diff --git a/cmds/servicemanager/ServiceManager.cpp b/cmds/servicemanager/ServiceManager.cpp
index 1f9892a..e80c321 100644
--- a/cmds/servicemanager/ServiceManager.cpp
+++ b/cmds/servicemanager/ServiceManager.cpp
@@ -213,17 +213,18 @@
         return Status::fromExceptionCode(Status::EX_ILLEGAL_STATE);
     }
 
-    auto entry = mNameToService.emplace(name, Service {
+    // Overwrite the old service if it exists
+    mNameToService[name] = Service {
         .binder = binder,
         .allowIsolated = allowIsolated,
         .dumpPriority = dumpPriority,
         .debugPid = ctx.debugPid,
-    });
+    };
 
     auto it = mNameToRegistrationCallback.find(name);
     if (it != mNameToRegistrationCallback.end()) {
         for (const sp<IServiceCallback>& cb : it->second) {
-            entry.first->second.guaranteeClient = true;
+            mNameToService[name].guaranteeClient = true;
             // permission checked in registerForNotifications
             cb->onRegistration(name, binder);
         }
diff --git a/cmds/servicemanager/test_sm.cpp b/cmds/servicemanager/test_sm.cpp
index 25245be..fb9f9df 100644
--- a/cmds/servicemanager/test_sm.cpp
+++ b/cmds/servicemanager/test_sm.cpp
@@ -135,6 +135,26 @@
         IServiceManager::DUMP_FLAG_PRIORITY_DEFAULT).isOk());
 }
 
+TEST(AddService, OverwriteExistingService) {
+    auto sm = getPermissiveServiceManager();
+    sp<IBinder> serviceA = getBinder();
+    EXPECT_TRUE(sm->addService("foo", serviceA, false /*allowIsolated*/,
+        IServiceManager::DUMP_FLAG_PRIORITY_DEFAULT).isOk());
+
+    sp<IBinder> outA;
+    EXPECT_TRUE(sm->getService("foo", &outA).isOk());
+    EXPECT_EQ(serviceA, outA);
+
+    // serviceA should be overwritten by serviceB
+    sp<IBinder> serviceB = getBinder();
+    EXPECT_TRUE(sm->addService("foo", serviceB, false /*allowIsolated*/,
+        IServiceManager::DUMP_FLAG_PRIORITY_DEFAULT).isOk());
+
+    sp<IBinder> outB;
+    EXPECT_TRUE(sm->getService("foo", &outB).isOk());
+    EXPECT_EQ(serviceB, outB);
+}
+
 TEST(AddService, NoPermissions) {
     std::unique_ptr<MockAccess> access = std::make_unique<NiceMock<MockAccess>>();
 
diff --git a/libs/binder/IPCThreadState.cpp b/libs/binder/IPCThreadState.cpp
index d67ce15..7d008e2 100644
--- a/libs/binder/IPCThreadState.cpp
+++ b/libs/binder/IPCThreadState.cpp
@@ -860,6 +860,10 @@
             err = FAILED_TRANSACTION;
             goto finish;
 
+        case BR_FROZEN_REPLY:
+            err = FAILED_TRANSACTION;
+            goto finish;
+
         case BR_ACQUIRE_RESULT:
             {
                 ALOG_ASSERT(acquireResult != NULL, "Unexpected brACQUIRE_RESULT");
@@ -1316,6 +1320,26 @@
         }
 }
 
+status_t IPCThreadState::freeze(pid_t pid, bool enable, uint32_t timeout_ms) {
+    struct binder_freeze_info info;
+    int ret = 0;
+
+    info.pid = pid;
+    info.enable = enable;
+    info.timeout_ms = timeout_ms;
+
+
+#if defined(__ANDROID__)
+    if (ioctl(self()->mProcess->mDriverFD, BINDER_FREEZE, &info) < 0)
+        ret = -errno;
+#endif
+
+    //
+    // ret==-EAGAIN indicates that transactions have not drained.
+    // Call again to poll for completion.
+    //
+    return ret;
+}
 
 void IPCThreadState::freeBuffer(Parcel* parcel, const uint8_t* data,
                                 size_t /*dataSize*/,
diff --git a/libs/binder/Parcel.cpp b/libs/binder/Parcel.cpp
index d5d7435..a5549fc 100644
--- a/libs/binder/Parcel.cpp
+++ b/libs/binder/Parcel.cpp
@@ -2472,7 +2472,7 @@
 
     releaseObjects();
 
-    if (data) {
+    if (data || desired == 0) {
         LOG_ALLOC("Parcel %p: restart from %zu to %zu capacity", this, mDataCapacity, desired);
         pthread_mutex_lock(&gParcelGlobalAllocSizeLock);
         gParcelGlobalAllocSize += desired;
diff --git a/libs/binder/include/binder/AppOpsManager.h b/libs/binder/include/binder/AppOpsManager.h
index 6afcd77..d93935a 100644
--- a/libs/binder/include/binder/AppOpsManager.h
+++ b/libs/binder/include/binder/AppOpsManager.h
@@ -131,7 +131,11 @@
         OP_DEPRECATED_1 = 96,
         OP_AUTO_REVOKE_PERMISSIONS_IF_UNUSED = 97,
         OP_AUTO_REVOKE_MANAGED_BY_INSTALLER = 98,
-        _NUM_OP = 99
+        OP_NO_ISOLATED_STORAGE = 99,
+        OP_PHONE_CALL_MICROPHONE = 100,
+        OP_PHONE_CALL_CAMERA = 101,
+        OP_RECORD_AUDIO_HOTWORD = 102,
+        _NUM_OP = 103
     };
 
     AppOpsManager();
diff --git a/libs/binder/include/binder/IPCThreadState.h b/libs/binder/include/binder/IPCThreadState.h
index 4818889..7024f27 100644
--- a/libs/binder/include/binder/IPCThreadState.h
+++ b/libs/binder/include/binder/IPCThreadState.h
@@ -34,7 +34,22 @@
 public:
     static  IPCThreadState*     self();
     static  IPCThreadState*     selfOrNull();  // self(), but won't instantiate
-    
+
+    // Freeze or unfreeze the binder interface to a specific process. When freezing, this method
+    // will block up to timeout_ms to process pending transactions directed to pid. Unfreeze
+    // is immediate. Transactions to processes frozen via this method won't be delivered and the
+    // driver will return BR_FROZEN_REPLY to the client sending them. After unfreeze,
+    // transactions will be delivered normally.
+    //
+    // pid: id for the process for which the binder interface is to be frozen
+    // enable: freeze (true) or unfreeze (false)
+    // timeout_ms: maximum time this function is allowed to block the caller waiting for pending
+    // binder transactions to be processed.
+    //
+    // returns: 0 in case of success, a value < 0 in case of error
+    __attribute__((weak))
+    static  status_t            freeze(pid_t pid, bool enabled, uint32_t timeout_ms);
+
             sp<ProcessState>    process();
             
             status_t            clearLastError();
diff --git a/libs/binder/include/private/binder/binder_module.h b/libs/binder/include/private/binder/binder_module.h
index c22be9f..7898928 100644
--- a/libs/binder/include/private/binder/binder_module.h
+++ b/libs/binder/include/private/binder/binder_module.h
@@ -36,6 +36,37 @@
 #include <sys/ioctl.h>
 #include <linux/android/binder.h>
 
