind statck */
 kobject_init_and_add_fail:
-	kfree(edac_pci_top_main_kobj);
+	kobject_put(edac_pci_top_main_kobj);
 
 kzalloc_fail:
 	module_put(THIS_MODULE);
diff --git a/drivers/edac/ghes_edac.c b/drivers/edac/ghes_edac.c
index cb3dab56a875..da60c29468a7 100644
--- a/drivers/edac/ghes_edac.c
+++ b/drivers/edac/ghes_edac.c
@@ -15,9 +15,7 @@
 #include "edac_module.h"
 #include <ras/ras_event.h>
 
-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 */
@@ -32,7 +30,16 @@ static refcount_t ghes_refcount = REFCOUNT_INIT(0);
  * also provides the necessary (implicit) memory barrier for the SMP
  * case to make the pointer visible on another CPU.
  */
-static struct ghes_edac_pvt *ghes_pvt;
+static struct ghes_pvt *ghes_pvt;
+
+/*
+ * This driver's representation of the system hardware, as collected
+ * from DMI.
+ */
+struct ghes_hw_desc {
+	int num_dimms;
+	struct dimm_info *dimms;
+} ghes_hw;
 
 /* GHES registration mutex */
 static DEFINE_MUTEX(ghes_reg_mutex);
@@ -74,136 +81,165 @@ struct memdev_dmi_entry {
 	u16 conf_mem_clk_speed;
 } __attribute__((__packed__));
 
-struct ghes_edac_dimm_fill {
-	struct mem_ctl_info *mci;
-	unsigned int count;
-};
+static struct dimm_info *find_dimm_by_handle(struct mem_ctl_info *mci, u16 handle)
+{
+	struct dimm_info *dimm;
+
+	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)++;
+	/* both strings must be non-zero */
+	if (bank && *bank && device && *device)
+		snprintf(dimm->label, sizeof(dimm->label), "%s %s", bank, device);
 }
 
-static int get_dimm_smbios_index(struct mem_ctl_info *mci, u16 handle)
+static void assign_dmi_dimm_info(struct dimm_info *dimm, struct memdev_dmi_entry *entry)
 {
-	struct dimm_info *dimm;
+	u16 rdr_mask = BIT(7) | BIT(13);
 
-	mci_for_each_dimm(mci, dimm) {
-		if (dimm->smbios_handle == handle)
-			return dimm->idx;
+	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);
+	}
+
+	switch (entry->memory_type) {
+	case 0x12:
+		if (entry->type_detail & BIT(13))
+			dimm->mtype = MEM_RDDR;
+		else
+			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;
 	}
 
-	return -1;
+	/*
+	 * 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 ghes_edac_dmidecode(const struct dmi_header *dh, void *arg)
+static void enumerate_dimms(const struct dmi_header *dh, void *arg)
 {
-	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_get_dimm(mci, dimm_fill->count, 0, 0);
-		u16 rdr_mask = BIT(7) | BIT(13);
-
-		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 & BIT(15))
-				dimm->nr_pages = MiB_TO_PAGES((entry->size & 0x7fff) << 10);
-			else
-				dimm->nr_pages = MiB_TO_PAGES(entry->size);
-		}
+	struct memdev_dmi_entry *entry = (struct memdev_dmi_entry *)dh;
+	struct ghes_hw_desc *hw = (struct ghes_hw_desc *)arg;
+	struct dimm_info *d;
 
-		switch (entry->memory_type) {
-		case 0x12:
-			if (entry->type_detail & BIT(13))
-				dimm->mtype = MEM_RDDR;
-			else
-				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;
-		}
+	if (dh->type != DMI_ENTRY_MEM_DEVICE)
+		return;
 
-		/*
-		 * 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;
+	/* Enlarge the array with additional 16 */
+	if (!hw->num_dimms || !(hw->num_dimms % 16)) {
+		struct dimm_info *new;
 
-		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, 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);
+		new = krealloc(hw->dimms, (hw->num_dimms + 16) * sizeof(struct dimm_info),
+			        GFP_KERNEL);
+		if (!new) {
+			WARN_ON_ONCE(1);
+			return;
 		}
 
