1 files changed, 436 insertions, 113 deletions

diff --git a/arch/powerpc/kernel/mce.c b/arch/powerpc/kernel/mce.c
index e0e131e662ed..219f28637a3e 100644
--- a/arch/powerpc/kernel/mce.c
+++ b/arch/powerpc/kernel/mce.c
@@ -1,20 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
 /*
  * Machine check exception handling.
  *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
- *
- * 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.
- *
  * Copyright 2013 IBM Corporation
  * Author: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
  */
@@ -22,24 +9,42 @@
 #undef DEBUG
 #define pr_fmt(fmt) "mce: " fmt
 
+#include <linux/hardirq.h>
 #include <linux/types.h>
 #include <linux/ptrace.h>
 #include <linux/percpu.h>
 #include <linux/export.h>
 #include <linux/irq_work.h>
+#include <linux/extable.h>
+#include <linux/ftrace.h>
+#include <linux/memblock.h>
+#include <linux/of.h>
+
+#include <asm/interrupt.h>
+#include <asm/machdep.h>
 #include <asm/mce.h>
+#include <asm/nmi.h>
+
+#include "setup.h"
 
-static DEFINE_PER_CPU(int, mce_nest_count);
-static DEFINE_PER_CPU(struct machine_check_event[MAX_MC_EVT], mce_event);
+static void machine_check_ue_event(struct machine_check_event *evt);
+static void machine_process_ue_event(struct work_struct *work);
 
-/* Queue for delayed MCE events. */
-static DEFINE_PER_CPU(int, mce_queue_count);
-static DEFINE_PER_CPU(struct machine_check_event[MAX_MC_EVT], mce_event_queue);
+static DECLARE_WORK(mce_ue_event_work, machine_process_ue_event);
+
+static BLOCKING_NOTIFIER_HEAD(mce_notifier_list);
+
+int mce_register_notifier(struct notifier_block *nb)
+{
+	return blocking_notifier_chain_register(&mce_notifier_list, nb);
+}
+EXPORT_SYMBOL_GPL(mce_register_notifier);
 
-static void machine_check_process_queued_event(struct irq_work *work);
-static struct irq_work mce_event_process_work = {
-        .func = machine_check_process_queued_event,
-};
+int mce_unregister_notifier(struct notifier_block *nb)
+{
+	return blocking_notifier_chain_unregister(&mce_notifier_list, nb);
+}
+EXPORT_SYMBOL_GPL(mce_unregister_notifier);
 
 static void mce_set_error_info(struct machine_check_event *mce,
 			       struct mce_error_info *mce_err)
@@ -73,17 +78,25 @@ static void mce_set_error_info(struct machine_check_event *mce,
 	}
 }
 
+void mce_irq_work_queue(void)
+{
+	/* Raise decrementer interrupt */
+	arch_irq_work_raise();
+	set_mce_pending_irq_work();
+}
+
 /*
  * Decode and save high level MCE information into per cpu buffer which
  * is an array of machine_check_event structure.
  */
 void save_mce_event(struct pt_regs *regs, long handled,
 		    struct mce_error_info *mce_err,
-		    uint64_t nip, uint64_t addr)
+		    uint64_t nip, uint64_t addr, uint64_t phys_addr)
 {
-	int index = __this_cpu_inc_return(mce_nest_count) - 1;
-	struct machine_check_event *mce = this_cpu_ptr(&mce_event[index]);
+	int index = local_paca->mce_info->mce_nest_count++;
+	struct machine_check_event *mce;
 
+	mce = &local_paca->mce_info->mce_event[index];
 	/*
 	 * Return if we don't have enough space to log mce event.
 	 * mce_nest_count may go beyond MAX_MC_EVT but that's ok,
@@ -98,6 +111,7 @@ void save_mce_event(struct pt_regs *regs, long handled,
 	mce->srr1 = regs->msr;
 	mce->gpr3 = regs->gpr[3];
 	mce->in_use = 1;
+	mce->cpu = get_paca()->paca_index;
 
 	/* Mark it recovered if we have handled it and MSR(RI=1). */
 	if (handled && (regs->msr & MSR_RI))
@@ -107,11 +121,22 @@ void save_mce_event(struct pt_regs *regs, long handled,
 
 	mce->initiator = mce_err->initiator;
 	mce->severity = mce_err->severity;
+	mce->sync_error = mce_err->sync_error;
+	mce->error_class = mce_err->error_class;
 