+#ifndef BR_FROZEN_REPLY
+// Temporary definition of BR_FROZEN_REPLY. For production
+// this will come from UAPI binder.h
+#define BR_FROZEN_REPLY _IO('r', 18)
+#endif //BR_FROZEN_REPLY
+
+#ifndef BINDER_FREEZE
+/*
+ * Temporary definitions for freeze support. For the final version
+ * these will be defined in the UAPI binder.h file from upstream kernel.
+ */
+#define BINDER_FREEZE _IOW('b', 14, struct binder_freeze_info)
+
+struct binder_freeze_info {
+    //
+    // Group-leader PID of process to be frozen
+    //
+    uint32_t            pid;
+    //
+    // Enable(1) / Disable(0) freeze for given PID
+    //
+    uint32_t            enable;
+    //
+    // Timeout to wait for transactions to drain.
+    // 0: don't wait (ioctl will return EAGAIN if not drained)
+    // N: number of ms to wait
+    uint32_t            timeout_ms;
+};
+#endif //BINDER_FREEZE
+
+
 #ifdef __cplusplus
 }   // namespace android
 #endif
diff --git a/libs/binder/tests/binderLibTest.cpp b/libs/binder/tests/binderLibTest.cpp
index 917751e..145c099 100644
--- a/libs/binder/tests/binderLibTest.cpp
+++ b/libs/binder/tests/binderLibTest.cpp
@@ -16,6 +16,7 @@
 
 #include <errno.h>
 #include <fcntl.h>
+#include <fstream>
 #include <poll.h>
 #include <pthread.h>
 #include <stdio.h>
@@ -79,6 +80,8 @@
     BINDER_LIB_TEST_CREATE_BINDER_TRANSACTION,
     BINDER_LIB_TEST_GET_WORK_SOURCE_TRANSACTION,
     BINDER_LIB_TEST_GET_SCHEDULING_POLICY,
+    BINDER_LIB_TEST_NOP_TRANSACTION_WAIT,
+    BINDER_LIB_TEST_GETPID,
     BINDER_LIB_TEST_ECHO_VECTOR,
     BINDER_LIB_TEST_REJECT_BUF,
 };
@@ -399,6 +402,40 @@
     EXPECT_EQ(NO_ERROR, ret);
 }
 
+TEST_F(BinderLibTest, Freeze) {
+    status_t ret;
+    Parcel data, reply, replypid;
+    std::ifstream freezer_file("/sys/fs/cgroup/freezer/cgroup.freeze");
+
+    //Pass test on devices where the freezer is not supported
+    if (freezer_file.fail()) {
+        GTEST_SKIP();
+        return;
+    }
+
+    std::string freezer_enabled;
+    std::getline(freezer_file, freezer_enabled);
+
+    //Pass test on devices where the freezer is disabled
+    if (freezer_enabled != "1") {
+        GTEST_SKIP();
+        return;
+    }
+
+    ret = m_server->transact(BINDER_LIB_TEST_GETPID, data, &replypid);
+    int32_t pid = replypid.readInt32();
+    EXPECT_EQ(NO_ERROR, ret);
+    for (int i = 0; i < 10; i++) {
+        EXPECT_EQ(NO_ERROR, m_server->transact(BINDER_LIB_TEST_NOP_TRANSACTION_WAIT, data, &reply, TF_ONE_WAY));
+    }
+    EXPECT_EQ(-EAGAIN, IPCThreadState::self()->freeze(pid, 1, 0));
+    EXPECT_EQ(-EAGAIN, IPCThreadState::self()->freeze(pid, 1, 0));
+    EXPECT_EQ(NO_ERROR, IPCThreadState::self()->freeze(pid, 1, 1000));
+    EXPECT_EQ(FAILED_TRANSACTION, m_server->transact(BINDER_LIB_TEST_NOP_TRANSACTION, data, &reply));
+    EXPECT_EQ(NO_ERROR, IPCThreadState::self()->freeze(pid, 0, 0));
+    EXPECT_EQ(NO_ERROR, m_server->transact(BINDER_LIB_TEST_NOP_TRANSACTION, data, &reply));
+}
+
 TEST_F(BinderLibTest, SetError) {
     int32_t testValue[] = { 0, -123, 123 };
     for (size_t i = 0; i < ARRAY_SIZE(testValue); i++) {
@@ -1178,6 +1215,12 @@
                 pthread_mutex_unlock(&m_serverWaitMutex);
                 return ret;
             }
+            case BINDER_LIB_TEST_GETPID:
+                reply->writeInt32(getpid());
+                return NO_ERROR;
+            case BINDER_LIB_TEST_NOP_TRANSACTION_WAIT:
+                usleep(5000);
+                return NO_ERROR;
             case BINDER_LIB_TEST_NOP_TRANSACTION:
                 return NO_ERROR;
             case BINDER_LIB_TEST_DELAYED_CALL_BACK: {
diff --git a/libs/renderengine/gl/GLESRenderEngine.cpp b/libs/renderengine/gl/GLESRenderEngine.cpp
index 92e7e71..0285c2f 100644
--- a/libs/renderengine/gl/GLESRenderEngine.cpp
+++ b/libs/renderengine/gl/GLESRenderEngine.cpp
@@ -409,6 +409,23 @@
     mImageManager = std::make_unique<ImageManager>(this);
     mImageManager->initThread();
     mDrawingBuffer = createFramebuffer();
+    sp<GraphicBuffer> buf =
+            new GraphicBuffer(1, 1, PIXEL_FORMAT_RGBA_8888, 1,
+                              GRALLOC_USAGE_HW_RENDER | GRALLOC_USAGE_HW_TEXTURE, "placeholder");
+
+    const status_t err = buf->initCheck();
+    if (err != OK) {
+        ALOGE("Error allocating placeholder buffer: %d", err);
+        return;
+    }
+    mPlaceholderBuffer = buf.get();
+    EGLint attributes[] = {
+            EGL_NONE,
+    };
+    mPlaceholderImage = eglCreateImageKHR(mEGLDisplay, EGL_NO_CONTEXT, EGL_NATIVE_BUFFER_ANDROID,
+                                          mPlaceholderBuffer, attributes);
+    ALOGE_IF(mPlaceholderImage == EGL_NO_IMAGE_KHR, "Failed to create placeholder image: %#x",
+             eglGetError());
 }
 
 GLESRenderEngine::~GLESRenderEngine() {
@@ -423,6 +440,7 @@
         eglDestroyImageKHR(mEGLDisplay, expired);
         DEBUG_EGL_IMAGE_TRACKER_DESTROY();
     }
+    eglDestroyImageKHR(mEGLDisplay, mPlaceholderImage);
     mImageCache.clear();
     eglMakeCurrent(mEGLDisplay, EGL_NO_SURFACE, EGL_NO_SURFACE, EGL_NO_CONTEXT);
     eglTerminate(mEGLDisplay);
@@ -589,6 +607,9 @@
 }
 
 void GLESRenderEngine::deleteTextures(size_t count, uint32_t const* names) {
+    for (int i = 0; i < count; ++i) {
+        mTextureView.erase(names[i]);
+    }
     glDeleteTextures(count, names);
 }
 
@@ -646,6 +667,7 @@
         }
 
         bindExternalTextureImage(texName, *cachedImage->second);
+        mTextureView.insert_or_assign(texName, buffer->getId());
     }
 
     // Wait for the new buffer to be ready.
@@ -887,7 +909,7 @@
     glBindFramebuffer(GL_FRAMEBUFFER, 0);
 }
 
-bool GLESRenderEngine::cleanupPostRender() {
+bool GLESRenderEngine::cleanupPostRender(CleanupMode mode) {
     ATRACE_CALL();
 
     if (mPriorResourcesCleaned ||
@@ -896,6 +918,30 @@
         return false;
     }
 
+    // This is a bit of a band-aid fix for FrameCaptureProcessor, as we should
+    // not need to keep memory around if we don't need to do so.
+    if (mode == CleanupMode::CLEAN_ALL) {
+        // TODO: SurfaceFlinger memory utilization may benefit from resetting
+        // texture bindings as well. Assess if it does and there's no performance regression
+        // when rebinding the same image data to the same texture, and if so then its mode
+        // behavior can be tweaked.
+        if (mPlaceholderImage != EGL_NO_IMAGE_KHR) {
+            for (auto [textureName, bufferId] : mTextureView) {
+                if (bufferId && mPlaceholderImage != EGL_NO_IMAGE_KHR) {
+                    glBindTexture(GL_TEXTURE_EXTERNAL_OES, textureName);
+                    glEGLImageTargetTexture2DOES(GL_TEXTURE_EXTERNAL_OES,
+                                                 static_cast<GLeglImageOES>(mPlaceholderImage));
+                    mTextureView[textureName] = std::nullopt;
+                    checkErrors();
+                }
+            }
+        }
+        {
+            std::lock_guard<std::mutex> lock(mRenderingMutex);
+            mImageCache.clear();
+        }
+    }
+
     // Bind the texture to dummy data so that backing image data can be freed.
     GLFramebuffer* glFramebuffer = static_cast<GLFramebuffer*>(getFramebufferForDrawing());
     glFramebuffer->allocateBuffers(1, 1, mPlaceholderDrawBuffer);
@@ -1616,6 +1662,16 @@
     return cachedImage != mImageCache.end();
 }
 
