diff --git a/arch_init.c b/arch_init.c
index 0471cd5a6bf41c31bcb0461bc5d7e177fd4ad87a..e47e1399bb5fcf961c3c5d95a3bc4a7469b81842 100644
--- a/arch_init.c
+++ b/arch_init.c
@@ -1196,15 +1196,14 @@ static void mig_sleep_cpu(void *opq)
    much time in the VM. The migration thread will try to catchup.
    Workload will experience a performance drop.
 */
-static void mig_throttle_cpu_down(CPUState *cpu, void *data)
-{
-    async_run_on_cpu(cpu, mig_sleep_cpu, NULL);
-}
-
 static void mig_throttle_guest_down(void)
 {
+    CPUState *cpu;
+
     qemu_mutex_lock_iothread();
-    qemu_for_each_cpu(mig_throttle_cpu_down, NULL);
+    CPU_FOREACH(cpu) {
+        async_run_on_cpu(cpu, mig_sleep_cpu, NULL);
+    }
     qemu_mutex_unlock_iothread();
 }
 
diff --git a/cpus.c b/cpus.c
index 363d392cd99d087d3b30d29000ee9629ef095436..e566297bd30c12d81724221b0cccb83ba66c86ac 100644
--- a/cpus.c
+++ b/cpus.c
@@ -854,12 +854,6 @@ static void *qemu_dummy_cpu_thread_fn(void *arg)
 
 static void tcg_exec_all(void);
 
-static void tcg_signal_cpu_creation(CPUState *cpu, void *data)
-{
-    cpu->thread_id = qemu_get_thread_id();
-    cpu->created = true;
-}
-
 static void *qemu_tcg_cpu_thread_fn(void *arg)
 {
     CPUState *cpu = arg;
@@ -868,7 +862,10 @@ static void *qemu_tcg_cpu_thread_fn(void *arg)
     qemu_thread_get_self(cpu->thread);
 
     qemu_mutex_lock(&qemu_global_mutex);
-    qemu_for_each_cpu(tcg_signal_cpu_creation, NULL);
+    CPU_FOREACH(cpu) {
+        cpu->thread_id = qemu_get_thread_id();
+        cpu->created = true;
+    }
     qemu_cond_signal(&qemu_cpu_cond);
 
     /* wait for initial kick-off after machine start */
diff --git a/exec.c b/exec.c
index ca2a504f16788e9b99990600213ab609f9ed9530..87b0b39b904154bc1deb3ff640b9fb56e7ed09c6 100644
--- a/exec.c
+++ b/exec.c
@@ -362,15 +362,6 @@ CPUState *qemu_get_cpu(int index)
     return NULL;
 }
 
-void qemu_for_each_cpu(void (*func)(CPUState *cpu, void *data), void *data)
-{
-    CPUState *cpu;
-
-    CPU_FOREACH(cpu) {
-        func(cpu, data);
-    }
-}
-
 void cpu_exec_init(CPUArchState *env)
 {
     CPUState *cpu = ENV_GET_CPU(env);
diff --git a/hw/acpi/piix4.c b/hw/acpi/piix4.c
index 613d98736a12f793161b4a7c90e1eee4abc84a44..0b8d1d9da9c10b4559bf5d06a492a875bc3b018a 100644
--- a/hw/acpi/piix4.c
+++ b/hw/acpi/piix4.c
@@ -667,22 +667,14 @@ static void piix4_cpu_added_req(Notifier *n, void *opaque)
     piix4_cpu_hotplug_req(s, CPU(opaque), PLUG);
 }
 