 	/*
 	 * Populate the mce error_type and type-specific error_type.
 	 */
 	mce_set_error_info(mce, mce_err);
+	if (mce->error_type == MCE_ERROR_TYPE_UE)
+		mce->u.ue_error.ignore_event = mce_err->ignore_event;
+
+	/*
+	 * Raise irq work, So that we don't miss to log the error for
+	 * unrecoverable errors.
+	 */
+	if (mce->disposition == MCE_DISPOSITION_NOT_RECOVERED)
+		mce_irq_work_queue();
 
 	if (!addr)
 		return;
@@ -137,6 +162,11 @@ void save_mce_event(struct pt_regs *regs, long handled,
 	} else if (mce->error_type == MCE_ERROR_TYPE_UE) {
 		mce->u.ue_error.effective_address_provided = true;
 		mce->u.ue_error.effective_address = addr;
+		if (phys_addr != ULONG_MAX) {
+			mce->u.ue_error.physical_address_provided = true;
+			mce->u.ue_error.physical_address = phys_addr;
+			machine_check_ue_event(mce);
+		}
 	}
 	return;
 }
@@ -160,7 +190,7 @@ void save_mce_event(struct pt_regs *regs, long handled,
  */
 int get_mce_event(struct machine_check_event *mce, bool release)
 {
-	int index = __this_cpu_read(mce_nest_count) - 1;
+	int index = local_paca->mce_info->mce_nest_count - 1;
 	struct machine_check_event *mc_evt;
 	int ret = 0;
 
@@ -170,7 +200,7 @@ int get_mce_event(struct machine_check_event *mce, bool release)
 
 	/* Check if we have MCE info to process. */
 	if (index < MAX_MC_EVT) {
-		mc_evt = this_cpu_ptr(&mce_event[index]);
+		mc_evt = &local_paca->mce_info->mce_event[index];
 		/* Copy the event structure and release the original */
 		if (mce)
 			*mce = *mc_evt;
@@ -180,7 +210,7 @@ int get_mce_event(struct machine_check_event *mce, bool release)
 	}
 	/* Decrement the count to free the slot. */
 	if (release)
-		__this_cpu_dec(mce_nest_count);
+		local_paca->mce_info->mce_nest_count--;
 
 	return ret;
 }
@@ -190,6 +220,28 @@ void release_mce_event(void)
 	get_mce_event(NULL, true);
 }
 
+static void machine_check_ue_work(void)
+{
+	schedule_work(&mce_ue_event_work);
+}
+
+/*
+ * Queue up the MCE event which then can be handled later.
+ */
+static void machine_check_ue_event(struct machine_check_event *evt)
+{
+	int index;
+
+	index = local_paca->mce_info->mce_ue_count++;
+	/* If queue is full, just return for now. */
+	if (index >= MAX_MC_EVT) {
+		local_paca->mce_info->mce_ue_count--;
+		return;
+	}
+	memcpy(&local_paca->mce_info->mce_ue_event_queue[index],
+	       evt, sizeof(*evt));
+}
+
 /*
  * Queue up the MCE event which then can be handled later.
  */
@@ -201,25 +253,81 @@ void machine_check_queue_event(void)
 	if (!get_mce_event(&evt, MCE_EVENT_RELEASE))
 		return;
 
-	index = __this_cpu_inc_return(mce_queue_count) - 1;
+	index = local_paca->mce_info->mce_queue_count++;
 	/* If queue is full, just return for now. */
 	if (index >= MAX_MC_EVT) {
-		__this_cpu_dec(mce_queue_count);
+		local_paca->mce_info->mce_queue_count--;
 		return;
 	}
-	memcpy(this_cpu_ptr(&mce_event_queue[index]), &evt, sizeof(evt));
+	memcpy(&local_paca->mce_info->mce_event_queue[index],
+	       &evt, sizeof(evt));
 
-	/* Queue irq work to process this event later. */
-	irq_work_queue(&mce_event_process_work);
+	mce_irq_work_queue();
 }
 
