1 files changed, 179 insertions, 58 deletions

diff --git a/arch/powerpc/platforms/powernv/smp.c b/arch/powerpc/platforms/powernv/smp.c
index 40dae96f7e20..8f41ef364fc6 100644
--- a/arch/powerpc/platforms/powernv/smp.c
+++ b/arch/powerpc/platforms/powernv/smp.c
@@ -1,12 +1,8 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
 /*
  * SMP support for PowerNV machines.
  *
  * Copyright 2011 IBM Corp.
- *
- * 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.
  */
 
 #include <linux/kernel.h>
@@ -32,11 +28,15 @@
 #include <asm/xive.h>
 #include <asm/opal.h>
 #include <asm/runlatch.h>
-#include <asm/code-patching.h>
+#include <asm/text-patching.h>
 #include <asm/dbell.h>
 #include <asm/kvm_ppc.h>
 #include <asm/ppc-opcode.h>
 #include <asm/cpuidle.h>
+#include <asm/kexec.h>
+#include <asm/reg.h>
+#include <asm/powernv.h>
+#include <asm/systemcfg.h>
 
 #include "powernv.h"
 
@@ -44,11 +44,18 @@
 #include <asm/udbg.h>
 #define DBG(fmt...) udbg_printf(fmt)
 #else
-#define DBG(fmt...)
+#define DBG(fmt...) do { } while (0)
 #endif
 
 static void pnv_smp_setup_cpu(int cpu)
 {
+	/*
+	 * P9 workaround for CI vector load (see traps.c),
+	 * enable the corresponding HMI interrupt
+	 */
+	if (pvr_version_is(PVR_POWER9))
+		mtspr(SPRN_HMEER, mfspr(SPRN_HMEER) | PPC_BIT(17));
+
 	if (xive_enabled())
 		xive_smp_setup_cpu();
 	else if (cpu != boot_cpuid)
@@ -57,7 +64,7 @@ static void pnv_smp_setup_cpu(int cpu)
 
 static int pnv_smp_kick_cpu(int nr)
 {
-	unsigned int pcpu = get_hard_smp_processor_id(nr);
+	unsigned int pcpu;
 	unsigned long start_here =
 			__pa(ppc_function_entry(generic_secondary_smp_init));
 	long rc;
@@ -66,11 +73,12 @@ static int pnv_smp_kick_cpu(int nr)
 	if (nr < 0 || nr >= nr_cpu_ids)
 		return -EINVAL;
 
+	pcpu = get_hard_smp_processor_id(nr);
 	/*
 	 * If we already started or OPAL is not supported, we just
 	 * kick the CPU via the PACA
 	 */
-	if (paca[nr].cpu_start || !firmware_has_feature(FW_FEATURE_OPAL))
+	if (paca_ptrs[nr]->cpu_start || !firmware_has_feature(FW_FEATURE_OPAL))
 		goto kick;
 
 	/*
@@ -129,32 +137,42 @@ static int pnv_smp_cpu_disable(void)
 	 * the generic fixup_irqs. --BenH.
 	 */
 	set_cpu_online(cpu, false);
-	vdso_data->processorCount--;
+#ifdef CONFIG_PPC64_PROC_SYSTEMCFG
+	systemcfg->processorCount--;
+#endif
 	if (cpu == boot_cpuid)
 		boot_cpuid = cpumask_any(cpu_online_mask);
 	if (xive_enabled())
 		xive_smp_disable_cpu();
 	else
 		xics_migrate_irqs_away();
+
+	cleanup_cpu_mmu_context();
+
 	return 0;
 }
 
-static void pnv_smp_cpu_kill_self(void)
+static void pnv_flush_interrupts(void)
 {
+	if (cpu_has_feature(CPU_FTR_ARCH_300)) {
+		if (xive_enabled())
+			xive_flush_interrupt();
+		else
+			icp_opal_flush_interrupt();
+	} else {
+		icp_native_flush_interrupt();
+	}
+}
+
+static void pnv_cpu_offline_self(void)
+{
+	unsigned long srr1, unexpected_mask, wmask;
 	unsigned int cpu;
-	unsigned long srr1, wmask;
+	u64 lpcr_val;
 
 	/* Standard hot unplug procedure */
-	/*
-	 * This hard disables local interurpts, ensuring we have no lazy
-	 * irqs pending.
-	 */
-	WARN_ON(irqs_disabled());
-	hard_irq_disable();
-	WARN_ON(lazy_irq_pending());
 
 	idle_task_exit();
-	current->active_mm = NULL; /* for sanity */
 	cpu = smp_processor_id();
 	DBG("CPU%d offline\n", cpu);
 	generic_set_cpu_dead(cpu);
@@ -164,11 +182,39 @@ static void pnv_smp_cpu_kill_self(void)
 	if (cpu_has_feature(CPU_FTR_ARCH_207S))
 		wmask = SRR1_WAKEMASK_P8;
 