+bool GLESRenderEngine::isTextureNameKnownForTesting(uint32_t texName) {
+    const auto& entry = mTextureView.find(texName);
+    return entry != mTextureView.end();
+}
+
+std::optional<uint64_t> GLESRenderEngine::getBufferIdForTextureNameForTesting(uint32_t texName) {
+    const auto& entry = mTextureView.find(texName);
+    return entry != mTextureView.end() ? entry->second : std::nullopt;
+}
+
 bool GLESRenderEngine::isFramebufferImageCachedForTesting(uint64_t bufferId) {
     std::lock_guard<std::mutex> lock(mFramebufferImageCacheMutex);
     return std::any_of(mFramebufferImageCache.cbegin(), mFramebufferImageCache.cend(),
diff --git a/libs/renderengine/gl/GLESRenderEngine.h b/libs/renderengine/gl/GLESRenderEngine.h
index 42b8537..d5254a8 100644
--- a/libs/renderengine/gl/GLESRenderEngine.h
+++ b/libs/renderengine/gl/GLESRenderEngine.h
@@ -75,7 +75,7 @@
                         const std::vector<const LayerSettings*>& layers,
                         ANativeWindowBuffer* buffer, const bool useFramebufferCache,
                         base::unique_fd&& bufferFence, base::unique_fd* drawFence) override;
-    bool cleanupPostRender() override;
+    bool cleanupPostRender(CleanupMode mode) override;
 
     EGLDisplay getEGLDisplay() const { return mEGLDisplay; }
     // Creates an output image for rendering to
@@ -86,6 +86,12 @@
     // Test-only methods
     // Returns true iff mImageCache contains an image keyed by bufferId
     bool isImageCachedForTesting(uint64_t bufferId) EXCLUDES(mRenderingMutex);
+    // Returns true iff texName was previously generated by RenderEngine and was
+    // not destroyed.
+    bool isTextureNameKnownForTesting(uint32_t texName);
+    // Returns the buffer ID of the content bound to texName, or nullopt if no
+    // such mapping exists.
+    std::optional<uint64_t> getBufferIdForTextureNameForTesting(uint32_t texName);
     // Returns true iff mFramebufferImageCache contains an image keyed by bufferId
     bool isFramebufferImageCachedForTesting(uint64_t bufferId)
             EXCLUDES(mFramebufferImageCacheMutex);
@@ -224,6 +230,8 @@
 
     // Cache of GL images that we'll store per GraphicBuffer ID
     std::unordered_map<uint64_t, std::unique_ptr<Image>> mImageCache GUARDED_BY(mRenderingMutex);
+    std::unordered_map<uint32_t, std::optional<uint64_t>> mTextureView;
+
     // Mutex guarding rendering operations, so that:
     // 1. GL operations aren't interleaved, and
     // 2. Internal state related to rendering that is potentially modified by
@@ -237,6 +245,11 @@
     // ensure that we align on a word. Allocating 16 bytes will provide a
     // guarantee that we don't clobber memory.
     uint32_t mPlaceholderDrawBuffer[4];
+    // Placeholder buffer and image, similar to mPlaceholderDrawBuffer, but
+    // instead these are intended for cleaning up texture memory with the
+    // GL_TEXTURE_EXTERNAL_OES target.
+    ANativeWindowBuffer* mPlaceholderBuffer = nullptr;
+    EGLImage mPlaceholderImage = EGL_NO_IMAGE_KHR;
     sp<Fence> mLastDrawFence;
     // Store a separate boolean checking if prior resources were cleaned up, as
     // devices that don't support native sync fences can't rely on a last draw
diff --git a/libs/renderengine/include/renderengine/RenderEngine.h b/libs/renderengine/include/renderengine/RenderEngine.h
index e06e128..74bc44b 100644
--- a/libs/renderengine/include/renderengine/RenderEngine.h
+++ b/libs/renderengine/include/renderengine/RenderEngine.h
@@ -111,14 +111,25 @@
     // Returns NO_ERROR when binds successfully, NO_MEMORY when there's no memory for allocation.
     virtual status_t bindFrameBuffer(Framebuffer* framebuffer) = 0;
     virtual void unbindFrameBuffer(Framebuffer* framebuffer) = 0;
+
+    enum class CleanupMode {
+        CLEAN_OUTPUT_RESOURCES,
+        CLEAN_ALL,
+    };
     // Clean-up method that should be called on the main thread after the
     // drawFence returned by drawLayers fires. This method will free up
     // resources used by the most recently drawn frame. If the frame is still
     // being drawn, then this call is silently ignored.
     //
+    // If mode is CLEAN_OUTPUT_RESOURCES, then only resources related to the
+    // output framebuffer are cleaned up, including the sibling texture.
+    //
+    // If mode is CLEAN_ALL, then we also cleanup resources related to any input
+    // buffers.
+    //
     // Returns true if resources were cleaned up, and false if we didn't need to
     // do any work.
-    virtual bool cleanupPostRender() = 0;
+    virtual bool cleanupPostRender(CleanupMode mode = CleanupMode::CLEAN_OUTPUT_RESOURCES) = 0;
 
     // queries
     virtual size_t getMaxTextureSize() const = 0;
diff --git a/libs/renderengine/include/renderengine/mock/RenderEngine.h b/libs/renderengine/include/renderengine/mock/RenderEngine.h
index df0f17a..101cbb7 100644
--- a/libs/renderengine/include/renderengine/mock/RenderEngine.h
+++ b/libs/renderengine/include/renderengine/mock/RenderEngine.h
@@ -56,7 +56,7 @@
     MOCK_CONST_METHOD0(isProtected, bool());
     MOCK_CONST_METHOD0(supportsProtectedContent, bool());
     MOCK_METHOD1(useProtectedContext, bool(bool));
-    MOCK_METHOD0(cleanupPostRender, bool());
+    MOCK_METHOD1(cleanupPostRender, bool(CleanupMode mode));
     MOCK_METHOD6(drawLayers,
                  status_t(const DisplaySettings&, const std::vector<const LayerSettings*>&,
                           ANativeWindowBuffer*, const bool, base::unique_fd&&, base::unique_fd*));
diff --git a/libs/renderengine/tests/RenderEngineTest.cpp b/libs/renderengine/tests/RenderEngineTest.cpp
index 16a8a0d..f577eb3 100644
--- a/libs/renderengine/tests/RenderEngineTest.cpp
+++ b/libs/renderengine/tests/RenderEngineTest.cpp
@@ -81,6 +81,7 @@
         }
         for (uint32_t texName : mTexNames) {
             sRE->deleteTextures(1, &texName);
+            EXPECT_FALSE(sRE->isTextureNameKnownForTesting(texName));
         }
     }
 
