diff --git a/audio/core/all-versions/default/Device.cpp b/audio/core/all-versions/default/Device.cpp
index bec22df..1a9df21 100644
--- a/audio/core/all-versions/default/Device.cpp
+++ b/audio/core/all-versions/default/Device.cpp
@@ -378,7 +378,7 @@
 }
 
 Return<Result> Device::setConnectedState(const DeviceAddress& address, bool connected) {
-    auto key = connected ? AudioParameter::keyStreamConnect : AudioParameter::keyStreamDisconnect;
+    auto key = connected ? AudioParameter::keyDeviceConnect : AudioParameter::keyDeviceDisconnect;
     return setParam(key, address);
 }
 #endif
diff --git a/audio/core/all-versions/default/Stream.cpp b/audio/core/all-versions/default/Stream.cpp
index b995657..e62f6d3 100644
--- a/audio/core/all-versions/default/Stream.cpp
+++ b/audio/core/all-versions/default/Stream.cpp
@@ -243,8 +243,8 @@
 
 Return<Result> Stream::setConnectedState(const DeviceAddress& address, bool connected) {
     return setParam(
-        connected ? AudioParameter::keyStreamConnect : AudioParameter::keyStreamDisconnect,
-        address);
+            connected ? AudioParameter::keyDeviceConnect : AudioParameter::keyDeviceDisconnect,
+            address);
 }
 #elif MAJOR_VERSION >= 4
 Return<void> Stream::getDevices(getDevices_cb _hidl_cb) {
diff --git a/automotive/vehicle/2.0/default/impl/vhal_v2_0/DefaultConfig.h b/automotive/vehicle/2.0/default/impl/vhal_v2_0/DefaultConfig.h
index a46de24..fc441ed 100644
--- a/automotive/vehicle/2.0/default/impl/vhal_v2_0/DefaultConfig.h
+++ b/automotive/vehicle/2.0/default/impl/vhal_v2_0/DefaultConfig.h
@@ -671,7 +671,7 @@
      .initialValue = {.int32Values = {toInt(VehicleApPowerStateReq::ON), 0}}},
 
     {.config = {.prop = toInt(VehicleProperty::AP_POWER_STATE_REPORT),
-                .access = VehiclePropertyAccess::WRITE,
+                .access = VehiclePropertyAccess::READ_WRITE,
                 .changeMode = VehiclePropertyChangeMode::ON_CHANGE},
      .initialValue = {.int32Values = {toInt(VehicleApPowerStateReport::WAIT_FOR_VHAL), 0}}},
 
diff --git a/automotive/vehicle/2.0/types.hal b/automotive/vehicle/2.0/types.hal
index f6ebcdd..0f20dd1 100644
--- a/automotive/vehicle/2.0/types.hal
+++ b/automotive/vehicle/2.0/types.hal
@@ -1282,7 +1282,7 @@
 
      *
      * @change_mode VehiclePropertyChangeMode:ON_CHANGE
-     * @access VehiclePropertyAccess:WRITE
+     * @access VehiclePropertyAccess:READ_WRITE
      */
     AP_POWER_STATE_REPORT = (
         0x0A01
diff --git a/boot/1.1/Android.bp b/boot/1.1/Android.bp
new file mode 100644
index 0000000..6a8d57a
--- /dev/null
+++ b/boot/1.1/Android.bp
@@ -0,0 +1,18 @@
+// This file is autogenerated by hidl-gen -Landroidbp.
+
+hidl_interface {
+    name: "android.hardware.boot@1.1",
+    root: "android.hardware",
+    vndk: {
+        enabled: true,
+    },
+    srcs: [
+        "types.hal",
+        "IBootControl.hal",
+    ],
+    interfaces: [
+        "android.hardware.boot@1.0",
+        "android.hidl.base@1.0",
+    ],
+    gen_java: true,
+}
diff --git a/boot/1.1/IBootControl.hal b/boot/1.1/IBootControl.hal
new file mode 100644
index 0000000..939dfb3
--- /dev/null
+++ b/boot/1.1/IBootControl.hal
@@ -0,0 +1,66 @@
+/*
+ * Copyright 2019 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package android.hardware.boot@1.1;
+
+import @1.0::IBootControl;
+
+interface IBootControl extends @1.0::IBootControl {
+    /**
+     * Sets whether a snapshot-merge of any dynamic partition is in progress.
+     *
+     * After the merge status is set to a given value, subsequent calls to
+     * getSnapshotMergeStatus must return the set value.
+     *
+     * The merge status must be persistent across reboots. That is, getSnapshotMergeStatus
+     * must return the same value after a reboot if the merge status is not altered in any way
+     * (e.g. set by setSnapshotMergeStatus or set to CANCELLED by bootloader).
+     *
+     * Read/write access to the merge status must be atomic. When the HAL is processing a
+     * setSnapshotMergeStatus call, all subsequent calls to getSnapshotMergeStatus must block until
+     * setSnapshotMergeStatus has returned.
+     *
+     * A MERGING state indicates that dynamic partitions are partially comprised by blocks in the
+     * userdata partition.
+     *
+     * When the merge status is set to MERGING, the following operations must be prohibited from the
+     * bootloader:
+     *  - Flashing or erasing "userdata" or "metadata".
+     *
+     * The following operations may be prohibited when the status is set to MERGING. If not
+     * prohibited, it is recommended that the user receive a warning.
+     *  - Changing the active slot (e.g. via "fastboot set_active")
+     *
+     * @param status Merge status.
+     *
+     * @return success True on success, false otherwise.
+     */
+    setSnapshotMergeStatus(MergeStatus status) generates (bool success);
+
+    /**
+     * Returns whether a snapshot-merge of any dynamic partition is in progress.
+     *
+     * This function must return the merge status set by the last setSnapshotMergeStatus call and
+     * recorded by the bootloader with one exception. If the partitions are being flashed from the
+     * bootloader such that the pending merge must be canceled (for example, if the super partition
+     * is being flashed), this function must return CANCELLED.
+     *
+     * @return success True if the merge status is read successfully, false otherwise.
+     * @return status Merge status.
+     */
+    getSnapshotMergeStatus() generates (MergeStatus status);
+};
+
diff --git a/boot/1.1/default/Android.bp b/boot/1.1/default/Android.bp
new file mode 100644
index 0000000..dca5c26
--- /dev/null
+++ b/boot/1.1/default/Android.bp
@@ -0,0 +1,43 @@
+cc_library_shared {
+    name: "android.hardware.boot@1.1-impl",
+    defaults: [
+        "hidl_defaults",
+        "libboot_control_defaults",
+    ],
+    relative_install_path: "hw",
+    vendor: true,
+    recovery_available: true,
+    srcs: ["BootControl.cpp"],
+
+    shared_libs: [
+        "liblog",
+        "libhidlbase",
+        "libhardware",
+        "libutils",
+        "android.hardware.boot@1.0",
+        "android.hardware.boot@1.1",
+    ],
+    static_libs: [
+        "libboot_control",
+        "libfstab",
+    ],
+}
+
+cc_binary {
+    name: "android.hardware.boot@1.1-service",
+    defaults: ["hidl_defaults"],
+    relative_install_path: "hw",
+    vendor: true,
+    init_rc: ["android.hardware.boot@1.1-service.rc"],
+    srcs: ["service.cpp"],
+
+    shared_libs: [
+        "liblog",
+        "libhardware",
+        "libhidlbase",
+        "libutils",
+        "android.hardware.boot@1.0",
+        "android.hardware.boot@1.1",
+    ],
+
+}
diff --git a/boot/1.1/default/BootControl.cpp b/boot/1.1/default/BootControl.cpp
new file mode 100644
index 0000000..c9c62a4
--- /dev/null
+++ b/boot/1.1/default/BootControl.cpp
@@ -0,0 +1,129 @@
+/*
+ * Copyright (C) 2019 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+#define LOG_TAG "android.hardware.boot@1.1-impl"
+
+#include <memory>
+
+#include <log/log.h>
+
+#include "BootControl.h"
+
+namespace android {
+namespace hardware {
+namespace boot {
+namespace V1_1 {
+namespace implementation {
+
+using ::android::hardware::boot::V1_0::CommandResult;
+
+bool BootControl::Init() {
+    return impl_.Init();
+}
+
+// Methods from ::android::hardware::boot::V1_0::IBootControl follow.
+Return<uint32_t> BootControl::getNumberSlots() {
+    return impl_.GetNumberSlots();
+}
+
+Return<uint32_t> BootControl::getCurrentSlot() {
+    return impl_.GetCurrentSlot();
+}
+
+Return<void> BootControl::markBootSuccessful(markBootSuccessful_cb _hidl_cb) {
+    struct CommandResult cr;
+    if (impl_.MarkBootSuccessful()) {
+        cr.success = true;
+        cr.errMsg = "Success";
+    } else {
+        cr.success = false;
+        cr.errMsg = "Operation failed";
+    }
+    _hidl_cb(cr);
+    return Void();
+}
+
+Return<void> BootControl::setActiveBootSlot(uint32_t slot, setActiveBootSlot_cb _hidl_cb) {
+    struct CommandResult cr;
+    if (impl_.SetActiveBootSlot(slot)) {
+        cr.success = true;
+        cr.errMsg = "Success";
+    } else {
+        cr.success = false;
+        cr.errMsg = "Operation failed";
+    }
+    _hidl_cb(cr);
+    return Void();
+}
+
+Return<void> BootControl::setSlotAsUnbootable(uint32_t slot, setSlotAsUnbootable_cb _hidl_cb) {
+    struct CommandResult cr;
+    if (impl_.SetSlotAsUnbootable(slot)) {
+        cr.success = true;
+        cr.errMsg = "Success";
+    } else {
+        cr.success = false;
+        cr.errMsg = "Operation failed";
+    }
+    _hidl_cb(cr);
+    return Void();
+}
+
+Return<BoolResult> BootControl::isSlotBootable(uint32_t slot) {
+    if (!impl_.IsValidSlot(slot)) {
+        return BoolResult::INVALID_SLOT;
+    }
+    return impl_.IsSlotBootable(slot) ? BoolResult::TRUE : BoolResult::FALSE;
+}
+
+Return<BoolResult> BootControl::isSlotMarkedSuccessful(uint32_t slot) {
+    if (!impl_.IsValidSlot(slot)) {
+        return BoolResult::INVALID_SLOT;
+    }
+    return impl_.IsSlotMarkedSuccessful(slot) ? BoolResult::TRUE : BoolResult::FALSE;
+}
+
+Return<void> BootControl::getSuffix(uint32_t slot, getSuffix_cb _hidl_cb) {
+    hidl_string ans;
+    const char* suffix = impl_.GetSuffix(slot);
+    if (suffix) {
+        ans = suffix;
+    }
+    _hidl_cb(ans);
+    return Void();
+}
+
+Return<bool> BootControl::setSnapshotMergeStatus(MergeStatus status) {
+    return impl_.SetSnapshotMergeStatus(status);
+}
+
+Return<MergeStatus> BootControl::getSnapshotMergeStatus() {
+    return impl_.GetSnapshotMergeStatus();
+}
+
+IBootControl* HIDL_FETCH_IBootControl(const char* /* hal */) {
+    auto module = std::make_unique<BootControl>();
+    if (!module->Init()) {
+        ALOGE("Could not initialize BootControl module");
+        return nullptr;
+    }
+    return module.release();
+}
+
+}  // namespace implementation
+}  // namespace V1_1
+}  // namespace boot
+}  // namespace hardware
+}  // namespace android
diff --git a/boot/1.1/default/BootControl.h b/boot/1.1/default/BootControl.h
new file mode 100644
index 0000000..75511b6
--- /dev/null
+++ b/boot/1.1/default/BootControl.h
@@ -0,0 +1,64 @@
+/*
+ * Copyright (C) 2019 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#pragma once
+
+#include <android/hardware/boot/1.1/IBootControl.h>
+#include <hidl/MQDescriptor.h>
+#include <hidl/Status.h>
+#include <libboot_control/libboot_control.h>
+
+namespace android {
+namespace hardware {
+namespace boot {
+namespace V1_1 {
+namespace implementation {
+
+using ::android::hardware::Return;
+using ::android::hardware::Void;
+using ::android::hardware::boot::V1_0::BoolResult;
+using ::android::hardware::boot::V1_1::IBootControl;
+using ::android::hardware::boot::V1_1::MergeStatus;
+
+class BootControl : public IBootControl {
+  public:
+    bool Init();
+
+    // Methods from ::android::hardware::boot::V1_0::IBootControl follow.
+    Return<uint32_t> getNumberSlots() override;
+    Return<uint32_t> getCurrentSlot() override;
+    Return<void> markBootSuccessful(markBootSuccessful_cb _hidl_cb) override;
+    Return<void> setActiveBootSlot(uint32_t slot, setActiveBootSlot_cb _hidl_cb) override;
+    Return<void> setSlotAsUnbootable(uint32_t slot, setSlotAsUnbootable_cb _hidl_cb) override;
+    Return<BoolResult> isSlotBootable(uint32_t slot) override;
+    Return<BoolResult> isSlotMarkedSuccessful(uint32_t slot) override;
+    Return<void> getSuffix(uint32_t slot, getSuffix_cb _hidl_cb) override;
+
+    // Methods from ::android::hardware::boot::V1_1::IBootControl follow.
+    Return<bool> setSnapshotMergeStatus(MergeStatus status) override;
+    Return<MergeStatus> getSnapshotMergeStatus() override;
+
+  private:
+    android::bootable::BootControl impl_;
+};
+
+extern "C" IBootControl* HIDL_FETCH_IBootControl(const char* name);
+
+}  // namespace implementation
+}  // namespace V1_1
+}  // namespace boot
+}  // namespace hardware
+}  // namespace android
diff --git a/boot/1.1/default/android.hardware.boot@1.1-service.rc b/boot/1.1/default/android.hardware.boot@1.1-service.rc
new file mode 100644
index 0000000..83fa9d0
--- /dev/null
+++ b/boot/1.1/default/android.hardware.boot@1.1-service.rc
@@ -0,0 +1,6 @@
+service vendor.boot-hal-1-1 /vendor/bin/hw/android.hardware.boot@1.1-service
+    interface android.hardware.boot@1.0::IBootControl default
+    interface android.hardware.boot@1.1::IBootControl default
+    class early_hal
+    user root
+    group root
diff --git a/boot/1.1/default/service.cpp b/boot/1.1/default/service.cpp
new file mode 100644
index 0000000..b24b464
--- /dev/null
+++ b/boot/1.1/default/service.cpp
@@ -0,0 +1,26 @@
+/*
+ * Copyright (C) 2019 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+#define LOG_TAG "android.hardware.boot@1.1-service"
+
+#include <android/hardware/boot/1.1/IBootControl.h>
+#include <hidl/LegacySupport.h>
+
+using android::hardware::defaultPassthroughServiceImplementation;
+using ::android::hardware::boot::V1_1::IBootControl;
+
+int main(int /* argc */, char* /* argv */[]) {
+    return defaultPassthroughServiceImplementation<IBootControl>();
+}
diff --git a/boot/1.1/types.hal b/boot/1.1/types.hal
new file mode 100644
index 0000000..6346078
--- /dev/null
+++ b/boot/1.1/types.hal
@@ -0,0 +1,44 @@
+/*
+ * Copyright 2019 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package android.hardware.boot@1.1;
+
+enum MergeStatus : int32_t {
+    /**
+     * No snapshot or merge is in progress.
+     */
+    NONE = 0,
+
+    /**
+     * The merge status could not be determined.
+     */
+    UNKNOWN,
+
+    /**
+     * Partitions are being snapshotted, but no merge has been started.
+     */
+    SNAPSHOTTED,
+
+    /**
+     * At least one partition has merge is in progress.
+     */
+    MERGING,
+
+    /**
+     * A merge was in progress, but it was canceled by the bootloader.
+     */
+    CANCELLED,
+};
diff --git a/boot/1.1/vts/functional/Android.bp b/boot/1.1/vts/functional/Android.bp
new file mode 100644
index 0000000..49ea09a
--- /dev/null
+++ b/boot/1.1/vts/functional/Android.bp
@@ -0,0 +1,28 @@
+//
+// Copyright (C) 2019 The Android Open Source Project
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//      http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+//
+
+cc_test {
+    name: "VtsHalBootV1_1TargetTest",
+    defaults: ["VtsHalTargetTestDefaults"],
+    srcs: ["VtsHalBootV1_1TargetTest.cpp"],
+    static_libs: [
+        "android.hardware.boot@1.0",
+        "android.hardware.boot@1.1",
+        "libgmock",
+    ],
+    test_suites: ["device-tests"],
+}
+
diff --git a/boot/1.1/vts/functional/VtsHalBootV1_1TargetTest.cpp b/boot/1.1/vts/functional/VtsHalBootV1_1TargetTest.cpp
new file mode 100644
index 0000000..fba9a5e
--- /dev/null
+++ b/boot/1.1/vts/functional/VtsHalBootV1_1TargetTest.cpp
@@ -0,0 +1,86 @@
+/*
+ * Copyright (C) 2019 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#define LOG_TAG "boot_hidl_hal_test"
+
+#include <vector>
+
+#include <android-base/logging.h>
+#include <android/hardware/boot/1.1/IBootControl.h>
+#include <android/hardware/boot/1.1/types.h>
+#include <gmock/gmock.h>
+#include <gtest/gtest.h>
+#include <hidl/GtestPrinter.h>
+#include <hidl/ServiceManagement.h>
+
+#include <unistd.h>
+
+using ::android::sp;
+using ::android::hardware::hidl_enum_range;
+using ::android::hardware::hidl_vec;
+using ::android::hardware::Return;
+using ::android::hardware::Void;
+using ::android::hardware::boot::V1_1::IBootControl;
+using ::android::hardware::boot::V1_1::MergeStatus;
+using ::testing::Contains;
+
+class BootHidlTest : public testing::TestWithParam<std::string> {
+  public:
+    virtual void SetUp() override {
+        boot = IBootControl::getService(GetParam());
+        ASSERT_NE(boot, nullptr);
+
+        LOG(INFO) << "Test is remote " << boot->isRemote();
+    }
+
+    sp<IBootControl> boot;
+};
+
+static std::vector<MergeStatus> ValidMergeStatusValues() {
+    std::vector<MergeStatus> values;
+    for (const auto value : hidl_enum_range<MergeStatus>()) {
+        if (value == MergeStatus::UNKNOWN) {
+            continue;
+        }
+        values.push_back(value);
+    }
+    return values;
+}
+
+/**
+ * Ensure merge status can be retrieved.
+ */
+TEST_P(BootHidlTest, GetSnapshotMergeStatus) {
+    auto values = ValidMergeStatusValues();
+    auto status = (MergeStatus)boot->getSnapshotMergeStatus();
+    EXPECT_THAT(values, Contains(status));
+}
+
+/**
+ * Ensure merge status can be set to arbitrary value.
+ */
+TEST_P(BootHidlTest, SetSnapshotMergeStatus) {
+    for (const auto value : ValidMergeStatusValues()) {
+        EXPECT_TRUE(boot->setSnapshotMergeStatus(value).withDefault(false));
+        auto status = boot->getSnapshotMergeStatus();
+        EXPECT_EQ(status, value);
+    }
+}
+
+INSTANTIATE_TEST_SUITE_P(
+        , BootHidlTest,
+        testing::ValuesIn(android::hardware::getAllHalInstanceNames(IBootControl::descriptor)),
+        android::hardware::PrintInstanceNameToString);
diff --git a/compatibility_matrices/compatibility_matrix.current.xml b/compatibility_matrices/compatibility_matrix.current.xml
index be8d7ca..6a89dab 100644
--- a/compatibility_matrices/compatibility_matrix.current.xml
+++ b/compatibility_matrices/compatibility_matrix.current.xml
@@ -89,7 +89,7 @@
     </hal>
     <hal format="hidl" optional="true">
         <name>android.hardware.boot</name>
-        <version>1.0</version>
+        <version>1.1</version>
         <interface>
             <name>IBootControl</name>
             <instance>default</instance>
@@ -467,7 +467,7 @@
     </hal>
     <hal format="hidl" optional="true">
         <name>android.hardware.vibrator</name>
-        <version>1.0-3</version>
+        <version>1.0-4</version>
         <interface>
             <name>IVibrator</name>
             <instance>default</instance>
diff --git a/contexthub/1.0/default/OWNERS b/contexthub/1.0/default/OWNERS
index 5373073..90c2330 100644
--- a/contexthub/1.0/default/OWNERS
+++ b/contexthub/1.0/default/OWNERS
@@ -1,4 +1,3 @@
-aarossig@google.com
 arthuri@google.com
 bduddie@google.com
-bstack@google.com
+stange@google.com
diff --git a/contexthub/1.0/vts/functional/OWNERS b/contexthub/1.0/vts/functional/OWNERS
index ee01441..045cc4e 100644
--- a/contexthub/1.0/vts/functional/OWNERS
+++ b/contexthub/1.0/vts/functional/OWNERS
@@ -1,8 +1,7 @@
 #Context Hub team
-aarossig@google.com
 arthuri@google.com
 bduddie@google.com
-bstack@google.com
+stange@google.com
 
 #VTS team
 yim@google.com
diff --git a/current.txt b/current.txt
index 3fabd7c..972fe16 100644
--- a/current.txt
+++ b/current.txt
@@ -574,7 +574,15 @@
 # ABI preserving changes to HALs during Android R
 b69a7615c508acf5c5201efd1bfa3262167874fc3594e2db5a3ff93addd8ac75 android.hardware.keymaster@4.0::IKeymasterDevice
 eb2fa0c883c2185d514be0b84c179b283753ef0c1b77b45b4f359bd23bba8b75 android.hardware.neuralnetworks@1.0::IPreparedModel
+f1109cbb10297b7429a11fab42afa912710b303c9bf20bd5cdb8bd57b9c84186 android.hardware.neuralnetworks@1.0::types
+9d8ee57c490ffeaa28f702eaea8d198cb510e4bbfb99e6cb5f63e73341057c7c android.hardware.neuralnetworks@1.1::types
 fb382e986c10b8fbb797a8546e8f9ea6d1107bfe6f3fb7e57f6bbbf1f807a906 android.hardware.neuralnetworks@1.2::IDevice
 40e71cd693de5b832325c5d8f081f2ff20a7ba2b89d401cee5b4b3eb0e241681 android.hardware.neuralnetworks@1.2::IPreparedModel
+71c0f7127335e5b74d1615d5e7f129831b43ffbae5318ad0924d7d8d8910a859 android.hardware.neuralnetworks@1.2::types
+a785a57447a81e9c130eef6904c3a5c256076c6a04588c40620ebd6fa2660d77 android.hardware.radio@1.2::types
 1a6e2bd289f22931c526b21916910f1d4c436b7acb9556e4243de4ce8e6cc2e4 android.hardware.soundtrigger@2.0::ISoundTriggerHwCallback
 fd65298e1e09e0e3c781ab18305920d757dbe55a3b459ce17814ec5cf6dfee99 android.hardware.wifi@1.0::IWifiP2pIface
+
+# HALs released in Android R
+07d0a252b2d8fa35887908a996ba395cf392968395fc30afab791f46e0c22a52 android.hardware.boot@1.1::IBootControl
+74049a402be913963edfdd80828a53736570e9d8124a1bf18166b6ed46a6b0ab android.hardware.boot@1.1::types
diff --git a/dumpstate/1.0/default/DumpstateDevice.cpp b/dumpstate/1.0/default/DumpstateDevice.cpp
index 25d92b0..c57bf43 100644
--- a/dumpstate/1.0/default/DumpstateDevice.cpp
+++ b/dumpstate/1.0/default/DumpstateDevice.cpp
@@ -37,11 +37,6 @@
     // NOTE: this is just an example on how to use the DumpstateUtil.h functions to implement
     // this interface.
 
-    // Exit when dump is completed since this is a lazy HAL.
-    addPostCommandTask([]() {
-        exit(0);
-    });
-
     if (handle == nullptr || handle->numFds < 1) {
         ALOGE("no FDs\n");
         return Void();
diff --git a/dumpstate/1.0/default/service.cpp b/dumpstate/1.0/default/service.cpp
index 4f276b7..76c72b5 100644
--- a/dumpstate/1.0/default/service.cpp
+++ b/dumpstate/1.0/default/service.cpp
@@ -15,22 +15,26 @@
  */
 #define LOG_TAG "android.hardware.dumpstate@1.0-service"
 
+#include <hidl/HidlLazyUtils.h>
 #include <hidl/HidlSupport.h>
 #include <hidl/HidlTransportSupport.h>
 
 #include "DumpstateDevice.h"
 
-using ::android::hardware::configureRpcThreadpool;
-using ::android::hardware::dumpstate::V1_0::IDumpstateDevice;
-using ::android::hardware::dumpstate::V1_0::implementation::DumpstateDevice;
-using ::android::hardware::joinRpcThreadpool;
 using ::android::OK;
 using ::android::sp;
+using ::android::hardware::configureRpcThreadpool;
+using ::android::hardware::joinRpcThreadpool;
+using ::android::hardware::LazyServiceRegistrar;
+using ::android::hardware::dumpstate::V1_0::IDumpstateDevice;
+using ::android::hardware::dumpstate::V1_0::implementation::DumpstateDevice;
 
 int main(int /* argc */, char* /* argv */ []) {
     sp<IDumpstateDevice> dumpstate = new DumpstateDevice;
     configureRpcThreadpool(1, true /* will join */);
-    if (dumpstate->registerAsService() != OK) {
+
+    auto registrar = LazyServiceRegistrar::getInstance();
+    if (registrar.registerService(dumpstate) != OK) {
         ALOGE("Could not register service.");
         return 1;
     }
diff --git a/graphics/composer/2.2/vts/functional/Android.bp b/graphics/composer/2.2/vts/functional/Android.bp
index 9f7e1cd..a8e70ae 100644
--- a/graphics/composer/2.2/vts/functional/Android.bp
+++ b/graphics/composer/2.2/vts/functional/Android.bp
@@ -26,7 +26,6 @@
     shared_libs: [
         "libfmq",
         "libhidlbase",
-        "libhidltransport",
         "libsync",
     ],
     static_libs: [
diff --git a/graphics/mapper/2.0/utils/hal/include/mapper-hal/2.0/Mapper.h b/graphics/mapper/2.0/utils/hal/include/mapper-hal/2.0/Mapper.h
index 5ad2a65..8134174 100644
--- a/graphics/mapper/2.0/utils/hal/include/mapper-hal/2.0/Mapper.h
+++ b/graphics/mapper/2.0/utils/hal/include/mapper-hal/2.0/Mapper.h
@@ -80,17 +80,21 @@
     }
 
     Return<Error> freeBuffer(void* buffer) override {
-        native_handle_t* bufferHandle = removeImportedBuffer(buffer);
+        native_handle_t* bufferHandle = getImportedBuffer(buffer);
         if (!bufferHandle) {
             return Error::BAD_BUFFER;
         }
 
-        return mHal->freeBuffer(bufferHandle);
+        Error error = mHal->freeBuffer(bufferHandle);
+        if (error == Error::NONE) {
+            removeImportedBuffer(buffer);
+        }
+        return error;
     }
 
     Return<void> lock(void* buffer, uint64_t cpuUsage, const V2_0::IMapper::Rect& accessRegion,
                       const hidl_handle& acquireFence, IMapper::lock_cb hidl_cb) override {
-        const native_handle_t* bufferHandle = getImportedBuffer(buffer);
+        const native_handle_t* bufferHandle = getConstImportedBuffer(buffer);
         if (!bufferHandle) {
             hidl_cb(Error::BAD_BUFFER, nullptr);
             return Void();
@@ -112,7 +116,7 @@
     Return<void> lockYCbCr(void* buffer, uint64_t cpuUsage, const V2_0::IMapper::Rect& accessRegion,
                            const hidl_handle& acquireFence,
                            IMapper::lockYCbCr_cb hidl_cb) override {
-        const native_handle_t* bufferHandle = getImportedBuffer(buffer);
+        const native_handle_t* bufferHandle = getConstImportedBuffer(buffer);
         if (!bufferHandle) {
             hidl_cb(Error::BAD_BUFFER, YCbCrLayout{});
             return Void();
@@ -132,7 +136,7 @@
     }
 
     Return<void> unlock(void* buffer, IMapper::unlock_cb hidl_cb) override {
-        const native_handle_t* bufferHandle = getImportedBuffer(buffer);
+        const native_handle_t* bufferHandle = getConstImportedBuffer(buffer);
         if (!bufferHandle) {
             hidl_cb(Error::BAD_BUFFER, nullptr);
             return Void();
@@ -160,7 +164,11 @@
         return static_cast<native_handle_t*>(buffer);
     }
 
-    virtual const native_handle_t* getImportedBuffer(void* buffer) const {
+    virtual native_handle_t* getImportedBuffer(void* buffer) const {
+        return static_cast<native_handle_t*>(buffer);
+    }
+
+    virtual const native_handle_t* getConstImportedBuffer(void* buffer) const {
         return static_cast<const native_handle_t*>(buffer);
     }
 
diff --git a/graphics/mapper/2.0/utils/passthrough/include/mapper-passthrough/2.0/GrallocLoader.h b/graphics/mapper/2.0/utils/passthrough/include/mapper-passthrough/2.0/GrallocLoader.h
index e8b1b4b..85a91c3 100644
--- a/graphics/mapper/2.0/utils/passthrough/include/mapper-passthrough/2.0/GrallocLoader.h
+++ b/graphics/mapper/2.0/utils/passthrough/include/mapper-passthrough/2.0/GrallocLoader.h
@@ -68,7 +68,14 @@
         return mBufferHandles.erase(bufferHandle) == 1 ? bufferHandle : nullptr;
     }
 
-    const native_handle_t* get(void* buffer) {
+    native_handle_t* get(void* buffer) {
+        auto bufferHandle = static_cast<native_handle_t*>(buffer);
+
+        std::lock_guard<std::mutex> lock(mMutex);
+        return mBufferHandles.count(bufferHandle) == 1 ? bufferHandle : nullptr;
+    }
+
+    const native_handle_t* getConst(void* buffer) {
         auto bufferHandle = static_cast<const native_handle_t*>(buffer);
 
         std::lock_guard<std::mutex> lock(mMutex);
@@ -92,9 +99,13 @@
         return GrallocImportedBufferPool::getInstance().remove(buffer);
     }
 
-    const native_handle_t* getImportedBuffer(void* buffer) const override {
+    native_handle_t* getImportedBuffer(void* buffer) const override {
         return GrallocImportedBufferPool::getInstance().get(buffer);
     }
+
+    const native_handle_t* getConstImportedBuffer(void* buffer) const override {
+        return GrallocImportedBufferPool::getInstance().getConst(buffer);
+    }
 };
 
 class GrallocLoader {
diff --git a/graphics/mapper/2.1/utils/passthrough/include/mapper-passthrough/2.1/Gralloc0Hal.h b/graphics/mapper/2.1/utils/passthrough/include/mapper-passthrough/2.1/Gralloc0Hal.h
index 18fbb6d..8540068 100644
--- a/graphics/mapper/2.1/utils/passthrough/include/mapper-passthrough/2.1/Gralloc0Hal.h
+++ b/graphics/mapper/2.1/utils/passthrough/include/mapper-passthrough/2.1/Gralloc0Hal.h
@@ -37,6 +37,10 @@
      Error validateBufferSize(const native_handle_t* bufferHandle,
                               const IMapper::BufferDescriptorInfo& descriptorInfo,
                               uint32_t stride) override {
+         if (descriptorInfo.layerCount != 1) {
+             return Error::BAD_VALUE;
+         }
+
          if (!mModule->validateBufferSize) {
              return Error::NONE;
          }
diff --git a/keymaster/4.0/support/include/keymasterV4_0/keymaster_tags.h b/keymaster/4.0/support/include/keymasterV4_0/keymaster_tags.h
index 97dab68..cb29c64 100644
--- a/keymaster/4.0/support/include/keymasterV4_0/keymaster_tags.h
+++ b/keymaster/4.0/support/include/keymasterV4_0/keymaster_tags.h
@@ -280,39 +280,50 @@
  */
 template <typename ValueT>
 class NullOr {
-    template <typename T>
-    struct reference_initializer {
-#pragma GCC diagnostic push
-#pragma GCC diagnostic ignored "-Wnull-dereference"
-        static T&& init() { return *static_cast<std::remove_reference_t<T>*>(nullptr); }
-#pragma GCC diagnostic pop
-    };
-    template <typename T>
-    struct pointer_initializer {
-        static T init() { return nullptr; }
-    };
-    template <typename T>
-    struct value_initializer {
-        static T init() { return T(); }
-    };
-    template <typename T>
-    using initializer_t =
-        std::conditional_t<std::is_lvalue_reference<T>::value, reference_initializer<T>,
-                           std::conditional_t<std::is_pointer<T>::value, pointer_initializer<T>,
-                                              value_initializer<T>>>;
+    using internal_t = std::conditional_t<std::is_lvalue_reference<ValueT>::value,
+                                          std::remove_reference_t<ValueT>*, ValueT>;
 
-   public:
-    NullOr() : value_(initializer_t<ValueT>::init()), null_(true) {}
-    NullOr(ValueT&& value) : value_(std::forward<ValueT>(value)), null_(false) {}
+    struct pointer_initializer {
+        static std::nullptr_t init() { return nullptr; }
+    };
+    struct value_initializer {
+        static ValueT init() { return ValueT(); }
+    };
+    struct value_pointer_deref_t {
+        static ValueT& deref(ValueT& v) { return v; }
+    };
+    struct reference_deref_t {
+        static auto& deref(internal_t v) { return *v; }
+    };
+    using initializer_t = std::conditional_t<std::is_lvalue_reference<ValueT>::value ||
+                                                     std::is_pointer<ValueT>::value,
+                                             pointer_initializer, value_initializer>;
+    using deref_t = std::conditional_t<std::is_lvalue_reference<ValueT>::value, reference_deref_t,
+                                       value_pointer_deref_t>;
+
+  public:
+    NullOr() : value_(initializer_t::init()), null_(true) {}
+    template <typename T>
+    NullOr(T&& value, typename std::enable_if<
+                              !std::is_lvalue_reference<ValueT>::value &&
+                                      std::is_same<std::decay_t<ValueT>, std::decay_t<T>>::value,
+                              int>::type = 0)
+        : value_(std::forward<ValueT>(value)), null_(false) {}
+    template <typename T>
+    NullOr(T& value, typename std::enable_if<
+                             std::is_lvalue_reference<ValueT>::value &&
+                                     std::is_same<std::decay_t<ValueT>, std::decay_t<T>>::value,
+                             int>::type = 0)
+        : value_(&value), null_(false) {}
 
     bool isOk() const { return !null_; }
 
-    const ValueT& value() const & { return value_; }
-    ValueT& value() & { return value_; }
-    ValueT&& value() && { return std::move(value_); }
+    const ValueT& value() const& { return deref_t::deref(value_); }
+    ValueT& value() & { return deref_t::deref(value_); }
+    ValueT&& value() && { return std::move(deref_t::deref(value_)); }
 
-   private:
-    ValueT value_;
+  private:
+    internal_t value_;
     bool null_;
 };
 
diff --git a/keymaster/4.0/vts/functional/KeymasterHidlTest.cpp b/keymaster/4.0/vts/functional/KeymasterHidlTest.cpp
index 3af1df3..4838e7e 100644
--- a/keymaster/4.0/vts/functional/KeymasterHidlTest.cpp
+++ b/keymaster/4.0/vts/functional/KeymasterHidlTest.cpp
@@ -48,10 +48,11 @@
 SecurityLevel KeymasterHidlTest::securityLevel_;
 hidl_string KeymasterHidlTest::name_;
 hidl_string KeymasterHidlTest::author_;
+string KeymasterHidlTest::service_name_;
 
-void KeymasterHidlTest::SetUpTestCase() {
-    string service_name = KeymasterHidlEnvironment::Instance()->getServiceName<IKeymasterDevice>();
-    keymaster_ = ::testing::VtsHalHidlTargetTestBase::getService<IKeymasterDevice>(service_name);
+void KeymasterHidlTest::InitializeKeymaster() {
+    service_name_ = KeymasterHidlEnvironment::Instance()->getServiceName<IKeymasterDevice>();
+    keymaster_ = ::testing::VtsHalHidlTargetTestBase::getService<IKeymasterDevice>(service_name_);
     ASSERT_NE(keymaster_, nullptr);
 
     ASSERT_TRUE(keymaster_
@@ -62,18 +63,22 @@
                         author_ = author;
                     })
                     .isOk());
+}
+
+void KeymasterHidlTest::SetUpTestCase() {
+
+    InitializeKeymaster();
 
     os_version_ = ::keymaster::GetOsVersion();
     os_patch_level_ = ::keymaster::GetOsPatchlevel();
 
     auto service_manager = android::hidl::manager::V1_0::IServiceManager::getService();
     ASSERT_NE(nullptr, service_manager.get());
-
     all_keymasters_.push_back(keymaster_);
     service_manager->listByInterface(
         IKeymasterDevice::descriptor, [&](const hidl_vec<hidl_string>& names) {
             for (auto& name : names) {
-                if (name == service_name) continue;
+                if (name == service_name_) continue;
                 auto keymaster =
                     ::testing::VtsHalHidlTargetTestBase::getService<IKeymasterDevice>(name);
                 ASSERT_NE(keymaster, nullptr);
@@ -269,6 +274,13 @@
     return GetCharacteristics(key_blob, client_id, app_data, key_characteristics);
 }
 