+void mce_common_process_ue(struct pt_regs *regs,
+			   struct mce_error_info *mce_err)
+{
+	const struct exception_table_entry *entry;
+
+	entry = search_kernel_exception_table(regs->nip);
+	if (entry) {
+		mce_err->ignore_event = true;
+		regs_set_return_ip(regs, extable_fixup(entry));
+	}
+}
+
+/*
+ * process pending MCE event from the mce event queue. This function will be
+ * called during syscall exit.
+ */
+static void machine_process_ue_event(struct work_struct *work)
+{
+	int index;
+	struct machine_check_event *evt;
+
+	while (local_paca->mce_info->mce_ue_count > 0) {
+		index = local_paca->mce_info->mce_ue_count - 1;
+		evt = &local_paca->mce_info->mce_ue_event_queue[index];
+		blocking_notifier_call_chain(&mce_notifier_list, 0, evt);
+#ifdef CONFIG_MEMORY_FAILURE
+		/*
+		 * This should probably queued elsewhere, but
+		 * oh! well
+		 *
+		 * Don't report this machine check because the caller has a
+		 * asked us to ignore the event, it has a fixup handler which
+		 * will do the appropriate error handling and reporting.
+		 */
+		if (evt->error_type == MCE_ERROR_TYPE_UE) {
+			if (evt->u.ue_error.ignore_event) {
+				local_paca->mce_info->mce_ue_count--;
+				continue;
+			}
+
+			if (evt->u.ue_error.physical_address_provided) {
+				unsigned long pfn;
+
+				pfn = evt->u.ue_error.physical_address >>
+					PAGE_SHIFT;
+				memory_failure(pfn, 0);
+			} else
+				pr_warn("Failed to identify bad address from "
+					"where the uncorrectable error (UE) "
+					"was generated\n");
+		}
+#endif
+		local_paca->mce_info->mce_ue_count--;
+	}
+}
 /*
  * process pending MCE event from the mce event queue. This function will be
  * called during syscall exit.
  */
-static void machine_check_process_queued_event(struct irq_work *work)
+static void machine_check_process_queued_event(void)
 {
 	int index;
+	struct machine_check_event *evt;
 
 	add_taint(TAINT_MACHINE_CHECK, LOCKDEP_NOW_UNRELIABLE);
 
@@ -227,18 +335,49 @@ static void machine_check_process_queued_event(struct irq_work *work)
 	 * For now just print it to console.
 	 * TODO: log this error event to FSP or nvram.
 	 */
-	while (__this_cpu_read(mce_queue_count) > 0) {
-		index = __this_cpu_read(mce_queue_count) - 1;
-		machine_check_print_event_info(
-				this_cpu_ptr(&mce_event_queue[index]), false);
-		__this_cpu_dec(mce_queue_count);
+	while (local_paca->mce_info->mce_queue_count > 0) {
+		index = local_paca->mce_info->mce_queue_count - 1;
+		evt = &local_paca->mce_info->mce_event_queue[index];
+
+		if (evt->error_type == MCE_ERROR_TYPE_UE &&
+		    evt->u.ue_error.ignore_event) {
+			local_paca->mce_info->mce_queue_count--;
+			continue;
+		}
+		machine_check_print_event_info(evt, false, false);
+		local_paca->mce_info->mce_queue_count--;
+	}
+}
+
+void set_mce_pending_irq_work(void)
+{
+	local_paca->mce_pending_irq_work = 1;
+}
+
+void clear_mce_pending_irq_work(void)
+{
+	local_paca->mce_pending_irq_work = 0;
+}
+
+void mce_run_irq_context_handlers(void)
+{
+	if (unlikely(local_paca->mce_pending_irq_work)) {
+		if (ppc_md.machine_check_log_err)
+			ppc_md.machine_check_log_err();
+		machine_check_process_queued_event();
+		machine_check_ue_work();
+		clear_mce_pending_irq_work();
 	}
 }
 