@@ -1424,10 +1425,44 @@
     if (fd >= 0) {
         sync_wait(fd, -1);
     }
-
     // Only cleanup the first time.
-    EXPECT_TRUE(sRE->cleanupPostRender());
-    EXPECT_FALSE(sRE->cleanupPostRender());
+    EXPECT_TRUE(sRE->cleanupPostRender(
+            renderengine::RenderEngine::CleanupMode::CLEAN_OUTPUT_RESOURCES));
+    EXPECT_FALSE(sRE->cleanupPostRender(
+            renderengine::RenderEngine::CleanupMode::CLEAN_OUTPUT_RESOURCES));
+}
+
+TEST_F(RenderEngineTest, cleanupPostRender_whenCleaningAll_replacesTextureMemory) {
+    renderengine::DisplaySettings settings;
+    settings.physicalDisplay = fullscreenRect();
+    settings.clip = fullscreenRect();
+
+    std::vector<const renderengine::LayerSettings*> layers;
+    renderengine::LayerSettings layer;
+    layer.geometry.boundaries = fullscreenRect().toFloatRect();
+    BufferSourceVariant<ForceOpaqueBufferVariant>::fillColor(layer, 1.0f, 0.0f, 0.0f, this);
+    layer.alpha = 1.0;
+    layers.push_back(&layer);
+
+    base::unique_fd fence;
+    sRE->drawLayers(settings, layers, mBuffer->getNativeBuffer(), true, base::unique_fd(), &fence);
+
+    const int fd = fence.get();
+    if (fd >= 0) {
+        sync_wait(fd, -1);
+    }
+
+    uint64_t bufferId = layer.source.buffer.buffer->getId();
+    uint32_t texName = layer.source.buffer.textureName;
+    EXPECT_TRUE(sRE->isImageCachedForTesting(bufferId));
+    EXPECT_EQ(bufferId, sRE->getBufferIdForTextureNameForTesting(texName));
+
+    EXPECT_TRUE(sRE->cleanupPostRender(renderengine::RenderEngine::CleanupMode::CLEAN_ALL));
+
+    // Now check that our view of memory is good.
+    EXPECT_FALSE(sRE->isImageCachedForTesting(bufferId));
+    EXPECT_EQ(std::nullopt, sRE->getBufferIdForTextureNameForTesting(bufferId));
+    EXPECT_TRUE(sRE->isTextureNameKnownForTesting(texName));
 }
 
 } // namespace android
diff --git a/services/sensorservice/SensorDevice.cpp b/services/sensorservice/SensorDevice.cpp
index 8a282e2..e355594 100644
--- a/services/sensorservice/SensorDevice.cpp
+++ b/services/sensorservice/SensorDevice.cpp
@@ -153,12 +153,18 @@
                                     SensorDeviceUtils::defaultResolutionForType(sensor.type);
                         }
 
-                        double promotedResolution = sensor.resolution;
-                        double promotedMaxRange = sensor.maxRange;
-                        if (fmod(promotedMaxRange, promotedResolution) != 0) {
-                            ALOGW("%s's max range %f is not a multiple of the resolution %f",
-                                    sensor.name, sensor.maxRange, sensor.resolution);
-                            SensorDeviceUtils::quantizeValue(&sensor.maxRange, promotedResolution);
+                        // Some sensors don't have a default resolution and will be left at 0.
+                        // Don't crash in this case since CTS will verify that devices don't go to
+                        // production with a resolution of 0.
+                        if (sensor.resolution != 0) {
+                            double promotedResolution = sensor.resolution;
+                            double promotedMaxRange = sensor.maxRange;
+                            if (fmod(promotedMaxRange, promotedResolution) != 0) {
+                                ALOGW("%s's max range %f is not a multiple of the resolution %f",
+                                        sensor.name, sensor.maxRange, sensor.resolution);
+                                SensorDeviceUtils::quantizeValue(
+                                        &sensor.maxRange, promotedResolution);
+                            }
                         }
                     }
 
diff --git a/services/sensorservice/SensorEventConnection.cpp b/services/sensorservice/SensorEventConnection.cpp
index b4b5f98..d14a301 100644
--- a/services/sensorservice/SensorEventConnection.cpp
+++ b/services/sensorservice/SensorEventConnection.cpp
@@ -28,6 +28,12 @@
 #define UNUSED(x) (void)(x)
 
 namespace android {
+namespace {
+
+// Used as the default value for the target SDK until it's obtained via getTargetSdkVersion.
+constexpr int kTargetSdkUnknown = 0;
+
+}  // namespace
 
 SensorService::SensorEventConnection::SensorEventConnection(
         const sp<SensorService>& service, uid_t uid, String8 packageName, bool isDataInjectionMode,
@@ -35,9 +41,9 @@
     : mService(service), mUid(uid), mWakeLockRefCount(0), mHasLooperCallbacks(false),
       mDead(false), mDataInjectionMode(isDataInjectionMode), mEventCache(nullptr),
       mCacheSize(0), mMaxCacheSize(0), mTimeOfLastEventDrop(0), mEventsDropped(0),
-      mPackageName(packageName), mOpPackageName(opPackageName), mDestroyed(false) {
+      mPackageName(packageName), mOpPackageName(opPackageName), mTargetSdk(kTargetSdkUnknown),
+      mDestroyed(false) {
     mChannel = new BitTube(mService->mSocketBufferSize);
-    mTargetSdk = SensorService::getTargetSdkVersion(opPackageName);
 #if DEBUG_CONNECTIONS
     mEventsReceived = mEventsSentFromCache = mEventsSent = 0;
     mTotalAcksNeeded = mTotalAcksReceived = 0;
@@ -445,6 +451,14 @@
     bool success = true;
     const auto iter = mHandleToAppOp.find(event.sensor);
     if (iter != mHandleToAppOp.end()) {
+        if (mTargetSdk == kTargetSdkUnknown) {
+            // getTargetSdkVersion returns -1 if it fails so this operation should only be run once
+            // per connection and then cached. Perform this here as opposed to in the constructor to
+            // avoid log spam for NDK/VNDK clients that don't use sensors guarded with permissions
+            // and pass in invalid op package names.
+            mTargetSdk = SensorService::getTargetSdkVersion(mOpPackageName);
+        }
+
         // Special handling for step count/detect backwards compatibility: if the app's target SDK
         // is pre-Q, still permit delivering events to the app even if permission isn't granted
         // (since this permission was only introduced in Q)
diff --git a/services/sensorservice/SensorService.cpp b/services/sensorservice/SensorService.cpp
index 60f9cd9..3ca34bb 100644
--- a/services/sensorservice/SensorService.cpp
+++ b/services/sensorservice/SensorService.cpp
@@ -79,6 +79,8 @@
 bool SensorService::sHmacGlobalKeyIsValid = false;
 std::map<String16, int> SensorService::sPackageTargetVersion;
 Mutex SensorService::sPackageTargetVersionLock;
+String16 SensorService::sSensorInterfaceDescriptorPrefix =
+        String16("android.frameworks.sensorservice@");
 AppOpsManager SensorService::sAppOpsManager;
 
 #define SENSOR_SERVICE_DIR "/data/system/sensor_service"
@@ -1847,6 +1849,13 @@
 }
 
 int SensorService::getTargetSdkVersion(const String16& opPackageName) {
+    // Don't query the SDK version for the ISensorManager descriptor as it doesn't have one. This
+    // descriptor tends to be used for VNDK clients, but can technically be set by anyone so don't
+    // give it elevated privileges.
+    if (opPackageName.startsWith(sSensorInterfaceDescriptorPrefix)) {
+        return -1;
+    }
+
     Mutex::Autolock packageLock(sPackageTargetVersionLock);
     int targetSdkVersion = -1;
     auto entry = sPackageTargetVersion.find(opPackageName);
diff --git a/services/sensorservice/SensorService.h b/services/sensorservice/SensorService.h
index 3bb8421..052cbfe 100644
--- a/services/sensorservice/SensorService.h
+++ b/services/sensorservice/SensorService.h
@@ -424,6 +424,7 @@
     static AppOpsManager sAppOpsManager;
     static std::map<String16, int> sPackageTargetVersion;
     static Mutex sPackageTargetVersionLock;
+    static String16 sSensorInterfaceDescriptorPrefix;
 };
 