+ErrorCode KeymasterHidlTest::GetDebugInfo(DebugInfo* debug_info) {
+    EXPECT_TRUE(keymaster_->getDebugInfo([&](const DebugInfo& hidl_debug_info) {
+      *debug_info = hidl_debug_info;
+    }).isOk());
+    return ErrorCode::OK;
+}
+
 ErrorCode KeymasterHidlTest::Begin(KeyPurpose purpose, const HidlBuf& key_blob,
                                    const AuthorizationSet& in_params, AuthorizationSet* out_params,
                                    OperationHandle* op_handle) {
@@ -611,6 +623,20 @@
     return ciphertext;
 }
 
+string KeymasterHidlTest::EncryptMessage(const string& message, BlockMode block_mode,
+                                         PaddingMode padding, uint8_t mac_length_bits,
+                                         const HidlBuf& iv_in) {
+    SCOPED_TRACE("EncryptMessage");
+    auto params = AuthorizationSetBuilder()
+                          .BlockMode(block_mode)
+                          .Padding(padding)
+                          .Authorization(TAG_MAC_LENGTH, mac_length_bits)
+                          .Authorization(TAG_NONCE, iv_in);
+    AuthorizationSet out_params;
+    string ciphertext = EncryptMessage(message, params, &out_params);
+    return ciphertext;
+}
+
 string KeymasterHidlTest::DecryptMessage(const HidlBuf& key_blob, const string& ciphertext,
                                          const AuthorizationSet& params) {
     SCOPED_TRACE("DecryptMessage");
diff --git a/keymaster/4.0/vts/functional/KeymasterHidlTest.h b/keymaster/4.0/vts/functional/KeymasterHidlTest.h
index 015fc43..b09da45 100644
--- a/keymaster/4.0/vts/functional/KeymasterHidlTest.h
+++ b/keymaster/4.0/vts/functional/KeymasterHidlTest.h
@@ -37,6 +37,7 @@
 
 using ::android::sp;
 using ::std::string;
+using hidl::base::V1_0::DebugInfo;
 
 class HidlBuf : public hidl_vec<uint8_t> {
     typedef hidl_vec<uint8_t> super;
@@ -95,6 +96,7 @@
 
     // SetUpTestCase runs only once per test case, not once per test.
     static void SetUpTestCase();
+    static void InitializeKeymaster();
     static void TearDownTestCase() {
         keymaster_.clear();
         all_keymasters_.clear();
@@ -140,6 +142,8 @@
                                  const HidlBuf& app_data, KeyCharacteristics* key_characteristics);
     ErrorCode GetCharacteristics(const HidlBuf& key_blob, KeyCharacteristics* key_characteristics);
 
+    ErrorCode GetDebugInfo(DebugInfo* debug_info);
+
     ErrorCode Begin(KeyPurpose purpose, const HidlBuf& key_blob, const AuthorizationSet& in_params,
                     AuthorizationSet* out_params, OperationHandle* op_handle);
     ErrorCode Begin(KeyPurpose purpose, const AuthorizationSet& in_params,
@@ -201,6 +205,8 @@
                           HidlBuf* iv_out);
     string EncryptMessage(const string& message, BlockMode block_mode, PaddingMode padding,
                           const HidlBuf& iv_in);
+    string EncryptMessage(const string& message, BlockMode block_mode, PaddingMode padding,
+                          uint8_t mac_length_bits, const HidlBuf& iv_in);
 
     string DecryptMessage(const HidlBuf& key_blob, const string& ciphertext,
                           const AuthorizationSet& params);
@@ -235,6 +241,7 @@
     static SecurityLevel securityLevel_;
     static hidl_string name_;
     static hidl_string author_;
+    static string service_name_;
 };
 
 }  // namespace test
diff --git a/keymaster/4.0/vts/functional/keymaster_hidl_hal_test.cpp b/keymaster/4.0/vts/functional/keymaster_hidl_hal_test.cpp
index 9e6cce7..0ac7e48 100644
--- a/keymaster/4.0/vts/functional/keymaster_hidl_hal_test.cpp
+++ b/keymaster/4.0/vts/functional/keymaster_hidl_hal_test.cpp
@@ -18,6 +18,7 @@
 #include <cutils/log.h>
 
 #include <iostream>
+#include <signal.h>
 
 #include <openssl/evp.h>
 #include <openssl/mem.h>
@@ -2706,6 +2707,40 @@
 }
 
 /*
+ * EncryptionOperationsTest.AesWrongPurpose
+ *
+ * Verifies that AES encryption fails in the correct way when an unauthorized purpose is specified.
+ */
+TEST_F(EncryptionOperationsTest, AesWrongPurpose) {
+    auto err = GenerateKey(AuthorizationSetBuilder()
+                                   .Authorization(TAG_NO_AUTH_REQUIRED)
+                                   .AesKey(128)
+                                   .Authorization(TAG_PURPOSE, KeyPurpose::ENCRYPT)
+                                   .Authorization(TAG_BLOCK_MODE, BlockMode::GCM)
+                                   .Authorization(TAG_MIN_MAC_LENGTH, 128)
+                                   .Padding(PaddingMode::NONE));
+    ASSERT_EQ(ErrorCode::OK, err) << "Got " << err;
+
+    err = Begin(KeyPurpose::DECRYPT,
+                AuthorizationSetBuilder().BlockMode(BlockMode::GCM).Padding(PaddingMode::NONE));
+    EXPECT_EQ(ErrorCode::INCOMPATIBLE_PURPOSE, err) << "Got " << err;
+
+    CheckedDeleteKey();
+
+    ASSERT_EQ(ErrorCode::OK, GenerateKey(AuthorizationSetBuilder()
+                                                 .Authorization(TAG_NO_AUTH_REQUIRED)
+                                                 .AesKey(128)
+                                                 .Authorization(TAG_PURPOSE, KeyPurpose::DECRYPT)
+                                                 .Authorization(TAG_BLOCK_MODE, BlockMode::GCM)
+                                                 .Authorization(TAG_MIN_MAC_LENGTH, 128)
+                                                 .Padding(PaddingMode::NONE)));
+
+    err = Begin(KeyPurpose::ENCRYPT,
+                AuthorizationSetBuilder().BlockMode(BlockMode::GCM).Padding(PaddingMode::NONE));
+    EXPECT_EQ(ErrorCode::INCOMPATIBLE_PURPOSE, err) << "Got " << err;
+}
+
+/*
  * EncryptionOperationsTest.AesEcbNoPaddingWrongInputSize
  *
  * Verifies that AES encryption fails in the correct way when provided an input that is not a
@@ -3225,6 +3260,92 @@
 }
 
 /*
+ * EncryptionOperationsTest.AesGcmRoundTripWithDelaySuccess
+ *
+ * Verifies that AES GCM mode works, even when there's a long delay
+ * between operations.
+ */
+TEST_F(EncryptionOperationsTest, AesGcmRoundTripWithDelaySuccess) {
+    ASSERT_EQ(ErrorCode::OK, GenerateKey(AuthorizationSetBuilder()
+                                             .Authorization(TAG_NO_AUTH_REQUIRED)
+                                             .AesEncryptionKey(128)
+                                             .Authorization(TAG_BLOCK_MODE, BlockMode::GCM)
+                                             .Padding(PaddingMode::NONE)
+                                             .Authorization(TAG_MIN_MAC_LENGTH, 128)));
+
+    string aad = "foobar";
+    string message = "123456789012345678901234567890123456";
+
+    auto begin_params = AuthorizationSetBuilder()
+                            .BlockMode(BlockMode::GCM)
+                            .Padding(PaddingMode::NONE)
+                            .Authorization(TAG_MAC_LENGTH, 128);
+
+    auto update_params =
+        AuthorizationSetBuilder().Authorization(TAG_ASSOCIATED_DATA, aad.data(), aad.size());
+
+    // Encrypt
+    AuthorizationSet begin_out_params;
+    ASSERT_EQ(ErrorCode::OK, Begin(KeyPurpose::ENCRYPT, begin_params, &begin_out_params))
+        << "Begin encrypt";
+    string ciphertext;
+    AuthorizationSet update_out_params;
+    sleep(5);
+    ASSERT_EQ(ErrorCode::OK,
+              Finish(op_handle_, update_params, message, "", &update_out_params, &ciphertext));
+
+    ASSERT_EQ(ciphertext.length(), message.length() + 16);
+
+    // Grab nonce
+    begin_params.push_back(begin_out_params);
+
+    // Decrypt.
+    ASSERT_EQ(ErrorCode::OK, Begin(KeyPurpose::DECRYPT, begin_params)) << "Begin decrypt";
+    string plaintext;
+    size_t input_consumed;
+    sleep(5);
+    ASSERT_EQ(ErrorCode::OK, Update(op_handle_, update_params, ciphertext, &update_out_params,
+                                    &plaintext, &input_consumed));
+    EXPECT_EQ(ciphertext.size(), input_consumed);
+    sleep(5);
+    EXPECT_EQ(ErrorCode::OK, Finish("", &plaintext));
+    EXPECT_EQ(message.length(), plaintext.length());
+    EXPECT_EQ(message, plaintext);
+}
+
+/*
+ * EncryptionOperationsTest.AesGcmDifferentNonces
+ *
+ * Verifies that encrypting the same data with different nonces produces different outputs.
+ */
+TEST_F(EncryptionOperationsTest, AesGcmDifferentNonces) {
+    ASSERT_EQ(ErrorCode::OK, GenerateKey(AuthorizationSetBuilder()
+                                                 .Authorization(TAG_NO_AUTH_REQUIRED)
+                                                 .AesEncryptionKey(128)
+                                                 .Authorization(TAG_BLOCK_MODE, BlockMode::GCM)
+                                                 .Padding(PaddingMode::NONE)
+                                                 .Authorization(TAG_MIN_MAC_LENGTH, 128)
+                                                 .Authorization(TAG_CALLER_NONCE)));
+
+    string aad = "foobar";
+    string message = "123456789012345678901234567890123456";
+    string nonce1 = "000000000000";
+    string nonce2 = "111111111111";
+    string nonce3 = "222222222222";
+
+    string ciphertext1 =
+            EncryptMessage(message, BlockMode::GCM, PaddingMode::NONE, 128, HidlBuf(nonce1));
+    string ciphertext2 =
+            EncryptMessage(message, BlockMode::GCM, PaddingMode::NONE, 128, HidlBuf(nonce2));
+    string ciphertext3 =
+            EncryptMessage(message, BlockMode::GCM, PaddingMode::NONE, 128, HidlBuf(nonce3));
+
+    ASSERT_NE(ciphertext1, ciphertext2);
+    ASSERT_NE(ciphertext1, ciphertext3);
+    ASSERT_NE(ciphertext2, ciphertext3);
+}
+
+/*
  * EncryptionOperationsTest.AesGcmTooShortTag
  *
  * Verifies that AES GCM mode fails correctly when a too-short tag length is specified.
@@ -4456,6 +4577,84 @@
     EXPECT_EQ(result, std::make_pair(ErrorCode::OK, HidlBuf()));
 }
 
+
+using ClearOperationsTest = KeymasterHidlTest;
+
+/*
+ * ClearSlotsTest.TooManyOperations
+ *
+ * Verifies that TOO_MANY_OPERATIONS is returned after the max number of
+ * operations are started without being finished or aborted. Also verifies
+ * that aborting the operations clears the operations.
+ *
+ */
+TEST_F(ClearOperationsTest, TooManyOperations) {
+    ASSERT_EQ(ErrorCode::OK, GenerateKey(AuthorizationSetBuilder()
+                                             .Authorization(TAG_NO_AUTH_REQUIRED)
+                                             .RsaEncryptionKey(2048, 65537)
+                                             .Padding(PaddingMode::NONE)));
+
+    auto params = AuthorizationSetBuilder().Padding(PaddingMode::NONE);
+    int max_operations = SecLevel() == SecurityLevel::STRONGBOX ? 4 : 16;
+    OperationHandle op_handles[max_operations];
+    AuthorizationSet out_params;
+    for(int i=0; i<max_operations; i++) {
+        EXPECT_EQ(ErrorCode::OK, Begin(KeyPurpose::ENCRYPT, key_blob_, params, &out_params, &(op_handles[i])));
+    }
+    EXPECT_EQ(ErrorCode::TOO_MANY_OPERATIONS,
+         Begin(KeyPurpose::ENCRYPT, key_blob_, params, &out_params, &op_handle_));
+    // Try again just in case there's a weird overflow bug
+    EXPECT_EQ(ErrorCode::TOO_MANY_OPERATIONS,
+         Begin(KeyPurpose::ENCRYPT, key_blob_, params, &out_params, &op_handle_));
+    for(int i=0; i<max_operations; i++) {
+        EXPECT_EQ(ErrorCode::OK, Abort(op_handles[i]));
+    }
+    EXPECT_EQ(ErrorCode::OK,
+         Begin(KeyPurpose::ENCRYPT, key_blob_, params, &out_params, &op_handle_));
+    AbortIfNeeded();
+}
+
+/*
+ * ClearSlotsTest.ServiceDeath
+ *
+ * Verifies that the service is restarted after death and the ongoing
+ * operations are cleared.
+ */
+TEST_F(ClearOperationsTest, ServiceDeath) {
+
+    ASSERT_EQ(ErrorCode::OK, GenerateKey(AuthorizationSetBuilder()
+                                             .Authorization(TAG_NO_AUTH_REQUIRED)
+                                             .RsaEncryptionKey(2048, 65537)
+                                             .Padding(PaddingMode::NONE)));
+
+    auto params = AuthorizationSetBuilder().Padding(PaddingMode::NONE);
+    int max_operations = SecLevel() == SecurityLevel::STRONGBOX ? 4 : 16;
+    OperationHandle op_handles[max_operations];
+    AuthorizationSet out_params;
+    for(int i=0; i<max_operations; i++) {
+        EXPECT_EQ(ErrorCode::OK, Begin(KeyPurpose::ENCRYPT, key_blob_, params, &out_params, &(op_handles[i])));
+    }
+    EXPECT_EQ(ErrorCode::TOO_MANY_OPERATIONS,
+         Begin(KeyPurpose::ENCRYPT, key_blob_, params, &out_params, &op_handle_));
+
+    DebugInfo debug_info;
+    GetDebugInfo(&debug_info);
+    kill(debug_info.pid, SIGKILL);
+    // wait 1 second for keymaster to restart
+    sleep(1);
+    InitializeKeymaster();
+
+    for(int i=0; i<max_operations; i++) {
+        EXPECT_EQ(ErrorCode::OK, Begin(KeyPurpose::ENCRYPT, key_blob_, params, &out_params, &(op_handles[i])));
+    }
+    EXPECT_EQ(ErrorCode::TOO_MANY_OPERATIONS,
+         Begin(KeyPurpose::ENCRYPT, key_blob_, params, &out_params, &op_handle_));
+    for(int i=0; i<max_operations; i++) {
+        EXPECT_EQ(ErrorCode::OK, Abort(op_handles[i]));
+    }
+}
+
+
 }  // namespace test
 }  // namespace V4_0
 }  // namespace keymaster
diff --git a/light/utils/main.cpp b/light/utils/main.cpp
index d07e799..b834132 100644
--- a/light/utils/main.cpp
+++ b/light/utils/main.cpp
@@ -25,7 +25,7 @@
     std::cerr << msg << std::endl;
 }
 
-int main() {
+int main(int argc, char* argv[]) {
     using ::android::hardware::hidl_vec;
     using ::android::hardware::light::V2_0::Brightness;
     using ::android::hardware::light::V2_0::Flash;
@@ -41,10 +41,29 @@
         return -1;
     }
 
-    const static LightState off = {
-        .color = 0u, .flashMode = Flash::NONE, .brightnessMode = Brightness::USER,
+    static LightState off = {
+            .color = 0u,
+            .flashMode = Flash::NONE,
+            .brightnessMode = Brightness::USER,
     };
 
+    if (argc > 2) {
+        error("Usage: blank_screen [color]");
+        return -1;
+    }
+
+    if (argc > 1) {
+        char* col_ptr;
+        unsigned int col_new;
+
+        col_new = strtoul(argv[1], &col_ptr, 0);
+        if (*col_ptr != '\0') {
+            error("Failed to convert " + std::string(argv[1]) + " to number");
+            return -1;
+        }
+        off.color = col_new;
+    }
+
     service->getSupportedTypes([&](const hidl_vec<Type>& types) {
         for (Type type : types) {
             Status ret = service->setLight(type, off);
diff --git a/neuralnetworks/1.0/types.hal b/neuralnetworks/1.0/types.hal
index 02db063..ba9d068 100644
--- a/neuralnetworks/1.0/types.hal
+++ b/neuralnetworks/1.0/types.hal
@@ -25,25 +25,24 @@
  * with at least one dimension). Types not prefaced by TENSOR_* represent
  * scalar values and must have no dimensions.
  *
- * Although many types are defined, most operators accept just a few
+ * Although we define many types, most operators accept just a few
  * types. Most used are {@link OperandType::TENSOR_FLOAT32},
  * {@link OperandType::TENSOR_QUANT8_ASYMM},
  * and {@link OperandType::INT32}.
  */
 enum OperandType : int32_t {
     /** A 32 bit floating point scalar value. */
-    FLOAT32             = 0,
+    FLOAT32 = 0,
     /** A signed 32 bit integer scalar value. */
-    INT32               = 1,
+    INT32 = 1,
     /** An unsigned 32 bit integer scalar value. */
-    UINT32              = 2,
-
+    UINT32 = 2,
     /** A tensor of 32 bit floating point values. */
-    TENSOR_FLOAT32      = 3,
+    TENSOR_FLOAT32 = 3,
     /** A tensor of 32 bit integer values. */
-    TENSOR_INT32        = 4,
+    TENSOR_INT32 = 4,
     /**
-     * A tensor of 8 bit integers that represent real numbers.
+     * A tensor of 8 bit unsigned integers that represent real numbers.
      *
      * Attached to this tensor are two numbers that can be used to convert the
      * 8 bit integer to the real value and vice versa. These two numbers are:
@@ -51,21 +50,21 @@
      * - zeroPoint: a 32 bit integer, in range [0, 255].
      *
      * The formula is:
-     * real_value = (integer_value - zeroPoint) * scale.
+     *   real_value = (integer_value - zeroPoint) * scale.
      */
     TENSOR_QUANT8_ASYMM = 5,
 
     /**
-     * DEPRECATED. Since NNAPI 1.2, extensions are the preferred alternative to
-     * OEM operation and data types.
+     * DEPRECATED. Since HAL version 1.2, extensions are the preferred
+     * alternative to OEM operation and data types.
      *
      * OEM specific scalar value.
      */
     OEM                 = 10000,
 
     /**
-     * DEPRECATED. Since NNAPI 1.2, extensions are the preferred alternative to
-     * OEM operation and data types.
+     * DEPRECATED. Since HAL version 1.2, extensions are the preferred
+     * alternative to OEM operation and data types.
      *
      * A tensor of OEM specific values.
      */
@@ -78,7 +77,6 @@
  * The type of an operation in a model.
  */
 enum OperationType : int32_t {
-
     /**
      * Adds two tensors, element-wise.
      *
@@ -110,14 +108,16 @@
      * * 0: A tensor.
      * * 1: A tensor of the same {@link OperandType}, and compatible dimensions
      *      as input0.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scales and zeroPoint can be different from input0 scale and zeroPoint.
      * * 2: An {@link OperandType::INT32} scalar, and has to be one of the
      *      {@link FusedActivationFunc} values. Specifies the activation to
      *      invoke on the result.
      *
      * Outputs:
      * * 0: The sum, a tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint can be different from inputs' scale and zeroPoint.
      */
     ADD = 0,
 
@@ -187,8 +187,8 @@
      * Outputs:
      * * 0: The output 4-D tensor, of shape
      *      [batches, out_height, out_width, depth].
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     AVERAGE_POOL_2D = 1,
 
@@ -206,22 +206,23 @@
      *
      * Inputs:
      * * 0 ~ n-1: The list of n input tensors, of shape
-     *            [D0, D1, ..., Daxis(i), ..., Dm]. For inputs of
-     *            {@link OperandType::TENSOR_QUANT8_ASYMM}, all input tensors
-     *            must have the same scale and zeroPoint.
+     *            [D0, D1, ..., Daxis(i), ..., Dm].
+     *            All input tensors of
+     *            {@link OperandType::TENSOR_QUANT8_ASYMM}
+     *            must have the same scale and zeroPoint as the output tensor.
      * * n: An {@link OperandType::INT32} scalar, specifying the
      *      concatenation axis.
      *
      * Outputs:
      * * 0: The output, a tensor of the same {@link OperandType} as the input
      *      tensors. The output shape is [D0, D1, ..., sum(Daxis(i)), ..., Dm].
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor, the scale and zeroPoint
+     *      values must be the same as the input tensors'.
      */
     CONCATENATION = 2,
 
     /**
-     * Performs an 2-D convolution operation.
+     * Performs a 2-D convolution operation.
      *
      * The CONV_2D op sweeps a 2-D filter that can mix channels together over a
      * batch of images, applying the filter to each window of each image of the
@@ -238,11 +239,17 @@
      *             filter[channel, di, dj, k]
      *         ) + bias[channel]
      *
-     * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT32}
-     * * {@link OperandType::TENSOR_QUANT8_ASYMM}
+     * Supported tensor {@link OperandType} configurations:
+     * * 32 bit floating point:
+     * * * {@link OperandType::TENSOR_FLOAT32} for input, filter, output, and bias.
      *
-     * Supported tensor rank: 4, with "NHWC" data layout.
+     * * Quantized:
+     * * * {@link OperandType::TENSOR_QUANT8_ASYMM} for input, filter, and output.
+     * * * {@link OperandType::TENSOR_INT32} for bias (with scale set to
+     * * * input.scale * filter.scale).
+     *
+     * Supported tensor rank: 4, with "NHWC" (i.e., Num_samples, Height, Width,
+     * and Channels) data layout.
      *
      * Both explicit padding and implicit padding are supported.
      *
@@ -252,12 +259,12 @@
      * * 1: A 4-D tensor, of shape
      *      [depth_out, filter_height, filter_width, depth_in], specifying the
      *      filter.
-     * * 2: A 1-D tensor, of shape [depth_out], specifying the bias.
-     *      For input tensor of {@link OperandType::TENSOR_FLOAT32}, the bias
-     *      should also be of {@link OperandType::TENSOR_FLOAT32}. For input
-     *      tensor of {@link OperandType::TENSOR_QUANT8_ASYMM}, the bias
-     *      should be of {@link OperandType::TENSOR_INT32}, with zeroPoint of
-     *      0 and bias_scale == input_scale * filter_scale.
+     * * 2: A 1-D tensor, of shape [depth_out], specifying the bias. For input
+     *      tensor of type {@link OperandType::TENSOR_FLOAT32}
+     *      the bias must be of the same
+     *      type. For filter tensor of {@link OperandType::TENSOR_QUANT8_ASYMM},
+     *      the bias should be of {@link OperandType::TENSOR_INT32}, with zeroPoint
+     *      of 0 and bias_scale == input_scale * filter_scale.
      * * 3: An {@link OperandType::INT32} scalar, specifying the padding on
      *      the left, in the ‘width’ dimension.
      * * 4: An {@link OperandType::INT32} scalar, specifying the padding on
@@ -281,11 +288,11 @@
      *      [depth_out, filter_height, filter_width, depth_in], specifying the
      *      filter.
      * * 2: A 1-D tensor, of shape [depth_out], specifying the bias. For input
-     *      tensor of {@link OperandType::TENSOR_FLOAT32}, the bias should
-     *      also be of {@link OperandType::TENSOR_FLOAT32}. For input tensor
-     *      of {@link OperandType::TENSOR_QUANT8_ASYMM}, the bias should be
-     *      of {@link OperandType::TENSOR_INT32}, with zeroPoint of 0 and
-     *      bias_scale == input_scale * filter_scale.
+     *      tensor of type {@link OperandType::TENSOR_FLOAT32}
+     *      the bias must be of the same
+     *      type. For filter tensor of {@link OperandType::TENSOR_QUANT8_ASYMM},
+     *      the bias should be of {@link OperandType::TENSOR_INT32}, with zeroPoint
+     *      of 0 and bias_scale == input_scale * filter_scale.
      * * 3: An {@link OperandType::INT32} scalar, specifying the implicit
      *      padding scheme, has to be one of the
      *      following values: {0 (NONE), 1 (SAME), 2 (VALID)}.
@@ -299,11 +306,9 @@
      *
      * Outputs:
      * * 0: The output 4-D tensor, of shape
-     *      [batches, out_height, out_width, depth_out]. For output tensor of
-     *      {@link OperandType::TENSOR_QUANT8_ASYMM}, the following condition
-     *      must be satisfied: output_scale > input_scale * filter_scale.
-     *
-     * Available since API level 27.
+     *      [batches, out_height, out_width, depth_out].
+     *      For output tensor of {@link OperandType::TENSOR_QUANT8_ASYMM},
+     *      the following condition must be satisfied: output_scale > input_scale * filter_scale
      */
     CONV_2D = 3,
 
@@ -329,11 +334,17 @@
      *             filter[1, di, dj, k * channel_multiplier + q]
      *         ) + bias[k * channel_multiplier + q]
      *
-     * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT32}
-     * * {@link OperandType::TENSOR_QUANT8_ASYMM}
+     * Supported tensor {@link OperandType} configurations:
+     * * 32 bit floating point:
+     * * * {@link OperandType::TENSOR_FLOAT32} for input, filter, output, and bias.
      *
-     * Supported tensor rank: 4, with "NHWC" data layout.
+     * * Quantized:
+     * * * {@link OperandType::TENSOR_QUANT8_ASYMM} for input, filter, and output.
+     * * * {@link OperandType::TENSOR_INT32} for bias (with scale set to
+     * * * input.scale * filter.scale).
+     *
+     * Supported tensor rank: 4, with "NHWC" (i.e., Num_samples, Height, Width,
+     * and Channels) data layout.
      *
      * Both explicit padding and implicit padding are supported.
      *
@@ -343,11 +354,11 @@
      * * 1: A 4-D tensor, of shape [1, filter_height, filter_width, depth_out],
      *      specifying the filter.
      * * 2: A 1-D tensor, of shape [depth_out], specifying the bias. For input
-     *      tensor of {@link OperandType::TENSOR_FLOAT32}, the bias should
-     *      also be of {@link OperandType::TENSOR_FLOAT32}. For input tensor
-     *      of {@link OperandType::TENSOR_QUANT8_ASYMM}, the bias should be
-     *      of {@link OperandType::TENSOR_INT32}, with zeroPoint of 0 and
-     *      bias_scale == input_scale * filter_scale.
+     *      tensor of type {@link OperandType::TENSOR_FLOAT32}
+     *      the bias must be of the same
+     *      type. For filter tensor of {@link OperandType::TENSOR_QUANT8_ASYMM},
+     *      the bias should be of {@link OperandType::TENSOR_INT32}, with zeroPoint
+     *      of 0 and bias_scale == input_scale * filter_scale.
      * * 3: An {@link OperandType::INT32} scalar, specifying the padding on
      *      the left, in the ‘width’ dimension.
      * * 4: An {@link OperandType::INT32} scalar, specifying the padding on
@@ -372,11 +383,11 @@
      * * 1: A 4-D tensor, of shape [1, filter_height, filter_width, depth_out],
      *      specifying the filter.
      * * 2: A 1-D tensor, of shape [depth_out], specifying the bias. For input
-     *      tensor of {@link OperandType::TENSOR_FLOAT32}, the bias should
-     *      also be of {@link OperandType::TENSOR_FLOAT32}. For input tensor
-     *      of {@link OperandType::TENSOR_QUANT8_ASYMM}, the bias should be
-     *      of {@link OperandType::TENSOR_INT32}, with zeroPoint of 0 and
-     *      bias_scale == input_scale * filter_scale.
+     *      tensor of type {@link OperandType::TENSOR_FLOAT32}
+     *      the bias must be of the same
+     *      type. For filter tensor of {@link OperandType::TENSOR_QUANT8_ASYMM},
+     *      the bias should be of {@link OperandType::TENSOR_INT32}, with zeroPoint
+     *      of 0 and bias_scale == input_scale * filter_scale.
      * * 3: An {@link OperandType::INT32} scalar, specifying the implicit
      *      padding scheme, has to be one of the
      *      following values: {0 (NONE), 1 (SAME), 2 (VALID)}.
@@ -392,11 +403,10 @@
      *
      * Outputs:
      * * 0: The output 4-D tensor, of shape
-     *      [batches, out_height, out_width, depth_out]. For output tensor of
-     *      {@link OperandType::TENSOR_QUANT8_ASYMM}, the following condition
-     *      must be satisfied: output_scale > input_scale * filter_scale.
-     *
-     * Available since API level 27.
+     *      [batches, out_height, out_width, depth_out]. For
+     *      output tensor of {@link OperandType::TENSOR_QUANT8_ASYMM},
+     *      the following condition must be satisfied:
+     *      output_scale > input_scale * filter_scale
      */
     DEPTHWISE_CONV_2D = 4,
 
@@ -419,7 +429,8 @@
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
-     * Supported tensor rank: 4, with "NHWC" data layout.
+     * Supported tensor rank: 4, with "NHWC" (i.e., Num_samples, Height, Width,
+     * and Channels) data layout.
      *
      * Inputs:
      * * 0: A 4-D tensor, of shape [batches, height, width, depth_in],
@@ -431,8 +442,8 @@
      * Outputs:
      * * 0: The output 4-D tensor, of shape [batch, height*block_size,
      *      width*block_size, depth/(block_size*block_size)].
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     DEPTH_TO_SPACE = 5,
 
@@ -443,19 +454,19 @@
      *
      *     output = (input - zeroPoint) * scale.
      *
-     * Supported tensor {@link OperandType}:
+     * Supported input tensor {@link OperandType}:
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
+     * Supported output tensor {@link OperandType}:
+     * * {@link OperandType::TENSOR_FLOAT32}.
+     *
      * Supported tensor rank: up to 4
      *
      * Inputs:
-     * * 0: A tensor of {@link OperandType::TENSOR_QUANT8_ASYMM}.
+     * * 0: A tensor.
      *
      * Outputs:
-     * * 0: The output tensor of same shape as input0, but with
-     *      {@link OperandType::TENSOR_FLOAT32}.
-     *
-     * Available since API level 27.
+     * * 0: A tensor with the same shape as input0.
      */
     DEQUANTIZE = 6,
 
@@ -479,6 +490,13 @@
      * If a value in Lookups is out of bounds, the operation must fail
      * and an error must be reported.
      *
+     * Supported value tensor {@link OperandType}:
+     * * {@link OperandType::TENSOR_FLOAT32}
+     * * {@link OperandType::TENSOR_INT32}
+     * * {@link OperandType::TENSOR_QUANT8_ASYMM}
+     *
+     * Supported value tensor rank: from 2
+     *
      * Inputs:
      * * 0: Lookups. A 1-D tensor of {@link OperandType::TENSOR_INT32}.
      *      The values are indices into the first dimension of Values.
@@ -489,8 +507,8 @@
      * * 0: A n-D tensor with the same rank and shape as the Values
      *      tensor, except for the first dimension which has the same size
      *      as Lookups' only dimension.
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input1.
      */
     EMBEDDING_LOOKUP = 7,
 
@@ -508,8 +526,6 @@
      * Outputs:
      * * 0: The output tensor, of the same {@link OperandType} and dimensions as
      *      the input tensor.
-     *
-     * Available since API level 27.
      */
     FLOOR = 8,
 
@@ -549,12 +565,9 @@
      *      invoke on the result.
      *
      * Outputs:
-     * * 0: The output tensor, of shape [batch_size, num_units]. For output
-     *      tensor of {@link OperandType::TENSOR_QUANT8_ASYMM}, the following
-     *      condition must be satisfied:
-     *      output_scale > input_scale * filter_scale.
-     *
-     * Available since API level 27.
+     * * 0: The output tensor, of shape [batch_size, num_units]. For
+     *      output tensor of {@link OperandType::TENSOR_QUANT8_ASYMM}, the following
+     *      condition must be satisfied: output_scale > input_scale * filter_scale.
      */
     FULLY_CONNECTED = 9,
 
@@ -585,6 +598,13 @@
      * must be selected. If no entry in Keys has 123456, a slice of zeroes
      * must be concatenated.
      *
+     * Supported value tensor {@link OperandType}:
+     * * {@link OperandType::TENSOR_FLOAT32}
+     * * {@link OperandType::TENSOR_INT32}
+     * * {@link OperandType::TENSOR_QUANT8_ASYMM}
+     *
+     * Supported value tensor rank: from 2
+     *
      * Inputs:
      * * 0: Lookups. A 1-D {@link OperandType::TENSOR_INT32} tensor with
      *      shape [ k ].
@@ -598,13 +618,13 @@
      *
      * Outputs:
      * * 0: Output. A tensor with shape [ k …].
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input2.
      * * 1: Hits. A boolean tensor with shape [ k ] indicates whether the lookup
      *      hits (True) or not (False).
      *      Stored as {@link OperandType::TENSOR_QUANT8_ASYMM} with offset 0
      *      and scale 1.0f.
      *      A non-zero byte represents True, a hit. A zero indicates otherwise.
-     *
-     * Available since API level 27.
      */
     HASHTABLE_LOOKUP = 10,
 
@@ -617,9 +637,6 @@
      *         input[batch, row, col, channel] /
      *         sqrt(sum_{c} pow(input[batch, row, col, c], 2))
      *
-     * For input tensor with more dimensions, independently normalizes each 1-D
-     * slice along dimension dim.
-     *
      * Supported tensor {@link OperandType}:
      * * {@link OperandType::TENSOR_FLOAT32}
      *
@@ -627,13 +644,10 @@
      * Height, Width, and Channels).
      *
      * Inputs:
-     * * 0: A 4-D tensor, of shape [batches, height, width, depth].
+     * * 0: A 4-D tensor, specifying the tensor to be normalized.
      *
      * Outputs:
-     * * 0: The output 4-D tensor, of the same shape as input
-     *      [batches, height, width, depth].
-     *
-     * Available since API level 27.
+     * * 0: A tensor of the same {@link OperandType} and same shape as input0.
      */
     L2_NORMALIZATION = 11,
 
@@ -652,7 +666,8 @@
      * Supported tensor {@link OperandType}:
      * * {@link OperandType::TENSOR_FLOAT32}
      *
-     * Supported tensor rank: 4, with "NHWC" data layout.
+     * Supported tensor rank: 4, with "NHWC" (i.e., Num_samples, Height, Width,
+     * and Channels) data layout.
      *
      * Both explicit padding and implicit padding are supported.
      *
@@ -700,8 +715,6 @@
      * Outputs:
      * * 0: The output 4-D tensor, of shape
      *      [batches, out_height, out_width, depth].
-     *
-     * Available since API level 27.
      */
     L2_POOL_2D = 12,
 
@@ -729,17 +742,18 @@
      *      the input.
      * * 1: An {@link OperandType::INT32} scalar, specifying the radius of
      *      the normalization window.
-     * * 2: An {@link OperandType::FLOAT32} scalar, specifying the bias, must
-     *      not be zero.
-     * * 3: An {@link OperandType::FLOAT32} scalar, specifying the scale
-     *      factor, alpha.
-     * * 4: An {@link OperandType::FLOAT32} scalar, specifying the exponent,
-     *      beta.
+     * * 2: A scalar, specifying the bias, must not be zero.
+     *      For input tensor of {@link OperandType::TENSOR_FLOAT32}, the bias
+     *      value must be of {@link OperandType::FLOAT32}.
+     * * 3: A scalar, specifying the scale factor, alpha.
+     *      For input tensor of {@link OperandType::TENSOR_FLOAT32}, the
+     *      alpha value must be of {@link OperandType::FLOAT32}.
+     * * 4: A scalar, specifying the exponent, beta.
+     *      For input tensor of {@link OperandType::TENSOR_FLOAT32}, the beta
+     *      value must be of {@link OperandType::FLOAT32}.
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
-     *
-     * Available since API level 27.
      */
     LOCAL_RESPONSE_NORMALIZATION = 13,
 
@@ -763,45 +777,53 @@
      * * 0: The output tensor of same shape as input0.
      *      For {@link OperandType::TENSOR_QUANT8_ASYMM},
      *      the scale must be 1.f / 256 and the zeroPoint must be 0.
-     *
-     * Available since API level 27.
      */
     LOGISTIC = 14,
 