 void machine_check_print_event_info(struct machine_check_event *evt,
-				    bool user_mode)
+				    bool user_mode, bool in_guest)
 {
-	const char *level, *sevstr, *subtype;
+	const char *level, *sevstr, *subtype, *err_type, *initiator;
+	uint64_t ea = 0, pa = 0;
+	int n = 0;
+	char dar_str[50];
+	char pa_str[50];
 	static const char *mc_ue_types[] = {
 		"Indeterminate",
 		"Instruction fetch",
@@ -264,18 +403,19 @@ void machine_check_print_event_info(struct machine_check_event *evt,
 	static const char *mc_user_types[] = {
 		"Indeterminate",
 		"tlbie(l) invalid",
+		"scv invalid",
 	};
 	static const char *mc_ra_types[] = {
 		"Indeterminate",
 		"Instruction fetch (bad)",
-		"Instruction fetch (foreign)",
+		"Instruction fetch (foreign/control memory)",
 		"Page table walk ifetch (bad)",
-		"Page table walk ifetch (foreign)",
+		"Page table walk ifetch (foreign/control memory)",
 		"Load (bad)",
 		"Store (bad)",
 		"Page table walk Load/Store (bad)",
-		"Page table walk Load/Store (foreign)",
-		"Load/Store (foreign)",
+		"Page table walk Load/Store (foreign/control memory)",
+		"Load/Store (foreign/control memory)",
 	};
 	static const char *mc_link_types[] = {
 		"Indeterminate",
@@ -285,6 +425,13 @@ void machine_check_print_event_info(struct machine_check_event *evt,
 		"Store (timeout)",
 		"Page table walk Load/Store (timeout)",
 	};
+	static const char *mc_error_class[] = {
+		"Unknown",
+		"Hardware error",
+		"Probable Hardware error (some chance of software cause)",
+		"Software error",
+		"Probable Software error (some chance of hardware cause)",
+	};
 
 	/* Print things out */
 	if (evt->version != MCE_V1) {
@@ -299,9 +446,9 @@ void machine_check_print_event_info(struct machine_check_event *evt,
 		break;
 	case MCE_SEV_WARNING:
 		level = KERN_WARNING;
-		sevstr = "";
+		sevstr = "Warning";
 		break;
-	case MCE_SEV_ERROR_SYNC:
+	case MCE_SEV_SEVERE:
 		level = KERN_ERR;
 		sevstr = "Severe";
 		break;
@@ -312,137 +459,313 @@ void machine_check_print_event_info(struct machine_check_event *evt,
 		break;
 	}
 
-	printk("%s%s Machine check interrupt [%s]\n", level, sevstr,
-	       evt->disposition == MCE_DISPOSITION_RECOVERED ?
-	       "Recovered" : "Not recovered");
-
-	if (user_mode) {
-		printk("%s  NIP: [%016llx] PID: %d Comm: %s\n", level,
-			evt->srr0, current->pid, current->comm);
-	} else {
-		printk("%s  NIP [%016llx]: %pS\n", level, evt->srr0,
-		       (void *)evt->srr0);
+	switch(evt->initiator) {
+	case MCE_INITIATOR_CPU:
+		initiator = "CPU";
+		break;
+	case MCE_INITIATOR_PCI:
+		initiator = "PCI";
+		break;
+	case MCE_INITIATOR_ISA:
+		initiator = "ISA";
+		break;
+	case MCE_INITIATOR_MEMORY:
+		initiator = "Memory";
+		break;
+	case MCE_INITIATOR_POWERMGM:
+		initiator = "Power Management";
+		break;
+	case MCE_INITIATOR_UNKNOWN:
+	default:
+		initiator = "Unknown";
+		break;
 	}
 