 } // namespace android
diff --git a/services/stats/StatsHal.cpp b/services/stats/StatsHal.cpp
index 80c3b65..ae0a984 100644
--- a/services/stats/StatsHal.cpp
+++ b/services/stats/StatsHal.cpp
@@ -58,7 +58,7 @@
     std::vector<int32_t> buckets = chargeCycles.cycleBucket;
     int initialSize = buckets.size();
     for (int i = 0; i < 10 - initialSize; i++) {
-        buckets.push_back(-1); // Push -1 for buckets that do not exist.
+        buckets.push_back(0); // Push 0 for buckets that do not exist.
     }
     android::util::stats_write(android::util::CHARGE_CYCLES_REPORTED, buckets[0], buckets[1],
             buckets[2], buckets[3], buckets[4], buckets[5], buckets[6], buckets[7], buckets[8],
diff --git a/services/surfaceflinger/BufferQueueLayer.cpp b/services/surfaceflinger/BufferQueueLayer.cpp
index 07be791..6e4235e 100644
--- a/services/surfaceflinger/BufferQueueLayer.cpp
+++ b/services/surfaceflinger/BufferQueueLayer.cpp
@@ -324,9 +324,6 @@
         }
 
         uint64_t bufferID = mQueueItems[0].mGraphicBuffer->getId();
-        mFlinger->mFrameTracer->traceFence(layerId, bufferID, currentFrameNumber,
-                                           mQueueItems[0].mFenceTime,
-                                           FrameTracer::FrameEvent::ACQUIRE_FENCE);
         mFlinger->mTimeStats->setLatchTime(layerId, currentFrameNumber, latchTime);
         mFlinger->mFrameTracer->traceTimestamp(layerId, bufferID, currentFrameNumber, latchTime,
                                                FrameTracer::FrameEvent::LATCH);
@@ -393,8 +390,12 @@
 
 void BufferQueueLayer::onFrameAvailable(const BufferItem& item) {
     const int32_t layerId = getSequence();
-    mFlinger->mFrameTracer->traceTimestamp(layerId, item.mGraphicBuffer->getId(), item.mFrameNumber,
-                                           systemTime(), FrameTracer::FrameEvent::QUEUE);
+    const uint64_t bufferId = item.mGraphicBuffer->getId();
+    mFlinger->mFrameTracer->traceTimestamp(layerId, bufferId, item.mFrameNumber, systemTime(),
+                                           FrameTracer::FrameEvent::QUEUE);
+    mFlinger->mFrameTracer->traceFence(layerId, bufferId, item.mFrameNumber,
+                                       std::make_shared<FenceTime>(item.mFence),
+                                       FrameTracer::FrameEvent::ACQUIRE_FENCE);
 
     ATRACE_CALL();
     // Add this buffer from our internal queue tracker
@@ -460,8 +461,12 @@
     }
 
     const int32_t layerId = getSequence();
-    mFlinger->mFrameTracer->traceTimestamp(layerId, item.mGraphicBuffer->getId(), item.mFrameNumber,
-                                           systemTime(), FrameTracer::FrameEvent::QUEUE);
+    const uint64_t bufferId = item.mGraphicBuffer->getId();
+    mFlinger->mFrameTracer->traceTimestamp(layerId, bufferId, item.mFrameNumber, systemTime(),
+                                           FrameTracer::FrameEvent::QUEUE);
+    mFlinger->mFrameTracer->traceFence(layerId, bufferId, item.mFrameNumber,
+                                       std::make_shared<FenceTime>(item.mFence),
+                                       FrameTracer::FrameEvent::ACQUIRE_FENCE);
     mConsumer->onBufferAvailable(item);
 }
 
diff --git a/services/surfaceflinger/Scheduler/VSyncDispatchTimerQueue.cpp b/services/surfaceflinger/Scheduler/VSyncDispatchTimerQueue.cpp
index a596bce..2a6fd05 100644
--- a/services/surfaceflinger/Scheduler/VSyncDispatchTimerQueue.cpp
+++ b/services/surfaceflinger/Scheduler/VSyncDispatchTimerQueue.cpp
@@ -243,7 +243,8 @@
     std::vector<Invocation> invocations;
     {
         std::lock_guard<decltype(mMutex)> lk(mMutex);
-        mLastTimerCallback = mTimeKeeper->now();
+        auto const now = mTimeKeeper->now();
+        mLastTimerCallback = now;
         for (auto it = mCallbacks.begin(); it != mCallbacks.end(); it++) {
             auto& callback = it->second;
             auto const wakeupTime = callback->wakeupTime();
@@ -251,7 +252,8 @@
                 continue;
             }
 
-            if (*wakeupTime < mIntendedWakeupTime + mTimerSlack) {
+            auto const lagAllowance = std::max(now - mIntendedWakeupTime, static_cast<nsecs_t>(0));
+            if (*wakeupTime < mIntendedWakeupTime + mTimerSlack + lagAllowance) {
                 callback->executing();
                 invocations.emplace_back(
                         Invocation{callback, *callback->lastExecutedVsyncTarget(), *wakeupTime});
diff --git a/services/surfaceflinger/Scheduler/VSyncPredictor.cpp b/services/surfaceflinger/Scheduler/VSyncPredictor.cpp
index ab5773d..61f3fbb 100644
--- a/services/surfaceflinger/Scheduler/VSyncPredictor.cpp
+++ b/services/surfaceflinger/Scheduler/VSyncPredictor.cpp
@@ -255,21 +255,9 @@
     }
 }
 
-bool VSyncPredictor::needsMoreSamples(nsecs_t now) const {
-    using namespace std::literals::chrono_literals;
+bool VSyncPredictor::needsMoreSamples() const {
     std::lock_guard<std::mutex> lk(mMutex);
-    bool needsMoreSamples = true;
-    if (mTimestamps.size() >= kMinimumSamplesForPrediction) {
-        nsecs_t constexpr aLongTime =
-                std::chrono::duration_cast<std::chrono::nanoseconds>(500ms).count();
-        if (!(mLastTimestampIndex < 0 || mTimestamps.empty())) {
-            auto const lastTimestamp = mTimestamps[mLastTimestampIndex];
-            needsMoreSamples = !((lastTimestamp + aLongTime) > now);
-        }
-    }
-
-    ATRACE_INT("VSP-moreSamples", needsMoreSamples);
-    return needsMoreSamples;
+    return mTimestamps.size() < kMinimumSamplesForPrediction;
 }
 
 void VSyncPredictor::resetModel() {
diff --git a/services/surfaceflinger/Scheduler/VSyncPredictor.h b/services/surfaceflinger/Scheduler/VSyncPredictor.h
index 3ca878d..5f3c418 100644
--- a/services/surfaceflinger/Scheduler/VSyncPredictor.h
+++ b/services/surfaceflinger/Scheduler/VSyncPredictor.h
@@ -52,11 +52,10 @@
      */
     void setPeriod(nsecs_t period) final;
 