     /**
      * Projects an input to a bit vector via locality senstive hashing.
      *
+     * Supported input tensor {@link OperandType}:
+     * * {@link OperandType::TENSOR_FLOAT32}
+     * * {@link OperandType::TENSOR_INT32}
+     * * {@link OperandType::TENSOR_QUANT8_ASYMM}
+     *
+     * Supported input tensor rank: from 1
+     *
      * Inputs:
      * * 0: Hash functions. Dim.size == 2, DataType: Float.
-     *            Tensor[0].Dim[0]: Number of hash functions.
-     *            Tensor[0].Dim[1]: Number of seeds per hash functions.
-     *            Tensor[0].Dim[1] <= 32 in sparse case.
+     *      Tensor[0].Dim[0]: Number of hash functions.
+     *      Tensor[0].Dim[1]: Number of projected output bits generated by each
+     *      hash function.
+     *      If the projection type is Sparse:
+     *      Tensor[0].Dim[1] + ceil(log2(Tensor[0].Dim[0])) <= 32
      *
      * * 1: Input. Dim.size >= 1, no restriction on DataType.
      * * 2: Weight. Optional. Dim.size == 1, DataType: Float.
-     *     If not set, each input element is considered to have the same weight
-     *     of 1.0.
-     *     Tensor[1].Dim[0] == Tensor[2].Dim[0]
+     *      If not set, each input element is considered to have the same weight
+     *      of 1.0.
+     *      Tensor[1].Dim[0] == Tensor[2].Dim[0]
      * * 3: Type:
-     *        Sparse: Value LSHProjectionType_SPARSE(=1).
+     *        Sparse:
+     *          Value LSHProjectionType_SPARSE(=1).
      *          Computed bit vector is considered to be sparse.
      *          Each output element is an int32 made up of multiple bits
      *          computed from hash functions.
      *
-     *        Dense: Value LSHProjectionType_DENSE(=2).
+     *        Dense:
+     *          Value LSHProjectionType_DENSE(=2).
      *          Computed bit vector is considered to be dense. Each output
      *          element represents a bit and can take the value of either
      *          0 or 1.
      *
      * Outputs:
-     * * 0: If the projection type is sparse:
-     *        Output.Dim == { Tensor[0].Dim[0] }
-     *        A tensor of int32 that represents hash signatures.
-     *      If the projection type is Dense:
-     *        Output.Dim == { Tensor[0].Dim[0] * Tensor[0].Dim[1] }
-     *        A flattened tensor that represents projected bit vectors.
+     * * 0: If the projection type is Sparse:
+     *      Output.Dim == { Tensor[0].Dim[0] }
+     *      A tensor of int32 that represents hash signatures.
      *
-     * Available since API level 27.
+     *      If the projection type is Dense:
+     *      Output.Dim == { Tensor[0].Dim[0] * Tensor[0].Dim[1] }
+     *      A flattened tensor that represents projected bit vectors.
      */
     LSH_PROJECTION = 15,
 
@@ -901,71 +923,54 @@
      * Supported tensor {@link OperandType}:
      * * {@link OperandType::TENSOR_FLOAT32}
      *
+     * All input and output tensors must be of the same type.
+     *
      * Inputs:
      * * 0: The input (\f$x_t\f$).
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [batch_size, input_size], where “batch_size” corresponds to the
-     *      batching dimension, and “input_size” is the size of the input.
+     *      A 2-D tensor of shape [batch_size, input_size], where “batch_size”
+     *      corresponds to the batching dimension, and “input_size” is the size
+     *      of the input.
      * * 1: The input-to-input weights (\f$W_{xi}\f$). Optional.
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units, input_size], where “num_units” corresponds to the
-     *      number of cell units.
+     *      A 2-D tensor of shape [num_units, input_size], where “num_units”
+     *      corresponds to the number of cell units.
      * * 2: The input-to-forget weights (\f$W_{xf}\f$).
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units, input_size].
+     *      A 2-D tensor of shape [num_units, input_size].
      * * 3: The input-to-cell weights (\f$W_{xc}\f$).
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units, input_size].
+     *      A 2-D tensor of shape [num_units, input_size].
      * * 4: The input-to-output weights (\f$W_{xo}\f$).
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units, input_size].
+     *      A 2-D tensor of shape [num_units, input_size].
      * * 5: The recurrent-to-input weights (\f$W_{hi}\f$). Optional.
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units, output_size], where “output_size” corresponds to either
-     *      the number of cell units (i.e., “num_units”), or the second
-     *      dimension of the “projection_weights”, if defined.
+     *      A 2-D tensor of shape [num_units, output_size], where “output_size”
+     *      corresponds to either the number of cell units (i.e., “num_units”),
+     *      or the second dimension of the “projection_weights”, if defined.
      * * 6: The recurrent-to-forget weights (\f$W_{hf}\f$).
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units, output_size].
+     *      A 2-D tensor of shape [num_units, output_size].
      * * 7: The recurrent-to-cell weights (\f$W_{hc}\f$).
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units, output_size].
+     *      A 2-D tensor of shape [num_units, output_size].
      * * 8: The recurrent-to-output weights (\f$W_{ho}\f$).
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units, output_size].
+     *      A 2-D tensor of shape [num_units, output_size].
      * * 9: The cell-to-input weights (\f$W_{ci}\f$). Optional.
-     *      A 1-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units].
+     *      A 1-D tensor of shape [num_units].
      * * 10:The cell-to-forget weights (\f$W_{cf}\f$). Optional.
-     *      A 1-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units].
+     *      A 1-D tensor of shape [num_units].
      * * 11:The cell-to-output weights (\f$W_{co}\f$). Optional.
-     *      A 1-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units].
+     *      A 1-D tensor of shape [num_units].
      * * 12:The input gate bias (\f$b_i\f$). Optional.
-     *      A 1-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units].
+     *      A 1-D tensor of shape [num_units].
      * * 13:The forget gate bias (\f$b_f\f$).
-     *      A 1-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units].
+     *      A 1-D tensor of shape [num_units].
      * * 14:The cell bias (\f$b_c\f$).
-     *      A 1-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units].
+     *      A 1-D tensor of shape [num_units].
      * * 15:The output gate bias (\f$b_o\f$).
-     *      A 1-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units].
+     *      A 1-D tensor of shape [num_units].
      * * 16:The projection weights (\f$W_{proj}\f$). Optional.
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [output_size, num_units].
+     *      A 2-D tensor of shape [output_size, num_units].
      * * 17:The projection bias (\f$b_{proj}\f$). Optional.
-     *      A 1-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [output_size].
+     *      A 1-D tensor of shape [output_size].
      * * 18:The output state (in) (\f$h_{t-1}\f$).
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [batch_size, output_size].
+     *      A 2-D tensor of shape [batch_size, output_size].
      * * 19:The cell state (in) (\f$C_{t-1}\f$).
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [batch_size, num_units].
+     *      A 2-D tensor of shape [batch_size, num_units].
      * * 20:The activation function (\f$g\f$).
      *      A value indicating the activation function:
      *      <ul>
@@ -984,21 +989,15 @@
      *
      * Outputs:
      * * 0: The scratch buffer.
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [batch_size, num_units * 3] with CIFG, or
+     *      A 2-D tensor of shape [batch_size, num_units * 3] with CIFG, or
      *      [batch_size, num_units * 4] without CIFG.
      * * 1: The output state (out) (\f$h_t\f$).
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [batch_size, output_size].
+     *      A 2-D tensor of shape [batch_size, output_size].
      * * 2: The cell state (out) (\f$C_t\f$).
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [batch_size, num_units].
+     *      A 2-D tensor of shape [batch_size, num_units].
      * * 3: The output (\f$o_t\f$).
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [batch_size, output_size]. This is effectively the same as the
-     *      current “output state (out)” value.
-     *
-     * Available since API level 27.
+     *      A 2-D tensor of shape [batch_size, output_size]. This is effectively
+     *      the same as the current “output state (out)” value.
      */
     LSTM = 16,
 
@@ -1019,7 +1018,8 @@
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
-     * Supported tensor rank: 4, with "NHWC" data layout.
+     * Supported tensor rank: 4, with "NHWC" (i.e., Num_samples, Height, Width,
+     * and Channels) data layout.
      *
      * Both explicit padding and implicit padding are supported.
      *
@@ -1067,8 +1067,8 @@
      * Outputs:
      * * 0: The output 4-D tensor, of shape
      *      [batches, out_height, out_width, depth].
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     MAX_POOL_2D = 17,
 
@@ -1106,8 +1106,6 @@
      *      For output tensor of {@link OperandType::TENSOR_QUANT8_ASYMM},
      *      the following condition must be satisfied:
      *      output_scale > input1_scale * input2_scale.
-     *
-     * Available since API level 27.
      */
     MUL = 18,
 
@@ -1129,8 +1127,8 @@
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     RELU = 19,
 
@@ -1151,9 +1149,9 @@
      * * 0: A tensor, specifying the input.
      *
      * Outputs:
-     * * 0: The output tensor of same shape as input0.
-     *
-     * Available since API level 27.
+     * * 0: The output tensor of the same shape as input0.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     RELU1 = 20,
 
@@ -1175,8 +1173,8 @@
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     RELU6 = 21,
 
@@ -1205,8 +1203,8 @@
      *
      * Outputs:
      * * 0: The output tensor, of shape specified by the input shape.
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     RESHAPE = 22,
 
@@ -1220,9 +1218,10 @@
      * Supported tensor {@link OperandType}:
      * * {@link OperandType::TENSOR_FLOAT32}
      *
-     * Supported tensor rank: 4, with "NHWC" data layout.
+     * Supported tensor rank: 4, with "NHWC" (i.e., Num_samples, Height, Width,
+     * and Channels) data layout.
      *
-     * Inputs:
+     * Inputs (resizing by shape):
      * * 0: A 4-D tensor, of shape [batches, height, width, depth], specifying
      *      the input.
      * * 1: An {@link OperandType::INT32} scalar, specifying the output
@@ -1233,8 +1232,6 @@
      * Outputs:
      * * 0: The output 4-D tensor, of shape
      *      [batches, new_height, new_width, depth].
-     *
-     * Available since API level 27.
      */
     RESIZE_BILINEAR = 23,
 
@@ -1257,25 +1254,23 @@
      * Supported tensor {@link OperandType}:
      * * {@link OperandType::TENSOR_FLOAT32}
      *
+     * The input tensors must all be the same type.
+     *
      * Inputs:
      * * 0: input.
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32} of shape
-     *      [batch_size, input_size], where “batch_size” corresponds to the
-     *      batching dimension, and “input_size” is the size of the input.
+     *      A 2-D tensor of shape [batch_size, input_size], where “batch_size”
+     *      corresponds to the batching dimension, and “input_size” is the size
+     *      of the input.
      * * 1: weights.
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units, input_size], where “num_units” corresponds to the
-     *      number of units.
+     *      A 2-D tensor of shape [num_units, input_size], where “num_units”
+     *      corresponds to the number of units.
      * * 2: recurrent_weights.
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units, num_units], with columns corresponding to the weights
-     *      from each unit.
+     *      A 2-D tensor of shape [num_units, num_units], with columns
+     *      corresponding to the weights from each unit.
      * * 3: bias.
-     *      A 1-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units].
+     *      A 1-D tensor of shape [num_units].
      * * 4: hidden state (in).
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [batch_size, num_units].
+     *      A 2-D tensor of shape [batch_size, num_units].
      * * 5: fused_activation_function.
      *      An optional {@link FusedActivationFunc} value indicating the
      *      activation function. If “NONE” is specified then it results in a
@@ -1283,15 +1278,11 @@
      *
      * Outputs:
      * * 0: hidden state (out).
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [batch_size, num_units].
+     *      A 2-D tensor of shape [batch_size, num_units].
      *
      * * 1: output.
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [batch_size, num_units]. This is effectively the same as the
-     *      current state value.
-     *
-     * Available since API level 27.
+     *      A 2-D tensor of shape [batch_size, num_units]. This is effectively
+     *      the same as the current state value.
      */
     RNN = 24,
 
@@ -1306,6 +1297,9 @@
      *         exp((input[batch, i] - max(input[batch, :])) * beta) /
      *         sum_{k}{exp((input[batch, k] - max(input[batch, :])) * beta)}
      *
+     * For input tensor with rank other than 2, the activation will be applied
+     * independently on each 1-D slice along specified dimension.
+     *
      * Supported tensor {@link OperandType}:
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
@@ -1314,15 +1308,15 @@
      *
      * Inputs:
      * * 0: A 2-D or 4-D tensor, specifying the tensor to be reshaped.
-     * * 1: An {@link OperandType::FLOAT32} scalar, specifying the positive
-     *      scaling factor for the exponent, beta.
+     * * 1: A scalar, specifying the positive scaling factor for the exponent,
+     *      beta. If input0 is of {@link OperandType::TENSOR_FLOAT32} or
+     *      {@link OperandType::TENSOR_QUANT8_ASYMM}, the scalar must be of
+     *      {@link OperandType::FLOAT32}.
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
      *      For {@link OperandType::TENSOR_QUANT8_ASYMM},
      *      the scale must be 1.f / 256 and the zeroPoint must be 0.
-     *
-     * Available since API level 27.
      */
     SOFTMAX = 25,
 
@@ -1344,7 +1338,8 @@
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
-     * Supported tensor rank: 4, with "NHWC" data layout.
+     * Supported tensor rank: 4, with "NHWC" (i.e., Num_samples, Height, Width,
+     * and Channels) data layout.
      *
      * Inputs:
      * * 0: A 4-D tensor, of shape [batches, height, width, depth_in],
@@ -1356,8 +1351,8 @@
      * Outputs:
      * * 0: The output 4-D tensor, of shape [batches, height/block_size,
      *      width/block_size, depth_in*block_size*block_size].
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     SPACE_TO_DEPTH = 26,
 
@@ -1403,25 +1398,23 @@
      * Supported tensor {@link OperandType}:
      * * {@link OperandType::TENSOR_FLOAT32}
      *
+     * All input tensors must be the same type.
+     *
      * Inputs:
      * * 0: input.
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [batch_size, input_size], where “batch_size” corresponds to the
-     *      batching dimension, and “input_size” is the size of the input.
+     *      A 2-D tensor of shape [batch_size, input_size], where “batch_size”
+     *      corresponds to the batching dimension, and “input_size” is the size
+     *      of the input.
      * * 1: weights_feature.
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units, input_size], where “num_units” corresponds to the
-     *      number of units.
+     *      A 2-D tensor of shape [num_units, input_size], where “num_units”
+     *      corresponds to the number of units.
      * * 2: weights_time.
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [num_units, memory_size], where “memory_size” corresponds to the
-     *      fixed-size of the memory.
+     *      A 2-D tensor of shape [num_units, memory_size], where “memory_size”
+     *      corresponds to the fixed-size of the memory.
      * * 3: bias.
-     *      An optional 1-D tensor of {@link OperandType::TENSOR_FLOAT32},
-     *      of shape [num_units].
+     *      An optional 1-D tensor of shape [num_units].
      * * 4: state (in).
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
-     *      [batch_size, (memory_size - 1) * num_units * rank].
+     *      A 2-D tensor of shape [batch_size, (memory_size - 1) * num_units * rank].
      * * 5: rank.
      *      The rank of the SVD approximation.
      * * 6: fused_activation_function.
@@ -1431,13 +1424,11 @@
      *
      * Outputs:
      * * 0: state (out).
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
+     *      A 2-D tensor of the same {@link OperandType} as the inputs, with shape
      *      [batch_size, (memory_size - 1) * num_units * rank].
      * * 1: output.
-     *      A 2-D tensor of {@link OperandType::TENSOR_FLOAT32}, of shape
+     *      A 2-D tensor of the same {@link OperandType} as the inputs, with shape
      *      [batch_size, num_units].
-     *
-     * Available since API level 27.
      */
     SVDF = 27,
 
@@ -1458,8 +1449,6 @@
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
-     *
-     * Available since API level 27.
      */
     TANH = 28,
 
diff --git a/neuralnetworks/1.0/types.t b/neuralnetworks/1.0/types.t
new file mode 100644
index 0000000..d7b26aa
--- /dev/null
+++ b/neuralnetworks/1.0/types.t
@@ -0,0 +1,431 @@
+%% template file for generating types.hal.
+%% see frameworks/ml/nn/tools/api/README.md.
+/*
+ * Copyright (C) 2017 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package android.hardware.neuralnetworks@1.0;
+
+%insert Operand_1.0_Comment
+enum OperandType : int32_t {
+%insert Operand_1.0
+
+    /**
+     * DEPRECATED. Since HAL version 1.2, extensions are the preferred
+     * alternative to OEM operation and data types.
+     *
+     * OEM specific scalar value.
+     */
+    OEM                 = 10000,
+
+    /**
+     * DEPRECATED. Since HAL version 1.2, extensions are the preferred
+     * alternative to OEM operation and data types.
+     *
+     * A tensor of OEM specific values.
+     */
+    TENSOR_OEM_BYTE     = 10001,
+};
+
+%insert Operation_1.0_Comment
+enum OperationType : int32_t {
+%insert Operation_1.0
+
+    /**
+     * DEPRECATED. Since NNAPI 1.2, extensions are the preferred alternative to
+     * OEM operation and data types.
+     *
+     * This operation is OEM specific. It should only be used for OEM
+     * applications.
+     */
+    OEM_OPERATION = 10000,
+};
+
+/**
+ * Fused activation function types.
+ */
+enum FusedActivationFunc : int32_t {
+    NONE  = 0,
+    RELU  = 1,
+    RELU1 = 2,
+    RELU6 = 3,
+};
+
+/**
+ * How an operand is used.
+ */
+enum OperandLifeTime : int32_t {
+    /**
+     * The operand is internal to the model. It's created by an operation and
+     * consumed by other operations. It must be an output operand of
+     * exactly one operation.
+     */
+    TEMPORARY_VARIABLE,
+
+    /**
+     * The operand is an input of the model. It must not be an output
+     * operand of any operation.
+     *
+     * An operand can't be both input and output of a model.
+     */
+    MODEL_INPUT,
+
+    /**
+     * The operand is an output of the model. It must be an output
+     * operand of exactly one operation.
+     *
+     * An operand can't be both input and output of a model.
+     */
+    MODEL_OUTPUT,
+
+    /**
+     * The operand is a constant found in Model.operandValues. It must
+     * not be an output operand of any operation.
+     */
+    CONSTANT_COPY,
+
+    /**
+     * The operand is a constant that was specified via a Memory
+     * object. It must not be an output operand of any operation.
+     */
+    CONSTANT_REFERENCE,
+
+    /**
+     * The operand does not have a value. This is valid only for optional
+     * arguments of operations.
+     */
+    NO_VALUE,
+};
+
+/**
+ * Status of a device.
+ */
+enum DeviceStatus : int32_t {
+    AVAILABLE,
+    BUSY,
+    OFFLINE,
+    UNKNOWN,
+};
+
+/**
+ * Performance information for the reference workload.
+ *
+ * Used by a driver to report its performance characteristics.
+ */
+struct PerformanceInfo {
+    /**
+     * Ratio of the time taken by the driver to execute the
+     * workload compared to the time the CPU would take for the
+     * same workload. A lower number is better.
+     */
+    float execTime;
+
+    /**
+     * Ratio of the energy used by the driver compared to what
+     * the CPU would use for doing the same workload. A lower number
+     * is better.
+     */
+    float powerUsage;
+};
+
+/**
+ * The capabilities of a driver.
+ */
+struct Capabilities {
+    /**
+     * Driver performance when operating on float32 data.
+     */
+    PerformanceInfo float32Performance;
+
+    /**
+     * Driver performance when operating on asymmetric 8-bit quantized data.
+     */
+    PerformanceInfo quantized8Performance;
+};
+
+/**
+ * Describes the location of a data object.
+ */
+struct DataLocation {
+    /**
+     * The index of the memory pool where this location is found.
+     */
+    uint32_t poolIndex;
+
+    /**
+     * Offset in bytes from the start of the pool.
+     */
+    uint32_t offset;
+
+    /**
+     * The length of the data in bytes.
+     */
+    uint32_t length;
+};
+
+/**
+ * Describes one operand of the model's graph.
+ */
+struct Operand {
+    /**
+     * Data type of the operand.
+     */
+    OperandType type;
+
+    /**
+     * Dimensions of the operand.
+     *
+     * For a scalar operand, dimensions.size() must be 0.
+     *
+     * For a tensor operand, dimensions.size() must be at least 1;
+     * however, any of the dimensions may be unspecified.
+     *
+     * A tensor operand with all dimensions specified has "fully
+     * specified" dimensions. Whenever possible (i.e., whenever the
+     * dimensions are known at model construction time), a tensor
+     * operand should have (but is not required to have) fully
+     * specified dimensions, in order to enable the best possible
+     * performance.
+     *
+     * If a tensor operand's dimensions are not fully specified, the
+     * dimensions of the operand are deduced from the operand
+     * dimensions and values of the operation for which that operand
+     * is an output.
+     *
+     * In the following situations, a tensor operand's dimensions must
+     * be fully specified:
+     *
+     *     . The operand has lifetime CONSTANT_COPY or
+     *       CONSTANT_REFERENCE.
+     *
+     *     . The operand has lifetime MODEL_INPUT or MODEL_OUTPUT. Fully
+     *       specified dimensions must either be present in the
+     *       Operand or they must be provided in the corresponding
+     *       RequestArgument.
+     *       EXCEPTION: If the input or output is optional and omitted
+     *       (by setting the hasNoValue field of the corresponding
+     *       RequestArgument to true) then it need not have fully
+     *       specified dimensions.
+     *
+     * A tensor operand with some number of unspecified dimensions is
+     * represented by setting each unspecified dimension to 0.
+     */
+    vec<uint32_t> dimensions;
+
+    /**
+     * The number of times this operand appears as an operation input.
+     *
+     * (For example, if this operand appears once in one operation's
+     * input list, and three times in another operation's input list,
+     * then numberOfConsumers = 4.)
+     */
+    uint32_t numberOfConsumers;
+
+    /**
+     * Quantized scale of the operand.
+     *
+     * Only applicable if the operand is of type TENSOR_QUANT8_ASYMM or
+     * TENSOR_INT32.
+     */
+    float scale;
+
+    /**
+     * Quantized zero-point offset of the operand.
+     *
+     * Only applicable if the operand is of type TENSOR_QUANT8_ASYMM.
+     */
+    int32_t zeroPoint;
+
+    /**
+     * How the operand is used.
+     */
+    OperandLifeTime lifetime;
+
+    /**
+     * Where to find the data for this operand.
+     * If the lifetime is TEMPORARY_VARIABLE, MODEL_INPUT, MODEL_OUTPUT, or
+     * NO_VALUE:
+     * - All the fields must be 0.
+     * If the lifetime is CONSTANT_COPY:
+     * - location.poolIndex is 0.
+     * - location.offset is the offset in bytes into Model.operandValues.
+     * - location.length is set.
+     * If the lifetime is CONSTANT_REFERENCE:
+     * - location.poolIndex is set.
+     * - location.offset is the offset in bytes into the specified pool.
+     * - location.length is set.
+     */
+    DataLocation location;
+};
+
+/**
+ * Describes one operation of the model's graph.
+ */
+struct Operation {
+    /**
+     * The operation type.
+     */
+    OperationType type;
+
+    /**
+     * Describes the table that contains the indexes of the inputs of the
+     * operation. The offset is the index in the operandIndexes table.
+     */
+    vec<uint32_t> inputs;
+
+    /**
+     * Describes the table that contains the indexes of the outputs of the
+     * operation. The offset is the index in the operandIndexes table.
+     */
+    vec<uint32_t> outputs;
+};
+
+/**
+ * A Neural Network Model.
+ *
+ * This includes not only the execution graph, but also constant data such as
+ * weights or scalars added at construction time. The only information that
+ * might not be known is the shape of the input tensors.
+ */
+struct Model {
+    /**
+     * All operands included in the model.
+     */
+    vec<Operand> operands;
+
+    /**
+     * All operations included in the model.
+     *
+     * The operations are sorted into execution order. Every operand
+     * with lifetime MODEL_OUTPUT or TEMPORARY_VARIABLE must be
+     * written before it is read.
+     */
+    vec<Operation> operations;
+
+    /**
+     * Input indexes of the model. There must be at least one.
+     *
+     * Each value corresponds to the index of the operand in "operands".
+     */
+    vec<uint32_t> inputIndexes;
+
+    /**
+     * Output indexes of the model. There must be at least one.
+     *
+     * Each value corresponds to the index of the operand in "operands".
+     */
+    vec<uint32_t> outputIndexes;
+
+    /**
+     * A byte buffer containing operand data that were copied into the model.
+     *
+     * An operand's value must be located here if and only if Operand::lifetime
+     * equals OperandLifeTime::CONSTANT_COPY.
+     */
+    vec<uint8_t> operandValues;
+
+    /**
+     * A collection of shared memory pools containing operand values.
+     *
+     * An operand's value must be located here if and only if Operand::lifetime
+     * equals OperandLifeTime::CONSTANT_REFERENCE.
+     */
+    vec<memory> pools;
+};
+
+/**
+ * Metadata information specifying the location of the input or output data and
+ * any updates to the input or output operand.
+ */
+struct RequestArgument {
+    /**
+     * If true, the argument does not have a value. This can be used for
+     * operations that take optional arguments. If true, the fields of location
+     * are set to 0 and the dimensions vector is left empty.
+     */
+    bool hasNoValue;
+
+    /**
+     * The location within one of the memory pools passed in the Request.
+     */
+    DataLocation location;
+
+    /**
+     * Updated dimension information.
+     *
+     * If dimensions.size() > 0, dimension information was provided
+     * along with the argument. This can be the case for models that
+     * accept inputs of varying size. This can't change the rank, just
+     * the value of the dimensions that were unspecified in the
+     * model. If dimensions.size() > 0, then all dimensions must be
+     * specified here; and any dimension that was specified in the
+     * model must have the same value here.
+     *
+     * If the dimensions in the model are not fully specified, then
+     * they must be fully specified here, unless hasNoValue is set to
+     * true. If the dimensions in the model are fully specified, then
+     * either dimensions.size() may be 0, or the dimensions in the
+     * model must be identical to the dimensions here.
+     */
+    vec<uint32_t> dimensions;
+};
+
+/**
+ * Inputs to be sent to and outputs to be retrieved from a prepared model.
+ *
+ * A Request serves two primary tasks:
+ * 1) Provides the input and output data to be used when executing the model.
+ * 2) Specifies any updates to the input operand metadata that were left
+ *    unspecified at model preparation time.
+ *
+ * An output must not overlap with any other output, with an input, or
+ * with an operand of lifetime CONSTANT_REFERENCE.
+ */
+struct Request {
+    /**
+     * Input data and information to be used in the execution of a prepared
+     * model.
+     *
+     * The index of the input corresponds to the index in Model.inputIndexes.
+     *   E.g., input[i] corresponds to Model.inputIndexes[i].
+     */
+    vec<RequestArgument> inputs;
+
+    /**
+     * Output data and information to be used in the execution of a prepared
+     * model.
+     *
+     * The index of the output corresponds to the index in Model.outputIndexes.
+     *   E.g., output[i] corresponds to Model.outputIndexes[i].
+     */
+    vec<RequestArgument> outputs;
+
+    /**
+     * A collection of shared memory pools containing operand data for both the
+     * inputs and the outputs to a model.
+     */
+    vec<memory> pools;
+};
+
+/**
+ * Return status of a function.
+ */
+enum ErrorStatus : int32_t {
+    NONE,
+    DEVICE_UNAVAILABLE,
+    GENERAL_FAILURE,
+    OUTPUT_INSUFFICIENT_SIZE,
+    INVALID_ARGUMENT,
+};
diff --git a/neuralnetworks/1.1/types.hal b/neuralnetworks/1.1/types.hal
index 73705bb..3d78fb6 100644
--- a/neuralnetworks/1.1/types.hal
+++ b/neuralnetworks/1.1/types.hal
@@ -26,7 +26,6 @@
  * The type of an operation in a model.
  */
 enum OperationType : @1.0::OperationType {
-
     /**
      * BatchToSpace for N-dimensional tensors.
      *
@@ -41,7 +40,8 @@
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
-     * Supported tensor rank: 4
+     * Supported tensor rank: 4, with "NHWC" (i.e., Num_samples, Height, Width,
+     * and Channels) data layout.
      *
      * Inputs:
      * * 0: An n-D tensor, specifying the tensor to be reshaped
@@ -51,8 +51,8 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 28.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     BATCH_TO_SPACE_ND = 29,
 
@@ -91,8 +91,6 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 28.
      */
     DIV = 30,
 
@@ -126,8 +124,8 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 28.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be same as input0.
      */
     MEAN = 31,
 
@@ -138,7 +136,8 @@
      *
      * Supported tensor {@link OperandType}:
      * * {@link OperandType::TENSOR_FLOAT32}
-     * * {@link OperandType::TENSOR_QUANT8_ASYMM} (the pad value is undefined)
+     * * {@link OperandType::TENSOR_QUANT8_ASYMM}
+     *   (the pad value is undefined)
      *
      * Supported tensor rank: up to 4
      *
@@ -160,11 +159,8 @@
      *      of the padding:
      *          output0.dimension[i] =
      *              padding[i, 0] + input0.dimension[i] + padding[i, 1]
-     *
-     *      NOTE: The pad value for {@link ANEURALNETWORKS_TENSOR_QUANT8_ASYMM}
-     *      is undefined.
-     *
-     * Available since API level 28.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     PAD = 32,
 
@@ -182,8 +178,10 @@
      * Supported tensor {@link OperandType}:
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
+     *   (the pad value is undefined)
      *
-     * Supported tensor rank: 4
+     * Supported tensor rank: 4, with "NHWC" (i.e., Num_samples, Height, Width,
+     * and Channels) data layout.
      *
      * Inputs:
      * * 0: An n-D tensor, specifying the input.
@@ -201,8 +199,8 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 28.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     SPACE_TO_BATCH_ND = 33,
 
@@ -232,8 +230,8 @@
      * * 0: A tensor of the same {@link OperandType} as input0. Contains the
      *      same data as input, but has one or more dimensions of size 1
      *      removed.
-     *
-     * Available since API level 28.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     SQUEEZE = 34,
 
@@ -278,8 +276,8 @@
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0 and rank (n - k),
      *      where k is the number of bits set in shrink_axis_mask.
-     *
-     * Available since API level 28.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     STRIDED_SLICE = 35,
 
@@ -318,8 +316,6 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 28.
      */
     SUB = 36,
 
@@ -345,11 +341,10 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 28.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     TRANSPOSE = 37,
-
 };
 
 /**
diff --git a/neuralnetworks/1.1/types.t b/neuralnetworks/1.1/types.t
new file mode 100644
index 0000000..75ac2e7
--- /dev/null
+++ b/neuralnetworks/1.1/types.t
@@ -0,0 +1,158 @@
+%% template file for generating types.hal.
+%% see frameworks/ml/nn/tools/api/README.md.
+/*
+ * Copyright (C) 2018 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package android.hardware.neuralnetworks@1.1;
+
+import @1.0::Operand;
+import @1.0::OperationType;
+import @1.0::PerformanceInfo;
+
+/**
+ * Operation types.
+ *
+ * The type of an operation in a model.
+ */
+enum OperationType : @1.0::OperationType {
+%insert Operation_1.1
+};
+
+/**
+ * The capabilities of a driver.
+ */
+struct Capabilities {
+    /**
+     * Driver performance when operating on float32 data.
+     */
+    PerformanceInfo float32Performance;
+
+    /**
+     * Driver performance when operating on asymmetric 8-bit quantized data.
+     */
+    PerformanceInfo quantized8Performance;
+
+    /**
+     * Driver performance when operating on float32 data but performing
+     * calculations with range and/or precision as low as that of the IEEE
+     * 754 16-bit floating-point format.
+     */
+    PerformanceInfo relaxedFloat32toFloat16Performance;
+};
+
+/**
+ * Describes one operation of the model's graph.
+ */
+struct Operation {
+    /**
+     * The operation type.
+     */
+    OperationType type;
+
+    /**
+     * Describes the table that contains the indexes of the inputs of the
+     * operation. The offset is the index in the operandIndexes table.
+     */
+    vec<uint32_t> inputs;
+
+    /**
+     * Describes the table that contains the indexes of the outputs of the
+     * operation. The offset is the index in the operandIndexes table.
+     */
+    vec<uint32_t> outputs;
+};
+
+/**
+ * A Neural Network Model.
+ *
+ * This includes not only the execution graph, but also constant data such as
+ * weights or scalars added at construction time. The only information that
+ * may not be known is the shape of the input tensors.
+ */
+struct Model {
+    /**
+     * All operands included in the model.
+     */
+    vec<Operand> operands;
+
+    /**
+     * All operations included in the model.
+     *
+     * The operations are sorted into execution order. Every operand
+     * with lifetime MODEL_OUTPUT or TEMPORARY_VARIABLE must be
+     * written before it is read.
+     */
+    vec<Operation> operations;
+
+    /**
+     * Input indexes of the model. There must be at least one.
+     *
+     * Each value corresponds to the index of the operand in "operands".
+     */
+    vec<uint32_t> inputIndexes;
+
+    /**
+     * Output indexes of the model. There must be at least one.
+     *
+     * Each value corresponds to the index of the operand in "operands".
+     */
+    vec<uint32_t> outputIndexes;
+
+    /**
+     * A byte buffer containing operand data that were copied into the model.
+     *
+     * An operand's value must be located here if and only if Operand::lifetime
+     * equals OperandLifeTime::CONSTANT_COPY.
+     */
+    vec<uint8_t> operandValues;
+
+    /**
+     * A collection of shared memory pools containing operand values.
+     *
+     * An operand's value must be located here if and only if Operand::lifetime
+     * equals OperandLifeTime::CONSTANT_REFERENCE.
+     */
+    vec<memory> pools;
+
+    /**
+     * 'true' indicates TENSOR_FLOAT32 may be calculated with range and/or
+     * precision as low as that of the IEEE 754 16-bit floating-point format.
+     * 'false' indicates TENSOR_FLOAT32 must be calculated using at least the
+     * range and precision of the IEEE 754 32-bit floating-point format.
+     */
+    bool relaxComputationFloat32toFloat16;
+};
+
+/**
+ * Execution preferences.
+ */
+enum ExecutionPreference : int32_t {
+    /**
+     * Prefer executing in a way that minimizes battery drain.
+     * This is desirable for compilations that will be executed often.
+     */
+    LOW_POWER = 0,
+    /**
+     * Prefer returning a single answer as fast as possible, even if this causes
+     * more power consumption.
+     */
+    FAST_SINGLE_ANSWER = 1,
+    /**
+     * Prefer maximizing the throughput of successive frames, for example when
+     * processing successive frames coming from the camera.
+     */
+    SUSTAINED_SPEED = 2,
+};
diff --git a/neuralnetworks/1.2/types.hal b/neuralnetworks/1.2/types.hal
index f368ce2..837ced5 100644
--- a/neuralnetworks/1.2/types.hal
+++ b/neuralnetworks/1.2/types.hal
@@ -43,8 +43,6 @@
      *
      * Values of this operand type are either true or false. A zero value
      * represents false; any other value represents true.
-     *
-     * Available since API level 29.
      */
     BOOL = 6,
     /**
@@ -55,14 +53,10 @@
      * realValue = integerValue * scale.
      *
      * scale is a 32 bit floating point with value greater than zero.
-     *
-     * Available since API level 29.
      */
     TENSOR_QUANT16_SYMM = 7,
     /**
      * A tensor of IEEE 754 16 bit floating point values.
-     *
-     * Available since API level 29.
      */
     TENSOR_FLOAT16 = 8,
     /**
@@ -70,14 +64,10 @@
      *
      * Values of this operand type are either true or false. A zero value
      * represents false; any other value represents true.
-     *
-     * Available since API level 29.
      */
     TENSOR_BOOL8 = 9,
     /**
      * An IEEE 754 16 bit floating point scalar value.
-     *
-     * Available since API level 29.
      */
     FLOAT16 = 10,
     /**
@@ -90,14 +80,13 @@
      * - scales: an array of positive 32 bit floating point values.
      * The size of the scales array must be equal to dimensions[channelDim].
      *
+     *{@link SymmPerChannelQuantParams} must hold the parameters for an Operand of this type.
      * The channel dimension of this tensor must not be unknown (dimensions[channelDim] != 0).
      *
      * The formula is:
      * realValue[..., C, ...] =
      *     integerValue[..., C, ...] * scales[C]
      * where C is an index in the Channel dimension.
-     *
-     * Available since API level 29.
      */
     TENSOR_QUANT8_SYMM_PER_CHANNEL = 11,
     /**
@@ -110,8 +99,6 @@
      *
      * The formula is:
      * real_value = (integer_value - zeroPoint) * scale.
-     *
-     * Available since API level 29.
      */
     TENSOR_QUANT16_ASYMM = 12,
     /**
@@ -122,20 +109,19 @@
      * realValue = integerValue * scale.
      *
      * scale is a 32 bit floating point with value greater than zero.
-     *
-     * Available since API level 29.
      */
     TENSOR_QUANT8_SYMM = 13,
+
     /*
-     * DEPRECATED. Since NNAPI 1.2, extensions are the preferred alternative to
-     * OEM operation and data types.
+     * DEPRECATED. Since HAL version 1.2, extensions are the preferred
+     * alternative to OEM operation and data types.
      *
      * OEM specific scalar value.
      * OEM                 = 10000,
      */
     /*
-     * DEPRECATED. Since NNAPI 1.2, extensions are the preferred alternative to
-     * OEM operation and data types.
+     * DEPRECATED. Since HAL version 1.2, extensions are the preferred
+     * alternative to OEM operation and data types.
      *
      * A tensor of OEM specific values.
      * TENSOR_OEM_BYTE     = 10001,
@@ -166,6 +152,7 @@
  * The type of an operation in a model.
  */
 enum OperationType : int32_t {
+
     /**
      * Adds two tensors, element-wise.
      *
@@ -187,12 +174,12 @@
      *     input2.dimension = {5, 4, 3, 1}
      *     output.dimension = {5, 4, 3, 2}
      *
-     * Since API level 29, generic zero-sized input tensor is supported. Zero
+     * Since HAL version 1.2, generic zero-sized input tensor is supported. Zero
      * dimension is only compatible with 0 or 1. The size of the output
      * dimension is zero if either of corresponding input dimension is zero.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
@@ -202,14 +189,16 @@
      * * 0: A tensor.
      * * 1: A tensor of the same {@link OperandType}, and compatible dimensions
      *      as input0.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scales and zeroPoint can be different from input0 scale and zeroPoint.
      * * 2: An {@link OperandType::INT32} scalar, and has to be one of the
      *      {@link FusedActivationFunc} values. Specifies the activation to
      *      invoke on the result.
      *
      * Outputs:
      * * 0: The sum, a tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint can be different from inputs' scale and zeroPoint.
      */
     ADD = @1.1::OperationType:ADD,
 