-	printk("%s  Initiator: %s\n", level,
-	       evt->initiator == MCE_INITIATOR_CPU ? "CPU" : "Unknown");
 	switch (evt->error_type) {
 	case MCE_ERROR_TYPE_UE:
+		err_type = "UE";
 		subtype = evt->u.ue_error.ue_error_type <
 			ARRAY_SIZE(mc_ue_types) ?
 			mc_ue_types[evt->u.ue_error.ue_error_type]
 			: "Unknown";
-		printk("%s  Error type: UE [%s]\n", level, subtype);
 		if (evt->u.ue_error.effective_address_provided)
-			printk("%s    Effective address: %016llx\n",
-			       level, evt->u.ue_error.effective_address);
+			ea = evt->u.ue_error.effective_address;
 		if (evt->u.ue_error.physical_address_provided)
-			printk("%s      Physical address: %016llx\n",
-			       level, evt->u.ue_error.physical_address);
+			pa = evt->u.ue_error.physical_address;
 		break;
 	case MCE_ERROR_TYPE_SLB:
+		err_type = "SLB";
 		subtype = evt->u.slb_error.slb_error_type <
 			ARRAY_SIZE(mc_slb_types) ?
 			mc_slb_types[evt->u.slb_error.slb_error_type]
 			: "Unknown";
-		printk("%s  Error type: SLB [%s]\n", level, subtype);
 		if (evt->u.slb_error.effective_address_provided)
-			printk("%s    Effective address: %016llx\n",
-			       level, evt->u.slb_error.effective_address);
+			ea = evt->u.slb_error.effective_address;
 		break;
 	case MCE_ERROR_TYPE_ERAT:
+		err_type = "ERAT";
 		subtype = evt->u.erat_error.erat_error_type <
 			ARRAY_SIZE(mc_erat_types) ?
 			mc_erat_types[evt->u.erat_error.erat_error_type]
 			: "Unknown";
-		printk("%s  Error type: ERAT [%s]\n", level, subtype);
 		if (evt->u.erat_error.effective_address_provided)
-			printk("%s    Effective address: %016llx\n",
-			       level, evt->u.erat_error.effective_address);
+			ea = evt->u.erat_error.effective_address;
 		break;
 	case MCE_ERROR_TYPE_TLB:
+		err_type = "TLB";
 		subtype = evt->u.tlb_error.tlb_error_type <
 			ARRAY_SIZE(mc_tlb_types) ?
 			mc_tlb_types[evt->u.tlb_error.tlb_error_type]
 			: "Unknown";
-		printk("%s  Error type: TLB [%s]\n", level, subtype);
 		if (evt->u.tlb_error.effective_address_provided)
-			printk("%s    Effective address: %016llx\n",
-			       level, evt->u.tlb_error.effective_address);
+			ea = evt->u.tlb_error.effective_address;
 		break;
 	case MCE_ERROR_TYPE_USER:
+		err_type = "User";
 		subtype = evt->u.user_error.user_error_type <
 			ARRAY_SIZE(mc_user_types) ?
 			mc_user_types[evt->u.user_error.user_error_type]
 			: "Unknown";
-		printk("%s  Error type: User [%s]\n", level, subtype);
 		if (evt->u.user_error.effective_address_provided)
-			printk("%s    Effective address: %016llx\n",
-			       level, evt->u.user_error.effective_address);
+			ea = evt->u.user_error.effective_address;
