Merge branch 'fixes' into next

Merge our fixes branch to bring in some changes that conflict with
upcoming next content.

arch/powerpc/Kconfig

@@ -163,7 +163,6 @@ config PPC
 	select ARCH_WANT_IRQS_OFF_ACTIVATE_MM
 	select ARCH_WANT_LD_ORPHAN_WARN
 	select ARCH_WANTS_MODULES_DATA_IN_VMALLOC	if PPC_BOOK3S_32 || PPC_8xx
-	select ARCH_WANTS_NO_INSTR
 	select ARCH_WEAK_RELEASE_ACQUIRE
 	select BINFMT_ELF
 	select BUILDTIME_TABLE_SORT

arch/powerpc/boot/wrapper

@@ -210,6 +210,10 @@ ld_version()
 	gsub(".*version ", "");
 	gsub("-.*", "");
 	split($1,a, ".");
+	if( length(a[3]) == "8" )
+		# a[3] is probably a date of format yyyymmdd used for release snapshots. We
+		# can assume it to be zero as it does not signify a new version as such.
+		a[3] = 0;
 	print a[1]*100000000 + a[2]*1000000 + a[3]*10000;
 	exit
     }'

arch/powerpc/include/asm/book3s/64/tlbflush.h

@@ -97,6 +97,8 @@ static inline void tlb_flush(struct mmu_gather *tlb)
 {
 	if (radix_enabled())
 		radix__tlb_flush(tlb);
+
+	return hash__tlb_flush(tlb);
 }
 
 #ifdef CONFIG_SMP

arch/powerpc/include/asm/hw_irq.h

@@ -175,6 +175,15 @@ static inline notrace unsigned long irq_soft_mask_or_return(unsigned long mask)
 	return flags;
 }
 
+static inline notrace unsigned long irq_soft_mask_andc_return(unsigned long mask)
+{
+	unsigned long flags = irq_soft_mask_return();
+
+	irq_soft_mask_set(flags & ~mask);
+
+	return flags;
+}
+
 static inline unsigned long arch_local_save_flags(void)
 {
 	return irq_soft_mask_return();
@@ -194,7 +203,7 @@ static inline void arch_local_irq_enable(void)
 
 static inline unsigned long arch_local_irq_save(void)
 {
-	return irq_soft_mask_set_return(IRQS_DISABLED);
+	return irq_soft_mask_or_return(IRQS_DISABLED);
 }
 
 static inline bool arch_irqs_disabled_flags(unsigned long flags)
@@ -333,10 +342,11 @@ bool power_pmu_wants_prompt_pmi(void);
  * is a different soft-masked interrupt pending that requires hard
  * masking.
  */
-static inline bool should_hard_irq_enable(void)
+static inline bool should_hard_irq_enable(struct pt_regs *regs)
 {
 	if (IS_ENABLED(CONFIG_PPC_IRQ_SOFT_MASK_DEBUG)) {
-		WARN_ON(irq_soft_mask_return() == IRQS_ENABLED);
+		WARN_ON(irq_soft_mask_return() != IRQS_ALL_DISABLED);
+		WARN_ON(!(get_paca()->irq_happened & PACA_IRQ_HARD_DIS));
 		WARN_ON(mfmsr() & MSR_EE);
 	}
 
@@ -349,8 +359,17 @@ static inline bool should_hard_irq_enable(void)
 	 *
 	 * TODO: Add test for 64e
 	 */
-	if (IS_ENABLED(CONFIG_PPC_BOOK3S_64) && !power_pmu_wants_prompt_pmi())
-		return false;
+	if (IS_ENABLED(CONFIG_PPC_BOOK3S_64)) {
+		if (!power_pmu_wants_prompt_pmi())
+			return false;
+		/*
+		 * If PMIs are disabled then IRQs should be disabled as well,
+		 * so we shouldn't see this condition, check for it just in
+		 * case because we are about to enable PMIs.
+		 */
+		if (WARN_ON_ONCE(regs->softe & IRQS_PMI_DISABLED))
+			return false;
+	}
 
 	if (get_paca()->irq_happened & PACA_IRQ_MUST_HARD_MASK)
 		return false;
@@ -360,18 +379,16 @@ static inline bool should_hard_irq_enable(void)
 