@@ -227,7 +216,7 @@
      *         ) / sum(1)
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
@@ -235,13 +224,14 @@
      * With the default data layout NHWC, the data is stored in the order of:
      * [batch, height, width, channels]. Alternatively, the data layout could
      * be NCHW, the data storage order of: [batch, channels, height, width].
+     * NCHW is supported since HAL version 1.2.
      *
      * Both explicit padding and implicit padding are supported.
      *
      * Inputs (explicit padding):
      * * 0: A 4-D tensor, of shape [batches, height, width, depth], specifying
-     *      the input. Since API level 29, zero batches is supported for this
-     *      tensor.
+     *      the input.
+     *      Since HAL version 1.2, zero batches is supported for this tensor.
      * * 1: An {@link OperandType::INT32} scalar, specifying the padding on
      *      the left, in the ‘width’ dimension.
      * * 2: An {@link OperandType::INT32} scalar, specifying the padding on
@@ -263,12 +253,12 @@
      *      invoke on the result.
      * * 10: An optional {@link OperandType::BOOL} scalar, default to false.
      *       Set to true to specify NCHW data layout for input0 and output0.
-     *       Available since API level 29.
+     *       Available since HAL version 1.2.
      *
      * Inputs (implicit padding):
      * * 0: A 4-D tensor, of shape [batches, height, width, depth], specifying
-     *      the input. Since API level 29, zero batches is supported for this
-     *      tensor.
+     *      the input.
+     *      Since HAL version 1.2, zero batches is supported for this tensor.
      * * 1: An {@link OperandType::INT32} scalar, specifying the implicit
      *      padding scheme, has to be one of the
      *      following values: {0 (NONE), 1 (SAME), 2 (VALID)}.
@@ -285,13 +275,13 @@
      *      invoke on the result.
      * * 7: An optional {@link OperandType::BOOL} scalar, default to false.
      *      Set to true to specify NCHW data layout for input0 and output0.
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      *
      * Outputs:
      * * 0: The output 4-D tensor, of shape
      *      [batches, out_height, out_width, depth].
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     AVERAGE_POOL_2D = @1.1::OperationType:AVERAGE_POOL_2D,
 
@@ -302,33 +292,34 @@
      * dimensions except the dimension along the concatenation axis.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
-     * * {@link OperandType::TENSOR_QUANT8_ASYMM} (full support since API
-     *   level 29, see the input section)
+     * * {@link OperandType::TENSOR_QUANT8_ASYMM}
+     *   (full support since HAL version 1.2, see the input section)
      *
      * Supported tensor rank: up to 4
      *
      * Inputs:
      * * 0 ~ n-1: The list of n input tensors, of shape
      *            [D0, D1, ..., Daxis(i), ..., Dm].
-     *            Before API level 29, all input tensors of
+     *            Before HAL version 1.2, all input tensors of
      *            {@link OperandType::TENSOR_QUANT8_ASYMM}
      *            must have the same scale and zeroPoint as the output tensor.
-     *            Since API level 29, zero-sized tensors are supported.
+     *            Since HAL version 1.2, zero-sized tensors are supported.
      * * n: An {@link OperandType::INT32} scalar, specifying the
      *      concatenation axis.
      *
      * Outputs:
      * * 0: The output, a tensor of the same {@link OperandType} as the input
      *      tensors. The output shape is [D0, D1, ..., sum(Daxis(i)), ..., Dm].
-     *
-     * Available since API level 27.
+     *      Since HAL version 1.2, for a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint values can be different from
+     *      input tensors. Before HAL version 1.2 they have to be the same as for the input tensors.
      */
     CONCATENATION = @1.1::OperationType:CONCATENATION,
 
     /**
-     * Performs an 2-D convolution operation.
+     * Performs a 2-D convolution operation.
      *
      * The CONV_2D op sweeps a 2-D filter that can mix channels together over a
      * batch of images, applying the filter to each window of each image of the
@@ -354,7 +345,7 @@
      * * * {@link OperandType::TENSOR_INT32} for bias (with scale set to
      * * * input.scale * filter.scale).
      *
-     * Available since API level 29:
+     * Available since HAL version 1.2:
      * * 16 bit floating point:
      * * * {@link OperandType::TENSOR_FLOAT16} for input, filter, output, and bias.
      *
@@ -368,27 +359,29 @@
      * With the default data layout NHWC, the data is stored in the order of:
      * [batch, height, width, channels]. Alternatively, the data layout could
      * be NCHW, the data storage order of: [batch, channels, height, width].
+     * NCHW is supported since HAL version 1.2.
      *
      * Both explicit padding and implicit padding are supported.
      *
      * Inputs (explicit padding):
      * * 0: A 4-D tensor, of shape [batches, height, width, depth_in],
-     *      specifying the input. Since API level 29, zero batches is supported
-     *      for this tensor.
+     *      specifying the input.
+     *      Since HAL version 1.2, zero batches is supported for this tensor.
      * * 1: A 4-D tensor, of shape
      *      [depth_out, filter_height, filter_width, depth_in], specifying the
-     *      filter. For tensor of type
-     *      {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL} the channel
-     *      dimension (extraParams.channelQuant.channelDim) must be set to 0.
+     *      filter.
+     *      For tensor of type {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL}
+     *      the channel dimension (SymmPerChannelQuantParams::channelDim)
+     *      must be set to 0.
      * * 2: A 1-D tensor, of shape [depth_out], specifying the bias. For input
-     *      tensor of type {@link OperandType::TENSOR_FLOAT32} or
-     *      {@link OperandType::TENSOR_FLOAT16}, the bias must be of the same
+     *      tensor of type {@link OperandType::TENSOR_FLOAT32}
+     *      or {@link OperandType::TENSOR_FLOAT16} the bias must be of the same
      *      type. For filter tensor of {@link OperandType::TENSOR_QUANT8_ASYMM},
      *      the bias should be of {@link OperandType::TENSOR_INT32}, with zeroPoint
-     *      of 0 and bias_scale == input_scale * filter_scale. For filter tensor
-     *      of {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL}, the bias
-     *      should be of {@link OperandType::TENSOR_INT32}, with zeroPoint of
-     *      0 and bias_scale of 0. The actual scale of each value 'i' is equal to
+     *      of 0 and bias_scale == input_scale * filter_scale.
+     *      For filter tensor of {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL},
+     *      the bias should be of {@link OperandType::TENSOR_INT32}, with zeroPoint of 0
+     *      and bias_scale of 0. The actual scale of each value 'i' is equal to
      *      bias_scale[i] = input_scale * filter_scale[i].
      * * 3: An {@link OperandType::INT32} scalar, specifying the padding on
      *      the left, in the ‘width’ dimension.
@@ -407,36 +400,37 @@
      *      invoke on the result.
      * * 10: An optional {@link OperandType::BOOL} scalar, default to false.
      *      Set to true to specify NCHW data layout for input0 and output0.
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      * * 11: An optional {@link OperandType::INT32} scalar, specifying the dilation
      *      factor for width. Defaults to 1. If set to k > 1, there will be k-1 skipped
      *      cells between each filter element on width dimension. If this input is set,
      *      input 12 (dilation factor for height) must be specified as well.
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      * * 12: An optional {@link OperandType::INT32} scalar, specifying the dilation
      *      factor for height. Defaults to 1. If set to k > 1, there will be k-1 skipped
      *      cells between each filter element on height dimension. If this input is set,
      *      input 11 (dilation factor for width) must be specified as well.
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      *
      * Inputs (implicit padding):
      * * 0: A 4-D tensor, of shape [batches, height, width, depth_in],
-     *      specifying the input. Since API level 29, zero batches is supported
-     *      for this tensor.
+     *      specifying the input.
+     *      Since HAL version 1.2, zero batches is supported for this tensor.
      * * 1: A 4-D tensor, of shape
      *      [depth_out, filter_height, filter_width, depth_in], specifying the
-     *      filter. For tensor of type
-     *      {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL} the channel
-     *      dimension (extraParams.channelQuant.channelDim) must be set to 0.
+     *      filter.
+     *      For tensor of type {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL}
+     *      the channel dimension (SymmPerChannelQuantParams::channelDim)
+     *      must be set to 0.
      * * 2: A 1-D tensor, of shape [depth_out], specifying the bias. For input
-     *      tensor of type {@link OperandType::TENSOR_FLOAT32} or
-     *      {@link OperandType::TENSOR_FLOAT16}, the bias must be of the same
+     *      tensor of type {@link OperandType::TENSOR_FLOAT32}
+     *      or {@link OperandType::TENSOR_FLOAT16} the bias must be of the same
      *      type. For filter tensor of {@link OperandType::TENSOR_QUANT8_ASYMM},
      *      the bias should be of {@link OperandType::TENSOR_INT32}, with zeroPoint
-     *      of 0 and bias_scale == input_scale * filter_scale. For filter tensor
-     *      of {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL}, the bias
-     *      should be of {@link OperandType::TENSOR_INT32}, with zeroPoint of
-     *      0 and bias_scale of 0. The actual scale of each value 'i' is equal to
+     *      of 0 and bias_scale == input_scale * filter_scale.
+     *      For filter tensor of {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL},
+     *      the bias should be of {@link OperandType::TENSOR_INT32}, with zeroPoint of 0
+     *      and bias_scale of 0. The actual scale of each value 'i' is equal to
      *      bias_scale[i] = input_scale * filter_scale[i].
      * * 3: An {@link OperandType::INT32} scalar, specifying the implicit
      *      padding scheme, has to be one of the
@@ -450,26 +444,23 @@
      *      invoke on the result.
      * * 7: An optional {@link OperandType::BOOL} scalar, default to false.
      *      Set to true to specify NCHW data layout for input0 and output0.
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      * * 8: An optional {@link OperandType::INT32} scalar, specifying the dilation
      *      factor for width. Defaults to 1. If set to k > 1, there will be k-1 skipped
      *      cells between each filter element on width dimension. If this input is set,
      *      input 9 (dilation factor for height) must be specified as well.
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      * * 9: An optional {@link OperandType::INT32} scalar, specifying the dilation
      *      factor for height. Defaults to 1. If set to k > 1, there will be k-1 skipped
      *      cells between each filter element on height dimension. If this input is set,
      *      input 8 (dilation factor for width) must be specified as well.
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      *
      * Outputs:
      * * 0: The output 4-D tensor, of shape
-     *      [batches, out_height, out_width, depth_out]. Before API level 29,
-     *      for output tensor of {@link OperandType::TENSOR_QUANT8_ASYMM}, the
-     *      following condition must be satisfied:
-     *      output_scale > input_scale * filter_scale
-     *
-     * Available since API level 27.
+     *      [batches, out_height, out_width, depth_out].
+     *      Before HAL version 1.2, for output tensor of {@link OperandType::TENSOR_QUANT8_ASYMM},
+     *      the following condition must be satisfied: output_scale > input_scale * filter_scale
      */
     CONV_2D = @1.1::OperationType:CONV_2D,
 
@@ -504,7 +495,7 @@
      * * * {@link OperandType::TENSOR_INT32} for bias (with scale set to
      * * * input.scale * filter.scale).
      *
-     * Available since API level 29:
+     * Available since HAL version 1.2:
      * * 16 bit floating point:
      * * * {@link OperandType::TENSOR_FLOAT16} for input, filter, output, and bias.
      *
@@ -518,6 +509,7 @@
      * With the default data layout NHWC, the data is stored in the order of:
      * [batch, height, width, channels]. Alternatively, the data layout could
      * be NCHW, the data storage order of: [batch, channels, height, width].
+     * NCHW is supported since HAL version 1.2.
      *
      * Both explicit padding and implicit padding are supported.
      *
@@ -525,18 +517,19 @@
      * * 0: A 4-D tensor, of shape [batches, height, width, depth_in],
      *      specifying the input.
      * * 1: A 4-D tensor, of shape [1, filter_height, filter_width, depth_out],
-     *      specifying the filter. For tensor of type
-     *      {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL} the channel
-     *      dimension (extraParams.channelQuant.channelDim) must be set to 3.
+     *      specifying the filter.
+     *      For tensor of type {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL}
+     *      the channel dimension (SymmPerChannelQuantParams::channelDim)
+     *      must be set to 3.
      * * 2: A 1-D tensor, of shape [depth_out], specifying the bias. For input
-     *      tensor of type {@link OperandType::TENSOR_FLOAT32} or
-     *      {@link OperandType::TENSOR_FLOAT16}, the bias must be of the same
+     *      tensor of type {@link OperandType::TENSOR_FLOAT32}
+     *      or {@link OperandType::TENSOR_FLOAT16} the bias must be of the same
      *      type. For filter tensor of {@link OperandType::TENSOR_QUANT8_ASYMM},
      *      the bias should be of {@link OperandType::TENSOR_INT32}, with zeroPoint
-     *      of 0 and bias_scale == input_scale * filter_scale. For filter tensor
-     *      of {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL}, the bias
-     *      should be of {@link OperandType::TENSOR_INT32}, with zeroPoint of
-     *      0 and bias_scale of 0. The actual scale of each value 'i' is equal to
+     *      of 0 and bias_scale == input_scale * filter_scale.
+     *      For filter tensor of {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL},
+     *      the bias should be of {@link OperandType::TENSOR_INT32}, with zeroPoint of 0
+     *      and bias_scale of 0. The actual scale of each value 'i' is equal to
      *      bias_scale[i] = input_scale * filter_scale[i].
      * * 3: An {@link OperandType::INT32} scalar, specifying the padding on
      *      the left, in the ‘width’ dimension.
@@ -557,17 +550,17 @@
      *       invoke on the result.
      * * 11: An optional {@link OperandType::BOOL} scalar, default to false.
      *       Set to true to specify NCHW data layout for input0 and output0.
-     *       Available since API level 29.
+     *       Available since HAL version 1.2.
      * * 12: An optional {@link OperandType::INT32} scalar, specifying the dilation
      *      factor for width. Defaults to 1. If set to k > 1, there will be k-1 skipped
      *      cells between each filter element on width dimension. If this input is set,
      *      input 13 (dilation factor for height) must be specified as well.
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      * * 13: An optional {@link OperandType::INT32} scalar, specifying the dilation
      *      factor for height. Defaults to 1. If set to k > 1, there will be k-1 skipped
      *      cells between each filter element on height dimension. If this input is set,
      *      input 12 (dilation factor for width) must be specified as well.
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      *
      * Inputs (implicit padding):
      * * 0: A 4-D tensor, of shape [batches, height, width, depth_in],
@@ -575,14 +568,14 @@
      * * 1: A 4-D tensor, of shape [1, filter_height, filter_width, depth_out],
      *      specifying the filter.
      * * 2: A 1-D tensor, of shape [depth_out], specifying the bias. For input
-     *      tensor of type {@link OperandType::TENSOR_FLOAT32} or
-     *      {@link OperandType::TENSOR_FLOAT16}, the bias must be of the same
+     *      tensor of type {@link OperandType::TENSOR_FLOAT32}
+     *      or {@link OperandType::TENSOR_FLOAT16} the bias must be of the same
      *      type. For filter tensor of {@link OperandType::TENSOR_QUANT8_ASYMM},
      *      the bias should be of {@link OperandType::TENSOR_INT32}, with zeroPoint
-     *      of 0 and bias_scale == input_scale * filter_scale. For filter tensor
-     *      of {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL}, the bias
-     *      should be of {@link OperandType::TENSOR_INT32}, with zeroPoint of
-     *      0 and bias_scale of 0. The actual scale of each value 'i' is equal to
+     *      of 0 and bias_scale == input_scale * filter_scale.
+     *      For filter tensor of {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL},
+     *      the bias should be of {@link OperandType::TENSOR_INT32}, with zeroPoint of 0
+     *      and bias_scale of 0. The actual scale of each value 'i' is equal to
      *      bias_scale[i] = input_scale * filter_scale[i].
      * * 3: An {@link OperandType::INT32} scalar, specifying the implicit
      *      padding scheme, has to be one of the
@@ -598,27 +591,24 @@
      *      invoke on the result.
      * * 8: An optional {@link OperandType::BOOL} scalar, default to false.
      *      Set to true to specify NCHW data layout for input0 and output0.
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      * * 9: An optional {@link OperandType::INT32} scalar, specifying the dilation
      *      factor for width. Defaults to 1. If set to k > 1, there will be k-1 skipped
      *      cells between each filter element on width dimension. If this input is set,
      *      input 10 (dilation factor for height) must be specified as well.
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      * * 10: An optional {@link OperandType::INT32} scalar, specifying the dilation
      *      factor for height. Defaults to 1. If set to k > 1, there will be k-1 skipped
      *      cells between each filter element on height dimension. If this input is set,
      *      input 9 (dilation factor for width) must be specified as well.
-     *      Available since API level 29.
-
+     *      Available since HAL version 1.2.
      *
      * Outputs:
      * * 0: The output 4-D tensor, of shape
-     *      [batches, out_height, out_width, depth_out]. Before API level 29,
-     *      for output tensor of {@link OperandType::TENSOR_QUANT8_ASYMM}, the
-     *      following condition must be satisfied:
+     *      [batches, out_height, out_width, depth_out]. Before HAL version 1.2, for
+     *      output tensor of {@link OperandType::TENSOR_QUANT8_ASYMM},
+     *      the following condition must be satisfied:
      *      output_scale > input_scale * filter_scale
-     *
-     * Available since API level 27.
      */
     DEPTHWISE_CONV_2D = @1.1::OperationType:DEPTHWISE_CONV_2D,
 
@@ -638,7 +628,7 @@
      * be divisible by block_size * block_size
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
@@ -646,6 +636,7 @@
      * With the default data layout NHWC, the data is stored in the order of:
      * [batch, height, width, channels]. Alternatively, the data layout could
      * be NCHW, the data storage order of: [batch, channels, height, width].
+     * NCHW is supported since HAL version 1.2.
      *
      * Inputs:
      * * 0: A 4-D tensor, of shape [batches, height, width, depth_in],
@@ -655,13 +646,13 @@
      *      of the input depth.
      * * 2: An optional {@link OperandType::BOOL} scalar, default to false.
      *      Set to true to specify NCHW data layout for input0 and output0.
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      *
      * Outputs:
      * * 0: The output 4-D tensor, of shape [batch, height*block_size,
      *      width*block_size, depth/(block_size*block_size)].
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     DEPTH_TO_SPACE = @1.1::OperationType:DEPTH_TO_SPACE,
 
@@ -674,22 +665,21 @@
      *
      * Supported input tensor {@link OperandType}:
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
-     * * {@link OperandType::TENSOR_QUANT8_SYMM} (since API level 29)
-     * * {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL} (since API level 29)
+     * * {@link OperandType::TENSOR_QUANT8_SYMM} (since HAL version 1.2)
+     * * {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL} (since HAL version 1.2)
      *
      * Supported output tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}.
      *
      * Supported tensor rank: up to 4
      *
      * Inputs:
-     * * 0: A tensor. Since API level 29, this tensor may be zero-sized.
+     * * 0: A tensor.
+     *      Since HAL version 1.2, this tensor may be zero-sized.
      *
      * Outputs:
      * * 0: A tensor with the same shape as input0.
-     *
-     * Available since API level 27.
      */
     DEQUANTIZE = @1.1::OperationType:DEQUANTIZE,
 
@@ -730,8 +720,8 @@
      * * 0: A n-D tensor with the same rank and shape as the Values
      *      tensor, except for the first dimension which has the same size
      *      as Lookups' only dimension.
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input1.
      */
     EMBEDDING_LOOKUP = @1.1::OperationType:EMBEDDING_LOOKUP,
 
@@ -739,7 +729,7 @@
      * Computes element-wise floor() on the input tensor.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      *
      * Supported tensor rank: up to 4
@@ -750,8 +740,6 @@
      * Outputs:
      * * 0: The output tensor, of the same {@link OperandType} and dimensions as
      *      the input tensor.
-     *
-     * Available since API level 27.
      */
     FLOOR = @1.1::OperationType:FLOOR,
 
@@ -764,7 +752,7 @@
      *     outputs = activation(inputs * weights’ + bias)
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
@@ -777,8 +765,8 @@
      *      [batch_size, input_size], where "input_size" corresponds to the
      *      number of inputs to the layer, matching the second dimension of
      *      weights, and "batch_size" is calculated by dividing the number of
-     *      elements by "input_size". Since API level 29, zero batch_size is
-     *      supported for this tensor.
+     *      elements by "input_size".
+     *      Since HAL version 1.2, zero batch_size is supported for this tensor.
      * * 1: A 2-D tensor, specifying the weights, of shape
      *      [num_units, input_size], where "num_units" corresponds to the number
      *      of output nodes.
@@ -793,12 +781,9 @@
      *      invoke on the result.
      *
      * Outputs:
-     * * 0: The output tensor, of shape [batch_size, num_units]. Before API
-     *      level 29, For output tensor of {@link
-     *      OperandType::TENSOR_QUANT8_ASYMM}, the following condition must be
-     *      satisfied: output_scale > input_scale * filter_scale.
-     *
-     * Available since API level 27.
+     * * 0: The output tensor, of shape [batch_size, num_units]. Before HAL version 1.2, for
+     *      output tensor of {@link OperandType::TENSOR_QUANT8_ASYMM}, the following
+     *      condition must be satisfied: output_scale > input_scale * filter_scale.
      */
     FULLY_CONNECTED = @1.1::OperationType:FULLY_CONNECTED,
 
@@ -849,13 +834,13 @@
      *
      * Outputs:
      * * 0: Output. A tensor with shape [ k …].
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input2.
      * * 1: Hits. A boolean tensor with shape [ k ] indicates whether the lookup
      *      hits (True) or not (False).
      *      Stored as {@link OperandType::TENSOR_QUANT8_ASYMM} with offset 0
      *      and scale 1.0f.
      *      A non-zero byte represents True, a hit. A zero indicates otherwise.
-     *
-     * Available since API level 27.
      */
     HASHTABLE_LOOKUP = @1.1::OperationType:HASHTABLE_LOOKUP,
 
@@ -872,12 +857,12 @@
      * 1-D slice along dimension dim.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
-     * * {@link OperandType::TENSOR_QUANT8_ASYMM} (since API level 29)
+     * * {@link OperandType::TENSOR_QUANT8_ASYMM} (since HAL version 1.2)
      *
      * Supported tensor rank: up to 4
-     * Tensors with rank less than 4 are only supported since API level 29.
+     * Tensors with rank less than 4 are only supported since HAL version 1.2.
      *
      * Inputs:
      * * 0: An n-D tensor, specifying the tensor to be normalized.
@@ -885,14 +870,12 @@
      *      specifying the dimension normalization would be performed on.
      *      Negative index is used to specify axis from the end (e.g. -1 for
      *      the last axis). Must be in the range [-n, n).
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} and same shape as input0.
      *      For {@link OperandType::TENSOR_QUANT8_ASYMM},
      *      the scale must be 1.f / 128 and the zeroPoint must be 128.
-     *
-     * Available since API level 27.
      */
     L2_NORMALIZATION = @1.1::OperationType:L2_NORMALIZATION,
 
@@ -909,20 +892,21 @@
      *              sum(1))
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      *
      * Supported tensor rank: 4, with "NHWC" or "NCHW" data layout.
      * With the default data layout NHWC, the data is stored in the order of:
      * [batch, height, width, channels]. Alternatively, the data layout could
      * be NCHW, the data storage order of: [batch, channels, height, width].
+     * NCHW is supported since HAL version 1.2.
      *
      * Both explicit padding and implicit padding are supported.
      *
      * Inputs (explicit padding):
      * * 0: A 4-D tensor, of shape [batches, height, width, depth], specifying
-     *      the input. Since API level 29, zero batches is supported for this
-     *      tensor.
+     *      the input.
+     *      Since HAL version 1.2, zero batches is supported for this tensor.
      * * 1: An {@link OperandType::INT32} scalar, specifying the padding on
      *      the left, in the ‘width’ dimension.
      * * 2: An {@link OperandType::INT32} scalar, specifying the padding on
@@ -944,12 +928,12 @@
      *      invoke on the result.
      * * 10: An optional {@link OperandType::BOOL} scalar, default to false.
      *       Set to true to specify NCHW data layout for input0 and output0.
-     *       Available since API level 29.
+     *       Available since HAL version 1.2.
      *
      * Inputs (implicit padding):
      * * 0: A 4-D tensor, of shape [batches, height, width, depth], specifying
-     *      the input. Since API level 29, zero batches is supported for this
-     *      tensor.
+     *      the input.
+     *      Since HAL version 1.2, zero batches is supported for this tensor.
      * * 1: An {@link OperandType::INT32} scalar, specifying the implicit
      *      padding scheme, has to be one of the
      *      following values: {0 (NONE), 1 (SAME), 2 (VALID)}.
@@ -966,13 +950,11 @@
      *      invoke on the result.
      * * 7: An optional {@link OperandType::BOOL} scalar, default to false.
      *      Set to true to specify NCHW data layout for input0 and output0.
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      *
      * Outputs:
      * * 0: The output 4-D tensor, of shape
      *      [batches, out_height, out_width, depth].
-     *
-     * Available since API level 27.
      */
     L2_POOL_2D = @1.1::OperationType:L2_POOL_2D,
 
@@ -994,11 +976,11 @@
      * 1-D slice along specified dimension.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      *
      * Supported tensor rank: up to 4
-     * Tensors with rank less than 4 are only supported since API level 29.
+     * Tensors with rank less than 4 are only supported since HAL version 1.2.
      *
      * Inputs:
      * * 0: A 4-D tensor, of shape [batches, height, width, depth], specifying
@@ -1011,10 +993,10 @@
      *      For input tensor of {@link OperandType::TENSOR_FLOAT32}, the bias
      *      value must be of {@link OperandType::FLOAT32}.
      * * 3: A scalar, specifying the scale factor, alpha.
-     *      For input tensor of {@link OperandType::TENSOR_FLOAT16}, the alpha
-     *      value must be of {@link OperandType::FLOAT16}.
-     *      For input tensor of {@link OperandType::TENSOR_FLOAT32}, the alpha
-     *      value must be of {@link OperandType::FLOAT32}.
+     *      For input tensor of {@link OperandType::TENSOR_FLOAT16}, the
+     *      alpha value must be of {@link OperandType::FLOAT16}.
+     *      For input tensor of {@link OperandType::TENSOR_FLOAT32}, the
+     *      alpha value must be of {@link OperandType::FLOAT32}.
      * * 4: A scalar, specifying the exponent, beta.
      *      For input tensor of {@link OperandType::TENSOR_FLOAT16}, the beta
      *      value must be of {@link OperandType::FLOAT16}.
@@ -1024,12 +1006,10 @@
      *      specifying the dimension normalization would be performed on.
      *      Negative index is used to specify axis from the end (e.g. -1 for
      *      the last axis). Must be in the range [-n, n).
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
-     *
-     * Available since API level 27.
      */
     LOCAL_RESPONSE_NORMALIZATION = @1.1::OperationType:LOCAL_RESPONSE_NORMALIZATION,
 
@@ -1041,22 +1021,20 @@
      *     output = 1 / (1 + exp(-input))
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
      * Supported tensor rank: up to 4.
      *
      * Inputs:
-     * * 0: A tensor, specifying the input. Since API level 29, this tensor may
-     *      be zero-sized.
+     * * 0: A tensor, specifying the input.
+     *      Since HAL version 1.2, this tensor may be zero-sized.
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
      *      For {@link OperandType::TENSOR_QUANT8_ASYMM},
      *      the scale must be 1.f / 256 and the zeroPoint must be 0.
-     *
-     * Available since API level 27.
      */
     LOGISTIC = @1.1::OperationType:LOGISTIC,
 
@@ -1064,7 +1042,7 @@
      * Projects an input to a bit vector via locality senstive hashing.
      *
      * Supported input tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_INT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
@@ -1086,7 +1064,7 @@
      *      Tensor[1].Dim[0] == Tensor[2].Dim[0]
      * * 3: Type:
      *        Sparse:
-     *          Value LSHProjectionType_SPARSE(=3) (since API level 29).
+     *          Value LSHProjectionType_SPARSE(=3) (since HAL version 1.2).
      *          Computed bit vector is considered to be sparse.
      *          Each output element is an int32 made up of multiple bits
      *          computed from hash functions.
@@ -1107,14 +1085,12 @@
      * Outputs:
      * * 0: If the projection type is Sparse:
      *      Output.Dim == { Tensor[0].Dim[0] }
-     *      A tensor of int32 that represents hash signatures,
+     *      A tensor of int32 that represents hash signatures.
      *
      *      If the projection type is Dense:
      *      Output.Dim == { Tensor[0].Dim[0] * Tensor[0].Dim[1] }
      *      A flattened tensor that represents projected bit vectors.
-     *
-     * Available since API level 27.
-     * The offset value for sparse projections was added in API level 29.
+     * The offset value for sparse projections was added in HAL version 1.2.
      */
     LSH_PROJECTION = @1.1::OperationType:LSH_PROJECTION,
 
@@ -1170,7 +1146,7 @@
      *   matrix, each element of which is the product of the corresponding
      *   elements of the input matrices.
      *
-     * Since API level 29 LSTM supports layer normalization.
+     * Since HAL version 1.2 LSTM supports layer normalization.
      * In case layer normalization is used, the inputs to internal activation
      * functions (sigmoid and \f$g\f$) are normalized, rescaled and recentered
      * following an approach from section 3.1 from
@@ -1197,7 +1173,7 @@
      * * The projection bias (\f$b_{proj}\f$) may (but not required to) have a
      *   value if the recurrent projection layer exists, and should otherwise
      *   have no value.
-     * * (API level >= 29) The four layer normalization weights either all have
+     * * (HAL version 1.2 or later) The four layer normalization weights either all have
      *   values or none of them have values. Additionally, if CIFG is used,
      *   input layer normalization weights tensor is omitted and the other layer
      *   normalization weights either all have values or none of them have
@@ -1228,7 +1204,7 @@
      * Jimmy Ba et al. "Layer Normalization"
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      *
      * All input and output tensors must be of the same type.
@@ -1291,24 +1267,24 @@
      * * 21:The clipping threshold (\f$t_{cell}\f$) for the cell state, such
      *      that values are bound within [-cell_clip, cell_clip]. If set to 0.0
      *      then clipping is disabled.
-     *      Until API level 29 this scalar must be of type {@link
-     *      FLOAT32}. Since API level 29, if all the input
+     *      Until HAL version 1.2 this scalar must be of type {@link
+     *      OperandType::FLOAT32}. Since HAL version 1.2, if all the input
      *      tensors have type {@link OperandType::TENSOR_FLOAT32}, this
      *      scalar must be of the type {@link OperandType::FLOAT32},
      *      otherwise if all the input tensors have the type {@link
-     *      TENSOR_FLOAT16}, this scalar must be of type {@link
-     *      FLOAT16}.
+     *      OperandType::TENSOR_FLOAT16}, this scalar must be of type {@link
+     *      OperandType::FLOAT16}.
      * * 22:The clipping threshold (\f$t_{proj}\f$) for the output from the
      *      projection layer, such that values are bound within
      *      [-proj_clip, proj_clip]. If set to 0.0 then clipping is disabled.
-     *      Until API level 29 this scalar must be of type {@link
-     *      FLOAT32}. Since API level 29, if all the input
+     *      Until HAL version 1.2 this scalar must be of type {@link
+     *      OperandType::FLOAT32}. Since HAL version 1.2, if all the input
      *      tensors have type {@link OperandType::TENSOR_FLOAT32}, this
      *      scalar must be of the type {@link OperandType::FLOAT32},
      *      otherwise if all the input tensors have the type {@link
-     *      TENSOR_FLOAT16}, this scalar must be of type {@link
-     *      FLOAT16}.
-     * Since API level 29 there are additional inputs to this op:
+     *      OperandType::TENSOR_FLOAT16}, this scalar must be of type {@link
+     *      OperandType::FLOAT16}.
+     * Since HAL version 1.2 there are additional inputs to this op:
      * * 23:The input layer normalization weights.
      *      A 1-D tensor of shape [num_units]. Used to rescale normalized inputs
      *      to activation at input gate.
@@ -1333,8 +1309,6 @@
      * * 3: The output (\f$o_t\f$).
      *      A 2-D tensor of shape [batch_size, output_size]. This is effectively
      *      the same as the current “output state (out)” value.
-     *
-     * Available since API level 27.
      */
     LSTM = @1.1::OperationType:LSTM,
 
@@ -1352,7 +1326,7 @@
      *         )
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
@@ -1360,13 +1334,14 @@
      * With the default data layout NHWC, the data is stored in the order of:
      * [batch, height, width, channels]. Alternatively, the data layout could
      * be NCHW, the data storage order of: [batch, channels, height, width].
+     * NCHW is supported since HAL version 1.2.
      *
      * Both explicit padding and implicit padding are supported.
      *
      * Inputs (explicit padding):
      * * 0: A 4-D tensor, of shape [batches, height, width, depth], specifying
-     *      the input. Since API level 29, zero batches is supported for this
-     *      tensor.
+     *      the input.
+     *      Since HAL version 1.2, zero batches is supported for this tensor.
      * * 1: An {@link OperandType::INT32} scalar, specifying the padding on
      *      the left, in the ‘width’ dimension.
      * * 2: An {@link OperandType::INT32} scalar, specifying the padding on
@@ -1388,12 +1363,12 @@
      *      invoke on the result.
      * * 10: An optional {@link OperandType::BOOL} scalar, default to false.
      *       Set to true to specify NCHW data layout for input0 and output0.
-     *       Available since API level 29.
+     *       Available since HAL version 1.2.
      *
      * Inputs (implicit padding):
      * * 0: A 4-D tensor, of shape [batches, height, width, depth], specifying
-     *      the input. Since API level 29, zero batches is supported for this
-     *      tensor.
+     *      the input.
+     *      Since HAL version 1.2, zero batches is supported for this tensor.
      * * 1: An {@link OperandType::INT32} scalar, specifying the implicit
      *      padding scheme, has to be one of the
      *      following values: {0 (NONE), 1 (SAME), 2 (VALID)}.
@@ -1410,13 +1385,13 @@
      *      invoke on the result.
      * * 7: An optional {@link OperandType::BOOL} scalar, default to false.
      *      Set to true to specify NCHW data layout for input0 and output0.
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      *
      * Outputs:
      * * 0: The output 4-D tensor, of shape
      *      [batches, out_height, out_width, depth].
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     MAX_POOL_2D = @1.1::OperationType:MAX_POOL_2D,
 
@@ -1435,15 +1410,15 @@
      * of the input operands. It starts with the trailing dimensions, and works
      * its way forward.
      *
-     * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
-     * * {@link OperandType::TENSOR_FLOAT32}
-     * * {@link OperandType::TENSOR_QUANT8_ASYMM}
-     *
-     * Since API level 29, generic zero-sized input tensor is supported. Zero
+     * Since HAL version 1.2, generic zero-sized input tensor is supported. Zero
      * dimension is only compatible with 0 or 1. The size of the output
      * dimension is zero if either of corresponding input dimension is zero.
      *
+     * Supported tensor {@link OperandType}:
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
+     * * {@link OperandType::TENSOR_FLOAT32}
+     * * {@link OperandType::TENSOR_QUANT8_ASYMM}
+     *
      * Supported tensor rank: up to 4
      *
      * Inputs:
@@ -1459,8 +1434,6 @@
      *      For output tensor of {@link OperandType::TENSOR_QUANT8_ASYMM},
      *      the following condition must be satisfied:
      *      output_scale > input1_scale * input2_scale.
-     *
-     * Available since API level 27.
      */
     MUL = @1.1::OperationType:MUL,
 
@@ -1472,20 +1445,20 @@
      *     output = max(0, input)
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
      * Supported tensor rank: up to 4.
      *
      * Inputs:
-     * * 0: A tensor, specifying the input. Since API level 29, this tensor may
-     *      be zero-sized.
+     * * 0: A tensor, specifying the input.
+     *      Since HAL version 1.2, this tensor may be zero-sized.
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     RELU = @1.1::OperationType:RELU,
 
@@ -1497,20 +1470,20 @@
      *     output = min(1.f, max(-1.f, input))
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
      * Supported tensor rank: up to 4.
      *
      * Inputs:
-     * * 0: A tensor, specifying the input. Since API level 29, this tensor may
-     *      be zero-sized.
+     * * 0: A tensor, specifying the input.
+     *      Since HAL version 1.2, this tensor may be zero-sized.
      *
      * Outputs:
-     * * 0: The output tensor of same shape as input0.
-     *
-     * Available since API level 27.
+     * * 0: The output tensor of the same shape as input0.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     RELU1 = @1.1::OperationType:RELU1,
 
@@ -1522,20 +1495,20 @@
      *     output = min(6, max(0, input))
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
      * Supported tensor rank: up to 4.
      *
      * Inputs:
-     * * 0: A tensor, specifying the input. Since API level 29, this tensor may
-     *      be zero-sized.
+     * * 0: A tensor, specifying the input.
+     *      Since HAL version 1.2, this tensor may be zero-sized.
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     RELU6 = @1.1::OperationType:RELU6,
 
@@ -1546,7 +1519,7 @@
      * tensor, but with a newly specified shape.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
@@ -1565,8 +1538,8 @@
      *
      * Outputs:
      * * 0: The output tensor, of shape specified by the input shape.
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     RESHAPE = @1.1::OperationType:RESHAPE,
 