 		break;
 	case MCE_ERROR_TYPE_RA:
+		err_type = "Real address";
 		subtype = evt->u.ra_error.ra_error_type <
 			ARRAY_SIZE(mc_ra_types) ?
 			mc_ra_types[evt->u.ra_error.ra_error_type]
 			: "Unknown";
-		printk("%s  Error type: Real address [%s]\n", level, subtype);
 		if (evt->u.ra_error.effective_address_provided)
-			printk("%s    Effective address: %016llx\n",
-			       level, evt->u.ra_error.effective_address);
+			ea = evt->u.ra_error.effective_address;
 		break;
 	case MCE_ERROR_TYPE_LINK:
+		err_type = "Link";
 		subtype = evt->u.link_error.link_error_type <
 			ARRAY_SIZE(mc_link_types) ?
 			mc_link_types[evt->u.link_error.link_error_type]
 			: "Unknown";
-		printk("%s  Error type: Link [%s]\n", level, subtype);
 		if (evt->u.link_error.effective_address_provided)
-			printk("%s    Effective address: %016llx\n",
-			       level, evt->u.link_error.effective_address);
+			ea = evt->u.link_error.effective_address;
+		break;
+	case MCE_ERROR_TYPE_DCACHE:
+		err_type = "D-Cache";
+		subtype = "Unknown";
+		break;
+	case MCE_ERROR_TYPE_ICACHE:
+		err_type = "I-Cache";
+		subtype = "Unknown";
 		break;
 	default:
 	case MCE_ERROR_TYPE_UNKNOWN:
-		printk("%s  Error type: Unknown\n", level);
+		err_type = "Unknown";
+		subtype = "";
 		break;
 	}
+
+	dar_str[0] = pa_str[0] = '\0';
+	if (ea && evt->srr0 != ea) {
+		/* Load/Store address */
+		n = sprintf(dar_str, "DAR: %016llx ", ea);
+		if (pa)
+			sprintf(dar_str + n, "paddr: %016llx ", pa);
+	} else if (pa) {
+		sprintf(pa_str, " paddr: %016llx", pa);
+	}
+
+	printk("%sMCE: CPU%d: machine check (%s) %s %s %s %s[%s]\n",
+		level, evt->cpu, sevstr, in_guest ? "Guest" : "",
+		err_type, subtype, dar_str,
+		evt->disposition == MCE_DISPOSITION_RECOVERED ?
+		"Recovered" : "Not recovered");
+
+	if (in_guest || user_mode) {
+		printk("%sMCE: CPU%d: PID: %d Comm: %s %sNIP: [%016llx]%s\n",
+			level, evt->cpu, current->pid, current->comm,
+			in_guest ? "Guest " : "", evt->srr0, pa_str);
+	} else {
+		printk("%sMCE: CPU%d: NIP: [%016llx] %pS%s\n",
+			level, evt->cpu, evt->srr0, (void *)evt->srr0, pa_str);
+	}
+
+	printk("%sMCE: CPU%d: Initiator %s\n", level, evt->cpu, initiator);
+
+	subtype = evt->error_class < ARRAY_SIZE(mc_error_class) ?
+		mc_error_class[evt->error_class] : "Unknown";
+	printk("%sMCE: CPU%d: %s\n", level, evt->cpu, subtype);
+
+#ifdef CONFIG_PPC_64S_HASH_MMU
+	/* Display faulty slb contents for SLB errors. */
+	if (evt->error_type == MCE_ERROR_TYPE_SLB && !in_guest)
+		slb_dump_contents(local_paca->mce_faulty_slbs);
+#endif
 }
 EXPORT_SYMBOL_GPL(machine_check_print_event_info);
 
