Merge tag 'powerpc-6.2-4' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux

Pull powerpc fixes from Michael Ellerman:
 "It's a bit of a big batch for rc6, but just because I didn't send any
  fixes the last week or two while I was on vacation, next week should
  be quieter:

   - Fix a few objtool warnings since we recently enabled objtool.

   - Fix a deadlock with the hash MMU vs perf record.

   - Fix perf profiling of asynchronous interrupt handlers.

   - Revert the IMC PMU nest_init_lock to being a mutex.

   - Two commits fixing problems with the kexec_file FDT size
     estimation.

   - Two commits fixing problems with strict RWX vs kernels running at
     non-zero.

   - Reconnect tlb_flush() to hash__tlb_flush()

  Thanks to Kajol Jain, Nicholas Piggin, Sachin Sant Sathvika Vasireddy,
  and Sourabh Jain"

* tag 'powerpc-6.2-4' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
  powerpc/64s: Reconnect tlb_flush() to hash__tlb_flush()
  powerpc/kexec_file: Count hot-pluggable memory in FDT estimate
  powerpc/64s/radix: Fix RWX mapping with relocated kernel
  powerpc/64s/radix: Fix crash with unaligned relocated kernel
  powerpc/kexec_file: Fix division by zero in extra size estimation
  powerpc/imc-pmu: Revert nest_init_lock to being a mutex
  powerpc/64: Fix perf profiling asynchronous interrupt handlers
  powerpc/64s: Fix local irq disable when PMIs are disabled
  powerpc/kvm: Fix unannotated intra-function call warning
  powerpc/85xx: Fix unannotated intra-function call warning

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

@@ -173,6 +173,15 @@ 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();
@@ -192,7 +201,7 @@ 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)
@@ -331,10 +340,11 @@ 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);
 	}
 
@@ -347,8 +357,17 @@ 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())
+	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;
@@ -358,18 +377,16 @@ 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();
 }
@@ -452,7 +469,7 @@ 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/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/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

@@ -989,10 +989,13 @@ 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()) +
+	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/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

arch/powerpc/perf/imc-pmu.c

@@ -22,7 +22,7 @@
  * Used to avoid races in counting the nest-pmu units during hotplug
  * register and unregister
  */
-static DEFINE_SPINLOCK(nest_init_lock);
+static DEFINE_MUTEX(nest_init_lock);
 static DEFINE_PER_CPU(struct imc_pmu_ref *, local_nest_imc_refc);
 static struct imc_pmu **per_nest_pmu_arr;
 static cpumask_t nest_imc_cpumask;
@@ -1629,7 +1629,7 @@ static void imc_common_mem_free(struct imc_pmu *pmu_ptr)
 static void imc_common_cpuhp_mem_free(struct imc_pmu *pmu_ptr)
 {
 	if (pmu_ptr->domain == IMC_DOMAIN_NEST) {
-		spin_lock(&nest_init_lock);
+		mutex_lock(&nest_init_lock);
 		if (nest_pmus == 1) {
 			cpuhp_remove_state(CPUHP_AP_PERF_POWERPC_NEST_IMC_ONLINE);
 			kfree(nest_imc_refc);
@@ -1639,7 +1639,7 @@ static void imc_common_cpuhp_mem_free(struct imc_pmu *pmu_ptr)
 
 		if (nest_pmus > 0)
 			nest_pmus--;
-		spin_unlock(&nest_init_lock);
+		mutex_unlock(&nest_init_lock);
 	}
 
 	/* Free core_imc memory */
@@ -1796,11 +1796,11 @@ int init_imc_pmu(struct device_node *parent, struct imc_pmu *pmu_ptr, int pmu_id
 		* rest. To handle the cpuhotplug callback unregister, we track
 		* the number of nest pmus in "nest_pmus".
 		*/
-		spin_lock(&nest_init_lock);
+		mutex_lock(&nest_init_lock);
 		if (nest_pmus == 0) {
 			ret = init_nest_pmu_ref();
 			if (ret) {
-				spin_unlock(&nest_init_lock);
+				mutex_unlock(&nest_init_lock);
 				kfree(per_nest_pmu_arr);
 				per_nest_pmu_arr = NULL;
 				goto err_free_mem;
@@ -1808,7 +1808,7 @@ int init_imc_pmu(struct device_node *parent, struct imc_pmu *pmu_ptr, int pmu_id
 			/* Register for cpu hotplug notification. */
 			ret = nest_pmu_cpumask_init();
 			if (ret) {
-				spin_unlock(&nest_init_lock);
+				mutex_unlock(&nest_init_lock);
 				kfree(nest_imc_refc);
 				kfree(per_nest_pmu_arr);
 				per_nest_pmu_arr = NULL;
@@ -1816,7 +1816,7 @@ int init_imc_pmu(struct device_node *parent, struct imc_pmu *pmu_ptr, int pmu_id
 			}
 		}
 		nest_pmus++;
-		spin_unlock(&nest_init_lock);
+		mutex_unlock(&nest_init_lock);
 		break;
 	case IMC_DOMAIN_CORE:
 		ret = core_imc_pmu_cpumask_init();