-    /* Query if the model is in need of more samples to make a prediction at timePoint.
-     * \param [in] timePoint    The timePoint to inquire of.
+    /* Query if the model is in need of more samples to make a prediction.
      * \return  True, if model would benefit from more samples, False if not.
      */
-    bool needsMoreSamples(nsecs_t timePoint) const;
+    bool needsMoreSamples() const final;
 
     std::tuple<nsecs_t /* slope */, nsecs_t /* intercept */> getVSyncPredictionModel() const;
 
diff --git a/services/surfaceflinger/Scheduler/VSyncReactor.cpp b/services/surfaceflinger/Scheduler/VSyncReactor.cpp
index c743de0..efa8bab 100644
--- a/services/surfaceflinger/Scheduler/VSyncReactor.cpp
+++ b/services/surfaceflinger/Scheduler/VSyncReactor.cpp
@@ -233,6 +233,7 @@
 }
 
 void VSyncReactor::startPeriodTransition(nsecs_t newPeriod) {
+    ATRACE_CALL();
     mPeriodConfirmationInProgress = true;
     mPeriodTransitioningTo = newPeriod;
     mMoreSamplesNeeded = true;
@@ -240,8 +241,7 @@
 }
 
 void VSyncReactor::endPeriodTransition() {
-    setIgnorePresentFencesInternal(false);
-    mMoreSamplesNeeded = false;
+    ATRACE_CALL();
     mPeriodTransitioningTo.reset();
     mPeriodConfirmationInProgress = false;
     mLastHwVsync.reset();
@@ -254,6 +254,8 @@
 
     if (!mSupportKernelIdleTimer && period == getPeriod()) {
         endPeriodTransition();
+        setIgnorePresentFencesInternal(false);
+        mMoreSamplesNeeded = false;
     } else {
         startPeriodTransition(period);
     }
@@ -303,6 +305,7 @@
 
     std::lock_guard<std::mutex> lk(mMutex);
     if (periodConfirmed(timestamp, hwcVsyncPeriod)) {
+        ATRACE_NAME("VSR: period confirmed");
         if (mPeriodTransitioningTo) {
             mTracker->setPeriod(*mPeriodTransitioningTo);
             for (auto& entry : mCallbacks) {
@@ -310,17 +313,29 @@
             }
             *periodFlushed = true;
         }
+
+        if (mLastHwVsync) {
+            mTracker->addVsyncTimestamp(*mLastHwVsync);
+        }
+        mTracker->addVsyncTimestamp(timestamp);
+
         endPeriodTransition();
+        mMoreSamplesNeeded = mTracker->needsMoreSamples();
     } else if (mPeriodConfirmationInProgress) {
+        ATRACE_NAME("VSR: still confirming period");
         mLastHwVsync = timestamp;
         mMoreSamplesNeeded = true;
         *periodFlushed = false;
     } else {
-        mMoreSamplesNeeded = false;
+        ATRACE_NAME("VSR: adding sample");
         *periodFlushed = false;
+        mTracker->addVsyncTimestamp(timestamp);
+        mMoreSamplesNeeded = mTracker->needsMoreSamples();
     }
 
-    mTracker->addVsyncTimestamp(timestamp);
+    if (!mMoreSamplesNeeded) {
+        setIgnorePresentFencesInternal(false);
+    }
     return mMoreSamplesNeeded;
 }
 
diff --git a/services/surfaceflinger/Scheduler/VSyncTracker.h b/services/surfaceflinger/Scheduler/VSyncTracker.h
index 05a6fc3..107c540 100644
--- a/services/surfaceflinger/Scheduler/VSyncTracker.h
+++ b/services/surfaceflinger/Scheduler/VSyncTracker.h
@@ -66,6 +66,8 @@
     /* Inform the tracker that the samples it has are not accurate for prediction. */
     virtual void resetModel() = 0;
 
+    virtual bool needsMoreSamples() const = 0;
+
     virtual void dump(std::string& result) const = 0;
 
 protected:
diff --git a/services/surfaceflinger/SurfaceFlinger.cpp b/services/surfaceflinger/SurfaceFlinger.cpp
index 0ac3d32..148749e 100644
--- a/services/surfaceflinger/SurfaceFlinger.cpp
+++ b/services/surfaceflinger/SurfaceFlinger.cpp
@@ -1566,7 +1566,7 @@
     mEventQueue->refresh();
 }
 
-nsecs_t SurfaceFlinger::getVsyncPeriod() const {
+nsecs_t SurfaceFlinger::getVsyncPeriodFromHWC() const {
     const auto displayId = getInternalDisplayIdLocked();
     if (!displayId || !getHwComposer().isConnected(*displayId)) {
         return 0;
@@ -1773,7 +1773,7 @@
     setPowerModeInternal(display, currentDisplayPowerMode);
 
     // Reset the timing values to account for the period of the swapped in HWC
-    const nsecs_t vsyncPeriod = getVsyncPeriod();
+    const nsecs_t vsyncPeriod = mRefreshRateConfigs->getCurrentRefreshRate().getVsyncPeriod();
     mAnimFrameTracker.setDisplayRefreshPeriod(vsyncPeriod);
 
     // The present fences returned from vr_hwc are not an accurate
@@ -4192,8 +4192,7 @@
                         {});
 
     setPowerModeInternal(display, hal::PowerMode::ON);
-
-    const nsecs_t vsyncPeriod = getVsyncPeriod();
+    const nsecs_t vsyncPeriod = mRefreshRateConfigs->getCurrentRefreshRate().getVsyncPeriod();
     mAnimFrameTracker.setDisplayRefreshPeriod(vsyncPeriod);
 
     // Use phase of 0 since phase is not known.
@@ -4228,7 +4227,7 @@
     if (mInterceptor->isEnabled()) {
         mInterceptor->savePowerModeUpdate(display->getSequenceId(), static_cast<int32_t>(mode));
     }
-
+    const auto vsyncPeriod = mRefreshRateConfigs->getCurrentRefreshRate().getVsyncPeriod();
     if (currentMode == hal::PowerMode::OFF) {
         if (SurfaceFlinger::setSchedFifo(true) != NO_ERROR) {
             ALOGW("Couldn't set SCHED_FIFO on display on: %s\n", strerror(errno));
@@ -4237,7 +4236,7 @@
         if (display->isPrimary() && mode != hal::PowerMode::DOZE_SUSPEND) {
             getHwComposer().setVsyncEnabled(*displayId, mHWCVsyncPendingState);
             mScheduler->onScreenAcquired(mAppConnectionHandle);
-            mScheduler->resyncToHardwareVsync(true, getVsyncPeriod());
+            mScheduler->resyncToHardwareVsync(true, vsyncPeriod);
         }
 
         mVisibleRegionsDirty = true;
@@ -4264,7 +4263,7 @@
         getHwComposer().setPowerMode(*displayId, mode);
         if (display->isPrimary() && currentMode == hal::PowerMode::DOZE_SUSPEND) {
             mScheduler->onScreenAcquired(mAppConnectionHandle);
-            mScheduler->resyncToHardwareVsync(true, getVsyncPeriod());
+            mScheduler->resyncToHardwareVsync(true, vsyncPeriod);
         }
     } else if (mode == hal::PowerMode::DOZE_SUSPEND) {
         // Leave display going to doze
@@ -4377,7 +4376,7 @@
 }
 
 void SurfaceFlinger::dumpStatsLocked(const DumpArgs& args, std::string& result) const {
-    StringAppendF(&result, "%" PRId64 "\n", getVsyncPeriod());
+    StringAppendF(&result, "%" PRId64 "\n", getVsyncPeriodFromHWC());
 
     if (args.size() > 1) {
         const auto name = String8(args[1]);
@@ -4442,7 +4441,7 @@
     mPhaseConfiguration->dump(result);
     StringAppendF(&result,
                   "      present offset: %9" PRId64 " ns\t     VSYNC period: %9" PRId64 " ns\n\n",
-                  dispSyncPresentTimeOffset, getVsyncPeriod());
+                  dispSyncPresentTimeOffset, getVsyncPeriodFromHWC());
 
     scheduler::RefreshRateConfigs::Policy policy = mRefreshRateConfigs->getDisplayManagerPolicy();
     StringAppendF(&result,
diff --git a/services/surfaceflinger/SurfaceFlinger.h b/services/surfaceflinger/SurfaceFlinger.h
index ccaeb2d..c727574 100644
--- a/services/surfaceflinger/SurfaceFlinger.h
+++ b/services/surfaceflinger/SurfaceFlinger.h
@@ -852,7 +852,7 @@
     /* ------------------------------------------------------------------------
      * VSync
      */
-    nsecs_t getVsyncPeriod() const REQUIRES(mStateLock);
+    nsecs_t getVsyncPeriodFromHWC() const REQUIRES(mStateLock);
 