-uint64_t get_mce_fault_addr(struct machine_check_event *evt)
+/*
+ * This function is called in real mode. Strictly no printk's please.
+ *
+ * regs->nip and regs->msr contains srr0 and ssr1.
+ */
+DEFINE_INTERRUPT_HANDLER_NMI(machine_check_early)
 {
-	switch (evt->error_type) {
-	case MCE_ERROR_TYPE_UE:
-		if (evt->u.ue_error.effective_address_provided)
-			return evt->u.ue_error.effective_address;
-		break;
-	case MCE_ERROR_TYPE_SLB:
-		if (evt->u.slb_error.effective_address_provided)
-			return evt->u.slb_error.effective_address;
-		break;
-	case MCE_ERROR_TYPE_ERAT:
-		if (evt->u.erat_error.effective_address_provided)
-			return evt->u.erat_error.effective_address;
-		break;
-	case MCE_ERROR_TYPE_TLB:
-		if (evt->u.tlb_error.effective_address_provided)
-			return evt->u.tlb_error.effective_address;
+	long handled = 0;
+
+	hv_nmi_check_nonrecoverable(regs);
+
+	/*
+	 * See if platform is capable of handling machine check.
+	 */
+	if (ppc_md.machine_check_early)
+		handled = ppc_md.machine_check_early(regs);
+
+	return handled;
+}
+
+/* Possible meanings for HMER_DEBUG_TRIG bit being set on POWER9 */
+static enum {
+	DTRIG_UNKNOWN,
+	DTRIG_VECTOR_CI,	/* need to emulate vector CI load instr */
+	DTRIG_SUSPEND_ESCAPE,	/* need to escape from TM suspend mode */
+} hmer_debug_trig_function;
+
+static int init_debug_trig_function(void)
+{
+	int pvr;
+	struct device_node *cpun;
+	struct property *prop = NULL;
+	const char *str;
+
+	/* First look in the device tree */
+	preempt_disable();
+	cpun = of_get_cpu_node(smp_processor_id(), NULL);
+	if (cpun) {
+		of_property_for_each_string(cpun, "ibm,hmi-special-triggers",
+					    prop, str) {
+			if (strcmp(str, "bit17-vector-ci-load") == 0)
+				hmer_debug_trig_function = DTRIG_VECTOR_CI;
+			else if (strcmp(str, "bit17-tm-suspend-escape") == 0)
+				hmer_debug_trig_function = DTRIG_SUSPEND_ESCAPE;
+		}
+		of_node_put(cpun);
+	}
+	preempt_enable();
+
+	/* If we found the property, don't look at PVR */
+	if (prop)
+		goto out;
+
+	pvr = mfspr(SPRN_PVR);
+	/* Check for POWER9 Nimbus (scale-out) */
+	if ((PVR_VER(pvr) == PVR_POWER9) && (pvr & 0xe000) == 0) {
+		/* DD2.2 and later */
+		if ((pvr & 0xfff) >= 0x202)
+			hmer_debug_trig_function = DTRIG_SUSPEND_ESCAPE;
+		/* DD2.0 and DD2.1 - used for vector CI load emulation */
+		else if ((pvr & 0xfff) >= 0x200)
+			hmer_debug_trig_function = DTRIG_VECTOR_CI;
+	}
+
+ out:
+	switch (hmer_debug_trig_function) {
+	case DTRIG_VECTOR_CI:
+		pr_debug("HMI debug trigger used for vector CI load\n");
 		break;
-	case MCE_ERROR_TYPE_USER:
-		if (evt->u.user_error.effective_address_provided)
-			return evt->u.user_error.effective_address;
+	case DTRIG_SUSPEND_ESCAPE:
+		pr_debug("HMI debug trigger used for TM suspend escape\n");
 		break;
-	case MCE_ERROR_TYPE_RA:
-		if (evt->u.ra_error.effective_address_provided)
-			return evt->u.ra_error.effective_address;
+	default:
 		break;
-	case MCE_ERROR_TYPE_LINK:
-		if (evt->u.link_error.effective_address_provided)
-			return evt->u.link_error.effective_address;
+	}
+	return 0;
+}
+__initcall(init_debug_trig_function);
+
+/*
+ * Handle HMIs that occur as a result of a debug trigger.
+ * Return values:
+ * -1 means this is not a HMI cause that we know about
+ *  0 means no further handling is required
+ *  1 means further handling is required
+ */
+long hmi_handle_debugtrig(struct pt_regs *regs)
+{
+	unsigned long hmer = mfspr(SPRN_HMER);
+	long ret = 0;
+
+	/* HMER_DEBUG_TRIG bit is used for various workarounds on P9 */
+	if (!((hmer & HMER_DEBUG_TRIG)
+	      && hmer_debug_trig_function != DTRIG_UNKNOWN))
+		return -1;
+		
+	hmer &= ~HMER_DEBUG_TRIG;
+	/* HMER is a write-AND register */
+	mtspr(SPRN_HMER, ~HMER_DEBUG_TRIG);
+
+	switch (hmer_debug_trig_function) {
+	case DTRIG_VECTOR_CI:
+		/*
+		 * Now to avoid problems with soft-disable we
+		 * only do the emulation if we are coming from
+		 * host user space
+		 */
+		if (regs && user_mode(regs))
+			ret = local_paca->hmi_p9_special_emu = 1;
+
 		break;
+
 	default:
-	case MCE_ERROR_TYPE_UNKNOWN:
 		break;
 	}
-	return 0;
+
+	/*
+	 * See if any other HMI causes remain to be handled
+	 */
+	if (hmer & mfspr(SPRN_HMEER))
+		return -1;
+
+	return ret;
+}
+
+/*
+ * Return values:
+ */
+DEFINE_INTERRUPT_HANDLER_NMI(hmi_exception_realmode)
+{	
+	int ret;
+
+	local_paca->hmi_irqs++;
+
+	ret = hmi_handle_debugtrig(regs);
+	if (ret >= 0)
+		return ret;
+
+	wait_for_subcore_guest_exit();
+
+	if (ppc_md.hmi_exception_early)
+		ppc_md.hmi_exception_early(regs);
+
+	wait_for_tb_resync();
+
+	return 1;
+}
+
+void __init mce_init(void)
+{
+	struct mce_info *mce_info;
+	u64 limit;
+	int i;
+
+	limit = min(ppc64_bolted_size(), ppc64_rma_size);
+	for_each_possible_cpu(i) {
+		mce_info = memblock_alloc_try_nid(sizeof(*mce_info),
+						  __alignof__(*mce_info),
+						  MEMBLOCK_LOW_LIMIT,
+						  limit, early_cpu_to_node(i));
+		if (!mce_info)
+			goto err;
+		paca_ptrs[i]->mce_info = mce_info;
+	}
+	return;
+err:
+	panic("Failed to allocate memory for MCE event data\n");
 }
-EXPORT_SYMBOL(get_mce_fault_addr);