 /*
  * Do the hard enabling, only call this if should_hard_irq_enable is true.
+ * This allows PMI interrupts to profile irq handlers.
  */
 static inline void do_hard_irq_enable(void)
 {
-	if (IS_ENABLED(CONFIG_PPC_IRQ_SOFT_MASK_DEBUG)) {
-		WARN_ON(irq_soft_mask_return() == IRQS_ENABLED);
-		WARN_ON(get_paca()->irq_happened & PACA_IRQ_MUST_HARD_MASK);
-		WARN_ON(mfmsr() & MSR_EE);
-	}
 	/*
-	 * This allows PMI interrupts (and watchdog soft-NMIs) through.
-	 * There is no other reason to enable this way.
+	 * Asynch interrupts come in with IRQS_ALL_DISABLED,
+	 * PACA_IRQ_HARD_DIS, and MSR[EE]=0.
 	 */
+	if (IS_ENABLED(CONFIG_PPC_BOOK3S_64))
+		irq_soft_mask_andc_return(IRQS_PMI_DISABLED);
 	get_paca()->irq_happened &= ~PACA_IRQ_HARD_DIS;
 	__hard_irq_enable();
 }
@@ -454,7 +471,7 @@ static inline bool arch_irq_disabled_regs(struct pt_regs *regs)
 	return !(regs->msr & MSR_EE);
 }
 
-static __always_inline bool should_hard_irq_enable(void)
+static __always_inline bool should_hard_irq_enable(struct pt_regs *regs)
 {
 	return false;
 }

arch/powerpc/include/asm/imc-pmu.h

@@ -137,7 +137,7 @@ struct imc_pmu {
  * are inited.
  */
 struct imc_pmu_ref {
-	struct mutex lock;
+	spinlock_t lock;
 	unsigned int id;
 	int refc;
 };

arch/powerpc/kernel/dbell.c

@@ -27,7 +27,7 @@ DEFINE_INTERRUPT_HANDLER_ASYNC(doorbell_exception)
 
 	ppc_msgsync();
 
-	if (should_hard_irq_enable())
+	if (should_hard_irq_enable(regs))
 		do_hard_irq_enable();
 
 	kvmppc_clear_host_ipi(smp_processor_id());

arch/powerpc/kernel/head_85xx.S

@@ -864,7 +864,7 @@ _GLOBAL(load_up_spe)
  * SPE unavailable trap from kernel - print a message, but let
  * the task use SPE in the kernel until it returns to user mode.
  */
-KernelSPE:
+SYM_FUNC_START_LOCAL(KernelSPE)
 	lwz	r3,_MSR(r1)
 	oris	r3,r3,MSR_SPE@h
 	stw	r3,_MSR(r1)	/* enable use of SPE after return */
@@ -881,6 +881,7 @@ KernelSPE:
 #endif
 	.align	4,0
 
+SYM_FUNC_END(KernelSPE)
 #endif /* CONFIG_SPE */
 
 /*

arch/powerpc/kernel/interrupt.c

@@ -50,16 +50,18 @@ static inline bool exit_must_hard_disable(void)
  */
 static notrace __always_inline bool prep_irq_for_enabled_exit(bool restartable)
 {
+	bool must_hard_disable = (exit_must_hard_disable() || !restartable);
+
 	/* This must be done with RI=1 because tracing may touch vmaps */
 	trace_hardirqs_on();
 
-	if (exit_must_hard_disable() || !restartable)
+	if (must_hard_disable)
 		__hard_EE_RI_disable();
 
 #ifdef CONFIG_PPC64
 	/* This pattern matches prep_irq_for_idle */
 	if (unlikely(lazy_irq_pending_nocheck())) {
-		if (exit_must_hard_disable() || !restartable) {
+		if (must_hard_disable) {
 			local_paca->irq_happened |= PACA_IRQ_HARD_DIS;
 			__hard_RI_enable();
 		}

arch/powerpc/kernel/irq.c

@@ -238,7 +238,7 @@ static void __do_irq(struct pt_regs *regs, unsigned long oldsp)
 	irq = static_call(ppc_get_irq)();
 
 	/* We can hard enable interrupts now to allow perf interrupts */
-	if (should_hard_irq_enable())
+	if (should_hard_irq_enable(regs))
 		do_hard_irq_enable();
 