@@ -1578,30 +1551,31 @@
      * same as corner pixels of input.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
-     * * {@link OperandType::TENSOR_QUANT8_ASYMM} (since API level 29)
+     * * {@link OperandType::TENSOR_QUANT8_ASYMM} (since HAL version 1.2)
      *
      * Supported tensor rank: 4, with "NHWC" or "NCHW" data layout.
      * With the default data layout NHWC, the data is stored in the order of:
      * [batch, height, width, channels]. Alternatively, the data layout could
      * be NCHW, the data storage order of: [batch, channels, height, width].
+     * NCHW is supported since HAL version 1.2.
      *
      * Both resizing by shape and resizing by scale are supported.
      *
      * Inputs (resizing by shape):
      * * 0: A 4-D tensor, of shape [batches, height, width, depth], specifying
-     *      the input. Since API level 29, zero batches is supported for this
-     *      tensor.
+     *      the input.
+     *      Since HAL version 1.2, zero batches is supported for this tensor.
      * * 1: An {@link OperandType::INT32} scalar, specifying the output
      *      width of the output tensor.
      * * 2: An {@link OperandType::INT32} scalar, specifying the output
      *      height of the output tensor.
      * * 3: An optional {@link OperandType::BOOL} scalar, default to false.
      *      Set to true to specify NCHW data layout for input0 and output0.
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      *
-     * Inputs (resizing by scale, since API level 29):
+     * Inputs (resizing by scale, since HAL version 1.2):
      * * 0: A 4-D tensor, of shape [batches, height, width, depth], specifying
      *      the input. Zero batches is supported for this tensor.
      * * 1: A scalar, specifying width_scale, the scaling factor of the width
@@ -1622,8 +1596,8 @@
      * Outputs:
      * * 0: The output 4-D tensor, of shape
      *      [batches, new_height, new_width, depth].
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     RESIZE_BILINEAR = @1.1::OperationType:RESIZE_BILINEAR,
 
@@ -1644,7 +1618,7 @@
      *   argument (if not “NONE”).
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      *
      * The input tensors must all be the same type.
@@ -1676,8 +1650,6 @@
      * * 1: output.
      *      A 2-D tensor of shape [batch_size, num_units]. This is effectively
      *      the same as the current state value.
-     *
-     * Available since API level 27.
      */
     RNN = @1.1::OperationType:RNN,
 
@@ -1696,34 +1668,32 @@
      * independently on each 1-D slice along specified dimension.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
      * Supported tensor rank: up to 4.
-     * Tensors with rank other than 2 or 4 are only supported since API level 29.
+     * Tensors with rank other than 2 or 4 are only supported since HAL version 1.2.
      *
      * Inputs:
-     * * 0: A 2-D or 4-D tensor, specifying the tensor to be reshaped. Since
-     *      API level 29, this tensor may be zero-sized.
+     * * 0: A 2-D or 4-D tensor, specifying the tensor to be reshaped.
+     *      Since HAL version 1.2, this tensor may be zero-sized.
      * * 1: A scalar, specifying the positive scaling factor for the exponent,
      *      beta. If input0 is of {@link OperandType::TENSOR_FLOAT32} or
      *      {@link OperandType::TENSOR_QUANT8_ASYMM}, the scalar must be of
-     *      {@link OperandType::FLOAT32}. If input0 is of {@link
-     *      OperandType::TENSOR_FLOAT16}, then the scalar must be of {@link
-     *      OperandType::FLOAT16}.
+     *      {@link OperandType::FLOAT32}.
+     *      If input0 is of {@link OperandType::TENSOR_FLOAT16}, then the
+     *      scalar must be of {@link OperandType::FLOAT16}.
      * * 2: An optional {@link OperandType::INT32} scalar, default to -1,
      *      specifying the dimension the activation would be performed on.
      *      Negative index is used to specify axis from the end (e.g. -1 for
      *      the last axis). Must be in the range [-n, n).
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
      *      For {@link OperandType::TENSOR_QUANT8_ASYMM},
      *      the scale must be 1.f / 256 and the zeroPoint must be 0.
-     *
-     * Available since API level 27.
      */
     SOFTMAX = @1.1::OperationType:SOFTMAX,
 
@@ -1742,7 +1712,7 @@
      * The input tensor's height and width must be divisible by block_size.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
@@ -1750,6 +1720,7 @@
      * With the default data layout NHWC, the data is stored in the order of:
      * [batch, height, width, channels]. Alternatively, the data layout could
      * be NCHW, the data storage order of: [batch, channels, height, width].
+     * NCHW is supported since HAL version 1.2.
      *
      * Inputs:
      * * 0: A 4-D tensor, of shape [batches, height, width, depth_in],
@@ -1759,13 +1730,13 @@
      *      input height and width.
      * * 2: An optional {@link OperandType::BOOL} scalar, default to false.
      *      Set to true to specify NCHW data layout for input0 and output0.
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      *
      * Outputs:
      * * 0: The output 4-D tensor, of shape [batches, height/block_size,
      *      width/block_size, depth_in*block_size*block_size].
-     *
-     * Available since API level 27.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     SPACE_TO_DEPTH = @1.1::OperationType:SPACE_TO_DEPTH,
 
@@ -1809,7 +1780,7 @@
      * the filters.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      *
      * All input tensors must be the same type.
@@ -1843,8 +1814,6 @@
      * * 1: output.
      *      A 2-D tensor of the same {@link OperandType} as the inputs, with shape
      *      [batch_size, num_units].
-     *
-     * Available since API level 27.
      */
     SVDF = @1.1::OperationType:SVDF,
 
@@ -1856,22 +1825,20 @@
      *     output = tanh(input)
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
-     * * {@link OperandType::TENSOR_QUANT8_ASYMM} (since API level 29)
+     * * {@link OperandType::TENSOR_QUANT8_ASYMM} (since HAL version 1.2)
      *
      * Supported tensor rank: up to 4.
      *
      * Inputs:
-     * * 0: A tensor, specifying the input. Since API level 29, this tensor may
-     *      be zero-sized.
+     * * 0: A tensor, specifying the input.
+     *      Since HAL version 1.2, this tensor may be zero-sized.
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
      *      For {@link OperandType::TENSOR_QUANT8_ASYMM},
      *      the scale must be 1.f / 128 and the zeroPoint must be 128.
-     *
-     * Available since API level 27.
      */
     TANH = @1.1::OperationType:TANH,
 
@@ -1886,7 +1853,7 @@
      * This is the reverse of SpaceToBatch.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
@@ -1894,6 +1861,7 @@
      * With the default data layout NHWC, the data is stored in the order of:
      * [batch, height, width, channels]. Alternatively, the data layout could
      * be NCHW, the data storage order of: [batch, channels, height, width].
+     * NCHW is supported since HAL version 1.2.
      *
      * Inputs:
      * * 0: An n-D tensor, specifying the tensor to be reshaped
@@ -1906,8 +1874,8 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 28.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     BATCH_TO_SPACE_ND = @1.1::OperationType:BATCH_TO_SPACE_ND,
 
@@ -1931,12 +1899,12 @@
      *     input2.dimension = {5, 4, 3, 1}
      *     output.dimension = {5, 4, 3, 2}
      *
-     * Since API level 29, generic zero-sized input tensor is supported. Zero
+     * Since HAL version 1.2, generic zero-sized input tensor is supported. Zero
      * dimension is only compatible with 0 or 1. The size of the output
      * dimension is zero if either of corresponding input dimension is zero.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      *
      * Supported tensor rank: up to 4
@@ -1951,8 +1919,6 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 28.
      */
     DIV = @1.1::OperationType:DIV,
 
@@ -1965,7 +1931,7 @@
      * length 1.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
@@ -1987,21 +1953,21 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 28.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be same as input0.
      */
     MEAN = @1.1::OperationType:MEAN,
 
     /**
-     * Pads a tensor with zeros.
+     * Pads a tensor.
      *
      * This operation pads a tensor according to the specified paddings.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
-     * * {@link OperandType::TENSOR_QUANT8_ASYMM} (full support since API
-     *   level 29, see the output section)
+     * * {@link OperandType::TENSOR_QUANT8_ASYMM}
+     *   (full support since HAL version 1.2, see the output section)
      *
      * Supported tensor rank: up to 4
      *
@@ -2023,12 +1989,12 @@
      *      of the padding:
      *          output0.dimension[i] =
      *              padding[i, 0] + input0.dimension[i] + padding[i, 1]
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      *
-     *      NOTE: Before API level 29, the pad value for
-     *      {@link ANEURALNETWORKS_TENSOR_QUANT8_ASYMM} is undefined.
-     *      Since API level 29, the pad value is always the logical zero.
-     *
-     * Available since API level 28.
+     *      NOTE: Before HAL version 1.2, the pad value for
+     *      {@link OperandType::TENSOR_QUANT8_ASYMM} is undefined.
+     *      Since HAL version 1.2, the pad value is always the logical zero.
      */
     PAD = @1.1::OperationType:PAD,
 
@@ -2044,14 +2010,16 @@
      * dimensions of the input are optionally zero padded according to paddings.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
+     *   (full support since HAL version 1.2, see the output section)
      *
      * Supported tensor rank: 4, with "NHWC" or "NCHW" data layout.
      * With the default data layout NHWC, the data is stored in the order of:
      * [batch, height, width, channels]. Alternatively, the data layout could
      * be NCHW, the data storage order of: [batch, channels, height, width].
+     * NCHW is supported since HAL version 1.2.
      *
      * Inputs:
      * * 0: An n-D tensor, specifying the input.
@@ -2068,12 +2036,16 @@
      *      end of dimension i.
      * * 3: An optional {@link OperandType::BOOL} scalar, default to false.
      *      Set to true to specify NCHW data layout for input0 and output0.
-     *      Available since API level 29.
+     *      Available since HAL version 1.2.
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      *
-     * Available since API level 28.
+     *      NOTE: Before HAL version 1.2, the pad value for
+     *      {@link OperandType::TENSOR_QUANT8_ASYMM} is undefined.
+     *      Since HAL version 1.2, the pad value is always the logical zero.
      */
     SPACE_TO_BATCH_ND = @1.1::OperationType:SPACE_TO_BATCH_ND,
 
@@ -2086,7 +2058,7 @@
      * dimensions by specifying the axes (input1).
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
@@ -2104,8 +2076,8 @@
      * * 0: A tensor of the same {@link OperandType} as input0. Contains the
      *      same data as input, but has one or more dimensions of size 1
      *      removed.
-     *
-     * Available since API level 28.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     SQUEEZE = @1.1::OperationType:SQUEEZE,
 
@@ -2119,7 +2091,7 @@
      * reverse slice.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
@@ -2151,8 +2123,8 @@
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0 and rank (n - k),
      *      where k is the number of bits set in shrink_axis_mask.
-     *
-     * Available since API level 28.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     STRIDED_SLICE = @1.1::OperationType:STRIDED_SLICE,
 
@@ -2176,14 +2148,14 @@
      *     input2.dimension = {5, 4, 3, 1}
      *     output.dimension = {5, 4, 3, 2}
      *
-     * Since API level 29, generic zero-sized input tensor is supported. Zero
+     * Since HAL version 1.2, generic zero-sized input tensor is supported. Zero
      * dimension is only compatible with 0 or 1. The size of the output
      * dimension is zero if either of corresponding input dimension is zero.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
-     * * {@link OperandType::TENSOR_QUANT8_ASYMM} (since API level 29)
+     * * {@link OperandType::TENSOR_QUANT8_ASYMM} (since HAL version 1.2)
      *
      * Supported tensor rank: up to 4
      *
@@ -2197,8 +2169,8 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 28.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint can be different from inputs' scale and zeroPoint.
      */
     SUB = @1.1::OperationType:SUB,
 
@@ -2212,7 +2184,7 @@
      * regular matrix transpose on 2-D input Tensors.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16} (since HAL version 1.2)
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
@@ -2220,14 +2192,14 @@
      *
      * Inputs:
      * * 0: An n-D tensor, specifying the tensor to be transposed.
-     *      Since API level 29, this tensor may be zero-sized.
+     *      Since HAL version 1.2, this tensor may be zero-sized.
      * * 1: An optional 1-D Tensor of {@link OperandType::TENSOR_INT32},
      *      the permutation of the dimensions of the input tensor.
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 28.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     TRANSPOSE = @1.1::OperationType:TRANSPOSE,
 
@@ -2245,8 +2217,6 @@
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
-     *
-     * Available since API level 29.
      */
     ABS = 38,
 
@@ -2269,8 +2239,6 @@
      *
      * Outputs:
      * * 0: An (n - 1)-D {@link OperandType::TENSOR_INT32} tensor.
-     *
-     * Available since API level 29.
      */
     // There is no underscore in ARG_MAX to avoid name conflict with
     // the macro defined in libc/kernel/uapi/linux/limits.h.
@@ -2295,8 +2263,6 @@
      *
      * Outputs:
      * * 0: An (n - 1)-D {@link OperandType::TENSOR_INT32} tensor.
-     *
-     * Available since API level 29.
      */
     ARGMIN = 40,  // See ARGMAX for naming discussion.
 
@@ -2341,8 +2307,8 @@
      * * 0: A tensor of the same {@link OperandType} as input0, with shape
      *      [num_rois, num_classes * 4], specifying the coordinates of each
      *      output bounding box for each class, with format [x1, y1, x2, y2].
-     *
-     * Available since API level 29.
+     *      For type of {@link OperandType::TENSOR_QUANT16_ASYMM}, the
+     *      scale must be 0.125 and the zero point must be 0.
      */
     AXIS_ALIGNED_BBOX_TRANSFORM = 41,
 
@@ -2482,17 +2448,15 @@
      *       then clipping is disabled.
      *       If all the input tensors have type {@link OperandType::TENSOR_FLOAT32},
      *       this scalar must be of the type {@link OperandType::FLOAT32},
-     *       otherwise if all the input tensors have the type {@link
-     *       TENSOR_FLOAT16}, this scalar must be of type {@link
-     *       FLOAT16}.
+     *       otherwise if all the input tensors have the type {@link OperandType::TENSOR_FLOAT16},
+     *       this scalar must be of type {@link OperandType::FLOAT16}.
      * * 50: The clipping threshold for the output from the
      *       projection layer, such that values are bound within
      *       [-proj_clip, proj_clip]. If set to 0.0 then clipping is disabled.
      *       If all the input tensors have type {@link OperandType::TENSOR_FLOAT32},
      *       this scalar must be of the type {@link OperandType::FLOAT32},
-     *       otherwise if all the input tensors have the type {@link
-     *       TENSOR_FLOAT16}, this scalar must be of type {@link
-     *       FLOAT16}.
+     *       otherwise if all the input tensors have the type {@link OperandType::TENSOR_FLOAT16},
+     *       this scalar must be of type {@link OperandType::FLOAT16}.
      * * 51: merge_outputs
      *       An {@link OperandType::BOOL} scalar specifying if the outputs
      *       from forward and backward cells should be merged.
@@ -2539,8 +2503,6 @@
      *      A 3-D tensor of shape:
      *        If time-major: [max_time, batch_size, bw_output_size]
      *        If batch-major: [batch_size, max_time, bw_output_size]
-     *
-     * Available since API level 29.
      */
     BIDIRECTIONAL_SEQUENCE_LSTM = 42,
 
@@ -2658,8 +2620,6 @@
      *      (timeMajor). If it is set to true, then the shape is set to
      *      [maxTime, batchSize, bwNumUnits], otherwise the shape is set to
      *      [batchSize, maxTime, bwNumUnits].
-     *
-     * Available since API level 29.
      */
     BIDIRECTIONAL_SEQUENCE_RNN = 43,
 
@@ -2737,8 +2697,6 @@
      * * 3: A 1-D {@link OperandType::TENSOR_INT32} tensor, of shape
      *      [num_output_rois], specifying the batch index of each box. Boxes
      *      with the same batch index are grouped together.
-     *
-     * Available since API level 29.
      */
     BOX_WITH_NMS_LIMIT = 44,
 
@@ -2762,8 +2720,6 @@
      *
      * Outputs:
      * * 0: A tensor with the same shape as input0.
-     *
-     * Available since API level 29.
      */
     CAST = 45,
 
@@ -2800,8 +2756,8 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} and same shape as input0.
-     *
-     * Available since API level 29.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     CHANNEL_SHUFFLE = 46,
 
@@ -2856,14 +2812,14 @@
      * * 11: A scalar, score_threshold. Boxes with scores lower than the
      *       threshold are filtered before sending to the NMS algorithm. The
      *       scalar must be of {@link OperandType::FLOAT16} if input0 is of
-     *       {@link OperandType::TENSOR_FLOAT16} and of {@link
-     *       OperandType::FLOAT32} if input0 is of {@link
-     *       OperandType::TENSOR_FLOAT32}.
+     *       {@link OperandType::TENSOR_FLOAT16} and of
+     *       {@link OperandType::FLOAT32} if input0 is of
+     *       {@link OperandType::TENSOR_FLOAT32}.
      * * 12: A scalar, specifying the IoU threshold for hard NMS. The scalar
-     *       must be of {@link OperandType::FLOAT16} if input0 is of {@link
-     *       OperandType::TENSOR_FLOAT16} and of {@link
-     *       OperandType::FLOAT32} if input0 is of {@link
-     *       OperandType::TENSOR_FLOAT32}.
+     *       must be of {@link OperandType::FLOAT16} if input0 is of
+     *       {@link OperandType::TENSOR_FLOAT16} and of
+     *       {@link OperandType::FLOAT32} if input0 is of
+     *       {@link OperandType::TENSOR_FLOAT32}.
      * * 13: An {@link OperandType::BOOL} scalar, set to true to include
      *       background class in the list of label map for the output, set
      *       to false to not include the background. When the background
@@ -2882,8 +2838,6 @@
      *      output detection.
      * * 3: An 1-D {@link OperandType::TENSOR_INT32} tensor, of shape [batches],
      *      specifying the number of valid output detections for each batch.
-     *
-     * Available since API level 29.
      */
     DETECTION_POSTPROCESSING = 47,
 
@@ -2908,8 +2862,6 @@
      *
      * Outputs:
      * * 0: A tensor of {@link OperandType::TENSOR_BOOL8}.
-     *
-     * Available since API level 29.
      */
     EQUAL = 48,
 
@@ -2927,8 +2879,6 @@
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
-     *
-     * Available since API level 29.
      */
     EXP = 49,
 
@@ -2956,8 +2906,8 @@
      * Outputs:
      * * 0: An (n + 1)-D tensor with the same {@link OperandType} and data as
      *      input0.
-     *
-     * Available since API level 29.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     EXPAND_DIMS = 50,
 
@@ -2994,8 +2944,8 @@
      *
      * Outputs:
      * * 0: An (n + k - 1)-D tensor with the same {@link OperandType} as input0.
-     *
-     * Available since API level 29.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     GATHER = 51,
 
@@ -3074,8 +3024,6 @@
      * * 2: A 1-D {@link OperandType::TENSOR_INT32} tensor, of shape
      *      [num_output_rois], specifying the batch index of each box. Boxes
      *      with the same batch index are grouped together.
-     *
-     * Available since API level 29.
      */
     GENERATE_PROPOSALS = 52,
 
@@ -3100,8 +3048,6 @@
      *
      * Outputs:
      * * 0: A tensor of {@link OperandType::TENSOR_BOOL8}.
-     *
-     * Available since API level 29.
      */
     GREATER = 53,
     /**
@@ -3125,8 +3071,6 @@
      *
      * Outputs:
      * * 0: A tensor of {@link OperandType::TENSOR_BOOL8}.
-     *
-     * Available since API level 29.
      */
     GREATER_EQUAL = 54,
 
@@ -3191,7 +3135,8 @@
      *      [depth_out, filter_height, filter_width, depth_group], specifying
      *      the filter, where depth_out must be divisible by num_groups.  For
      *      tensor of type {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL}
-     *      the channel dimension must be set to 0.
+     *      the channel dimension (channelDim at
+     *      {@link SymmPerChannelQuantParams}) must be set to 0.
      * * 2: A 1-D tensor, of shape [depth_out], specifying the bias. For input
      *      tensor of type {@link OperandType::TENSOR_FLOAT32} or
      *      {@link OperandType::TENSOR_FLOAT16}, the bias must be of the same
@@ -3229,7 +3174,8 @@
      *      [depth_out, filter_height, filter_width, depth_group], specifying
      *      the filter, where depth_out must be divisible by num_groups.  For
      *      tensor of type {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL}
-     *      the channel dimension must be set to 0.
+     *      the channel dimension (SymmPerChannelQuantParams::channelDim)
+     *      must be set to 0.
      * * 2: A 1-D tensor, of shape [depth_out], specifying the bias. For input
      *      tensor of type {@link OperandType::TENSOR_FLOAT32} or
      *      {@link OperandType::TENSOR_FLOAT16}, the bias must be of the same
@@ -3258,8 +3204,8 @@
      * Outputs:
      * * 0: The output 4-D tensor, of shape
      *      [batches, out_height, out_width, depth_out].
-     *
-     * Available since API level 29.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint can be different from inputs' scale and zeroPoint.
      */
     GROUPED_CONV_2D = 55,
 
@@ -3300,12 +3246,14 @@
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0, with shape
      *      [num_boxes, num_keypoints], specifying score of the keypoints.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint can be different from input0 scale and zeroPoint.
      * * 1: A tensor of the same {@link OperandType} as input1, with shape
      *      [num_boxes, num_keypoints, 2], specifying the location of
      *      the keypoints, the second dimension is organized as
      *      [keypoint_x, keypoint_y].
-     *
-     * Available since API level 29.
+     *      For type of {@link OperandType::TENSOR_QUANT16_ASYMM}, the
+     *      scale must be 0.125 and the zero point must be 0.
      */
     HEATMAP_MAX_KEYPOINT = 56,
 
@@ -3339,26 +3287,24 @@
      * * 0: An n-D tensor, specifying the tensor to be normalized.
      * * 1: A scalar, specifying gamma, the scale applied to the normalized
      *      tensor. The scalar must be of {@link OperandType::FLOAT16} if
-     *      input0 is of {@link OperandType::TENSOR_FLOAT16} and of {@link
-     *      OperandType::FLOAT32} if input0 is of {@link
-     *      OperandType::TENSOR_FLOAT32}.
+     *      input0 is of {@link OperandType::TENSOR_FLOAT16} and of
+     *      {@link OperandType::FLOAT32} if input0 is of
+     *      {@link OperandType::TENSOR_FLOAT32}.
      * * 2: A scalar, specifying beta, the offset applied to the normalized
      *      tensor. The scalar must be of {@link OperandType::FLOAT16} if
-     *      input0 is of {@link OperandType::TENSOR_FLOAT16} and of {@link
-     *      OperandType::FLOAT32} if input0 is of {@link
-     *      OperandType::TENSOR_FLOAT32}.
+     *      input0 is of {@link OperandType::TENSOR_FLOAT16} and of
+     *      {@link OperandType::FLOAT32} if input0 is of
+     *      {@link OperandType::TENSOR_FLOAT32}.
      * * 3: A scalar, specifying epsilon, the small value added to variance to
      *      avoid dividing by zero. The scalar must be of {@link OperandType::FLOAT16} if
-     *      input0 is of {@link OperandType::TENSOR_FLOAT16} and of {@link
-     *      OperandType::FLOAT32} if input0 is of {@link
-     *      OperandType::TENSOR_FLOAT32}.
+     *      input0 is of {@link OperandType::TENSOR_FLOAT16} and of
+     *      {@link OperandType::FLOAT32} if input0 is of
+     *      {@link OperandType::TENSOR_FLOAT32}.
      * * 4: An {@link OperandType::BOOL} scalar, set to true to specify
      *      NCHW data layout for input0 and output0. Set to false for NHWC.
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} and same shape as input0.
-     *
-     * Available since API level 29.
      */
     INSTANCE_NORMALIZATION = 57,
 
@@ -3383,8 +3329,6 @@
      *
      * Outputs:
      * * 0: A tensor of {@link OperandType::TENSOR_BOOL8}.
-     *
-     * Available since API level 29.
      */
     LESS = 58,
 
@@ -3409,8 +3353,6 @@
      *
      * Outputs:
      * * 0: A tensor of {@link OperandType::TENSOR_BOOL8}.
-     *
-     * Available since API level 29.
      */
     LESS_EQUAL = 59,
 
@@ -3428,8 +3370,6 @@
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
-     *
-     * Available since API level 29.
      */
     LOG = 60,
 
@@ -3450,8 +3390,6 @@
      *
      * Outputs:
      * * 0: A tensor of {@link OperandType::TENSOR_BOOL8}.
-     *
-     * Available since API level 29.
      */
     LOGICAL_AND = 61,
 
@@ -3468,8 +3406,6 @@
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
-     *
-     * Available since API level 29.
      */
     LOGICAL_NOT = 62,
 
@@ -3490,8 +3426,6 @@
      *
      * Outputs:
      * * 0: A tensor of {@link OperandType::TENSOR_BOOL8}.
-     *
-     * Available since API level 29.
      */
     LOGICAL_OR = 63,
 
@@ -3523,8 +3457,6 @@
      * Outputs:
      * * 0: The output tensor of the same {@link OperandType} and shape as
      *      input0.
-     *
-     * Available since API level 29.
      */
     LOG_SOFTMAX = 64,
 
@@ -3543,11 +3475,13 @@
      * * 0: A tensor.
      * * 1: A tensor of the same {@link OperandType} and compatible dimensions
      *      with input0.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scales and zeroPoint can be different from input0 scale and zeroPoint.
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 29.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint can be different from inputs' scale and zeroPoint.
      */
     MAXIMUM = 65,
 
@@ -3566,11 +3500,13 @@
      * * 0: A tensor.
      * * 1: A tensor of the same {@link OperandType} and compatible dimensions
      *      with input0.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scales and zeroPoint can be different from input0 scale and zeroPoint.
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 29.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint can be different from inputs' scale and zeroPoint.
      */
     MINIMUM = 66,
 
@@ -3589,8 +3525,6 @@
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
-     *
-     * Available since API level 29.
      */
     NEG = 67,
 
@@ -3615,8 +3549,6 @@
      *
      * Outputs:
      * * 0: A tensor of {@link OperandType::TENSOR_BOOL8}.
-     *
-     * Available since API level 29.
      */
     NOT_EQUAL = 68,
 
@@ -3657,8 +3589,8 @@
      *      of the padding:
      *          output0.dimension[i] =
      *              padding[i, 0] + input0.dimension[i] + padding[i, 1]
-     *
-     * Available since API level 29.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     PAD_V2 = 69,
 
@@ -3689,8 +3621,6 @@
      *
      * Outputs:
      * * 0: An output tensor.
-     *
-     * Available since API level 29.
      */
     POW = 70,
 
@@ -3728,8 +3658,8 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 29.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint can be diffent from the input0 scale and zeroPoint.
      */
     PRELU = 71,
 
@@ -3752,8 +3682,6 @@
      * Outputs:
      * * 0: The output tensor of same shape as input0, but with
      *      {@link OperandType::TENSOR_QUANT8_ASYMM}.
-     *
-     * Available since API level 29.
      */
     QUANTIZE = 72,
 
@@ -3879,8 +3807,6 @@
      * Outputs:
      * * 0: A 2-D {@link OperandType::TENSOR_INT32} tensor with shape
      *      [batches, samples], containing the drawn samples.
-     *
-     * Available since API level 29.
      */
     RANDOM_MULTINOMIAL = 74,
 
@@ -3906,8 +3832,6 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 29.
      */
     REDUCE_ALL = 75,
 
@@ -3933,8 +3857,6 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 29.
      */
     REDUCE_ANY = 76,
 
@@ -3962,8 +3884,8 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 29.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     REDUCE_MAX = 77,
 
@@ -3991,8 +3913,8 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 29.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     REDUCE_MIN = 78,
 
@@ -4018,8 +3940,6 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 29.
      */
     REDUCE_PROD = 79,
 
@@ -4045,8 +3965,6 @@
      *
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0.
-     *
-     * Available since API level 29.
      */
     REDUCE_SUM = 80,
 
@@ -4064,7 +3982,7 @@
      * interpolation.
      *
      * Supported tensor {@link OperandType}:
-     * * {@link OperandType::TENSOR_FLOAT16} (since API level 29)
+     * * {@link OperandType::TENSOR_FLOAT16}
      * * {@link OperandType::TENSOR_FLOAT32}
      * * {@link OperandType::TENSOR_QUANT8_ASYMM}
      *
@@ -4105,8 +4023,8 @@
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0. The output
      *      shape is [num_rois, out_height, out_width, depth].
-     *
-     * Available since API level 29.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint can be different from the input0 scale and zeroPoint.
      */
     ROI_ALIGN = 81,
 
@@ -4156,8 +4074,8 @@
      * Outputs:
      * * 0: A tensor of the same {@link OperandType} as input0. The output
      *      shape is [num_rois, out_height, out_width, depth].
-     *
-     * Available since API level 29.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     ROI_POOLING = 82,
 
@@ -4175,8 +4093,6 @@
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
-     *
-     * Available since API level 29.
      */
     RSQRT = 83,
 
@@ -4201,9 +4117,13 @@
      *      true) or input2 (if false).
      * * 1: An input tensor of the same shape as input0.
      * * 2: An input tensor of the same shape and type as input1.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scales and zeroPoint can be different from input1 scale and zeroPoint.
      *
      * Outputs:
      * * 0: A tensor of the same type and shape as input1 and input2.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint can be different from inputs' scale and zeroPoint.
      *
      */
     SELECT = 84,
@@ -4222,8 +4142,6 @@
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
-     *
-     * Available since API level 29.
      */
     SIN = 85,
 
@@ -4235,7 +4153,6 @@
      * for each dimension. The size is specified as a 1-D tensor containing
      * either size of a slice along corresponding dimension or -1. In the latter
      * case, all the remaining elements in dimension are included in the slice.
-     * Slice size in each dimension cannot be zero.
      *
      * A sum of begin offset and a size of a slice must not exceed size of a
      * corresponding dimension.
@@ -4257,8 +4174,8 @@
      *
      * Outputs:
      * * 0: An n-D tensor of the same type as the input containing the slice.
-     *
-     * Available since API level 29.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      its scale and zeroPoint has to be same as the input0 scale and zeroPoint.
      */
     SLICE = 86,
 
@@ -4282,8 +4199,8 @@
      *
      * Outputs:
      * * 0 ~ (num_splits - 1): Resulting subtensors.
-     *
-     * Available since API level 29.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     SPLIT = 87,
 
@@ -4301,8 +4218,6 @@
      *
      * Outputs:
      * * 0: The output tensor of same shape as input0.
-     *
-     * Available since API level 29.
      */
     SQRT = 88,
 
@@ -4330,8 +4245,8 @@
      *
      * Outputs:
      * * 0: A tiled tensor of the same {@link OperandType} and rank as `input`.
-     *
-     * Available since API level 29.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     TILE = 89,
 
@@ -4357,10 +4272,10 @@
      * Outputs:
      * * 0: An n-D tensor of the same type as the input, containing the k
      *      largest elements along each last dimensional slice.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      * * 1: An n-D tensor of type {@link OperandType::TENSOR_INT32}
      *      containing the indices of values within the last dimension of input.
-     *
-     * Available since API level 29.
      */
     TOPK_V2 = 90,
 
@@ -4374,7 +4289,7 @@
      * The output dimensions are functions of the filter dimensions, stride, and
      * padding.
      *
-     * Supported tensor {@link OperandCode} configurations:
+     * Supported tensor {@link OperandType} configurations:
      * * 16 bit floating point:
      * * * {@link OperandType::TENSOR_FLOAT16} for input, filter, output, and bias.
      *
@@ -4406,7 +4321,7 @@
      *      [depth_out, filter_height, filter_width, depth_in], specifying the
      *      filter. For tensor of type
      *      {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL} the channel
-     *      dimension (extraParams.channelQuant.channelDim) must be set to 0.
+     *      dimension (SymmPerChannelQuantParams::channelDim) must be set to 0.
      * * 2: A 1-D tensor, of shape [depth_out], specifying the bias. For input
      *      tensor of type {@link OperandType::TENSOR_FLOAT32} or
      *      {@link OperandType::TENSOR_FLOAT16}, the bias should be of the
@@ -4443,7 +4358,7 @@
      *      [depth_out, filter_height, filter_width, depth_in], specifying the
      *      filter. For tensor of type
      *      {@link OperandType::TENSOR_QUANT8_SYMM_PER_CHANNEL} the channel
-     *      dimension (extraParams.channelQuant.channelDim) must be set to 0.
+     *      dimension (SymmPerChannelQuantParams::channelDim) must be set to 0.
      * * 2: A 1-D tensor, of shape [depth_out], specifying the bias. For input
      *      tensor of type {@link OperandType::TENSOR_FLOAT32} or
      *      {@link OperandType::TENSOR_FLOAT16}, the bias should be of the
@@ -4473,8 +4388,8 @@
      * Outputs:
      * * 0: The output 4-D tensor, of shape
      *      [batches, out_height, out_width, depth_out].
-     *
-     * Available since API level 29.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint can be different from inputs' scale and zeroPoint.
      */
     TRANSPOSE_CONV_2D = 91,
 
@@ -4584,8 +4499,6 @@
      *      A 3-D tensor of shape:
      *        If time-major: [max_time, batch_size, output_size]
      *        If batch-major: [batch_size, max_time, output_size]
-     *
-     * Available since API level 29.
      */
     UNIDIRECTIONAL_SEQUENCE_LSTM = 92,
 
@@ -4641,8 +4554,6 @@
      *      it is set to 1, then the output has a shape [maxTime, batchSize,
      *      numUnits], otherwise the output has a shape [batchSize, maxTime,
      *      numUnits].
-     *
-     * Available since API level 29.
      */
     UNIDIRECTIONAL_SEQUENCE_RNN = 93,
 