-		dimm->smbios_handle = entry->handle;
-
-		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)
+{
+	static bool scanned;
+
+	if (scanned)
+		return;
+
+	dmi_walk(enumerate_dimms, &ghes_hw);
+
+	scanned = true;
 }
 
 void ghes_edac_report_mem_error(int sev, struct cper_sec_mem_err *mem_err)
 {
 	struct edac_raw_error_desc *e;
 	struct mem_ctl_info *mci;
-	struct ghes_edac_pvt *pvt;
+	struct ghes_pvt *pvt;
 	unsigned long flags;
 	char *p;
 
@@ -228,7 +264,6 @@ void ghes_edac_report_mem_error(int sev, struct cper_sec_mem_err *mem_err)
 	memset(e, 0, sizeof (*e));
 	e->error_count = 1;
 	e->grain = 1;
-	strcpy(e->label, "unknown label");
 	e->msg = pvt->msg;
 	e->other_detail = pvt->other_detail;
 	e->top_layer = -1;
@@ -345,7 +380,7 @@ void ghes_edac_report_mem_error(int sev, struct cper_sec_mem_err *mem_err)
 		p += sprintf(p, "bit_pos:%d ", mem_err->bit_pos);
 	if (mem_err->validation_bits & CPER_MEM_VALID_MODULE_HANDLE) {
 		const char *bank = NULL, *device = NULL;
-		int index = -1;
+		struct dimm_info *dimm;
 
 		dmi_memdev_name(mem_err->mem_dev_handle, &bank, &device);
 		if (bank != NULL && device != NULL)
@@ -354,13 +389,18 @@ void ghes_edac_report_mem_error(int sev, struct cper_sec_mem_err *mem_err)
 			p += sprintf(p, "DIMM DMI handle: 0x%.4x ",
 				     mem_err->mem_dev_handle);
 
-		index = get_dimm_smbios_index(mci, mem_err->mem_dev_handle);
-		if (index >= 0)
-			e->top_layer = index;
+		dimm = find_dimm_by_handle(mci, mem_err->mem_dev_handle);
+		if (dimm) {
+			e->top_layer = dimm->idx;
+			strcpy(e->label, dimm->label);
+		}
 	}
 	if (p > e->location)
 		*(p - 1) = '\0';
 
+	if (!*e->label)
+		strcpy(e->label, "unknown memory");
+
 	/* All other fields are mapped on e->other_detail */
 	p = pvt->other_detail;
 	p += snprintf(p, sizeof(pvt->other_detail),
@@ -455,13 +495,12 @@ static struct acpi_platform_list plat_list[] = {
 int ghes_edac_register(struct ghes *ghes, struct device *dev)
 {
 	bool fake = false;
-	int rc = 0, num_dimm = 0;
 	struct mem_ctl_info *mci;
-	struct ghes_edac_pvt *pvt;
+	struct ghes_pvt *pvt;
 	struct edac_mc_layer layers[1];
-	struct ghes_edac_dimm_fill dimm_fill;
 	unsigned long flags;
 	int idx = -1;
+	int rc = 0;
 
 	if (IS_ENABLED(CONFIG_X86)) {
 		/* Check if safe to enable on this system */
@@ -481,20 +520,19 @@ int ghes_edac_register(struct ghes *ghes, struct device *dev)
 	if (refcount_inc_not_zero(&ghes_refcount))
 		goto unlock;
 
-	/* Get the number of DIMMs */
-	dmi_walk(ghes_edac_count_dimms, &num_dimm);
+	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;
 
-	mci = edac_mc_alloc(0, ARRAY_SIZE(layers), layers, sizeof(struct ghes_edac_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");
 		rc = -ENOMEM;
@@ -502,7 +540,6 @@ int ghes_edac_register(struct ghes *ghes, struct device *dev)
 	}
 
 	pvt		= mci->pvt_info;
-	pvt->ghes	= ghes;
 	pvt->mci	= mci;
 
 	mci->pdev = dev;
@@ -523,13 +560,34 @@ int ghes_edac_register(struct ghes *ghes, struct device *dev)
 		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);
+		pr_info("This system has %d DIMM sockets.\n", ghes_hw.num_dimms);
 	}
 
 	if (!fake) {
-		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_get_dimm(mci, 0, 0, 0);
 
