diff -drupN a/arch/arm64/kernel/smp.c b/arch/arm64/kernel/smp.c
--- a/arch/arm64/kernel/smp.c	2018-08-06 17:23:04.000000000 +0300
+++ b/arch/arm64/kernel/smp.c	2022-06-12 05:28:14.000000000 +0300
@@ -58,6 +58,9 @@
 #define CREATE_TRACE_POINTS
 #include <trace/events/ipi.h>
 
+DEFINE_PER_CPU_READ_MOSTLY(int, cpu_number);
+EXPORT_PER_CPU_SYMBOL(cpu_number);
+
 /*
  * as from 2.5, kernels no longer have an init_tasks structure
  * so we need some other way of telling a new secondary core
@@ -146,6 +149,7 @@ int __cpu_up(unsigned int cpu, struct ta
 	 * We need to tell the secondary core where to find its stack and the
 	 * page tables.
 	 */
+	secondary_data.task = idle;
 	secondary_data.stack = task_stack_page(idle) + THREAD_START_SP;
 	update_cpu_boot_status(CPU_MMU_OFF);
 	__flush_dcache_area(&secondary_data, sizeof(secondary_data));
@@ -170,6 +174,7 @@ int __cpu_up(unsigned int cpu, struct ta
 		pr_err("CPU%u: failed to boot: %d\n", cpu, ret);
 	}
 
+	secondary_data.task = NULL;
 	secondary_data.stack = NULL;
 	status = READ_ONCE(secondary_data.status);
 	if (ret && status) {
@@ -208,7 +213,10 @@ int __cpu_up(unsigned int cpu, struct ta
 asmlinkage void secondary_start_kernel(void)
 {
 	struct mm_struct *mm = &init_mm;
-	unsigned int cpu = smp_processor_id();
+	unsigned int cpu;
+
+	cpu = task_cpu(current);
+	set_my_cpu_offset(per_cpu_offset(cpu));
 
 	/*
 	 * All kernel threads share the same mm context; grab a
@@ -217,8 +225,6 @@ asmlinkage void secondary_start_kernel(v
 	atomic_inc(&mm->mm_count);
 	current->active_mm = mm;
 
-	set_my_cpu_offset(per_cpu_offset(smp_processor_id()));
-
 	/*
 	 * TTBR0 is only used for the identity mapping at this stage. Make it
 	 * point to zero page to avoid speculatively fetching new entries.
@@ -718,6 +724,8 @@ void __init smp_prepare_cpus(unsigned in
 	 */
 	for_each_possible_cpu(cpu) {
 
+		per_cpu(cpu_number, cpu) = cpu;
+
 		if (cpu == smp_processor_id())
 			continue;
 
@@ -815,6 +823,8 @@ static void ipi_cpu_stop(unsigned int cp
 
 	local_irq_disable();
 
+	flush_cache_all();
+
 	while (1)
 		cpu_relax();
 }