-	/* We don't want to take decrementer interrupts while we are offline,
-	 * so clear LPCR:PECE1. We keep PECE2 (and LPCR_PECE_HVEE on P9)
-	 * enabled as to let IPIs in.
+	/*
+	 * This turns the irq soft-disabled state we're called with, into a
+	 * hard-disabled state with pending irq_happened interrupts cleared.
+	 *
+	 * PACA_IRQ_DEC   - Decrementer should be ignored.
+	 * PACA_IRQ_HMI   - Can be ignored, processing is done in real mode.
+	 * PACA_IRQ_DBELL, EE, PMI - Unexpected.
 	 */
-	mtspr(SPRN_LPCR, mfspr(SPRN_LPCR) & ~(u64)LPCR_PECE1);
+	hard_irq_disable();
+	if (generic_check_cpu_restart(cpu))
+		goto out;
+
+	unexpected_mask = ~(PACA_IRQ_DEC | PACA_IRQ_HMI | PACA_IRQ_HARD_DIS);
+	if (local_paca->irq_happened & unexpected_mask) {
+		if (local_paca->irq_happened & PACA_IRQ_EE)
+			pnv_flush_interrupts();
+		DBG("CPU%d Unexpected exit while offline irq_happened=%lx!\n",
+				cpu, local_paca->irq_happened);
+	}
+	local_paca->irq_happened = PACA_IRQ_HARD_DIS;
+
+	/*
+	 * We don't want to take decrementer interrupts while we are
+	 * offline, so clear LPCR:PECE1. We keep PECE2 (and
+	 * LPCR_PECE_HVEE on P9) enabled so as to let IPIs in.
+	 *
+	 * If the CPU gets woken up by a special wakeup, ensure that
+	 * the SLW engine sets LPCR with decrementer bit cleared, else
+	 * the CPU will come back to the kernel due to a spurious
+	 * wakeup.
+	 */
+	lpcr_val = mfspr(SPRN_LPCR) & ~(u64)LPCR_PECE1;
+	pnv_program_cpu_hotplug_lpcr(cpu, lpcr_val);
 
 	while (!generic_check_cpu_restart(cpu)) {
 		/*
@@ -178,10 +224,11 @@ static void pnv_smp_cpu_kill_self(void)
 		 * for coming online, which are handled via
 		 * generic_check_cpu_restart() calls.
 		 */
-		kvmppc_set_host_ipi(cpu, 0);
+		kvmppc_clear_host_ipi(cpu);
 
 		srr1 = pnv_cpu_offline(cpu);
 
+		WARN_ON_ONCE(!irqs_disabled());
 		WARN_ON(lazy_irq_pending());
 
 		/*
@@ -197,19 +244,36 @@ static void pnv_smp_cpu_kill_self(void)
 		 */
 		if (((srr1 & wmask) == SRR1_WAKEEE) ||
 		    ((srr1 & wmask) == SRR1_WAKEHVI)) {
-			if (cpu_has_feature(CPU_FTR_ARCH_300)) {
-				if (xive_enabled())
-					xive_flush_interrupt();
-				else
-					icp_opal_flush_interrupt();
-			} else
-				icp_native_flush_interrupt();
+			pnv_flush_interrupts();
 		} else if ((srr1 & wmask) == SRR1_WAKEHDBELL) {
 			unsigned long msg = PPC_DBELL_TYPE(PPC_DBELL_SERVER);
 			asm volatile(PPC_MSGCLR(%0) : : "r" (msg));
+		} else if ((srr1 & wmask) == SRR1_WAKERESET) {
+			irq_set_pending_from_srr1(srr1);
+			/* Does not return */
 		}
+
 		smp_mb();
 
+		/*
+		 * For kdump kernels, we process the ipi and jump to
+		 * crash_ipi_callback
+		 */
+		if (kdump_in_progress()) {
+			/*
+			 * If we got to this point, we've not used
+			 * NMI's, otherwise we would have gone
+			 * via the SRR1_WAKERESET path. We are
+			 * using regular IPI's for waking up offline
+			 * threads.
+			 */
+			struct pt_regs regs;
+
+			ppc_save_regs(&regs);
+			crash_ipi_callback(&regs);
+			/* Does not return */
+		}
+
 		if (cpu_core_split_required())
 			continue;
 
@@ -219,8 +283,16 @@ static void pnv_smp_cpu_kill_self(void)
 
 	}
 
-	/* Re-enable decrementer interrupts */
-	mtspr(SPRN_LPCR, mfspr(SPRN_LPCR) | LPCR_PECE1);
+	/*
+	 * Re-enable decrementer interrupts in LPCR.
+	 *
+	 * Further, we want stop states to be woken up by decrementer
+	 * for non-hotplug cases. So program the LPCR via stop api as
+	 * well.
+	 */
+	lpcr_val = mfspr(SPRN_LPCR) | (u64)LPCR_PECE1;
+	pnv_program_cpu_hotplug_lpcr(cpu, lpcr_val);
+out:
 	DBG("CPU%d coming online...\n", cpu);
 }
 
