Merge tag 'pull-migration-20220720c' of https://gitlab.com/dagrh/qemu into staging

Migration pull 2022-07-20

This replaces yesterdays pull and:
  a) Fixes some test build errors without TLS
  b) Reenabled the zlib acceleration on s390
     now that we have Ilya's fix

  Hyman's dirty page rate limit set
  Ilya's fix for zlib vs migration
  Peter's postcopy-preempt
  Cleanup from Dan
  zero-copy tidy ups from Leo
  multifd doc fix from Juan
  Revert disable of zlib acceleration on s390x

Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>

# gpg: Signature made Wed 20 Jul 2022 12:18:56 BST
# gpg:                using RSA key 45F5C71B4A0CB7FB977A9FA90516331EBC5BFDE7
# gpg: Good signature from "Dr. David Alan Gilbert (RH2) <dgilbert@redhat.com>" [full]
# Primary key fingerprint: 45F5 C71B 4A0C B7FB 977A  9FA9 0516 331E BC5B FDE7

* tag 'pull-migration-20220720c' of https://gitlab.com/dagrh/qemu: (30 commits)
  Revert "gitlab: disable accelerated zlib for s390x"
  migration: Avoid false-positive on non-supported scenarios for zero-copy-send
  multifd: Document the locking of MultiFD{Send/Recv}Params
  migration/multifd: Report to user when zerocopy not working
  Add dirty-sync-missed-zero-copy migration stat
  QIOChannelSocket: Fix zero-copy flush returning code 1 when nothing sent
  migration: remove unreachable code after reading data
  tests: Add postcopy preempt tests
  tests: Add postcopy tls recovery migration test
  tests: Add postcopy tls migration test
  tests: Move MigrateCommon upper
  migration: Respect postcopy request order in preemption mode
  migration: Enable TLS for preempt channel
  migration: Export tls-[creds|hostname|authz] params to cmdline too
  migration: Add helpers to detect TLS capability
  migration: Add property x-postcopy-preempt-break-huge
  migration: Create the postcopy preempt channel asynchronously
  migration: Postcopy recover with preempt enabled
  migration: Postcopy preemption enablement
  migration: Postcopy preemption preparation on channel creation
  ...

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

2 files changed, 41 insertions, 10 deletions

diff --git a/accel/kvm/kvm-all.c b/accel/kvm/kvm-all.c
index ed8b6b896e..3187656570 100644
--- a/accel/kvm/kvm-all.c
+++ b/accel/kvm/kvm-all.c
@@ -45,6 +45,7 @@
 #include "qemu/guest-random.h"
 #include "sysemu/hw_accel.h"
 #include "kvm-cpus.h"
+#include "sysemu/dirtylimit.h"
 
 #include "hw/boards.h"
 #include "monitor/stats.h"
@@ -477,6 +478,7 @@ int kvm_init_vcpu(CPUState *cpu, Error **errp)
     cpu->kvm_state = s;
     cpu->vcpu_dirty = true;
     cpu->dirty_pages = 0;
+    cpu->throttle_us_per_full = 0;
 
     mmap_size = kvm_ioctl(s, KVM_GET_VCPU_MMAP_SIZE, 0);
     if (mmap_size < 0) {
@@ -757,17 +759,20 @@ static uint32_t kvm_dirty_ring_reap_one(KVMState *s, CPUState *cpu)
 }
 
 /* Must be with slots_lock held */
-static uint64_t kvm_dirty_ring_reap_locked(KVMState *s)
+static uint64_t kvm_dirty_ring_reap_locked(KVMState *s, CPUState* cpu)
 {
     int ret;
-    CPUState *cpu;
     uint64_t total = 0;
     int64_t stamp;
 
     stamp = get_clock();
 
-    CPU_FOREACH(cpu) {
-        total += kvm_dirty_ring_reap_one(s, cpu);
+    if (cpu) {
+        total = kvm_dirty_ring_reap_one(s, cpu);
+    } else {
+        CPU_FOREACH(cpu) {
+            total += kvm_dirty_ring_reap_one(s, cpu);
+        }
     }
 
     if (total) {
@@ -788,7 +793,7 @@ static uint64_t kvm_dirty_ring_reap_locked(KVMState *s)
  * Currently for simplicity, we must hold BQL before calling this.  We can
  * consider to drop the BQL if we're clear with all the race conditions.
  */
-static uint64_t kvm_dirty_ring_reap(KVMState *s)
+static uint64_t kvm_dirty_ring_reap(KVMState *s, CPUState *cpu)
 {
     uint64_t total;
 