@@ -4696,8 +4607,8 @@
      * Outputs:
      * * 0: The output 4-D tensor, of shape
      *      [batches, new_height, new_width, depth].
-     *
-     * Available since API level 29.
+     *      For a {@link OperandType::TENSOR_QUANT8_ASYMM} tensor,
+     *      the scale and zeroPoint must be the same as input0.
      */
     RESIZE_NEAREST_NEIGHBOR = 94,
 
diff --git a/neuralnetworks/1.2/types.t b/neuralnetworks/1.2/types.t
new file mode 100644
index 0000000..cab330d
--- /dev/null
+++ b/neuralnetworks/1.2/types.t
@@ -0,0 +1,745 @@
+%% template file for generating types.hal.
+%% see frameworks/ml/nn/tools/api/README.md.
+/*
+ * Copyright (C) 2018 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package android.hardware.neuralnetworks@1.2;
+
+import @1.0::DataLocation;
+import @1.0::ErrorStatus;
+import @1.0::OperandLifeTime;
+import @1.0::OperandType;
+import @1.0::PerformanceInfo;
+import @1.1::OperationType;
+
+import android.hidl.safe_union@1.0::Monostate;
+
+enum Constant : uint32_t {
+    /**
+     * The byte size of the cache token.
+     */
+    BYTE_SIZE_OF_CACHE_TOKEN = 32,
+
+    /**
+     * The maximum number of files for each type of cache in compilation caching.
+     */
+    MAX_NUMBER_OF_CACHE_FILES = 32,
+};
+
+enum OperandType : @1.0::OperandType {
+%insert Operand_1.2
+
+    /*
+     * DEPRECATED. Since HAL version 1.2, extensions are the preferred
+     * alternative to OEM operation and data types.
+     *
+     * OEM specific scalar value.
+     * OEM                 = 10000,
+     */
+    /*
+     * DEPRECATED. Since HAL version 1.2, extensions are the preferred
+     * alternative to OEM operation and data types.
+     *
+     * A tensor of OEM specific values.
+     * TENSOR_OEM_BYTE     = 10001,
+     */
+    /* ADDING A NEW FUNDAMENTAL TYPE REQUIRES UPDATING THE VALUE OF
+     * OperandTypeRange::FUNDAMENTAL_MAX.
+     */
+    /* ADDING A NEW OEM TYPE REQUIRES UPDATING THE VALUE OF
+     * OperandTypeRange::OEM_MAX.
+     */
+};
+
+/**
+ * The range of operand values in the OperandType enum.
+ */
+enum OperandTypeRange : uint32_t {
+    BASE_MIN        = 0,
+    FUNDAMENTAL_MIN = 0,
+%insert Operand_1.2_MAX
+    OEM_MIN         = 10000,
+    OEM_MAX         = 10001,
+    BASE_MAX        = 0xFFFF,
+};
+
+/**
+ * Operation types.
+ *
+ * The type of an operation in a model.
+ */
+enum OperationType : int32_t {
+
+%insert Operation_1.0
+
+%insert Operation_1.1
+
+%insert Operation_1.2
+
+    /**
+     * DEPRECATED. Since NNAPI 1.2, extensions are the preferred alternative to
+     * OEM operation and data types.
+     *
+     * This operation is OEM specific. It should only be used for OEM
+     * applications.
+     */
+    OEM_OPERATION = @1.1::OperationType:OEM_OPERATION,
+    /* ADDING A NEW FUNDAMENTAL OPERATION REQUIRES UPDATING THE VALUE OF
+     * OperationTypeRange::FUNDAMENTAL_MAX.
+     */
+    /* ADDING A NEW OEM OPERATION REQUIRES UPDATING THE VALUE OF
+     * OperationTypeRange::OEM_MAX.
+     */
+};
+
+/**
+ * The range of values in the OperationType enum.
+ */
+enum OperationTypeRange : uint32_t {
+    BASE_MIN        = 0,
+    FUNDAMENTAL_MIN = 0,
+%insert Operation_1.2_MAX
+    OEM_MIN         = 10000,
+    OEM_MAX         = 10000,
+    BASE_MAX        = 0xFFFF,
+};
+
+/**
+ * Device types.
+ *
+ * The type of NNAPI device.
+ */
+enum DeviceType : int32_t {
+    // Leaving 0 unused as it means unknown type in NDK NNAPI. There is no
+    // HAL equivalent of unknown type and a 1.2 HAL implementation must belong
+    // to one of the categories below.
+    /** The device does not fall into any category below. */
+    OTHER             = 1,
+    /** The device runs NNAPI models on single or multi-core CPU. */
+    CPU               = 2,
+    /** The device can run NNAPI models and also accelerate graphics APIs such
+      * as OpenGL ES and Vulkan. */
+    GPU               = 3,
+    /** Dedicated accelerator for Machine Learning workloads. */
+    ACCELERATOR       = 4,
+};
+
+/**
+ * The capabilities of a driver.
+ *
+ * Performance of an operation comes from the type of its first operand.
+ * This represents performance for non extension operand types.
+ */
+struct Capabilities {
+    /**
+     * Driver performance when operating on float32 data but performing
+     * calculations with range and/or precision as low as that of the IEEE
+     * 754 16-bit floating-point format.
+     */
+    PerformanceInfo relaxedFloat32toFloat16PerformanceScalar;
+    PerformanceInfo relaxedFloat32toFloat16PerformanceTensor;
+
+    /**
+     * Driver performance when operating on a particular data type.
+     * In the case of float32 data, this is used when the calculations
+     * are not relaxed.
+     */
+    struct OperandPerformance {
+        OperandType type;
+        PerformanceInfo info;
+    };
+
+    /**
+     * Performance by operand type. Must be sorted by OperandType.
+     * If a particular OperandType is not present in operandPerformance,
+     * its performance is treated as { .execTime = FLT_MAX, .powerUsage = FLT_MAX }.
+     */
+    vec<OperandPerformance> operandPerformance;
+};
+
+/**
+ * Describes one operation of the model's graph.
+ */
+struct Operation {
+    /**
+     * The operation type.
+     *
+     * Besides the values listed in {@link OperationType}, any value above
+     * {@link OperationTypeRange::BASE_MAX} is possible and should be interpreted
+     * as an extension type according to {@link Model::extensionNameToPrefix}.
+     */
+    OperationType type;
+
+    /**
+     * Describes the table that contains the indexes of the inputs of the
+     * operation. The offset is the index in the operandIndexes table.
+     */
+    vec<uint32_t> inputs;
+
+    /**
+     * Describes the table that contains the indexes of the outputs of the
+     * operation. The offset is the index in the operandIndexes table.
+     */
+    vec<uint32_t> outputs;
+};
+
+/**
+ * Parameters for TENSOR_QUANT8_SYMM_PER_CHANNEL operand.
+ */
+struct SymmPerChannelQuantParams {
+    /** Array of scaling values for each channel. Each value must be greater than zero. */
+    vec<float> scales;
+    /** Index of the channel dimension */
+    uint32_t channelDim;
+};
+
+/**
+ * Describes one operand of the model's graph.
+ */
+struct Operand {
+    /**
+     * The data type.
+     *
+     * Besides the values listed in {@link OperandType}, any value above
+     * {@link OperandTypeRange::BASE_MAX} is possible and should be interpreted
+     * as an extension type according to {@link Model::extensionNameToPrefix}.
+     */
+    OperandType type;
+
+    /**
+     * Dimensions of the operand.
+     *
+     * For a scalar operand, dimensions.size() must be 0.
+     *
+     * A tensor operand with all dimensions specified has "fully
+     * specified" dimensions. Whenever possible (i.e., whenever the
+     * dimensions are known at model construction time), a tensor
+     * operand should have (but is not required to have) fully
+     * specified dimensions, in order to enable the best possible
+     * performance.
+     *
+     * If a tensor operand's dimensions are not fully specified, the
+     * dimensions of the operand are deduced from the operand
+     * dimensions and values of the operation for which that operand
+     * is an output.
+     *
+     * In the following situations, a tensor operand's dimensions must
+     * be fully specified:
+     *
+     *     . The operand has lifetime CONSTANT_COPY or
+     *       CONSTANT_REFERENCE.
+     *
+     *     . The operand has lifetime MODEL_INPUT. Fully
+     *       specified dimensions must either be present in the
+     *       Operand or they must be provided in the corresponding
+     *       RequestArgument.
+     *       EXCEPTION: If the input is optional and omitted
+     *       (by setting the hasNoValue field of the corresponding
+     *       RequestArgument to true) then it need not have fully
+     *       specified dimensions.
+     *
+     * A tensor operand with some number of unspecified dimensions is
+     * represented by setting each unspecified dimension to 0.
+     *
+     * A tensor operand with unspecified rank is represented by providing
+     * an empty dimensions vector.
+     */
+    vec<uint32_t> dimensions;
+
+    /**
+     * The number of times this operand appears as an operation input.
+     *
+     * (For example, if this operand appears once in one operation's
+     * input list, and three times in another operation's input list,
+     * then numberOfConsumers = 4.)
+     */
+    uint32_t numberOfConsumers;
+
+    /**
+     * Quantized scale of the operand.
+     *
+     * Only applicable if the operand is of type TENSOR_QUANT8_ASYMM or
+     * TENSOR_INT32.
+     */
+    float scale;
+
+    /**
+     * Quantized zero-point offset of the operand.
+     *
+     * Only applicable if the operand is of type TENSOR_QUANT8_ASYMM.
+     */
+    int32_t zeroPoint;
+
+    /**
+     * How the operand is used.
+     */
+    OperandLifeTime lifetime;
+
+    /**
+     * Where to find the data for this operand.
+     * If the lifetime is TEMPORARY_VARIABLE, MODEL_INPUT, MODEL_OUTPUT, or
+     * NO_VALUE:
+     * - All the fields must be 0.
+     * If the lifetime is CONSTANT_COPY:
+     * - location.poolIndex is 0.
+     * - location.offset is the offset in bytes into Model.operandValues.
+     * - location.length is set.
+     * If the lifetime is CONSTANT_REFERENCE:
+     * - location.poolIndex is set.
+     * - location.offset is the offset in bytes into the specified pool.
+     * - location.length is set.
+     */
+    DataLocation location;
+
+    /**
+     * Additional parameters specific to a particular operand type.
+     */
+    safe_union ExtraParams {
+       /**
+        * No additional parameters.
+        */
+       Monostate none;
+
+       /**
+        * Symmetric per-channel quantization parameters.
+        *
+        * Only applicable to operands of type TENSOR_QUANT8_SYMM_PER_CHANNEL.
+        */
+       SymmPerChannelQuantParams channelQuant;
+
+       /**
+        * Extension operand parameters.
+        *
+        * The framework treats this as an opaque data blob.
+        * The format is up to individual extensions.
+        */
+       vec<uint8_t> extension;
+    } extraParams;
+};
+
+/**
+ * A Neural Network Model.
+ *
+ * This includes not only the execution graph, but also constant data such as
+ * weights or scalars added at construction time. The only information that
+ * may not be known is the shape of the input tensors.
+ */
+struct Model {
+    /**
+     * All operands included in the model.
+     */
+    vec<Operand> operands;
+
+    /**
+     * All operations included in the model.
+     *
+     * The operations are sorted into execution order. Every operand
+     * with lifetime MODEL_OUTPUT or TEMPORARY_VARIABLE must be
+     * written before it is read.
+     */
+    vec<Operation> operations;
+
+    /**
+     * Input indexes of the model. There must be at least one.
+     *
+     * Each value corresponds to the index of the operand in "operands".
+     */
+    vec<uint32_t> inputIndexes;
+
+    /**
+     * Output indexes of the model. There must be at least one.
+     *
+     * Each value corresponds to the index of the operand in "operands".
+     */
+    vec<uint32_t> outputIndexes;
+
+    /**
+     * A byte buffer containing operand data that were copied into the model.
+     *
+     * An operand's value must be located here if and only if Operand::lifetime
+     * equals OperandLifeTime::CONSTANT_COPY.
+     */
+    vec<uint8_t> operandValues;
+
+    /**
+     * A collection of shared memory pools containing operand values.
+     *
+     * An operand's value must be located here if and only if Operand::lifetime
+     * equals OperandLifeTime::CONSTANT_REFERENCE.
+     */
+    vec<memory> pools;
+
+    /**
+     * 'true' indicates TENSOR_FLOAT32 may be calculated with range and/or
+     * precision as low as that of the IEEE 754 16-bit floating-point format.
+     * 'false' indicates TENSOR_FLOAT32 must be calculated using at least the
+     * range and precision of the IEEE 754 32-bit floating-point format.
+     */
+    bool relaxComputationFloat32toFloat16;
+
+    /**
+     * The mapping between extension names and prefixes of operand and
+     * operation type values.
+     *
+     * An operand or operation whose numeric type value is above
+     * {@link OperandTypeRange::BASE_MAX} or
+     * {@link OperationTypeRange::BASE_MAX} respectively should be interpreted
+     * as an extension operand. The low
+     * {@link Model::ExtensionTypeEncoding::LOW_BITS_TYPE} bits of the value
+     * correspond to the type ID within the extension and the high
+     * {@link Model::ExtensionTypeEncoding::HIGH_BITS_PREFIX} bits encode
+     * the "prefix", which maps uniquely to the extension name.
+     *
+     * For example, if a model contains an operation whose value is
+     * 0xAAAABBBB and extensionNameToPrefix contains an entry with
+     * prefix=0xAAAA and name="vendor.test.test_extension", then
+     * the operation should be interpreted as the operation 0xBBBB
+     * of the extension named vendor.test.test_extension.
+     *
+     * This is a one-to-one correspondence. That is, there must be at most one
+     * prefix corresponding to each extension name and at most one extension
+     * name corresponding to each prefix.
+     */
+    vec<ExtensionNameAndPrefix> extensionNameToPrefix;
+
+    /**
+     * A correspondence between an extension name and a prefix of operand and
+     * operation type values.
+     */
+    struct ExtensionNameAndPrefix {
+        /**
+         * The extension name.
+         *
+         * See {@link Extension::name} for the format specification.
+         */
+        string name;
+
+        /**
+         * The unique extension identifier within the model.
+         *
+         * See {@link Model::extensionNameToPrefix}.
+         */
+        uint16_t prefix;
+    };
+
+    /**
+     * Numeric values of extension operand and operation types have the
+     * following structure:
+     * - 16 high bits represent the "prefix", which corresponds uniquely to the
+     *   extension name.
+     * - 16 low bits represent the type ID within the extension.
+     */
+    enum ExtensionTypeEncoding : uint8_t {
+        HIGH_BITS_PREFIX = 16,
+        LOW_BITS_TYPE = 16,
+    };
+};
+
+/**
+ * Describes the shape information of an output operand after execution.
+ */
+struct OutputShape {
+    /**
+     * Dimensions of the operand.
+     */
+    vec<uint32_t> dimensions;
+
+    /**
+     * Whether the provided buffer size is sufficient for the output.
+     */
+    bool isSufficient;
+};
+
+/**
+ * Specifies whether or not to measure timing information during execution.
+ */
+enum MeasureTiming : int32_t {
+    NO  = 0,
+    YES = 1,
+};
+
+/**
+
+ * Timing information measured during execution. Each time is a duration from
+ * the beginning of some task to the end of that task, including time when that
+ * task is not active (for example, preempted by some other task, or
+ * waiting for some resource to become available).
+ *
+ * Times are measured in microseconds.
+ * When a time is not available, it must be reported as UINT64_MAX.
+ */
+struct Timing {
+    /** Execution time on device (not driver, which runs on host processor). */
+    uint64_t timeOnDevice;
+    /** Execution time in driver (including time on device). */
+    uint64_t timeInDriver;
+};
+
+/**
+ * FmqRequestDatum is a single element of a serialized representation of an
+ * execution request (a {@link @1.0::Request} object and a {@link MeasureTiming}
+ * value) which is sent across FastMessageQueue.
+ *
+ * The serialized representation for a particular execution is referred to later
+ * in these descriptions as a 'packet'.
+ *
+ * FastMessageQueue can only pass HIDL-defined types that do not involve nested
+ * buffers, handles, or interfaces.
+ *
+ * The request is serialized as follows:
+ * 1) 'packetInformation'
+ * 2) For each input operand:
+ *    2.1) 'inputOperandInformation'
+ *    2.2) For each dimension element of the operand:
+ *         2.2.1) 'inputOperandDimensionValue'
+ * 3) For each output operand:
+ *    3.1) 'outputOperandInformation'
+ *    3.2) For each dimension element of the operand:
+ *         3.2.1) 'outputOperandDimensionValue'
+ * 4) For each pool:
+ *    4.1) 'poolIdentifier'
+ * 5) 'measureTiming'
+ */
+safe_union FmqRequestDatum {
+    /**
+     * Type to describe the high-level layout of the packet.
+     */
+    struct PacketInformation {
+        /**
+         * How many elements the packet contains, including the
+         * "packetInformation" datum.
+         */
+        uint32_t packetSize;
+
+        /**
+         * Number of input operands.
+         */
+        uint32_t numberOfInputOperands;
+
+        /**
+         * Number of output operands.
+         */
+        uint32_t numberOfOutputOperands;
+
+        /**
+         * Number of pool identifiers.
+         */
+        uint32_t numberOfPools;
+    };
+
+    /**
+     * Type representing the information for each operand.
+     */
+    struct OperandInformation {
+        /**
+         * If true, the argument does not have a value. This can be used for
+         * operations that take optional arguments. If true, the fields of
+         * 'location' are set to 0, 'numberOfDimensions' is set to 0,  and the
+         * dimensions information is omitted from the serialization.
+         */
+        bool hasNoValue;
+
+        /**
+         * The location within one of the memory pools passed in the Request.
+         */
+        DataLocation location;
+
+        /**
+         * Number of subsequent elements that belong to the dimensions vector.
+         */
+        uint32_t numberOfDimensions;
+    };
+
+    /**
+     * packetInformation is the first element of the packet and describes the
+     * remainder of the packet.
+     */
+    PacketInformation packetInformation;
+
+    /**
+     * Information for each input operand.
+     */
+    OperandInformation inputOperandInformation;
+
+    /**
+     * Element of the dimensions vector.
+     */
+    uint32_t inputOperandDimensionValue;
+
+    /**
+     * Information for each output operand.
+     */
+    OperandInformation outputOperandInformation;
+
+    /**
+     * Element of the dimensions vector.
+     */
+    uint32_t outputOperandDimensionValue;
+
+    /**
+     * Unique identifier for a pool.
+     *
+     * A {@link @1.0::Request} passes across one or more pools of shared memory
+     * for the inputs and outputs of an execution. However, these memory pools
+     * are not able to be sent across FastMessageQueue directly. Instead, the
+     * producing side of the FMQ represents each different pool with a unique
+     * identifier, and sends this identifier across the FMQ. Whenever the
+     * consuming side of the FMQ needs the memory corresponding to this unique
+     * identifier, it can pass the identifier to
+     * {@link IBurstCallback::getMemories} to retreive the memory. Although this
+     * HIDL Binder call is expensive compared to communication across FMQ, it is
+     * only needed in the cases when the consumer does not recognize the unique
+     * identifier.
+     */
+    int32_t poolIdentifier;
+
+    /**
+     * Specifies whether or not to measure duration of the execution. The
+     * duration runs from the time the driver dequeues the request from a
+     * FastMessageQueue to the time the driver enqueues results to a
+     * FastMessageQueue.
+     */
+    MeasureTiming measureTiming;
+};
+
+/**
+ * FmqResultDatum is a single element of a serialized representation of the
+ * values returned from an execution ({@link @1.0::ErrorStatus},
+ * vec<{@link OutputShape}>, and {@link Timing}) which is returned via
+ * FastMessageQueue.
+ *
+ * The serialized representation for a particular execution is referred to later
+ * in these descriptions as a 'packet'.
+ *
+ * FastMessageQueue can only pass HIDL-defined types that do not involve nested
+ * buffers, handles, or interfaces.
+ *
+ * The execution return values ({@link @1.0::ErrorStatus} and
+ * vec<{@link OutputShape}>) are serialized as follows:
+ * 1) 'packetInformation'
+ * 2) For each returned operand:
+ *    2.1) 'operandInformation'
+ *    2.2) For each dimension element of the operand:
+ *         2.2.1) 'operandDimensionValue'
+ * 3) 'executionTiming'
+ */
+safe_union FmqResultDatum {
+    /**
+     * Type to describe the high-level layout of the packet.
+     */
+    struct PacketInformation {
+        /**
+         * How many elements the packet contains, including the
+         * "packetInformation" datum.
+         */
+        uint32_t packetSize;
+
+        /**
+         * Status of the execution.
+         */
+        ErrorStatus errorStatus;
+
+        /**
+         * Number of returned operands.
+         */
+        uint32_t numberOfOperands;
+    };
+
+    /**
+     * Type representing the information for each operand.
+     */
+    struct OperandInformation {
+        /**
+         * Indicates whether the operand's output buffer is large enough to
+         * store the operand's result data.
+         */
+        bool isSufficient;
+
+        /**
+         * Number of subsequent elements that belong to the dimensions vector.
+         */
+        uint32_t numberOfDimensions;
+    };
+
+    /**
+     * packetInformation is the first element of the packet and describes the
+     * remainder of the packet. It additionally includes the status of the
+     * execution.
+     */
+    PacketInformation packetInformation;
+
+    /**
+     * Information for each returned operand.
+     */
+    OperandInformation operandInformation;
+
+    /**
+     * Element of the dimensions vector.
+     */
+    uint32_t operandDimensionValue;
+
+    /**
+     * Duration of execution. Unless measurement was requested and execution
+     * succeeds, all times must be reported as UINT64_MAX. A driver may choose
+     * to report any time as UINT64_MAX, indicating that measurement is not
+     * available.
+     */
+    Timing executionTiming;
+};
+
+/**
+ * Information about an extension.
+ */
+struct Extension {
+    /**
+     * The extension name.
+     *
+     * The name must consist of lowercase latin letters, numbers, periods, and
+     * underscore signs. The name must contain at least one period.
+     *
+     * The name must start with the reverse domain name of the vendor.
+     *
+     * Example: com.google.test_extension
+     */
+    string name;
+
+    /**
+     * Information about an extension operand type.
+     */
+    struct OperandTypeInformation {
+        /**
+         * The extension operand type.
+         */
+        uint16_t type;
+
+        /**
+         * Indicates whether the extension operand type represents a tensor or
+         * a scalar.
+         */
+        bool isTensor;
+
+        /**
+         * The byte size of the operand (if scalar) or of a single element (if
+         * tensor).
+         */
+        uint32_t byteSize;
+    };
+
+    /**
+     * Information about operand types defined by the extension.
+     */
+    vec<OperandTypeInformation> operandTypes;
+};
diff --git a/power/stats/1.0/default/PowerStats.cpp b/power/stats/1.0/default/PowerStats.cpp
index 78766f2..68275ce 100644
--- a/power/stats/1.0/default/PowerStats.cpp
+++ b/power/stats/1.0/default/PowerStats.cpp
@@ -87,7 +87,7 @@
     std::string railFileName;
     std::string spsFileName;
     uint32_t index = 0;
-    uint32_t samplingRate;
+    unsigned long samplingRate;
     for (const auto& path : mPm.devicePaths) {
         railFileName = path + "/enabled_rails";
         spsFileName = path + "/sampling_rate";
@@ -109,10 +109,11 @@
         while (std::getline(railNames, line)) {
             std::vector<std::string> words = android::base::Split(line, ":");
             if (words.size() == 2) {
-                mPm.railsInfo.emplace(words[0], RailData{.devicePath = path,
-                                                         .index = index,
-                                                         .subsysName = words[1],
-                                                         .samplingRate = samplingRate});
+                mPm.railsInfo.emplace(
+                        words[0], RailData{.devicePath = path,
+                                           .index = index,
+                                           .subsysName = words[1],
+                                           .samplingRate = static_cast<uint32_t>(samplingRate)});
                 index++;
             } else {
                 ALOGW("Unexpected format in file: %s", railFileName.c_str());
diff --git a/radio/1.2/types.hal b/radio/1.2/types.hal
index dffebd3..f10d753 100644
--- a/radio/1.2/types.hal
+++ b/radio/1.2/types.hal
@@ -161,7 +161,8 @@
     ScanType type;
 
     /**
-     * Time interval in seconds between periodic scans, only valid when type = PERIODIC
+     * Time interval in seconds between the completion of one scan and the start of a subsequent scan.
+     * This field is only valid when 'type' is 'PERIODIC'.
      * Range: ScanIntervalRange:MIN to ScanIntervalRange:MAX
      */
     int32_t interval;
diff --git a/radio/1.2/vts/functional/radio_hidl_hal_api.cpp b/radio/1.2/vts/functional/radio_hidl_hal_api.cpp
index 5184ef9..a98f22a 100644
--- a/radio/1.2/vts/functional/radio_hidl_hal_api.cpp
+++ b/radio/1.2/vts/functional/radio_hidl_hal_api.cpp
@@ -46,7 +46,10 @@
     ::android::hardware::radio::V1_2::NetworkScanRequest request = {
             .type = ScanType::ONE_SHOT,
             .interval = 60,
-            .specifiers = {::GERAN_SPECIFIER_P900, ::GERAN_SPECIFIER_850}};
+            .specifiers = {::GERAN_SPECIFIER_P900, ::GERAN_SPECIFIER_850},
+            .maxSearchTime = 60,
+            .incrementalResults = false,
+            .incrementalResultsPeriodicity = 1};
 
     Return<void> res = radio_v1_2->startNetworkScan_1_2(serial, request);
     ASSERT_OK(res);
diff --git a/radio/1.4/vts/functional/radio_hidl_hal_api.cpp b/radio/1.4/vts/functional/radio_hidl_hal_api.cpp
index b2d19a2..d3012bb 100644
--- a/radio/1.4/vts/functional/radio_hidl_hal_api.cpp
+++ b/radio/1.4/vts/functional/radio_hidl_hal_api.cpp
@@ -178,7 +178,12 @@
                                       .channels = {1, 2}};
 
     ::android::hardware::radio::V1_2::NetworkScanRequest request = {
-            .type = ScanType::ONE_SHOT, .interval = 60, .specifiers = {specifier}};
+            .type = ScanType::ONE_SHOT,
+            .interval = 60,
+            .specifiers = {specifier},
+            .maxSearchTime = 60,
+            .incrementalResults = false,
+            .incrementalResultsPeriodicity = 1};
 
     Return<void> res = radio_v1_4->startNetworkScan_1_4(serial, request);
     ASSERT_OK(res);
diff --git a/tv/tuner/1.0/IDemux.hal b/tv/tuner/1.0/IDemux.hal
index e03095b..7ead34b 100644
--- a/tv/tuner/1.0/IDemux.hal
+++ b/tv/tuner/1.0/IDemux.hal
@@ -1,3 +1,19 @@
+/*
+ * Copyright (C) 2019 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
 package android.hardware.tv.tuner@1.0;
 
 import IDemuxCallback;
@@ -236,7 +252,7 @@
      *         INVALID_STATE if failed for wrong state.
      *         UNKNOWN_ERROR if failed for other reasons.
      */
-    attachOutputTsFilter(DemuxFilterId filterId) generates (Result result);
+    attachOutputFilter(DemuxFilterId filterId) generates (Result result);
 
     /**
      * Detach one filter from the demux's output.
@@ -250,7 +266,7 @@
      *         INVALID_STATE if failed for wrong state.
      *         UNKNOWN_ERROR if failed for other reasons.
      */
-    detachOutputTsFilter(DemuxFilterId filterId) generates (Result result);
+    detachOutputFilter(DemuxFilterId filterId) generates (Result result);
 
     /**
      * Start to take data to the demux's output.
diff --git a/tv/tuner/1.0/IDemuxCallback.hal b/tv/tuner/1.0/IDemuxCallback.hal
index 55e8420..7bce9ef 100644
--- a/tv/tuner/1.0/IDemuxCallback.hal
+++ b/tv/tuner/1.0/IDemuxCallback.hal
@@ -1,3 +1,19 @@
+/*
+ * Copyright (C) 2019 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
 package android.hardware.tv.tuner@1.0;
 
 interface IDemuxCallback {
diff --git a/tv/tuner/1.0/IDescrambler.hal b/tv/tuner/1.0/IDescrambler.hal
index d078657..61ff1df 100644
--- a/tv/tuner/1.0/IDescrambler.hal
+++ b/tv/tuner/1.0/IDescrambler.hal
@@ -1,3 +1,19 @@
+/*
+ * Copyright (C) 2019 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
 package android.hardware.tv.tuner@1.0;
 /**
  * Descrambler is used to descramble input data.
diff --git a/tv/tuner/1.0/IFrontend.hal b/tv/tuner/1.0/IFrontend.hal
index f9f78ef..962e4ba 100644
--- a/tv/tuner/1.0/IFrontend.hal
+++ b/tv/tuner/1.0/IFrontend.hal
@@ -143,7 +143,7 @@
      *         cable frontend.
      *         UNKNOWN_ERROR if failed for other reasons.
      */
-    setLnb(ILnb lnb) generates (Result result);
+    setLnb(LnbId lnbId) generates (Result result);
 
     /**
      * Enble or Disable Low Noise Amplifier (LNA).
@@ -156,22 +156,4 @@
      *         UNKNOWN_ERROR if failed for other reasons.
      */
     setLna(bool bEnable) generates (Result result);
-
-    /**
-     *  Sends DiSEqC (Digital Satellite Equipment Control) message.
-     *
-     * Client sends DiSeqc message to DiSEqc compatible device through the
-     * frontend. The response message from the device comes back to the client
-     * through frontend's callback onDiseqcMessage.
-     *
-     * @param diseqcMessage a byte array of data for DiSEqC message which is
-     *        specified by EUTELSAT Bus Functional Specification Version 4.2.
-     *
-     * @return result Result status of the operation.
-     *         SUCCESS if successful,
-     *         INVALID_STATE if the frontend can't send DiSEqc Message, such as
-     *         cable frontend.
-     *         UNKNOWN_ERROR if failed for other reasons.
-     */
-    sendDiseqcMessage(vec<uint8_t> diseqcMessage) generates (Result result);
 };
diff --git a/tv/tuner/1.0/ILnb.hal b/tv/tuner/1.0/ILnb.hal
index 49fc3b4..6b7119e 100644
--- a/tv/tuner/1.0/ILnb.hal
+++ b/tv/tuner/1.0/ILnb.hal
@@ -55,6 +55,24 @@
     setSatellitePosition(FrontendLnbPosition position) generates (Result result);
 
     /**
+     *  Sends DiSEqC (Digital Satellite Equipment Control) message.
+     *
+     * Client sends DiSeqc message to DiSEqc to LNB. The response message from
+     * the device comes back to the client through frontend's callback
+     * onDiseqcMessage.
+     *
+     * @param diseqcMessage a byte array of data for DiSEqC message which is
+     *        specified by EUTELSAT Bus Functional Specification Version 4.2.
+     *
+     * @return result Result status of the operation.
+     *         SUCCESS if successful,
+     *         INVALID_STATE if the frontend can't send DiSEqc Message, such as
+     *         cable frontend.
+     *         UNKNOWN_ERROR if failed for other reasons.
+     */
+    sendDiseqcMessage(vec<uint8_t> diseqcMessage) generates (Result result);
+
+    /**
      * Releases the LNB instance
      *
      * Associated resources are released.  close may be called more than once.
diff --git a/tv/tuner/1.0/ITuner.hal b/tv/tuner/1.0/ITuner.hal
index f1a8617..0d63442 100644
--- a/tv/tuner/1.0/ITuner.hal
+++ b/tv/tuner/1.0/ITuner.hal
@@ -68,6 +68,16 @@
          generates (Result result, DemuxId demuxId, IDemux demux);
 
     /**
+     * Retrieve the Demux's Capabilities.
+     *
+     * @return result Result status of the operation.
+     *         SUCCESS if successful,
+     *         UNKNOWN_ERROR if the inquiry failed for other reasons.
+     * @return caps the Demux's Capabilities.
+     */
+    getDemuxCaps() generates (Result result, DemuxCapabilities caps);
+
+    /**
      * Create a new instance of Descrambler.
      *
      * It is used by the client to create a Descrambler instance.
@@ -81,14 +91,13 @@
          generates (Result result, IDescrambler descrambler);
 
     /**
-     * Create a new instance of Descrambler.
+     * Retrieve the frontend's information.
      *
-     * It is used by the client to create a Descrambler instance.
-     *
+     * @param frontendId the id of the frontend to be inquiried.
      * @return result Result status of the operation.
      *         SUCCESS if successful,
-     *         UNKNOWN_ERROR if creation failed for other reasons.
-     * @return descrambler the newly created descrambler interface.
+     *         UNKNOWN_ERROR if the inquiry failed for other reasons.
+     * @return info the frontend's information.
      */
     getFrontendInfo(FrontendId frontendId)
         generates (Result result, FrontendInfo info);
@@ -119,6 +128,5 @@
      */
     openLnbById(LnbId lnbId)
         generates (Result result, ILnb lnb);
-
 };
 
diff --git a/tv/tuner/1.0/default/Demux.cpp b/tv/tuner/1.0/default/Demux.cpp
index 889e42e..04382b0 100644
--- a/tv/tuner/1.0/default/Demux.cpp
+++ b/tv/tuner/1.0/default/Demux.cpp
@@ -67,8 +67,9 @@
         0x73, 0x63, 0x65, 0x6e, 0x65,
 };
 
-Demux::Demux(uint32_t demuxId) {
+Demux::Demux(uint32_t demuxId, sp<Tuner> tuner) {
     mDemuxId = demuxId;
+    mTunerService = tuner;
 }
 
 Demux::~Demux() {}
@@ -76,9 +77,20 @@
 Return<Result> Demux::setFrontendDataSource(uint32_t frontendId) {
     ALOGV("%s", __FUNCTION__);
 
-    mSourceFrontendId = frontendId;
+    if (mTunerService == nullptr) {
+        return Result::NOT_INITIALIZED;
+    }
 
-    return Result::SUCCESS;
+    mFrontend = mTunerService->getFrontendById(frontendId);
+
+    if (mFrontend == nullptr) {
+        return Result::INVALID_STATE;
+    }
+
+    mFrontendSourceFile = mFrontend->getSourceFile();
+
+    mTunerService->setFrontendAsDemuxSource(frontendId, mDemuxId);
+    return startBroadcastInputLoop();
 }
 
 Return<void> Demux::addFilter(DemuxFilterType type, uint32_t bufferSize,
@@ -100,6 +112,8 @@
         mFilterEventFlags.resize(filterId + 1);
         mFilterThreadRunning.resize(filterId + 1);
         mFilterThreads.resize(filterId + 1);
+        mFilterPids.resize(filterId + 1);
+        mFilterOutputs.resize(filterId + 1);
     }
 
     mUsedFilterIds.insert(filterId);
@@ -142,10 +156,34 @@
     return Void();
 }
 
-Return<Result> Demux::configureFilter(uint32_t /* filterId */,
-                                      const DemuxFilterSettings& /* settings */) {
+Return<Result> Demux::configureFilter(uint32_t filterId, const DemuxFilterSettings& settings) {
     ALOGV("%s", __FUNCTION__);
 
+    switch (mFilterEvents[filterId].filterType) {
+        case DemuxFilterType::SECTION:
+            mFilterPids[filterId] = settings.section().tpid;
+            break;
+        case DemuxFilterType::PES:
+            mFilterPids[filterId] = settings.pesData().tpid;
+            break;
+        case DemuxFilterType::TS:
+            mFilterPids[filterId] = settings.ts().tpid;
+            break;
+        case DemuxFilterType::AUDIO:
+            mFilterPids[filterId] = settings.audio().tpid;
+            break;
+        case DemuxFilterType::VIDEO:
+            mFilterPids[filterId] = settings.video().tpid;
+            break;
+        case DemuxFilterType::RECORD:
+            mFilterPids[filterId] = settings.record().tpid;
+            break;
+        case DemuxFilterType::PCR:
+            mFilterPids[filterId] = settings.pcr().tpid;
+            break;
+        default:
+            return Result::UNKNOWN_ERROR;
+    }
     return Result::SUCCESS;
 }
 
@@ -158,36 +196,18 @@
         return Result::INVALID_ARGUMENT;
     }
 
-    switch (mFilterEvents[filterId].filterType) {
-        case DemuxFilterType::SECTION:
-            result = startFilterLoop(filterId);
-            break;
-        case DemuxFilterType::PES:
-            result = startPesFilterHandler(filterId);
-            break;
-        case DemuxFilterType::TS:
-            result = startTsFilterHandler();
-            return Result::SUCCESS;
-        case DemuxFilterType::AUDIO:
-        case DemuxFilterType::VIDEO:
-            result = startMediaFilterHandler(filterId);
-            break;
-        case DemuxFilterType::RECORD:
-            result = startRecordFilterHandler(filterId);
-            break;
-        case DemuxFilterType::PCR:
-            result = startPcrFilterHandler();
-            return Result::SUCCESS;
-        default:
-            return Result::UNKNOWN_ERROR;
-    }
+    result = startFilterLoop(filterId);
 
     return result;
 }
 
-Return<Result> Demux::stopFilter(uint32_t /* filterId */) {
+Return<Result> Demux::stopFilter(uint32_t filterId) {
     ALOGV("%s", __FUNCTION__);
 
+    mFilterThreadRunning[filterId] = false;
+
+    std::lock_guard<std::mutex> lock(mFilterThreadLock);
+
     return Result::SUCCESS;
 }
 
@@ -238,6 +258,8 @@
     mFilterMQs.clear();
     mFilterEvents.clear();
     mFilterEventFlags.clear();
+    mFilterOutputs.clear();
+    mFilterPids.clear();
     mLastUsedFilterId = -1;
 
     return Result::SUCCESS;
@@ -277,19 +299,21 @@
     return Void();
 }
 
-Return<Result> Demux::configureOutput(const DemuxOutputSettings& /* settings */) {
+Return<Result> Demux::configureOutput(const DemuxOutputSettings& settings) {
+    ALOGV("%s", __FUNCTION__);
+
+    mOutputConfigured = true;
+    mOutputSettings = settings;
+    return Result::SUCCESS;
+}
+
+Return<Result> Demux::attachOutputFilter(uint32_t /*filterId*/) {
     ALOGV("%s", __FUNCTION__);
 
     return Result::SUCCESS;
 }
 
-Return<Result> Demux::attachOutputTsFilter(uint32_t /*filterId*/) {
-    ALOGV("%s", __FUNCTION__);
-
-    return Result::SUCCESS;
-}
-
-Return<Result> Demux::detachOutputTsFilter(uint32_t /* filterId */) {
+Return<Result> Demux::detachOutputFilter(uint32_t /* filterId */) {
     ALOGV("%s", __FUNCTION__);
 
     return Result::SUCCESS;
@@ -353,15 +377,26 @@
     return Void();
 }
 
-Return<Result> Demux::configureInput(const DemuxInputSettings& /* settings */) {
+Return<Result> Demux::configureInput(const DemuxInputSettings& settings) {
     ALOGV("%s", __FUNCTION__);
 
+    mInputConfigured = true;
+    mInputSettings = settings;
+
     return Result::SUCCESS;
 }
 
 Return<Result> Demux::startInput() {
     ALOGV("%s", __FUNCTION__);
 
+    if (!mInputCallback) {
+        return Result::NOT_INITIALIZED;
+    }
+
+    if (!mInputConfigured) {
+        return Result::INVALID_STATE;
+    }
+
     pthread_create(&mInputThread, NULL, __threadLoopInput, this);
     pthread_setname_np(mInputThread, "demux_input_waiting_loop");
 
@@ -373,6 +408,10 @@
 Return<Result> Demux::stopInput() {
     ALOGV("%s", __FUNCTION__);
 
+    mInputThreadRunning = false;
+
+    std::lock_guard<std::mutex> lock(mInputThreadLock);
+
     return Result::SUCCESS;
 }
 
@@ -403,36 +442,52 @@
     return Result::SUCCESS;
 }
 
-Result Demux::startSectionFilterHandler(uint32_t filterId, vector<uint8_t> data) {
-    if (!writeSectionsAndCreateEvent(filterId, data)) {
+Result Demux::startSectionFilterHandler(uint32_t filterId) {
+    if (mFilterOutputs[filterId].empty()) {
+        return Result::SUCCESS;
+    }
+    if (!writeSectionsAndCreateEvent(filterId, mFilterOutputs[filterId])) {
         ALOGD("[Demux] filter %d fails to write into FMQ. Ending thread", filterId);
         return Result::UNKNOWN_ERROR;
     }
 
+    mFilterOutputs[filterId].clear();
+
     return Result::SUCCESS;
 }
 
 Result Demux::startPesFilterHandler(uint32_t filterId) {
-    // TODO generate multiple events in one event callback
+    std::lock_guard<std::mutex> lock(mFilterEventLock);
     DemuxFilterPesEvent pesEvent;
-    pesEvent = {
-            // temp dump meta data
-            .streamId = 0,
-            .dataLength = 530,
-    };
-    mFilterEvents[filterId].events.resize(1);
-    mFilterEvents[filterId].events[0].pes(pesEvent);
-    /*pthread_create(&mThreadId, NULL, __threadLoop, this);
-    pthread_setname_np(mThreadId, "demux_section_filter_waiting_loop");*/
-    if (!writeDataToFilterMQ(fakeDataInputBuffer, filterId)) {
-        return Result::INVALID_STATE;
+    if (mFilterOutputs[filterId].empty()) {
+        return Result::SUCCESS;
     }
 