-static void piix4_init_cpu_status(CPUState *cpu, void *data)
-{
-    CPUStatus *g = (CPUStatus *)data;
-    CPUClass *k = CPU_GET_CLASS(cpu);
-    int64_t id = k->get_arch_id(cpu);
-
-    g_assert((id / 8) < PIIX4_PROC_LEN);
-    g->sts[id / 8] |= (1 << (id % 8));
-}
-
 static int piix4_device_hotplug(DeviceState *qdev, PCIDevice *dev,
                                 PCIHotplugState state);
 
 static void piix4_acpi_system_hot_add_init(MemoryRegion *parent,
                                            PCIBus *bus, PIIX4PMState *s)
 {
+    CPUState *cpu;
+
     memory_region_init_io(&s->io_gpe, OBJECT(s), &piix4_gpe_ops, s,
                           "acpi-gpe0", GPE_LEN);
     memory_region_add_subregion(parent, GPE_BASE, &s->io_gpe);
@@ -693,7 +685,13 @@ static void piix4_acpi_system_hot_add_init(MemoryRegion *parent,
                                 &s->io_pci);
     pci_bus_hotplug(bus, piix4_device_hotplug, DEVICE(s));
 
-    qemu_for_each_cpu(piix4_init_cpu_status, &s->gpe_cpu);
+    CPU_FOREACH(cpu) {
+        CPUClass *cc = CPU_GET_CLASS(cpu);
+        int64_t id = cc->get_arch_id(cpu);
+
+        g_assert((id / 8) < PIIX4_PROC_LEN);
+        s->gpe_cpu.sts[id / 8] |= (1 << (id % 8));
+    }
     memory_region_init_io(&s->io_cpu, OBJECT(s), &cpu_hotplug_ops, s,
                           "acpi-cpu-hotplug", PIIX4_PROC_LEN);
     memory_region_add_subregion(parent, PIIX4_PROC_BASE, &s->io_cpu);
diff --git a/include/qom/cpu.h b/include/qom/cpu.h
index 79f7c8709d32c8f724c0fab58a2646b80af209a0..7739e000679f85dd0f7f2a15f4c3d4dace9fabb9 100644
--- a/include/qom/cpu.h
+++ b/include/qom/cpu.h
@@ -402,15 +402,6 @@ void run_on_cpu(CPUState *cpu, void (*func)(void *data), void *data);
  */
 void async_run_on_cpu(CPUState *cpu, void (*func)(void *data), void *data);
 
-/**
- * qemu_for_each_cpu:
- * @func: The function to be executed.
- * @data: Data to pass to the function.
- *
- * Executes @func for each CPU.
- */
-void qemu_for_each_cpu(void (*func)(CPUState *cpu, void *data), void *data);
-
 /**
  * qemu_get_cpu:
  * @index: The CPUState@cpu_index value of the CPU to obtain.
diff --git a/qom/cpu.c b/qom/cpu.c
index e71e57bd6b2634a19c8dd4c05ee5511b47dee3d4..fa7ec6b199c4204006d321fdda22af8f3217d812 100644
--- a/qom/cpu.c
+++ b/qom/cpu.c
@@ -25,30 +25,18 @@
 #include "qemu/log.h"
 #include "sysemu/sysemu.h"
 
-typedef struct CPUExistsArgs {
-    int64_t id;
-    bool found;
-} CPUExistsArgs;
-
-static void cpu_exist_cb(CPUState *cpu, void *data)
-{
-    CPUClass *klass = CPU_GET_CLASS(cpu);
-    CPUExistsArgs *arg = data;
-
-    if (klass->get_arch_id(cpu) == arg->id) {
-        arg->found = true;
-    }
-}
-
 bool cpu_exists(int64_t id)
 {
-    CPUExistsArgs data = {
-        .id = id,
-        .found = false,
-    };
+    CPUState *cpu;
+
+    CPU_FOREACH(cpu) {
+        CPUClass *cc = CPU_GET_CLASS(cpu);
 
-    qemu_for_each_cpu(cpu_exist_cb, &data);
-    return data.found;
+        if (cc->get_arch_id(cpu) == id) {
+            return true;
+        }
+    }
+    return false;
 }
 
 bool cpu_paging_enabled(const CPUState *cpu)