 	/* And finally process it */

arch/powerpc/kernel/time.c

@@ -515,7 +515,7 @@ DEFINE_INTERRUPT_HANDLER_ASYNC(timer_interrupt)
 	}
 
 	/* Conditionally hard-enable interrupts. */
-	if (should_hard_irq_enable()) {
+	if (should_hard_irq_enable(regs)) {
 		/*
 		 * Ensure a positive value is written to the decrementer, or
 		 * else some CPUs will continue to take decrementer exceptions.

arch/powerpc/kexec/file_load_64.c

@@ -26,6 +26,7 @@
 #include <asm/firmware.h>
 #include <asm/kexec_ranges.h>
 #include <asm/crashdump-ppc64.h>
+#include <asm/mmzone.h>
 #include <asm/prom.h>
 
 struct umem_info {
@@ -989,10 +990,13 @@ unsigned int kexec_extra_fdt_size_ppc64(struct kimage *image)
 	 * linux,drconf-usable-memory properties. Get an approximate on the
 	 * number of usable memory entries and use for FDT size estimation.
 	 */
-	usm_entries = ((memblock_end_of_DRAM() / drmem_lmb_size()) +
-		       (2 * (resource_size(&crashk_res) / drmem_lmb_size())));
-
-	extra_size = (unsigned int)(usm_entries * sizeof(u64));
+	if (drmem_lmb_size()) {
+		usm_entries = ((memory_hotplug_max() / drmem_lmb_size()) +
+			       (2 * (resource_size(&crashk_res) / drmem_lmb_size())));
+		extra_size = (unsigned int)(usm_entries * sizeof(u64));
+	} else {
+		extra_size = 0;
+	}
 
 	/*
 	 * Get the number of CPU nodes in the current DT. This allows to

arch/powerpc/kvm/booke.c

@@ -912,16 +912,15 @@ static int kvmppc_handle_debug(struct kvm_vcpu *vcpu)
 
 static void kvmppc_fill_pt_regs(struct pt_regs *regs)
 {
-	ulong r1, ip, msr, lr;
+	ulong r1, msr, lr;
 
 	asm("mr %0, 1" : "=r"(r1));
 	asm("mflr %0" : "=r"(lr));
 	asm("mfmsr %0" : "=r"(msr));
-	asm("bl 1f; 1: mflr %0" : "=r"(ip));
 
 	memset(regs, 0, sizeof(*regs));
 	regs->gpr[1] = r1;
-	regs->nip = ip;
+	regs->nip = _THIS_IP_;
 	regs->msr = msr;
 	regs->link = lr;
 }

arch/powerpc/mm/book3s64/hash_utils.c

@@ -1012,7 +1012,7 @@ static void __init hash_init_partition_table(phys_addr_t hash_table,
 
 void hpt_clear_stress(void);
 static struct timer_list stress_hpt_timer;
-void stress_hpt_timer_fn(struct timer_list *timer)
+static void stress_hpt_timer_fn(struct timer_list *timer)
 {
 	int next_cpu;
 

arch/powerpc/mm/book3s64/radix_pgtable.c

@@ -234,6 +234,14 @@ void radix__mark_rodata_ro(void)
 	end = (unsigned long)__end_rodata;
 
 	radix__change_memory_range(start, end, _PAGE_WRITE);
+
+	for (start = PAGE_OFFSET; start < (unsigned long)_stext; start += PAGE_SIZE) {
+		end = start + PAGE_SIZE;
+		if (overlaps_interrupt_vector_text(start, end))
+			radix__change_memory_range(start, end, _PAGE_WRITE);
+		else
+			break;
+	}
 }
 
 void radix__mark_initmem_nx(void)
@@ -262,6 +270,22 @@ print_mapping(unsigned long start, unsigned long end, unsigned long size, bool e
 static unsigned long next_boundary(unsigned long addr, unsigned long end)
 {
 #ifdef CONFIG_STRICT_KERNEL_RWX
+	unsigned long stext_phys;
+
+	stext_phys = __pa_symbol(_stext);
+
+	// Relocatable kernel running at non-zero real address
+	if (stext_phys != 0) {
+		// The end of interrupts code at zero is a rodata boundary
+		unsigned long end_intr = __pa_symbol(__end_interrupts) - stext_phys;
+		if (addr < end_intr)
+			return end_intr;
+
+		// Start of relocated kernel text is a rodata boundary
+		if (addr < stext_phys)
+			return stext_phys;
+	}
+
 	if (addr < __pa_symbol(__srwx_boundary))
 		return __pa_symbol(__srwx_boundary);
 #endif