-    if (mDemuxCallbacks[filterId] == nullptr) {
-        return Result::NOT_INITIALIZED;
+    for (int i = 0; i < mFilterOutputs[filterId].size(); i += 188) {
+        uint8_t pusi = mFilterOutputs[filterId][i + 1] & 0x40;
+        uint8_t adaptFieldControl = (mFilterOutputs[filterId][i + 3] & 0x30) >> 4;
+        ALOGD("[Demux] pusi %d, adaptFieldControl %d", pusi, adaptFieldControl);
+        if (pusi && (adaptFieldControl == 0x01)) {
+            vector<uint8_t>::const_iterator first = mFilterOutputs[filterId].begin() + i + 4;
+            vector<uint8_t>::const_iterator last = mFilterOutputs[filterId].begin() + i + 187;
+            vector<uint8_t> filterOutData(first, last);
+            if (!writeDataToFilterMQ(filterOutData, filterId)) {
+                mFilterOutputs[filterId].clear();
+                return Result::INVALID_STATE;
+            }
+            pesEvent = {
+                    // temp dump meta data
+                    .streamId = filterOutData[3],
+                    .dataLength = static_cast<uint16_t>(filterOutData.size()),
+            };
+            int size = mFilterEvents[filterId].events.size();
+            mFilterEvents[filterId].events.resize(size + 1);
+            mFilterEvents[filterId].events[size].pes(pesEvent);
+        }
     }
 
-    mDemuxCallbacks[filterId]->onFilterEvent(mFilterEvents[filterId]);
+    mFilterOutputs[filterId].clear();
+
     return Result::SUCCESS;
 }
 
@@ -451,6 +506,8 @@
     };
     mFilterEvents[filterId].events.resize(1);
     mFilterEvents[filterId].events[0].media() = mediaEvent;
+
+    mFilterOutputs[filterId].clear();
     // TODO handle write FQM for media stream
     return Result::SUCCESS;
 }
@@ -465,6 +522,8 @@
     recordEvent.indexMask.tsIndexMask() = 0x01;
     mFilterEvents[filterId].events.resize(1);
     mFilterEvents[filterId].events[0].ts() = recordEvent;
+
+    mFilterOutputs[filterId].clear();
     return Result::SUCCESS;
 }
 
@@ -499,18 +558,18 @@
 bool Demux::writeSectionsAndCreateEvent(uint32_t filterId, vector<uint8_t> data) {
     // TODO check how many sections has been read
     std::lock_guard<std::mutex> lock(mFilterEventLock);
-    int size = mFilterEvents[filterId].events.size();
-    mFilterEvents[filterId].events.resize(size + 1);
     if (!writeDataToFilterMQ(data, filterId)) {
         return false;
     }
+    int size = mFilterEvents[filterId].events.size();
+    mFilterEvents[filterId].events.resize(size + 1);
     DemuxFilterSectionEvent secEvent;
     secEvent = {
             // temp dump meta data
             .tableId = 0,
             .version = 1,
             .sectionNum = 1,
-            .dataLength = 530,
+            .dataLength = static_cast<uint16_t>(data.size()),
     };
     mFilterEvents[filterId].events[size].section(secEvent);
     return true;
@@ -524,21 +583,44 @@
     return false;
 }
 
-bool Demux::filterAndOutputData() {
-    ALOGD("[Demux] start to dispatch data to filters");
+bool Demux::readInputFMQ() {
     // Read input data from the input FMQ
     int size = mInputMQ->availableToRead();
+    int inputPacketSize = mInputSettings.packetSize;
     vector<uint8_t> dataOutputBuffer;
-    dataOutputBuffer.resize(size);
-    mInputMQ->read(dataOutputBuffer.data(), size);
+    dataOutputBuffer.resize(inputPacketSize);
 
-    Result result;
-    // Filter the data and feed the output to each filter
+    // Dispatch the packet to the PID matching filter output buffer
+    for (int i = 0; i < size / inputPacketSize; i++) {
+        if (!mInputMQ->read(dataOutputBuffer.data(), inputPacketSize)) {
+            return false;
+        }
+        startTsFilter(dataOutputBuffer);
+    }
+
+    return true;
+}
+
+void Demux::startTsFilter(vector<uint8_t> data) {
     set<uint32_t>::iterator it;
     for (it = mUsedFilterIds.begin(); it != mUsedFilterIds.end(); it++) {
+        uint16_t pid = ((data[1] & 0x1f) << 8) | ((data[2] & 0xff));
+        ALOGW("start ts filter pid: %d", pid);
+        if (pid == mFilterPids[*it]) {
+            mFilterOutputs[*it].insert(mFilterOutputs[*it].end(), data.begin(), data.end());
+        }
+    }
+}
+
+bool Demux::startFilterDispatcher() {
+    Result result;
+    set<uint32_t>::iterator it;
+
+    // Handle the output data per filter type
+    for (it = mUsedFilterIds.begin(); it != mUsedFilterIds.end(); it++) {
         switch (mFilterEvents[*it].filterType) {
             case DemuxFilterType::SECTION:
-                result = startSectionFilterHandler(*it, dataOutputBuffer);
+                result = startSectionFilterHandler(*it);
                 break;
             case DemuxFilterType::PES:
                 result = startPesFilterHandler(*it);
@@ -578,6 +660,7 @@
 
 void Demux::filterThreadLoop(uint32_t filterId) {
     ALOGD("[Demux] filter %d threadLoop start.", filterId);
+    std::lock_guard<std::mutex> lock(mFilterThreadLock);
     mFilterThreadRunning[filterId] = true;
 
     // For the first time of filter output, implementation needs to send the filter
@@ -640,6 +723,7 @@
 
 void Demux::inputThreadLoop() {
     ALOGD("[Demux] input threadLoop start.");
+    std::lock_guard<std::mutex> lock(mInputThreadLock);
     mInputThreadRunning = true;
 
     while (mInputThreadRunning) {
@@ -653,16 +737,110 @@
         }
         // Our current implementation filter the data and write it into the filter FMQ immedaitely
         // after the DATA_READY from the VTS/framework
-        if (!filterAndOutputData()) {
+        if (!readInputFMQ() || !startFilterDispatcher()) {
             ALOGD("[Demux] input data failed to be filtered. Ending thread");
             break;
         }
+
+        maySendInputStatusCallback();
     }
 
     mInputThreadRunning = false;
     ALOGD("[Demux] input thread ended.");
 }
 
+void Demux::maySendInputStatusCallback() {
+    std::lock_guard<std::mutex> lock(mInputStatusLock);
+    int availableToRead = mInputMQ->availableToRead();
+    int availableToWrite = mInputMQ->availableToWrite();
+
+    DemuxInputStatus newStatus =
+            checkStatusChange(availableToWrite, availableToRead, mInputSettings.highThreshold,
+                              mInputSettings.lowThreshold);
+    if (mIntputStatus != newStatus) {
+        mInputCallback->onInputStatus(newStatus);
+        mIntputStatus = newStatus;
+    }
+}
+
+DemuxInputStatus Demux::checkStatusChange(uint32_t availableToWrite, uint32_t availableToRead,
+                                          uint32_t highThreshold, uint32_t lowThreshold) {
+    if (availableToWrite == 0) {
+        return DemuxInputStatus::SPACE_FULL;
+    } else if (availableToRead > highThreshold) {
+        return DemuxInputStatus::SPACE_ALMOST_FULL;
+    } else if (availableToRead < lowThreshold) {
+        return DemuxInputStatus::SPACE_ALMOST_EMPTY;
+    } else if (availableToRead == 0) {
+        return DemuxInputStatus::SPACE_EMPTY;
+    }
+    return mIntputStatus;
+}
+
+Result Demux::startBroadcastInputLoop() {
+    pthread_create(&mBroadcastInputThread, NULL, __threadLoopBroadcast, this);
+    pthread_setname_np(mBroadcastInputThread, "broadcast_input_thread");
+
+    return Result::SUCCESS;
+}
+
+void* Demux::__threadLoopBroadcast(void* user) {
+    Demux* const self = static_cast<Demux*>(user);
+    self->broadcastInputThreadLoop();
+    return 0;
+}
+
+void Demux::broadcastInputThreadLoop() {
+    std::lock_guard<std::mutex> lock(mBroadcastInputThreadLock);
+    mBroadcastInputThreadRunning = true;
+    mKeepFetchingDataFromFrontend = true;
+
+    // open the stream and get its length
+    std::ifstream inputData(mFrontendSourceFile, std::ifstream::binary);
+    // TODO take the packet size from the frontend setting
+    int packetSize = 188;
+    int writePacketAmount = 6;
+    char* buffer = new char[packetSize];
+    ALOGW("[Demux] broadcast input thread loop start %s", mFrontendSourceFile.c_str());
+    if (!inputData.is_open()) {
+        mBroadcastInputThreadRunning = false;
+        ALOGW("[Demux] Error %s", strerror(errno));
+    }
+
+    while (mBroadcastInputThreadRunning) {
+        // move the stream pointer for packet size * 6 every read until the end
+        while (mKeepFetchingDataFromFrontend) {
+            for (int i = 0; i < writePacketAmount; i++) {
+                inputData.read(buffer, packetSize);
+                if (!inputData) {
+                    mBroadcastInputThreadRunning = false;
+                    break;
+                }
+                // filter and dispatch filter output
+                vector<uint8_t> byteBuffer;
+                byteBuffer.resize(sizeof(buffer));
+                for (int index = 0; index < byteBuffer.size(); index++) {
+                    byteBuffer[index] = static_cast<uint8_t>(buffer[index]);
+                }
+                startTsFilter(byteBuffer);
+                inputData.seekg(packetSize, inputData.cur);
+            }
+            startFilterDispatcher();
+            sleep(1);
+        }
+    }
+
+    ALOGW("[Demux] Broadcast Input thread end.");
+    delete[] buffer;
+    inputData.close();
+}
+
+void Demux::stopBroadcastInput() {
+    mKeepFetchingDataFromFrontend = false;
+    mBroadcastInputThreadRunning = false;
+    std::lock_guard<std::mutex> lock(mBroadcastInputThreadLock);
+}
+
 }  // namespace implementation
 }  // namespace V1_0
 }  // namespace tuner
diff --git a/tv/tuner/1.0/default/Demux.h b/tv/tuner/1.0/default/Demux.h
index 2fdde8d..e4a4e2b 100644
--- a/tv/tuner/1.0/default/Demux.h
+++ b/tv/tuner/1.0/default/Demux.h
@@ -20,6 +20,8 @@
 #include <android/hardware/tv/tuner/1.0/IDemux.h>
 #include <fmq/MessageQueue.h>
 #include <set>
+#include "Frontend.h"
+#include "Tuner.h"
 
 using namespace std;
 
@@ -40,9 +42,12 @@
 
 using FilterMQ = MessageQueue<uint8_t, kSynchronizedReadWrite>;
 
+class Tuner;
+class Frontend;
+
 class Demux : public IDemux {
   public:
-    Demux(uint32_t demuxId);
+    Demux(uint32_t demuxId, sp<Tuner> tuner);
 
     ~Demux();
 
@@ -91,9 +96,9 @@
 
     virtual Return<Result> configureOutput(const DemuxOutputSettings& settings) override;
 
-    virtual Return<Result> attachOutputTsFilter(uint32_t filterId) override;
+    virtual Return<Result> attachOutputFilter(uint32_t filterId) override;
 
-    virtual Return<Result> detachOutputTsFilter(uint32_t filterId) override;
+    virtual Return<Result> detachOutputFilter(uint32_t filterId) override;
 
     virtual Return<Result> startOutput() override;
 
@@ -103,7 +108,17 @@
 
     virtual Return<Result> removeOutput() override;
 
+    // Functions interacts with Tuner Service
+    void stopBroadcastInput();
+
   private:
+    // Tuner service
+    sp<Tuner> mTunerService;
+
+    // Frontend source
+    sp<Frontend> mFrontend;
+    string mFrontendSourceFile;
+
     // A struct that passes the arguments to a newly created filter thread
     struct ThreadArgs {
         Demux* user;
@@ -115,13 +130,14 @@
      * They are also responsible to write the filtered output into the filter FMQ
      * and update the filterEvent bound with the same filterId.
      */
-    Result startSectionFilterHandler(uint32_t filterId, vector<uint8_t> data);
+    Result startSectionFilterHandler(uint32_t filterId);
     Result startPesFilterHandler(uint32_t filterId);
     Result startTsFilterHandler();
     Result startMediaFilterHandler(uint32_t filterId);
     Result startRecordFilterHandler(uint32_t filterId);
     Result startPcrFilterHandler();
     Result startFilterLoop(uint32_t filterId);
+    Result startBroadcastInputLoop();
 
     /**
      * To create a FilterMQ with the the next available Filter ID.
@@ -136,18 +152,24 @@
     bool writeDataToFilterMQ(const std::vector<uint8_t>& data, uint32_t filterId);
     bool readDataFromMQ();
     bool writeSectionsAndCreateEvent(uint32_t filterId, vector<uint8_t> data);
+    void maySendInputStatusCallback();
+    DemuxInputStatus checkStatusChange(uint32_t availableToWrite, uint32_t availableToRead,
+                                       uint32_t highThreshold, uint32_t lowThreshold);
     /**
      * A dispatcher to read and dispatch input data to all the started filters.
      * Each filter handler handles the data filtering/output writing/filterEvent updating.
      */
-    bool filterAndOutputData();
+    bool readInputFMQ();
+    void startTsFilter(vector<uint8_t> data);
+    bool startFilterDispatcher();
     static void* __threadLoopFilter(void* data);
     static void* __threadLoopInput(void* user);
+    static void* __threadLoopBroadcast(void* user);
     void filterThreadLoop(uint32_t filterId);
     void inputThreadLoop();
+    void broadcastInputThreadLoop();
 
     uint32_t mDemuxId;
-    uint32_t mSourceFrontendId;
     /**
      * Record the last used filter id. Initial value is -1.
      * Filter Id starts with 0.
@@ -169,6 +191,8 @@
      * A list of created FilterMQ ptrs.
      * The array number is the filter ID.
      */
+    vector<uint16_t> mFilterPids;
+    vector<vector<uint8_t>> mFilterOutputs;
     vector<unique_ptr<FilterMQ>> mFilterMQs;
     vector<EventFlag*> mFilterEventFlags;
     vector<DemuxFilterEvent> mFilterEvents;
@@ -182,15 +206,26 @@
     vector<sp<IDemuxCallback>> mDemuxCallbacks;
     sp<IDemuxCallback> mInputCallback;
     sp<IDemuxCallback> mOutputCallback;
+    bool mInputConfigured = false;
+    bool mOutputConfigured = false;
+    DemuxInputSettings mInputSettings;
+    DemuxOutputSettings mOutputSettings;
+
     // Thread handlers
     pthread_t mInputThread;
     pthread_t mOutputThread;
+    pthread_t mBroadcastInputThread;
     vector<pthread_t> mFilterThreads;
+
+    // FMQ status local records
+    DemuxInputStatus mIntputStatus;
     /**
      * If a specific filter's writing loop is still running
      */
     vector<bool> mFilterThreadRunning;
     bool mInputThreadRunning;
+    bool mBroadcastInputThreadRunning;
+    bool mKeepFetchingDataFromFrontend;
     /**
      * Lock to protect writes to the FMQs
      */
@@ -198,8 +233,16 @@
     /**
      * Lock to protect writes to the filter event
      */
+    // TODO make each filter separate event lock
     std::mutex mFilterEventLock;
     /**
+     * Lock to protect writes to the input status
+     */
+    std::mutex mInputStatusLock;
+    std::mutex mBroadcastInputThreadLock;
+    std::mutex mFilterThreadLock;
+    std::mutex mInputThreadLock;
+    /**
      * How many times a filter should write
      * TODO make this dynamic/random/can take as a parameter
      */
diff --git a/tv/tuner/1.0/default/Frontend.cpp b/tv/tuner/1.0/default/Frontend.cpp
index 0609d05..1e07edd 100644
--- a/tv/tuner/1.0/default/Frontend.cpp
+++ b/tv/tuner/1.0/default/Frontend.cpp
@@ -27,14 +27,10 @@
 namespace V1_0 {
 namespace implementation {
 
-Frontend::Frontend() {
-    // Init callback to nullptr
-    mCallback = nullptr;
-}
-
-Frontend::Frontend(FrontendType type, FrontendId id) {
+Frontend::Frontend(FrontendType type, FrontendId id, sp<Tuner> tuner) {
     mType = type;
     mId = id;
+    mTunerService = tuner;
     // Init callback to nullptr
     mCallback = nullptr;
 }
@@ -67,13 +63,18 @@
         return Result::INVALID_STATE;
     }
 
-    mCallback->onEvent(FrontendEventType::NO_SIGNAL);
+    // TODO dynamically allocate file to the source file
+    mSourceStreamFile = FRONTEND_STREAM_FILE;
+
+    mCallback->onEvent(FrontendEventType::LOCKED);
     return Result::SUCCESS;
 }
 
 Return<Result> Frontend::stopTune() {
     ALOGV("%s", __FUNCTION__);
 
+    mTunerService->frontendStopTune(mId);
+
     return Result::SUCCESS;
 }
 
@@ -105,13 +106,7 @@
     return Result::SUCCESS;
 }
 
-Return<Result> Frontend::setLnb(const sp<ILnb>& /* lnb */) {
-    ALOGV("%s", __FUNCTION__);
-
-    return Result::SUCCESS;
-}
-
-Return<Result> Frontend::sendDiseqcMessage(const hidl_vec<uint8_t>& /* diseqcMessage */) {
+Return<Result> Frontend::setLnb(uint32_t /* lnb */) {
     ALOGV("%s", __FUNCTION__);
 
     return Result::SUCCESS;
@@ -125,6 +120,10 @@
     return mId;
 }
 
+string Frontend::getSourceFile() {
+    return mSourceStreamFile;
+}
+
 }  // namespace implementation
 }  // namespace V1_0
 }  // namespace tuner
diff --git a/tv/tuner/1.0/default/Frontend.h b/tv/tuner/1.0/default/Frontend.h
index fc586b5..07fa7b9 100644
--- a/tv/tuner/1.0/default/Frontend.h
+++ b/tv/tuner/1.0/default/Frontend.h
@@ -18,7 +18,9 @@
 #define ANDROID_HARDWARE_TV_TUNER_V1_0_FRONTEND_H_
 
 #include <android/hardware/tv/tuner/1.0/IFrontend.h>
-#include <android/hardware/tv/tuner/1.0/ITuner.h>
+#include <fstream>
+#include <iostream>
+#include "Tuner.h"
 
 using namespace std;
 
@@ -35,11 +37,11 @@
 using ::android::hardware::tv::tuner::V1_0::IFrontendCallback;
 using ::android::hardware::tv::tuner::V1_0::Result;
 
+class Tuner;
+
 class Frontend : public IFrontend {
   public:
-    Frontend();
-
-    Frontend(FrontendType type, FrontendId id);
+    Frontend(FrontendType type, FrontendId id, sp<Tuner> tuner);
 
     virtual Return<Result> close() override;
 
@@ -56,21 +58,26 @@
     virtual Return<void> getStatus(const hidl_vec<FrontendStatusType>& statusTypes,
                                    getStatus_cb _hidl_cb) override;
 
-    virtual Return<Result> sendDiseqcMessage(const hidl_vec<uint8_t>& diseqcMessage) override;
-
     virtual Return<Result> setLna(bool bEnable) override;
 
-    virtual Return<Result> setLnb(const sp<ILnb>& lnb) override;
+    virtual Return<Result> setLnb(uint32_t lnb) override;
 
     FrontendType getFrontendType();
 
     FrontendId getFrontendId();
 
+    string getSourceFile();
+
   private:
     virtual ~Frontend();
     sp<IFrontendCallback> mCallback;
+    sp<Tuner> mTunerService;
     FrontendType mType = FrontendType::UNDEFINED;
     FrontendId mId = 0;
+
+    const string FRONTEND_STREAM_FILE = "/vendor/etc/test1.ts";
+    string mSourceStreamFile;
+    std::ifstream mFrontendData;
 };
 
 }  // namespace implementation
diff --git a/tv/tuner/1.0/default/Lnb.cpp b/tv/tuner/1.0/default/Lnb.cpp
index b81bb15..1446f7f 100644
--- a/tv/tuner/1.0/default/Lnb.cpp
+++ b/tv/tuner/1.0/default/Lnb.cpp
@@ -48,6 +48,12 @@
     return Result::SUCCESS;
 }
 
+Return<Result> Lnb::sendDiseqcMessage(const hidl_vec<uint8_t>& /* diseqcMessage */) {
+    ALOGV("%s", __FUNCTION__);
+
+    return Result::SUCCESS;
+}
+
 Return<Result> Lnb::close() {
     ALOGV("%s", __FUNCTION__);
 
diff --git a/tv/tuner/1.0/default/Lnb.h b/tv/tuner/1.0/default/Lnb.h
index df7e0fe..4c251f7 100644
--- a/tv/tuner/1.0/default/Lnb.h
+++ b/tv/tuner/1.0/default/Lnb.h
@@ -38,12 +38,14 @@
   public:
     Lnb();
 
-    virtual Return<Result> setVoltage(FrontendLnbVoltage voltage);
+    virtual Return<Result> setVoltage(FrontendLnbVoltage voltage) override;
 
     virtual Return<Result> setTone(FrontendLnbTone tone) override;
 
     virtual Return<Result> setSatellitePosition(FrontendLnbPosition position) override;
 
+    virtual Return<Result> sendDiseqcMessage(const hidl_vec<uint8_t>& diseqcMessage) override;
+
     virtual Return<Result> close() override;
 
   private:
diff --git a/tv/tuner/1.0/default/Tuner.cpp b/tv/tuner/1.0/default/Tuner.cpp
index 00831ae..f86b28d 100644
--- a/tv/tuner/1.0/default/Tuner.cpp
+++ b/tv/tuner/1.0/default/Tuner.cpp
@@ -38,14 +38,14 @@
     // Array index matches their FrontendId in the default impl
     mFrontendSize = 8;
     mFrontends.resize(mFrontendSize);
-    mFrontends[0] = new Frontend();
-    mFrontends[1] = new Frontend(FrontendType::ATSC, 1);
-    mFrontends[2] = new Frontend(FrontendType::DVBC, 2);
-    mFrontends[3] = new Frontend(FrontendType::DVBS, 3);
-    mFrontends[4] = new Frontend(FrontendType::DVBT, 4);
-    mFrontends[5] = new Frontend(FrontendType::ISDBT, 5);
-    mFrontends[6] = new Frontend(FrontendType::ANALOG, 6);
-    mFrontends[7] = new Frontend(FrontendType::ATSC, 7);
+    mFrontends[0] = new Frontend(FrontendType::DVBT, 0, this);
+    mFrontends[1] = new Frontend(FrontendType::ATSC, 1, this);
+    mFrontends[2] = new Frontend(FrontendType::DVBC, 2, this);
+    mFrontends[3] = new Frontend(FrontendType::DVBS, 3, this);
+    mFrontends[4] = new Frontend(FrontendType::DVBT, 4, this);
+    mFrontends[5] = new Frontend(FrontendType::ISDBT, 5, this);
+    mFrontends[6] = new Frontend(FrontendType::ANALOG, 6, this);
+    mFrontends[7] = new Frontend(FrontendType::ATSC, 7, this);
 }
 
 Tuner::~Tuner() {}
@@ -81,12 +81,22 @@
 
     DemuxId demuxId = mLastUsedId + 1;
     mLastUsedId += 1;
-    sp<IDemux> demux = new Demux(demuxId);
+    sp<Demux> demux = new Demux(demuxId, this);
+    mDemuxes[demuxId] = demux;
 
     _hidl_cb(Result::SUCCESS, demuxId, demux);
     return Void();
 }
 
+Return<void> Tuner::getDemuxCaps(getDemuxCaps_cb _hidl_cb) {
+    ALOGV("%s", __FUNCTION__);
+
+    DemuxCapabilities caps;
+
+    _hidl_cb(Result::SUCCESS, caps);
+    return Void();
+}
+
 Return<void> Tuner::openDescrambler(openDescrambler_cb _hidl_cb) {
     ALOGV("%s", __FUNCTION__);
 
@@ -123,6 +133,25 @@
     return Void();
 }
 
+sp<Frontend> Tuner::getFrontendById(uint32_t frontendId) {
+    ALOGV("%s", __FUNCTION__);
+
+    return mFrontends[frontendId];
+}
+
+void Tuner::setFrontendAsDemuxSource(uint32_t frontendId, uint32_t demuxId) {
+    mFrontendToDemux[frontendId] = demuxId;
+}
+
+void Tuner::frontendStopTune(uint32_t frontendId) {
+    map<uint32_t, uint32_t>::iterator it = mFrontendToDemux.find(frontendId);
+    uint32_t demuxId;
+    if (it != mFrontendToDemux.end()) {
+        demuxId = it->second;
+        mDemuxes[demuxId]->stopBroadcastInput();
+    }
+}
+
 }  // namespace implementation
 }  // namespace V1_0
 }  // namespace tuner
diff --git a/tv/tuner/1.0/default/Tuner.h b/tv/tuner/1.0/default/Tuner.h
index 62227ee..96da257 100644
--- a/tv/tuner/1.0/default/Tuner.h
+++ b/tv/tuner/1.0/default/Tuner.h
@@ -18,6 +18,8 @@
 #define ANDROID_HARDWARE_TV_TUNER_V1_0_TUNER_H_
 
 #include <android/hardware/tv/tuner/1.0/ITuner.h>
+#include <map>
+#include "Demux.h"
 #include "Frontend.h"
 
 using namespace std;
@@ -29,6 +31,9 @@
 namespace V1_0 {
 namespace implementation {
 
+class Frontend;
+class Demux;
+
 class Tuner : public ITuner {
   public:
     Tuner();
@@ -39,6 +44,8 @@
 
     virtual Return<void> openDemux(openDemux_cb _hidl_cb) override;
 
+    virtual Return<void> getDemuxCaps(getDemuxCaps_cb _hidl_cb) override;
+
     virtual Return<void> openDescrambler(openDescrambler_cb _hidl_cb) override;
 
     virtual Return<void> getFrontendInfo(FrontendId frontendId,
@@ -48,10 +55,18 @@
 
     virtual Return<void> openLnbById(LnbId lnbId, openLnbById_cb _hidl_cb) override;
 
+    sp<Frontend> getFrontendById(uint32_t frontendId);
+
+    void setFrontendAsDemuxSource(uint32_t frontendId, uint32_t demuxId);
+
+    void frontendStopTune(uint32_t frontendId);
+
   private:
     virtual ~Tuner();
     // Static mFrontends array to maintain local frontends information
     vector<sp<Frontend>> mFrontends;
+    std::map<uint32_t, uint32_t> mFrontendToDemux;
+    std::map<uint32_t, sp<Demux>> mDemuxes;
     // To maintain how many Frontends we have
     int mFrontendSize;
     // The last used demux id. Initial value is -1.
diff --git a/tv/tuner/1.0/types.hal b/tv/tuner/1.0/types.hal
index a1153e9..897818b 100644
--- a/tv/tuner/1.0/types.hal
+++ b/tv/tuner/1.0/types.hal
@@ -43,7 +43,7 @@
     ANALOG,
     /* Advanced Television Systems Committee (ATSC) Standard A/72. */
     ATSC,
-    /* Advanced Television Systems Committee (ATSC 3.0) Standard A/330. */
+    /* Advanced Television Systems Committee (ATSC 3.0) Standard A/300. */
     ATSC3,
     /**
      * Digital Video Broadcasting - Cable
@@ -63,11 +63,11 @@
      */
     DVBT,
     /* Integrated Services Digital Broadcasting-Satellite (ISDB-S)
-     * ARIB SDT-B20 is technical document of ISDB-S.
+     * ARIB STD-B20 is technical document of ISDB-S.
      */
     ISDBS,
     /* Integrated Services Digital Broadcasting-Satellite (ISDB-S)
-     * ARIB TR-B15 is technical document of ISDB-S3.
+     * ARIB STD-B44 is technical document of ISDB-S3.
      */
     ISDBS3,
     /* Integrated Services Digital Broadcasting-Terrestrial (ISDB-T or SBTVD)
@@ -164,8 +164,10 @@
 @export
 enum FrontendAtscModulation : uint32_t {
     UNDEFINED = 0,
-    MOD_8VSB = 1 << 0,
-    MOD_16VSB = 1 << 1,
+    /** hardware is able to detect and set modulation automatically */
+    AUTO      = 1 << 0,
+    MOD_8VSB  = 1 << 2,
+    MOD_16VSB = 1 << 3,
 };
 
 /**
@@ -191,12 +193,14 @@
 @export
 enum FrontendAtsc3Modulation : uint32_t {
     UNDEFINED = 0,
-    MOD_QPSK = 1 << 0,
-    MOD_16QAM = 1 << 1,
-    MOD_64QAM = 1 << 2,
-    MOD_256QAM = 1 << 3,
-    MOD_1024QAM = 1 << 4,
-    MOD_4096QAM = 1 << 5,
+    /** hardware is able to detect and set modulation automatically */
+    AUTO        = 1 << 0,
+    MOD_QPSK    = 1 << 1,
+    MOD_16QAM   = 1 << 2,
+    MOD_64QAM   = 1 << 3,
+    MOD_256QAM  = 1 << 4,
+    MOD_1024QAM = 1 << 5,
+    MOD_4096QAM = 1 << 6,
 };
 
 /**
@@ -205,9 +209,11 @@
 @export
 enum FrontendAtsc3Bandwidth : uint32_t {
     UNDEFINED = 0,
-    BANDWIDTH_8MHZ = 1 << 0,
-    BANDWIDTH_7MHZ = 1 << 1,
-    BANDWIDTH_6MHZ = 1 << 2,
+    /** hardware is able to detect and set bandwidth automatically */
+    AUTO = 1 << 0,
+    BANDWIDTH_6MHZ = 1 << 1,
+    BANDWIDTH_7MHZ = 1 << 2,
+    BANDWIDTH_8MHZ = 1 << 3,
 };
 
 /**
@@ -215,9 +221,11 @@
  */
 @export
 enum FrontendAtsc3TimeInterleaveMode : uint32_t {
-    UNDEFINED,
-    CTI,
-    HTI,
+    UNDEFINED = 0,
+    /** hardware is able to detect and set TimeInterleaveMode automatically */
+    AUTO = 1 << 0,
+    CTI  = 1 << 1,
+    HTI  = 1 << 2,
 };
 
 /**
@@ -247,13 +255,39 @@
  */
 @export
 enum FrontendAtsc3Fec : uint32_t {
-    UNDEFINED,
-    BCH_LDPC_16K,
-    BCH_LDPC_64K,
-    CRC_LDPC_16K,
-    CRC_LDPC_64K,
-    LDPC_16K,
-    LDPC_64K,
+    UNDEFINED = 0,
+    /** hardware is able to detect and set FEC automatically */
+    AUTO     = 1 << 0,
+    BCH_LDPC_16K = 1 << 1,
+    BCH_LDPC_64K = 1 << 2,
+    CRC_LDPC_16K = 1 << 3,
+    CRC_LDPC_64K = 1 << 4,
+    LDPC_16K     = 1 << 5,
+    LDPC_64K     = 1 << 6,
+};
+
+/**
+ *  Demodulator Output Format for an ATSC3 Frontend.
+ */
+@export
+enum FrontendAtsc3DemodOutputFormat : uint8_t {
+    /** Dummy. Scan uses this. */
+    UNDEFINED = 0,
+    /** ALP format. Typically used in US region. */
+    ATSC3_LINKLAYER_PACKET = 1 << 0,
+    /** BaseBand packet format. Typically used in Korea region. */
+    BASEBAND_PACKET        = 1 << 1,
+};
+
+/**
+ *  PLP basis Signal Settings for an ATSC3 Frontend.
+ */
+struct FrontendAtsc3PlpSettings {
+    uint8_t plpId;
+    FrontendAtsc3Modulation modulation;
+    FrontendAtsc3TimeInterleaveMode interleaveMode;
+    FrontendAtsc3CodeRate codeRate;
+    FrontendAtsc3Fec fec;
 };
 
 /**
@@ -262,21 +296,28 @@
 struct FrontendAtsc3Settings {
     /** Signal frequency in Hertz */
     uint32_t frequency;
+    /** Bandwidth of tuning band. */
     FrontendAtsc3Bandwidth bandwidth;
-    FrontendAtsc3TimeInterleaveMode interleaveMode;
-    FrontendAtsc3CodeRate codeRate;
-    FrontendAtsc3Fec fec;
-    vec<uint8_t> plpIdList;
+    FrontendAtsc3DemodOutputFormat demodOutputFormat;
+    vec<FrontendAtsc3PlpSettings> plpSettings;
 };
 
 /**
  *  Capabilities for ATSC3 Frontend.
  */
 struct FrontendAtsc3Capabilities {
-    /** Modulation capability */
-    bitfield<FrontendAtsc3Modulation> modulationCap;
     /** Bandwidth capability */
     bitfield<FrontendAtsc3Bandwidth> bandwidthCap;
+    /** Modulation capability */
+    bitfield<FrontendAtsc3Modulation> modulationCap;
+    /** TimeInterleaveMode capability */
+    bitfield<FrontendAtsc3TimeInterleaveMode> timeInterleaveModeCap;
+    /** CodeRate capability */
+    bitfield<FrontendAtsc3CodeRate> codeRateCap;
+    /** FEC capability */
+    bitfield<FrontendAtsc3Fec> fecCap;
+    /** Demodulator Output Format capability */
+    bitfield<FrontendAtsc3DemodOutputFormat> demodOutputFormatCap;
 };
 
 /**
@@ -647,7 +688,7 @@
 @export
 enum FrontendIsdbsStreamIdType : uint32_t {
     STREAM_ID,
-    RELATIVE_STREAM_ID,
+    RELATIVE_STREAM_NUMBER,
 };
 
 /**
@@ -845,6 +886,7 @@
     M_EIA_J = 1 << 13,
     I_NICAM = 1 << 14,
     L_NICAM = 1 << 15,
+    L_PRIME = 1 << 16,
 };
 
 /**
@@ -907,10 +949,20 @@
     PLP_IDS,
     /** Locked group Ids for DVBT2 frontend. */
     GROUP_IDS,
-    /** Locked the number of the Plps. */
-   INPUT_STREAM_IDS,
+    /** Stream Ids. */
+    INPUT_STREAM_IDS,
     /** Locked signal stardard.  */
     STANDARD,
+    /** PLP status in a tuned frequency band for ATSC3 frontend. */
+    ATSC3_PLP_INFO,
+};
+
+/**
+ *  ATSC3.0 PLP information for scan
+ */
+struct FrontendScanAtsc3PlpInfo {
+    uint8_t plpId;
+    bool bLlsFlag;
 };
 
 /**
@@ -927,11 +979,13 @@
     uint32_t symbolRate;
     vec<uint8_t> plpIds;
     vec<uint8_t> groupIds;
-    vec<uint8_t> inputStreamIds;
+    vec<uint16_t> inputStreamIds;
     safe_union standard {
         FrontendDvbsStandard sStd;
         FrontendDvbtStandard tStd;
     } std;
+    /** A list of PLP status in a tuned frequency band for ATSC3 frontend. */
+    vec<FrontendScanAtsc3PlpInfo> atsc3PlpInfos;
 };
 
 /**
@@ -940,17 +994,17 @@
 @export
 enum FrontendEventType : uint32_t {
     /**
-     * If frontend locked the signal which is specified by tune method, HAL sent
+     * If frontend locked the signal which is specified by tune method, HAL sends
      * Locked event.
      */
     LOCKED,
     /**
      * If frontend can't locked the signal which is specified by tune method,
-     * HAL sent NO_SIGNAL event.
+     * HAL sends NO_SIGNAL event.
      */
     NO_SIGNAL,
     /**
-     * If frontend detect that the locked signal get lost, HAL sent LOST_LOCK
+     * If frontend detect that the locked signal get lost, HAL sends LOST_LOCK
      * event.
      */
     LOST_LOCK,
@@ -977,15 +1031,15 @@
  */
 @export
 enum FrontendStatusType : uint32_t {
-    /** Lock status for RF or Demod. */
-    LOCK,
+    /** Lock status for Demod. */
+    DEMOD_LOCK,
     /** Signal to Noise Ratio. */
     SNR,
     /** Bit Error Ratio. */
     BER,
     /** Packages Error Ratio. */
     PER,
-    /** Bit Error Ratio befor FEC. */
+    /** Bit Error Ratio before FEC. */
     PRE_BER,
     /*
      * Signal Quality (0..100). Good data over total data in percent can be
@@ -993,7 +1047,7 @@
      */
     SIGNAL_QUALITY,
     /** Signal Strength. */
-    SIGGAL_STRENGTH,
+    SIGNAL_STRENGTH,
     /** Symbol Rate. */
     SYMBOL_RATE,
     /** Forward Error Correction Type. */
@@ -1008,21 +1062,62 @@
     PLP_ID,
     /** Status for Emergency Warning Broadcasting System. */
     EWBS,
+    /** Automatic Gain Control. */
+    AGC,
+    /** Low Noise Amplifier. */
+    LNA,
+    /** Lock status for stream. */
+    STREAM_LOCK,
+    /** Error status by layer. */
+    LAYER_ERROR,
+    /** CN value by VBER. */
+    VBER_CN,
+    /** CN value by LBER. */
+    LBER_CN,
+    /** CN value by XER. */
+    XER_CN,
+    /** Moduration Error Ratio. */
+    MER,
+    /** Difference between tuning frequency and actual locked frequency. */
+    FREQ_OFFSET,
+    /* Hierarchy for DVBT. */
+    HIERARCHY,
+    /** Lock status for RF. */
+    RF_LOCK,
+    /** PLP information in a frequency band for ATSC3.0 frontend. */
+    ATSC3_PLP_INFO,
 };
 