@@ -808,7 +813,7 @@ static uint64_t kvm_dirty_ring_reap(KVMState *s)
      *     reset below.
      */
     kvm_slots_lock();
-    total = kvm_dirty_ring_reap_locked(s);
+    total = kvm_dirty_ring_reap_locked(s, cpu);
     kvm_slots_unlock();
 
     return total;
@@ -855,7 +860,7 @@ static void kvm_dirty_ring_flush(void)
      * vcpus out in a synchronous way.
      */
     kvm_cpu_synchronize_kick_all();
-    kvm_dirty_ring_reap(kvm_state);
+    kvm_dirty_ring_reap(kvm_state, NULL);
     trace_kvm_dirty_ring_flush(1);
 }
 
@@ -1399,7 +1404,7 @@ static void kvm_set_phys_mem(KVMMemoryListener *kml,
                  * Not easy.  Let's cross the fingers until it's fixed.
                  */
                 if (kvm_state->kvm_dirty_ring_size) {
-                    kvm_dirty_ring_reap_locked(kvm_state);
+                    kvm_dirty_ring_reap_locked(kvm_state, NULL);
                 } else {
                     kvm_slot_get_dirty_log(kvm_state, mem);
                 }
@@ -1467,11 +1472,16 @@ static void *kvm_dirty_ring_reaper_thread(void *data)
          */
         sleep(1);
 
+        /* keep sleeping so that dirtylimit not be interfered by reaper */
+        if (dirtylimit_in_service()) {
+            continue;
+        }
+
         trace_kvm_dirty_ring_reaper("wakeup");
         r->reaper_state = KVM_DIRTY_RING_REAPER_REAPING;
 
         qemu_mutex_lock_iothread();
-        kvm_dirty_ring_reap(s);
+        kvm_dirty_ring_reap(s, NULL);
         qemu_mutex_unlock_iothread();
 
         r->reaper_iteration++;
@@ -2315,6 +2325,11 @@ static void query_stats_cb(StatsResultList **result, StatsTarget target,
                            strList *names, strList *targets, Error **errp);
 static void query_stats_schemas_cb(StatsSchemaList **result, Error **errp);
 
+uint32_t kvm_dirty_ring_size(void)
+{
+    return kvm_state->kvm_dirty_ring_size;
+}
+
 static int kvm_init(MachineState *ms)
 {
     MachineClass *mc = MACHINE_GET_CLASS(ms);
@@ -2967,8 +2982,19 @@ int kvm_cpu_exec(CPUState *cpu)
              */
             trace_kvm_dirty_ring_full(cpu->cpu_index);
             qemu_mutex_lock_iothread();
-            kvm_dirty_ring_reap(kvm_state);
+            /*
+             * We throttle vCPU by making it sleep once it exit from kernel
+             * due to dirty ring full. In the dirtylimit scenario, reaping
+             * all vCPUs after a single vCPU dirty ring get full result in
+             * the miss of sleep, so just reap the ring-fulled vCPU.
+             */
+            if (dirtylimit_in_service()) {
+                kvm_dirty_ring_reap(kvm_state, cpu);
+            } else {
+                kvm_dirty_ring_reap(kvm_state, NULL);
+            }
             qemu_mutex_unlock_iothread();
+            dirtylimit_vcpu_execute(cpu);
             ret = 0;
             break;
         case KVM_EXIT_SYSTEM_EVENT:
diff --git a/accel/stubs/kvm-stub.c b/accel/stubs/kvm-stub.c
index 3345882d85..2ac5f9c036 100644
--- a/accel/stubs/kvm-stub.c
+++ b/accel/stubs/kvm-stub.c
@@ -148,3 +148,8 @@ bool kvm_dirty_ring_enabled(void)
 {
     return false;
 }
+
+uint32_t kvm_dirty_ring_size(void)
+{
+    return 0;
+}