     // Sets the refresh rate by switching active configs, if they are available for
     // the desired refresh rate.
diff --git a/services/surfaceflinger/tests/unittests/DisplayTransactionTest.cpp b/services/surfaceflinger/tests/unittests/DisplayTransactionTest.cpp
index ce5f35c..06bdcdc 100644
--- a/services/surfaceflinger/tests/unittests/DisplayTransactionTest.cpp
+++ b/services/surfaceflinger/tests/unittests/DisplayTransactionTest.cpp
@@ -85,7 +85,7 @@
 using HotplugEvent = TestableSurfaceFlinger::HotplugEvent;
 using HWC2Display = TestableSurfaceFlinger::HWC2Display;
 
-constexpr int32_t DEFAULT_REFRESH_RATE = 16'666'666;
+constexpr int32_t DEFAULT_REFRESH_RATE = 16'666'667;
 constexpr int32_t DEFAULT_DPI = 320;
 constexpr int DEFAULT_VIRTUAL_DISPLAY_SURFACE_FORMAT = HAL_PIXEL_FORMAT_RGB_565;
 
diff --git a/services/surfaceflinger/tests/unittests/VSyncDispatchRealtimeTest.cpp b/services/surfaceflinger/tests/unittests/VSyncDispatchRealtimeTest.cpp
index be49ef3..c2a7752 100644
--- a/services/surfaceflinger/tests/unittests/VSyncDispatchRealtimeTest.cpp
+++ b/services/surfaceflinger/tests/unittests/VSyncDispatchRealtimeTest.cpp
@@ -51,6 +51,7 @@
 
     void setPeriod(nsecs_t) final {}
     void resetModel() final {}
+    bool needsMoreSamples() const final { return false; }
     void dump(std::string&) const final {}
 
 private:
@@ -86,6 +87,7 @@
 
     void setPeriod(nsecs_t) final {}
     void resetModel() final {}
+    bool needsMoreSamples() const final { return false; }
     void dump(std::string&) const final {}
 
 private:
diff --git a/services/surfaceflinger/tests/unittests/VSyncDispatchTimerQueueTest.cpp b/services/surfaceflinger/tests/unittests/VSyncDispatchTimerQueueTest.cpp
index d940dc5..373f6a5 100644
--- a/services/surfaceflinger/tests/unittests/VSyncDispatchTimerQueueTest.cpp
+++ b/services/surfaceflinger/tests/unittests/VSyncDispatchTimerQueueTest.cpp
@@ -47,6 +47,7 @@
     MOCK_CONST_METHOD0(currentPeriod, nsecs_t());
     MOCK_METHOD1(setPeriod, void(nsecs_t));
     MOCK_METHOD0(resetModel, void());
+    MOCK_CONST_METHOD0(needsMoreSamples, bool());
     MOCK_CONST_METHOD1(dump, void(std::string&));
 
     nsecs_t nextVSyncTime(nsecs_t timePoint) const {
@@ -115,11 +116,15 @@
 
     operator VSyncDispatch::CallbackToken() const { return mToken; }
 
-    void counter(nsecs_t time, nsecs_t) { mCalls.push_back(time); }
+    void counter(nsecs_t time, nsecs_t wakeup_time) {
+        mCalls.push_back(time);
+        mWakeupTime.push_back(wakeup_time);
+    }
 
     VSyncDispatch& mDispatch;
     VSyncDispatch::CallbackToken mToken;
     std::vector<nsecs_t> mCalls;
+    std::vector<nsecs_t> mWakeupTime;
 };
 
 class PausingCallback {
@@ -747,6 +752,31 @@
     EXPECT_THAT(cb2.mCalls.size(), Eq(1));
 }
 
+TEST_F(VSyncDispatchTimerQueueTest, laggedTimerGroupsCallbacksWithinLag) {
+    CountingCallback cb1(mDispatch);
+    CountingCallback cb2(mDispatch);
+
+    Sequence seq;
+    EXPECT_CALL(mStubTracker, nextAnticipatedVSyncTimeFrom(1000))
+            .InSequence(seq)
+            .WillOnce(Return(1000));
+    EXPECT_CALL(mMockClock, alarmIn(_, 600)).InSequence(seq);
+    EXPECT_CALL(mStubTracker, nextAnticipatedVSyncTimeFrom(1000))
+            .InSequence(seq)
+            .WillOnce(Return(1000));
+
+    EXPECT_EQ(mDispatch.schedule(cb1, 400, 1000), ScheduleResult::Scheduled);
+    EXPECT_EQ(mDispatch.schedule(cb2, 390, 1000), ScheduleResult::Scheduled);
+
+    mMockClock.setLag(100);
+    mMockClock.advanceBy(700);
+
+    ASSERT_THAT(cb1.mWakeupTime.size(), Eq(1));
+    EXPECT_THAT(cb1.mWakeupTime[0], Eq(600));
+    ASSERT_THAT(cb2.mWakeupTime.size(), Eq(1));
+    EXPECT_THAT(cb2.mWakeupTime[0], Eq(610));
+}
+
 class VSyncDispatchTimerQueueEntryTest : public testing::Test {
 protected:
     nsecs_t const mPeriod = 1000;
diff --git a/services/surfaceflinger/tests/unittests/VSyncPredictorTest.cpp b/services/surfaceflinger/tests/unittests/VSyncPredictorTest.cpp
index fc39235..d4cd11d 100644
--- a/services/surfaceflinger/tests/unittests/VSyncPredictorTest.cpp
+++ b/services/surfaceflinger/tests/unittests/VSyncPredictorTest.cpp
@@ -73,27 +73,27 @@
 
 TEST_F(VSyncPredictorTest, reportsSamplesNeededWhenHasNoDataPoints) {
     for (auto i = 0u; i < kMinimumSamplesForPrediction; i++) {
-        EXPECT_TRUE(tracker.needsMoreSamples(mNow += mPeriod));
-        tracker.addVsyncTimestamp(mNow);
+        EXPECT_TRUE(tracker.needsMoreSamples());
+        tracker.addVsyncTimestamp(mNow += mPeriod);
     }
-    EXPECT_FALSE(tracker.needsMoreSamples(mNow));
+    EXPECT_FALSE(tracker.needsMoreSamples());
 }
 
 TEST_F(VSyncPredictorTest, reportsSamplesNeededAfterExplicitRateChange) {
     for (auto i = 0u; i < kMinimumSamplesForPrediction; i++) {
         tracker.addVsyncTimestamp(mNow += mPeriod);
     }
-    EXPECT_FALSE(tracker.needsMoreSamples(mNow));
+    EXPECT_FALSE(tracker.needsMoreSamples());
 
     auto const changedPeriod = mPeriod * 2;
     tracker.setPeriod(changedPeriod);
-    EXPECT_TRUE(tracker.needsMoreSamples(mNow));
+    EXPECT_TRUE(tracker.needsMoreSamples());
 
     for (auto i = 0u; i < kMinimumSamplesForPrediction; i++) {
-        EXPECT_TRUE(tracker.needsMoreSamples(mNow += changedPeriod));
-        tracker.addVsyncTimestamp(mNow);
+        EXPECT_TRUE(tracker.needsMoreSamples());
+        tracker.addVsyncTimestamp(mNow += changedPeriod);
     }
-    EXPECT_FALSE(tracker.needsMoreSamples(mNow));
+    EXPECT_FALSE(tracker.needsMoreSamples());
 }
 
 TEST_F(VSyncPredictorTest, transitionsToModelledPointsAfterSynthetic) {
@@ -320,20 +320,6 @@
     EXPECT_THAT(intercept, Eq(0));
 }
 