@@ -542,7 +600,7 @@ 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);
 		rc = -ENODEV;
 		goto unlock;
@@ -556,6 +614,11 @@ int ghes_edac_register(struct ghes *ghes, struct device *dev)
 	refcount_set(&ghes_refcount, 1);
 
 unlock:
+
+	/* Not needed anymore */
+	kfree(ghes_hw.dimms);
+	ghes_hw.dimms = NULL;
+
 	mutex_unlock(&ghes_reg_mutex);
 
 	return rc;
diff --git a/drivers/edac/i10nm_base.c b/drivers/edac/i10nm_base.c
index 9b0044cd21cd..c8d11da85bec 100644
--- a/drivers/edac/i10nm_base.c
+++ b/drivers/edac/i10nm_base.c
@@ -135,9 +135,11 @@ static struct res_config i10nm_cfg1 = {
 };
 
 static const struct x86_cpu_id i10nm_cpuids[] = {
-	X86_MATCH_INTEL_FAM6_MODEL(ATOM_TREMONT_D,	&i10nm_cfg0),
-	X86_MATCH_INTEL_FAM6_MODEL(ICELAKE_X,		&i10nm_cfg0),
-	X86_MATCH_INTEL_FAM6_MODEL(ICELAKE_D,		&i10nm_cfg1),
+	X86_MATCH_INTEL_FAM6_MODEL_STEPPINGS(ATOM_TREMONT_D,	X86_STEPPINGS(0x0, 0x3), &i10nm_cfg0),
+	X86_MATCH_INTEL_FAM6_MODEL_STEPPINGS(ATOM_TREMONT_D,	X86_STEPPINGS(0x4, 0xf), &i10nm_cfg1),
+	X86_MATCH_INTEL_FAM6_MODEL_STEPPINGS(ICELAKE_X,		X86_STEPPINGS(0x0, 0x3), &i10nm_cfg0),
+	X86_MATCH_INTEL_FAM6_MODEL_STEPPINGS(ICELAKE_X,		X86_STEPPINGS(0x4, 0xf), &i10nm_cfg1),
+	X86_MATCH_INTEL_FAM6_MODEL_STEPPINGS(ICELAKE_D,		X86_STEPPINGS(0x0, 0xf), &i10nm_cfg1),
 	{}
 };
 MODULE_DEVICE_TABLE(x86cpu, i10nm_cpuids);
@@ -264,10 +266,6 @@ static int __init i10nm_init(void)
 
 	cfg = (struct res_config *)id->driver_data;
 
-	/* Newer steppings have different offset for ATOM_TREMONT_D/ICELAKE_X */
-	if (boot_cpu_data.x86_stepping >= 4)
-		cfg->busno_cfg_offset = 0xd0;
-
 	rc = skx_get_hi_lo(0x09a2, off, &tolm, &tohm);
 	if (rc)
 		return rc;
diff --git a/drivers/edac/pnd2_edac.c b/drivers/edac/pnd2_edac.c
index c1f2e6deb021..fd363746f5b0 100644
--- a/drivers/edac/pnd2_edac.c
+++ b/drivers/edac/pnd2_edac.c
@@ -1432,6 +1432,7 @@ static int pnd2_mce_check_error(struct notifier_block *nb, unsigned long val, vo
 
 static struct notifier_block pnd2_mce_dec = {
 	.notifier_call	= pnd2_mce_check_error,
+	.priority	= MCE_PRIO_EDAC,
 };
 
 #ifdef CONFIG_EDAC_DEBUG
diff --git a/drivers/edac/skx_base.c b/drivers/edac/skx_base.c
index b907a0f4ece6..2c7db95df326 100644
--- a/drivers/edac/skx_base.c
+++ b/drivers/edac/skx_base.c
@@ -164,7 +164,7 @@ static struct res_config skx_cfg = {
 };
 
 static const struct x86_cpu_id skx_cpuids[] = {
-	X86_MATCH_INTEL_FAM6_MODEL(SKYLAKE_X,	&skx_cfg),
+	X86_MATCH_INTEL_FAM6_MODEL_STEPPINGS(SKYLAKE_X, X86_STEPPINGS(0x0, 0xf), &skx_cfg),
 	{ }
 };
 MODULE_DEVICE_TABLE(x86cpu, skx_cpuids);