@@ -258,23 +330,6 @@ static void pnv_cause_ipi(int cpu)
 	ic_cause_ipi(cpu);
 }
 
-static void pnv_p9_dd1_cause_ipi(int cpu)
-{
-	int this_cpu = get_cpu();
-
-	/*
-	 * POWER9 DD1 has a global addressed msgsnd, but for now we restrict
-	 * IPIs to same core, because it requires additional synchronization
-	 * for inter-core doorbells which we do not implement.
-	 */
-	if (cpumask_test_cpu(cpu, cpu_sibling_mask(this_cpu)))
-		doorbell_global_ipi(cpu);
-	else
-		ic_cause_ipi(cpu);
-
-	put_cpu();
-}
-
 static void __init pnv_smp_probe(void)
 {
 	if (xive_enabled())
@@ -286,15 +341,74 @@ static void __init pnv_smp_probe(void)
 		ic_cause_ipi = smp_ops->cause_ipi;
 		WARN_ON(!ic_cause_ipi);
 
-		if (cpu_has_feature(CPU_FTR_ARCH_300)) {
-			if (cpu_has_feature(CPU_FTR_POWER9_DD1))
-				smp_ops->cause_ipi = pnv_p9_dd1_cause_ipi;
-			else
-				smp_ops->cause_ipi = doorbell_global_ipi;
-		} else {
+		if (cpu_has_feature(CPU_FTR_ARCH_300))
+			smp_ops->cause_ipi = doorbell_global_ipi;
+		else
 			smp_ops->cause_ipi = pnv_cause_ipi;
+	}
+}
+
+noinstr static int pnv_system_reset_exception(struct pt_regs *regs)
+{
+	if (smp_handle_nmi_ipi(regs))
+		return 1;
+	return 0;
+}
+
+static int pnv_cause_nmi_ipi(int cpu)
+{
+	int64_t rc;
+
+	if (cpu >= 0) {
+		int h = get_hard_smp_processor_id(cpu);
+
+		if (opal_check_token(OPAL_QUIESCE))
+			opal_quiesce(QUIESCE_HOLD, h);
+
+		rc = opal_signal_system_reset(h);
+
+		if (opal_check_token(OPAL_QUIESCE))
+			opal_quiesce(QUIESCE_RESUME, h);
+
+		if (rc != OPAL_SUCCESS)
+			return 0;
+		return 1;
+
+	} else if (cpu == NMI_IPI_ALL_OTHERS) {
+		bool success = true;
+		int c;
+
+		if (opal_check_token(OPAL_QUIESCE))
+			opal_quiesce(QUIESCE_HOLD, -1);
+
+		/*
+		 * We do not use broadcasts (yet), because it's not clear
+		 * exactly what semantics Linux wants or the firmware should
+		 * provide.
+		 */
+		for_each_online_cpu(c) {
+			if (c == smp_processor_id())
+				continue;
+
+			rc = opal_signal_system_reset(
+						get_hard_smp_processor_id(c));
+			if (rc != OPAL_SUCCESS)
+				success = false;
 		}
+
+		if (opal_check_token(OPAL_QUIESCE))
+			opal_quiesce(QUIESCE_RESUME, -1);
+
+		if (success)
+			return 1;
+
+		/*
+		 * Caller will fall back to doorbells, which may pick
+		 * up the remainders.
+		 */
 	}
+
+	return 0;
 }
 
 static struct smp_ops_t pnv_smp_ops = {
@@ -309,15 +423,22 @@ static struct smp_ops_t pnv_smp_ops = {
 #ifdef CONFIG_HOTPLUG_CPU
 	.cpu_disable	= pnv_smp_cpu_disable,
 	.cpu_die	= generic_cpu_die,
+	.cpu_offline_self = pnv_cpu_offline_self,
 #endif /* CONFIG_HOTPLUG_CPU */
 };
 
 /* This is called very early during platform setup_arch */
 void __init pnv_smp_init(void)
 {
+	if (opal_check_token(OPAL_SIGNAL_SYSTEM_RESET)) {
+		ppc_md.system_reset_exception = pnv_system_reset_exception;
+		pnv_smp_ops.cause_nmi_ipi = pnv_cause_nmi_ipi;
+	}
 	smp_ops = &pnv_smp_ops;
 
 #ifdef CONFIG_HOTPLUG_CPU
-	ppc_md.cpu_die	= pnv_smp_cpu_kill_self;
+#ifdef CONFIG_CRASH_DUMP
+	crash_wake_offline = 1;
+#endif
 #endif
 }