-TEST_F(VSyncPredictorTest, willBecomeInaccurateAfterA_longTimeWithNoSamples) {
-    auto const simulatedVsyncs = generateVsyncTimestamps(kMinimumSamplesForPrediction, mPeriod, 0);
-
-    for (auto const& timestamp : simulatedVsyncs) {
-        tracker.addVsyncTimestamp(timestamp);
-    }
-    auto const mNow = *simulatedVsyncs.rbegin();
-    EXPECT_FALSE(tracker.needsMoreSamples(mNow));
-
-    // TODO: would be better to decay this as a result of the variance of the samples
-    static auto constexpr aLongTimeOut = 1000000000;
-    EXPECT_TRUE(tracker.needsMoreSamples(mNow + aLongTimeOut));
-}
-
 TEST_F(VSyncPredictorTest, idealModelPredictionsBeforeRegressionModelIsBuilt) {
     auto const simulatedVsyncs =
             generateVsyncTimestamps(kMinimumSamplesForPrediction + 1, mPeriod, 0);
@@ -443,7 +429,7 @@
         // When VsyncPredictor returns the period it means that it doesn't know how to predict and
         // it needs to get more samples
         if (slope == mPeriod && intercept == 0) {
-            EXPECT_TRUE(tracker.needsMoreSamples(now));
+            EXPECT_TRUE(tracker.needsMoreSamples());
         }
     }
 }
diff --git a/services/surfaceflinger/tests/unittests/VSyncReactorTest.cpp b/services/surfaceflinger/tests/unittests/VSyncReactorTest.cpp
index a972562..6856612 100644
--- a/services/surfaceflinger/tests/unittests/VSyncReactorTest.cpp
+++ b/services/surfaceflinger/tests/unittests/VSyncReactorTest.cpp
@@ -41,6 +41,7 @@
     MOCK_CONST_METHOD0(currentPeriod, nsecs_t());
     MOCK_METHOD1(setPeriod, void(nsecs_t));
     MOCK_METHOD0(resetModel, void());
+    MOCK_CONST_METHOD0(needsMoreSamples, bool());
     MOCK_CONST_METHOD1(dump, void(std::string&));
 };
 
@@ -57,6 +58,7 @@
     nsecs_t currentPeriod() const final { return mTracker->currentPeriod(); }
     void setPeriod(nsecs_t period) final { mTracker->setPeriod(period); }
     void resetModel() final { mTracker->resetModel(); }
+    bool needsMoreSamples() const final { return mTracker->needsMoreSamples(); }
     void dump(std::string& result) const final { mTracker->dump(result); }
 
 private:
@@ -455,6 +457,83 @@
     EXPECT_FALSE(mReactor.addPresentFence(generateSignalledFenceWithTime(0)));
 }
 
+TEST_F(VSyncReactorTest, hwVsyncIsRequestedForTracker) {
+    auto time = 0;
+    bool periodFlushed = false;
+    nsecs_t const newPeriod = 4000;
+    mReactor.setPeriod(newPeriod);
+
+    static auto constexpr numSamplesWithNewPeriod = 4;
+    Sequence seq;
+    EXPECT_CALL(*mMockTracker, needsMoreSamples())
+            .Times(numSamplesWithNewPeriod - 2)
+            .InSequence(seq)
+            .WillRepeatedly(Return(true));
+    EXPECT_CALL(*mMockTracker, needsMoreSamples())
+            .Times(1)
+            .InSequence(seq)
+            .WillRepeatedly(Return(false));
+    EXPECT_CALL(*mMockTracker, addVsyncTimestamp(_)).Times(numSamplesWithNewPeriod);
+
+    EXPECT_TRUE(mReactor.addResyncSample(time += period, std::nullopt, &periodFlushed));
+
+    EXPECT_TRUE(mReactor.addResyncSample(time += period, std::nullopt, &periodFlushed));
+    // confirmed period, but predictor wants numRequest samples. This one and prior are valid.
+    EXPECT_TRUE(mReactor.addResyncSample(time += newPeriod, std::nullopt, &periodFlushed));
+    EXPECT_TRUE(mReactor.addResyncSample(time += newPeriod, std::nullopt, &periodFlushed));
+    EXPECT_FALSE(mReactor.addResyncSample(time += newPeriod, std::nullopt, &periodFlushed));
+}
+
+TEST_F(VSyncReactorTest, hwVsyncturnsOffOnConfirmationWhenTrackerDoesntRequest) {
+    auto time = 0;
+    bool periodFlushed = false;
+    nsecs_t const newPeriod = 4000;
+    mReactor.setPeriod(newPeriod);
+
+    Sequence seq;
+    EXPECT_CALL(*mMockTracker, needsMoreSamples())
+            .Times(1)
+            .InSequence(seq)
+            .WillRepeatedly(Return(false));
+    EXPECT_CALL(*mMockTracker, addVsyncTimestamp(_)).Times(2);
+
+    EXPECT_TRUE(mReactor.addResyncSample(time += period, std::nullopt, &periodFlushed));
+    EXPECT_TRUE(mReactor.addResyncSample(time += period, std::nullopt, &periodFlushed));
+    EXPECT_FALSE(mReactor.addResyncSample(time += newPeriod, std::nullopt, &periodFlushed));
+}
+
+TEST_F(VSyncReactorTest, hwVsyncIsRequestedForTrackerMultiplePeriodChanges) {
+    auto time = 0;
+    bool periodFlushed = false;
+    nsecs_t const newPeriod1 = 4000;
+    nsecs_t const newPeriod2 = 7000;
+
+    mReactor.setPeriod(newPeriod1);
+
+    Sequence seq;
+    EXPECT_CALL(*mMockTracker, needsMoreSamples())
+            .Times(4)
+            .InSequence(seq)
+            .WillRepeatedly(Return(true));
+    EXPECT_CALL(*mMockTracker, needsMoreSamples())
+            .Times(1)
+            .InSequence(seq)
+            .WillRepeatedly(Return(false));
+    EXPECT_CALL(*mMockTracker, addVsyncTimestamp(_)).Times(7);
+
+    EXPECT_TRUE(mReactor.addResyncSample(time += period, std::nullopt, &periodFlushed));
+    EXPECT_TRUE(mReactor.addResyncSample(time += period, std::nullopt, &periodFlushed));
+    // confirmed period, but predictor wants numRequest samples. This one and prior are valid.
+    EXPECT_TRUE(mReactor.addResyncSample(time += newPeriod1, std::nullopt, &periodFlushed));
+    EXPECT_TRUE(mReactor.addResyncSample(time += newPeriod1, std::nullopt, &periodFlushed));
+
+    mReactor.setPeriod(newPeriod2);
+    EXPECT_TRUE(mReactor.addResyncSample(time += newPeriod1, std::nullopt, &periodFlushed));
+    EXPECT_TRUE(mReactor.addResyncSample(time += newPeriod2, std::nullopt, &periodFlushed));
+    EXPECT_TRUE(mReactor.addResyncSample(time += newPeriod2, std::nullopt, &periodFlushed));
+    EXPECT_FALSE(mReactor.addResyncSample(time += newPeriod2, std::nullopt, &periodFlushed));
+}
+
 static nsecs_t computeWorkload(nsecs_t period, nsecs_t phase) {
     return period - phase;
 }
@@ -648,7 +727,7 @@
 
 TEST_F(VSyncReactorTest, periodChangeWithGivenVsyncPeriod) {
     bool periodFlushed = true;
-    EXPECT_CALL(*mMockTracker, addVsyncTimestamp(_)).Times(3);
+    EXPECT_CALL(*mMockTracker, addVsyncTimestamp(_)).Times(2);
     mReactor.setIgnorePresentFences(true);
 
     nsecs_t const newPeriod = 5000;
@@ -672,7 +751,7 @@
                                     kPendingLimit, true /* supportKernelIdleTimer */);
 
     bool periodFlushed = true;
-    EXPECT_CALL(*mMockTracker, addVsyncTimestamp(_)).Times(5);
+    EXPECT_CALL(*mMockTracker, addVsyncTimestamp(_)).Times(4);
     idleReactor.setIgnorePresentFences(true);
 
     // First, set the same period, which should only be confirmed when we receive two