 /**
+ * Status for each tuning PLPs
+ */
+struct FrontendStatusAtsc3PlpInfo {
+    /** PLP Id value. */
+    uint8_t plpId;
+    /** Demod Lock/Unlock status of this particular PLP. */
+    bool isLocked;
+    /** Uncorrectable Error Counts (UEC) of this particular PLP since last tune operation. */
+    uint32_t uec;
+};
+
+
+/**
  * Modulation Type for Frontend's status.
  */
 safe_union FrontendModulationStatus {
+    FrontendDvbcModulation dvbc;
     FrontendDvbsModulation dvbs;
-    FrontendAtsc3Modulation atsc3;
+    FrontendIsdbsModulation isdbs;
+    FrontendIsdbs3Modulation isdbs3;
+    FrontendIsdbtModulation isdbt;
 };
 
 /**
  *  The status for Frontend.
  */
 safe_union FrontendStatus {
-    bool isLocked;
+    /** Lock status for Demod in True/False. */
+    bool isDemodLocked;
     /** SNR value measured by 0.001 dB. */
     int32_t snr;
     /** The number of error bit per 1 billion bits. */
@@ -1043,6 +1138,25 @@
     FrontendLnbVoltage lnbVoltage;
     uint8_t plpId;
     bool isEWBS;
+    /** AGC value is normalized from 0 to 255. */
+    uint8_t agc;
+    bool isLnaOn;
+    bool isStreamLock;
+    vec<bool> isLayerError;
+    /** CN value by VBER measured by 0.001 dB */
+    int32_t vberCn;
+    /** CN value by LBER measured by 0.001 dB */
+    int32_t lberCn;
+    /** CN value by XER measured by 0.001 dB */
+    int32_t xerCn;
+    /** MER value measured by 0.001 dB */
+    int32_t mer;
+    /** Frequency difference in Hertz. */
+    int32_t freqOffset;
+    FrontendDvbtHierarchy hierarchy;
+    bool isRfLocked;
+    /** A list of PLP status for tuned PLPs for ATSC3 frontend. */
+    vec<FrontendStatusAtsc3PlpInfo> plpInfo;
 };
 
 /**
@@ -1121,7 +1235,6 @@
     POSITION_B,
 };
 
-
 /* Demux ID is used to associate with a hardware demux resource. */
 typedef uint32_t DemuxId;
 
@@ -1150,7 +1263,7 @@
      */
     AUDIO,
     /**
-     * A filter to filter Vidoe Metadata out from input stream.
+     * A filter to filter Video Metadata out from input stream.
      */
     VIDEO,
     /**
@@ -1475,6 +1588,8 @@
     PES,
     /* Data is Elementary Stream. */
     ES,
+    /* Data is TLV (type-length-value) Stream for JP SHV */
+    SHV_TLV,
 };
 
 /**
@@ -1534,6 +1649,10 @@
     SPACE_FULL         = 1 << 3,
 };
 
+/**
+ *  The Settings for the demux's input.
+ */
+@export
 struct DemuxInputSettings {
     /**
      * Register for interested status events so that the HAL can send these
@@ -1559,3 +1678,30 @@
      */
     uint8_t packetSize;
 };
+
+/**
+ *  Capabilities for Demux.
+ */
+@export
+struct DemuxCapabilities {
+    /* The number of Demux to be supported. */
+    uint32_t numDemux;
+    /* The number of Input to be supported. */
+    uint32_t numInput;
+    /* The number of Output to be supported. */
+    uint32_t numOutput;
+    /* The number of TS Filter to be supported. */
+    uint32_t numTsFilter;
+    /* The number of Section Filter to be supported. */
+    uint32_t numSectionFilter;
+    /* The number of Audio Filter to be supported. */
+    uint32_t numAudioFilter;
+    /* The number of Video Filter to be supported. */
+    uint32_t numVideoFilter;
+    /* The number of PES Filter to be supported. */
+    uint32_t numPesFilter;
+    /* The number of PCR Filter to be supported. */
+    uint32_t numPcrFilter;
+    /* The maximum number of bytes is supported in the mask of Section Filter. */
+    uint32_t numBytesInSectionFilter;
+};
diff --git a/tv/tuner/1.0/vts/functional/VtsHalTvTunerV1_0TargetTest.cpp b/tv/tuner/1.0/vts/functional/VtsHalTvTunerV1_0TargetTest.cpp
index 7256cc4..7936185 100644
--- a/tv/tuner/1.0/vts/functional/VtsHalTvTunerV1_0TargetTest.cpp
+++ b/tv/tuner/1.0/vts/functional/VtsHalTvTunerV1_0TargetTest.cpp
@@ -39,6 +39,7 @@
 #include <map>
 
 #define WAIT_TIMEOUT 3000000000
+#define WAIT_TIMEOUT_data_ready 3000000000 * 4
 
 using android::Condition;
 using android::IMemory;
@@ -58,8 +59,10 @@
 using android::hardware::Void;
 using android::hardware::tv::tuner::V1_0::DemuxDataFormat;
 using android::hardware::tv::tuner::V1_0::DemuxFilterEvent;
+using android::hardware::tv::tuner::V1_0::DemuxFilterPesDataSettings;
 using android::hardware::tv::tuner::V1_0::DemuxFilterPesEvent;
 using android::hardware::tv::tuner::V1_0::DemuxFilterSectionEvent;
+using android::hardware::tv::tuner::V1_0::DemuxFilterSectionSettings;
 using android::hardware::tv::tuner::V1_0::DemuxFilterSettings;
 using android::hardware::tv::tuner::V1_0::DemuxFilterStatus;
 using android::hardware::tv::tuner::V1_0::DemuxFilterType;
@@ -130,9 +133,6 @@
 
 const uint16_t FMQ_SIZE_4K = 0x1000;
 const uint32_t FMQ_SIZE_1M = 0x100000;
-// Equal to SECTION_WRITE_COUNT on the HAL impl side
-// The HAL impl will repeatedly write to the FMQ the count times
-const uint16_t SECTION_READ_COUNT = 10;
 
 struct FilterConf {
     DemuxFilterType type;
@@ -214,11 +214,15 @@
 class DemuxCallback : public IDemuxCallback {
   public:
     virtual Return<void> onFilterEvent(const DemuxFilterEvent& filterEvent) override {
-        ALOGW("[VTS] FILTER EVENT %d", filterEvent.filterId);
         android::Mutex::Autolock autoLock(mMsgLock);
-        mFilterEventReceived = true;
+        // Temprarily we treat the first coming back filter data on the matching pid a success
+        // once all of the MQ are cleared, means we got all the expected output
         mFilterIdToEvent[filterEvent.filterId] = filterEvent;
-        startFilterEventThread(filterEvent);
+        readFilterEventData(filterEvent.filterId);
+        mPidFilterOutputCount++;
+        // mFilterIdToMQ.erase(filterEvent.filterId);
+
+        // startFilterEventThread(filterEvent);
         mMsgCondition.signal();
         return Void();
     }
@@ -232,13 +236,16 @@
 
     virtual Return<void> onInputStatus(DemuxInputStatus status) override {
         // android::Mutex::Autolock autoLock(mMsgLock);
+        ALOGW("[vts] input status %d", status);
         switch (status) {
             case DemuxInputStatus::SPACE_EMPTY:
             case DemuxInputStatus::SPACE_ALMOST_EMPTY:
+                ALOGW("[vts] keep inputing %d", status);
                 mKeepWritingInputFMQ = true;
                 break;
             case DemuxInputStatus::SPACE_ALMOST_FULL:
             case DemuxInputStatus::SPACE_FULL:
+                ALOGW("[vts] stop inputing %d", status);
                 mKeepWritingInputFMQ = false;
                 break;
         }
@@ -246,78 +253,64 @@
     }
 
     void testOnFilterEvent(uint32_t filterId);
-    void testOnSectionFilterEvent(sp<IDemux>& demux, uint32_t filterId, MQDesc& filterMQDescriptor,
-                                  MQDesc& inputMQDescriptor);
     void testFilterDataOutput();
-    // Legacy
-    bool readAndCompareSectionEventData(uint32_t filterId);
+    void stopInputThread();
 
     void startPlaybackInputThread(InputConf inputConf, MQDesc& inputMQDescriptor);
     void startFilterEventThread(DemuxFilterEvent event);
     static void* __threadLoopInput(void* threadArgs);
     static void* __threadLoopFilter(void* threadArgs);
-    void inputThreadLoop(InputConf inputConf, bool* keepWritingInputFMQ, MQDesc& inputMQDescriptor);
+    void inputThreadLoop(InputConf* inputConf, bool* keepWritingInputFMQ);
     void filterThreadLoop(DemuxFilterEvent& event);
 
     void updateFilterMQ(uint32_t filterId, MQDesc& filterMQDescriptor);
     void updateGoldenOutputMap(uint32_t filterId, string goldenOutputFile);
+    bool readFilterEventData(uint32_t filterId);
 
   private:
     struct InputThreadArgs {
         DemuxCallback* user;
-        InputConf inputConf;
+        InputConf* inputConf;
         bool* keepWritingInputFMQ;
-        MQDesc& inputMQDesc;
     };
     struct FilterThreadArgs {
         DemuxCallback* user;
-        DemuxFilterEvent& event;
+        DemuxFilterEvent event;
     };
     uint16_t mDataLength = 0;
     std::vector<uint8_t> mDataOutputBuffer;
 
     bool mFilterEventReceived;
     std::map<uint32_t, string> mFilterIdToGoldenOutput;
-    std::map<uint32_t, DemuxFilterEvent> mFilterIdToEvent;
 
     std::map<uint32_t, std::unique_ptr<FilterMQ>> mFilterIdToMQ;
     std::unique_ptr<FilterMQ> mInputMQ;
     std::map<uint32_t, EventFlag*> mFilterIdToMQEventFlag;
+    std::map<uint32_t, DemuxFilterEvent> mFilterIdToEvent;
     EventFlag* mInputMQEventFlag;
 
     android::Mutex mMsgLock;
     android::Mutex mFilterOutputLock;
+    android::Mutex mInputThreadLock;
     android::Condition mMsgCondition;
     android::Condition mFilterOutputCondition;
 
-    bool mKeepWritingInputFMQ;
+    bool mKeepWritingInputFMQ = true;
     bool mInputThreadRunning;
     pthread_t mInputThread;
     pthread_t mFilterThread;
+
+    int mPidFilterOutputCount = 0;
 };
 
-// Legacy
-void DemuxCallback::testOnFilterEvent(uint32_t filterId) {
-    android::Mutex::Autolock autoLock(mMsgLock);
-    while (!mFilterEventReceived) {
-        if (-ETIMEDOUT == mMsgCondition.waitRelative(mMsgLock, WAIT_TIMEOUT)) {
-            EXPECT_TRUE(false) << "filter event not received within timeout";
-            return;
-        }
-    }
-    // Reset the filter event recieved flag
-    mFilterEventReceived = false;
-    // Check if filter id match
-    EXPECT_TRUE(filterId == mFilterIdToEvent[filterId].filterId) << "filter id match";
-}
-
 void DemuxCallback::startPlaybackInputThread(InputConf inputConf, MQDesc& inputMQDescriptor) {
+    mInputMQ = std::make_unique<FilterMQ>(inputMQDescriptor, true /* resetPointers */);
+    EXPECT_TRUE(mInputMQ);
     struct InputThreadArgs* threadArgs =
             (struct InputThreadArgs*)malloc(sizeof(struct InputThreadArgs));
     threadArgs->user = this;
-    threadArgs->inputConf = inputConf;
+    threadArgs->inputConf = &inputConf;
     threadArgs->keepWritingInputFMQ = &mKeepWritingInputFMQ;
-    threadArgs->inputMQDesc = inputMQDescriptor;
 
     pthread_create(&mInputThread, NULL, __threadLoopInput, (void*)threadArgs);
     pthread_setname_np(mInputThread, "test_playback_input_loop");
@@ -334,72 +327,22 @@
 }
 
 void DemuxCallback::testFilterDataOutput() {
-    android::Mutex::Autolock autoLock(mFilterOutputLock);
-    while (!mFilterIdToMQ.empty()) {
-        if (-ETIMEDOUT == mFilterOutputCondition.waitRelative(mFilterOutputLock, WAIT_TIMEOUT)) {
-            EXPECT_TRUE(false) << "filter output does not match golden output within timeout";
+    android::Mutex::Autolock autoLock(mMsgLock);
+    while (mPidFilterOutputCount < 1) {
+        if (-ETIMEDOUT == mMsgCondition.waitRelative(mMsgLock, WAIT_TIMEOUT)) {
+            EXPECT_TRUE(false) << "filter output matching pid does not output within timeout";
             return;
         }
     }
+    mPidFilterOutputCount = 0;
+    ALOGW("[vts] pass and stop");
 }
 
-// Legacy
-void DemuxCallback::testOnSectionFilterEvent(sp<IDemux>& demux, uint32_t filterId,
-                                             MQDesc& filterMQDescriptor,
-                                             MQDesc& inputMQDescriptor) {
-    Result status;
-    // Create MQ to read the output into the local buffer
-    mFilterIdToMQ[filterId] =
-            std::make_unique<FilterMQ>(filterMQDescriptor, true /* resetPointers */);
-    EXPECT_TRUE(mFilterIdToMQ[filterId]);
-    // Get the MQ to write the input to the HAL
-    mInputMQ = std::make_unique<FilterMQ>(inputMQDescriptor, true /* resetPointers */);
-    EXPECT_TRUE(mInputMQ);
-    // Create the EventFlag that is used to signal the HAL impl that data have been
-    // read the Filter FMQ
-    EXPECT_TRUE(EventFlag::createEventFlag(mFilterIdToMQ[filterId]->getEventFlagWord(),
-                                           &mFilterIdToMQEventFlag[filterId]) == android::OK);
-    // Create the EventFlag that is used to signal the HAL impl that data have been
-    // written into the Input FMQ
-    EXPECT_TRUE(EventFlag::createEventFlag(mInputMQ->getEventFlagWord(), &mInputMQEventFlag) ==
-                android::OK);
-    // Start filter
-    status = demux->startFilter(filterId);
-    status = demux->startInput();
+void DemuxCallback::stopInputThread() {
+    mInputThreadRunning = false;
+    mKeepWritingInputFMQ = false;
 
-    EXPECT_EQ(status, Result::SUCCESS);
-    // Test start filter and receive callback event
-    for (int i = 0; i < SECTION_READ_COUNT; i++) {
-        // Write input FMQ and notify the Tuner Implementation
-        EXPECT_TRUE(mInputMQ->write(goldenDataOutputBuffer.data(), goldenDataOutputBuffer.size()));
-        mInputMQEventFlag->wake(static_cast<uint32_t>(DemuxQueueNotifyBits::DATA_READY));
-        testOnFilterEvent(filterId);
-        // checksum of mDataOutputBuffer and Input golden input
-        if (readAndCompareSectionEventData(filterId) && i < SECTION_READ_COUNT - 1) {
-            mFilterIdToMQEventFlag[filterId]->wake(
-                    static_cast<uint32_t>(DemuxQueueNotifyBits::DATA_CONSUMED));
-        }
-    }
-}
-
-// Legacy
-bool DemuxCallback::readAndCompareSectionEventData(uint32_t filterId) {
-    bool result = false;
-    DemuxFilterEvent filterEvent = mFilterIdToEvent[filterId];
-    for (int i = 0; i < filterEvent.events.size(); i++) {
-        DemuxFilterSectionEvent event = filterEvent.events[i].section();
-        mDataLength = event.dataLength;
-        EXPECT_TRUE(mDataLength == goldenDataOutputBuffer.size()) << "buffer size does not match";
-
-        mDataOutputBuffer.resize(mDataLength);
-        result = mFilterIdToMQ[filterId]->read(mDataOutputBuffer.data(), mDataLength);
-        EXPECT_TRUE(result) << "can't read from Filter MQ";
-
-        for (int i = 0; i < mDataLength; i++) {
-            EXPECT_TRUE(goldenDataOutputBuffer[i] == mDataOutputBuffer[i]) << "data does not match";
-        }
-    }
-    return result;
+    android::Mutex::Autolock autoLock(mInputThreadLock);
 }
 
 void DemuxCallback::updateFilterMQ(uint32_t filterId, MQDesc& filterMQDescriptor) {
@@ -418,58 +361,60 @@
     DemuxCallback* const self =
             static_cast<DemuxCallback*>(((struct InputThreadArgs*)threadArgs)->user);
     self->inputThreadLoop(((struct InputThreadArgs*)threadArgs)->inputConf,
-                          ((struct InputThreadArgs*)threadArgs)->keepWritingInputFMQ,
-                          ((struct InputThreadArgs*)threadArgs)->inputMQDesc);
+                          ((struct InputThreadArgs*)threadArgs)->keepWritingInputFMQ);
     return 0;
 }
 
-void DemuxCallback::inputThreadLoop(InputConf inputConf, bool* keepWritingInputFMQ,
-                                    MQDesc& inputMQDescriptor) {
+void DemuxCallback::inputThreadLoop(InputConf* inputConf, bool* keepWritingInputFMQ) {
+    android::Mutex::Autolock autoLock(mInputThreadLock);
     mInputThreadRunning = true;
 
-    std::unique_ptr inputMQ =
-            std::make_unique<FilterMQ>(inputMQDescriptor, true /* resetPointers */);
-    EXPECT_TRUE(inputMQ);
-
     // Create the EventFlag that is used to signal the HAL impl that data have been
     // written into the Input FMQ
     EventFlag* inputMQEventFlag;
-    EXPECT_TRUE(EventFlag::createEventFlag(inputMQ->getEventFlagWord(), &inputMQEventFlag) ==
+    EXPECT_TRUE(EventFlag::createEventFlag(mInputMQ->getEventFlagWord(), &inputMQEventFlag) ==
                 android::OK);
 
     // open the stream and get its length
-    std::ifstream inputData(inputConf.inputDataFile /*"ts/test1.ts"*/, std::ifstream::binary);
-    int writeSize = inputConf.setting.packetSize * 6;
+    std::ifstream inputData(inputConf->inputDataFile, std::ifstream::binary);
+    int writeSize = inputConf->setting.packetSize * 6;
     char* buffer = new char[writeSize];
-    if (!inputData) {
-        // log
+    ALOGW("[vts] input thread loop start %s", inputConf->inputDataFile.c_str());
+    if (!inputData.is_open()) {
         mInputThreadRunning = false;
+        ALOGW("[vts] Error %s", strerror(errno));
     }
 
     while (mInputThreadRunning) {
-        // move the stream pointer for packet size * 2k? every read until end
+        // move the stream pointer for packet size * 6 every read until the end
         while (*keepWritingInputFMQ) {
             inputData.read(buffer, writeSize);
             if (!inputData) {
                 int leftSize = inputData.gcount();
+                if (leftSize == 0) {
+                    mInputThreadRunning = false;
+                    break;
+                }
                 inputData.clear();
                 inputData.read(buffer, leftSize);
                 // Write the left over of the input data and quit the thread
                 if (leftSize > 0) {
-                    EXPECT_TRUE(inputMQ->write((unsigned char*)&buffer[0],
-                                               leftSize / inputConf.setting.packetSize));
+                    EXPECT_TRUE(mInputMQ->write((unsigned char*)&buffer[0], leftSize));
                     inputMQEventFlag->wake(static_cast<uint32_t>(DemuxQueueNotifyBits::DATA_READY));
                 }
                 mInputThreadRunning = false;
                 break;
             }
             // Write input FMQ and notify the Tuner Implementation
-            EXPECT_TRUE(inputMQ->write((unsigned char*)&buffer[0], 6));
+            EXPECT_TRUE(mInputMQ->write((unsigned char*)&buffer[0], writeSize));
             inputMQEventFlag->wake(static_cast<uint32_t>(DemuxQueueNotifyBits::DATA_READY));
             inputData.seekg(writeSize, inputData.cur);
+            sleep(1);
         }
     }
 
+    ALOGW("[vts] Input thread end.");
+
     delete[] buffer;
     inputData.close();
 }
@@ -481,9 +426,9 @@
     return 0;
 }
 
-void DemuxCallback::filterThreadLoop(DemuxFilterEvent& /*event*/) {
+void DemuxCallback::filterThreadLoop(DemuxFilterEvent& /* event */) {
     android::Mutex::Autolock autoLock(mFilterOutputLock);
-    // Read from MQ[event.filterId] per event and filter type
+    // Read from mFilterIdToMQ[event.filterId] per event and filter type
 
     // Assemble to filterOutput[filterId]
 
@@ -494,6 +439,30 @@
     // end thread
 }
 
+bool DemuxCallback::readFilterEventData(uint32_t filterId) {
+    bool result = false;
+    DemuxFilterEvent filterEvent = mFilterIdToEvent[filterId];
+    ALOGW("[vts] reading from filter FMQ %d", filterId);
+    // todo separate filter handlers
+    for (int i = 0; i < filterEvent.events.size(); i++) {
+        DemuxFilterPesEvent event = filterEvent.events[i].pes();
+        mDataLength = event.dataLength;
+        // EXPECT_TRUE(mDataLength == goldenDataOutputBuffer.size()) << "buffer size does not
+        // match";
+
+        mDataOutputBuffer.resize(mDataLength);
+        result = mFilterIdToMQ[filterId]->read(mDataOutputBuffer.data(), mDataLength);
+        EXPECT_TRUE(result) << "can't read from Filter MQ";
+
+        /*for (int i = 0; i < mDataLength; i++) {
+            EXPECT_TRUE(goldenDataOutputBuffer[i] == mDataOutputBuffer[i]) << "data does not match";
+        }*/
+    }
+    mFilterIdToMQEventFlag[filterId]->wake(
+            static_cast<uint32_t>(DemuxQueueNotifyBits::DATA_CONSUMED));
+    return result;
+}
+
 // Test environment for Tuner HIDL HAL.
 class TunerHidlEnvironment : public ::testing::VtsHalHidlTargetTestEnvBase {
   public:
@@ -528,6 +497,7 @@
     sp<DemuxCallback> mDemuxCallback;
     MQDesc mFilterMQDescriptor;
     MQDesc mInputMQDescriptor;
+    vector<uint32_t> mUsedFilterIds;
 
     uint32_t mDemuxId;
     uint32_t mFilterId;
@@ -540,7 +510,8 @@
     ::testing::AssertionResult stopTuneFrontend(int32_t frontendId);
     ::testing::AssertionResult closeFrontend(int32_t frontendId);
     ::testing::AssertionResult createDemux();
-    ::testing::AssertionResult createDemuxWithFrontend(int32_t frontendId);
+    ::testing::AssertionResult createDemuxWithFrontend(int32_t frontendId,
+                                                       FrontendSettings settings);
     ::testing::AssertionResult getInputMQDescriptor();
     ::testing::AssertionResult addInputToDemux(DemuxInputSettings setting);
     ::testing::AssertionResult addFilterToDemux(DemuxFilterType type, DemuxFilterSettings setting);
@@ -552,10 +523,8 @@
     ::testing::AssertionResult playbackDataFlowTest(vector<FilterConf> filterConf,
                                                     InputConf inputConf,
                                                     vector<string> goldenOutputFiles);
-
-    // Legacy
-    ::testing::AssertionResult addSectionFilterToDemux();
-    ::testing::AssertionResult readSectionFilterDataOutput();
+    ::testing::AssertionResult broadcastDataFlowTest(vector<FilterConf> filterConf,
+                                                     vector<string> goldenOutputFiles);
 };
 
 ::testing::AssertionResult TunerHidlTest::createFrontend(int32_t frontendId) {
@@ -586,7 +555,7 @@
             .frequency = 0,
             .modulation = FrontendAtscModulation::UNDEFINED,
     };
-    frontendSettings.atsc() = frontendAtscSettings;
+    frontendSettings.atsc(frontendAtscSettings);
     mFrontendCallback->testOnEvent(mFrontend, frontendSettings);
 
     FrontendDvbtSettings frontendDvbtSettings{
@@ -630,7 +599,8 @@
     return ::testing::AssertionResult(status == Result::SUCCESS);
 }
 
-::testing::AssertionResult TunerHidlTest::createDemuxWithFrontend(int32_t frontendId) {
+::testing::AssertionResult TunerHidlTest::createDemuxWithFrontend(int32_t frontendId,
+                                                                  FrontendSettings settings) {
     Result status;
 
     if (!mDemux && createDemux() == ::testing::AssertionFailure()) {
@@ -641,6 +611,8 @@
         return ::testing::AssertionFailure();
     }
 
+    mFrontendCallback->testOnEvent(mFrontend, settings);
+
     status = mDemux->setFrontendDataSource(frontendId);
 
     return ::testing::AssertionResult(status == Result::SUCCESS);
@@ -705,7 +677,7 @@
         mDemuxCallback = new DemuxCallback();
     }
 
-    // Add section filter to the local demux
+    // Add playback input to the local demux
     status = mDemux->addInput(FMQ_SIZE_1M, mDemuxCallback);
 
     if (status != Result::SUCCESS) {
@@ -732,29 +704,6 @@
     return ::testing::AssertionResult(status == Result::SUCCESS);
 }
 
-// Legacy
-::testing::AssertionResult TunerHidlTest::addSectionFilterToDemux() {
-    Result status;
-
-    if (!mDemux && createDemux() == ::testing::AssertionFailure()) {
-        return ::testing::AssertionFailure();
-    }
-
-    // Create demux callback
-    if (!mDemuxCallback) {
-        mDemuxCallback = new DemuxCallback();
-    }
-
-    // Add section filter to the local demux
-    mDemux->addFilter(DemuxFilterType::SECTION, FMQ_SIZE_4K, mDemuxCallback,
-                      [&](Result result, uint32_t filterId) {
-                          mFilterId = filterId;
-                          status = result;
-                      });
-
-    return ::testing::AssertionResult(status == Result::SUCCESS);
-}
-
 ::testing::AssertionResult TunerHidlTest::addFilterToDemux(DemuxFilterType type,
                                                            DemuxFilterSettings setting) {
     Result status;
@@ -800,36 +749,10 @@
     return ::testing::AssertionResult(status == Result::SUCCESS);
 }
 
-// Legacy
-::testing::AssertionResult TunerHidlTest::readSectionFilterDataOutput() {
-    // Filter Configuration Module
-    DemuxInputSettings setting{
-            .statusMask = 0xf,
-            .lowThreshold = 0x1000,
-            .highThreshold = 0x100000,
-            .dataFormat = DemuxDataFormat::TS,
-            .packetSize = 188,
-    };
-    if (addSectionFilterToDemux() == ::testing::AssertionFailure() ||
-        getFilterMQDescriptor(mFilterId) == ::testing::AssertionFailure() ||
-        addInputToDemux(setting) == ::testing::AssertionFailure() ||
-        getInputMQDescriptor() == ::testing::AssertionFailure()) {
-        return ::testing::AssertionFailure();
-    }
-
-    // Data Verify Module
-    // Test start filter and read the output data
-    mDemuxCallback->testOnSectionFilterEvent(mDemux, mFilterId, mFilterMQDescriptor,
-                                             mInputMQDescriptor);
-
-    // Clean Up Module
-    return closeDemux();  //::testing::AssertionSuccess();
-}
-
-::testing::AssertionResult TunerHidlTest::playbackDataFlowTest(vector<FilterConf> filterConf,
-                                                               InputConf inputConf,
-                                                               vector<string> goldenOutputFiles) {
+::testing::AssertionResult TunerHidlTest::playbackDataFlowTest(
+        vector<FilterConf> filterConf, InputConf inputConf, vector<string> /*goldenOutputFiles*/) {
     Result status;
+    int filterIdsSize;
     // Filter Configuration Module
     for (int i = 0; i < filterConf.size(); i++) {
         if (addFilterToDemux(filterConf[i].type, filterConf[i].setting) ==
@@ -838,8 +761,11 @@
             getFilterMQDescriptor(mFilterId) == ::testing::AssertionFailure()) {
             return ::testing::AssertionFailure();
         }
+        filterIdsSize = mUsedFilterIds.size();
+        mUsedFilterIds.resize(filterIdsSize + 1);
+        mUsedFilterIds[filterIdsSize] = mFilterId;
         mDemuxCallback->updateFilterMQ(mFilterId, mFilterMQDescriptor);
-        mDemuxCallback->updateGoldenOutputMap(mFilterId, goldenOutputFiles[i]);
+        // mDemuxCallback->updateGoldenOutputMap(mFilterId, goldenOutputFiles[i]);
         status = mDemux->startFilter(mFilterId);
         if (status != Result::SUCCESS) {
             return ::testing::AssertionFailure();
@@ -860,8 +786,76 @@
 
     // Data Verify Module
     mDemuxCallback->testFilterDataOutput();
+    mDemuxCallback->stopInputThread();
 
     // Clean Up Module
+    for (int i = 0; i <= filterIdsSize; i++) {
+        if (mDemux->stopFilter(mUsedFilterIds[i]) != Result::SUCCESS) {
+            return ::testing::AssertionFailure();
+        }
+    }
+    if (mDemux->stopInput() != Result::SUCCESS) {
+        return ::testing::AssertionFailure();
+    }
+    return closeDemux();
+}
+
+::testing::AssertionResult TunerHidlTest::broadcastDataFlowTest(
+        vector<FilterConf> filterConf, vector<string> /*goldenOutputFiles*/) {
+    Result status;
+    hidl_vec<FrontendId> feIds;
+
+    mService->getFrontendIds([&](Result result, const hidl_vec<FrontendId>& frontendIds) {
+        status = result;
+        feIds = frontendIds;
+    });
+
+    if (feIds.size() == 0) {
+        ALOGW("[   WARN   ] Frontend isn't available");
+        return ::testing::AssertionFailure();
+    }
+
+    FrontendDvbtSettings dvbt{
+            .frequency = 1000,
+    };
+    FrontendSettings settings;
+    settings.dvbt(dvbt);
+
+    if (createDemuxWithFrontend(feIds[0], settings) != ::testing::AssertionSuccess()) {
+        return ::testing::AssertionFailure();
+    }
+
+    int filterIdsSize;
+    // Filter Configuration Module
+    for (int i = 0; i < filterConf.size(); i++) {
+        if (addFilterToDemux(filterConf[i].type, filterConf[i].setting) ==
+                    ::testing::AssertionFailure() ||
+            // TODO use a map to save the FMQs/EvenFlags and pass to callback
+            getFilterMQDescriptor(mFilterId) == ::testing::AssertionFailure()) {
+            return ::testing::AssertionFailure();
+        }
+        filterIdsSize = mUsedFilterIds.size();
+        mUsedFilterIds.resize(filterIdsSize + 1);
+        mUsedFilterIds[filterIdsSize] = mFilterId;
+        mDemuxCallback->updateFilterMQ(mFilterId, mFilterMQDescriptor);
+        status = mDemux->startFilter(mFilterId);
+        if (status != Result::SUCCESS) {
+            return ::testing::AssertionFailure();
+        }
+    }
+
+    // Data Verify Module
+    mDemuxCallback->testFilterDataOutput();
+
+    // Clean Up Module
+    for (int i = 0; i <= filterIdsSize; i++) {
+        if (mDemux->stopFilter(mUsedFilterIds[i]) != Result::SUCCESS) {
+            return ::testing::AssertionFailure();
+        }
+    }
+    if (mFrontend->stopTune() != Result::SUCCESS) {
+        return ::testing::AssertionFailure();
+    }
     return closeDemux();
 }
 
@@ -948,7 +942,7 @@
     }
 }
 
-TEST_F(TunerHidlTest, CreateDemuxWithFrontend) {
+/*TEST_F(TunerHidlTest, CreateDemuxWithFrontend) {
     Result status;
     hidl_vec<FrontendId> feIds;
 
@@ -963,10 +957,17 @@
         return;
     }
 
+    FrontendDvbtSettings dvbt{
+        .frequency = 1000,
+    };
+    FrontendSettings settings;
+    settings.dvbt(dvbt);
+
     for (size_t i = 0; i < feIds.size(); i++) {
-        ASSERT_TRUE(createDemuxWithFrontend(feIds[i]));
+        ASSERT_TRUE(createDemuxWithFrontend(feIds[i], settings));
+        mFrontend->stopTune();
     }
-}
+}*/
 
 TEST_F(TunerHidlTest, CreateDemux) {
     description("Create Demux");
@@ -991,9 +992,63 @@
 /*
  * DATA FLOW TESTS
  */
-TEST_F(TunerHidlTest, ReadSectionFilterOutput) {
-    description("Read data output from FMQ of a Section Filter");
-    ASSERT_TRUE(readSectionFilterDataOutput());
+TEST_F(TunerHidlTest, PlaybackDataFlowWithPesFilterTest) {
+    description("Feed ts data from playback and configure pes filter to get output");
+
+    // todo modulize the filter conf parser
+    vector<FilterConf> filterConf;
+    filterConf.resize(1);
+
+    DemuxFilterSettings filterSetting;
+    DemuxFilterPesDataSettings pesFilterSetting{
+            .tpid = 18,
+    };
+    filterSetting.pesData(pesFilterSetting);
+    FilterConf pesFilterConf{
+            .type = DemuxFilterType::PES,
+            .setting = filterSetting,
+    };
+    filterConf[0] = pesFilterConf;
+
+    DemuxInputSettings inputSetting{
+            .statusMask = 0xf,
+            .lowThreshold = 0x1000,
+            .highThreshold = 0x07fff,
+            .dataFormat = DemuxDataFormat::TS,
+            .packetSize = 188,
+    };
+
+    InputConf inputConf{
+            .inputDataFile = "/vendor/etc/test1.ts",
+            .setting = inputSetting,
+    };
+
+    vector<string> goldenOutputFiles;
+
+    ASSERT_TRUE(playbackDataFlowTest(filterConf, inputConf, goldenOutputFiles));
+}
+
+TEST_F(TunerHidlTest, BroadcastDataFlowWithPesFilterTest) {
+    description("Feed ts data from frontend and test with PES filter");
+
+    // todo modulize the filter conf parser
+    vector<FilterConf> filterConf;
+    filterConf.resize(1);
+
+    DemuxFilterSettings filterSetting;
+    DemuxFilterPesDataSettings pesFilterSetting{
+            .tpid = 18,
+    };
+    filterSetting.pesData(pesFilterSetting);
+    FilterConf pesFilterConf{
+            .type = DemuxFilterType::PES,
+            .setting = filterSetting,
+    };
+    filterConf[0] = pesFilterConf;
+
+    vector<string> goldenOutputFiles;
+
+    ASSERT_TRUE(broadcastDataFlowTest(filterConf, goldenOutputFiles));
 }
 
 }  // namespace
diff --git a/vibrator/1.4/Android.bp b/vibrator/1.4/Android.bp
new file mode 100644
index 0000000..cf31fcd
--- /dev/null
+++ b/vibrator/1.4/Android.bp
@@ -0,0 +1,22 @@
+// This file is autogenerated by hidl-gen -Landroidbp.
+
+hidl_interface {
+    name: "android.hardware.vibrator@1.4",
+    root: "android.hardware",
+    vndk: {
+        enabled: true,
+    },
+    srcs: [
+        "types.hal",
+        "IVibrator.hal",
+        "IVibratorCallback.hal",
+    ],
+    interfaces: [
+        "android.hardware.vibrator@1.0",
+        "android.hardware.vibrator@1.1",
+        "android.hardware.vibrator@1.2",
+        "android.hardware.vibrator@1.3",
+        "android.hidl.base@1.0",
+    ],
+    gen_java: true,
+}
diff --git a/vibrator/1.4/IVibrator.hal b/vibrator/1.4/IVibrator.hal
new file mode 100644
index 0000000..913abe3
--- /dev/null
+++ b/vibrator/1.4/IVibrator.hal
@@ -0,0 +1,57 @@
+/*
+ * Copyright (C) 2019 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package android.hardware.vibrator@1.4;
+
+import @1.0::EffectStrength;
+import @1.3::Effect;
+import @1.0::Status;
+import @1.3::IVibrator;
+import IVibratorCallback;
+
+interface IVibrator extends @1.3::IVibrator {
+    /**
+     * Determine capabilities of the vibrator HAL.
+     */
+    getCapabilities() generates (bitfield<Capabilities> capabilities);
+
+    /**
+     * Turn on vibrator
+     *
+     * This function must only be called after the previous timeout has expired or
+     * was canceled (through off()).
+     * @param timeoutMs number of milliseconds to vibrate.
+     * @param callback A callback used to inform Frameworks of state change, if supported.
+     * @return vibratorOnRet whether vibrator command was successful or not.
+     */
+    on_1_4(uint32_t timeoutMs, IVibratorCallback callback) generates (Status vibratorOnRet);
+
+    /**
+     * Fire off a predefined haptic event.
+     *
+     * @param effect The type of haptic event to trigger.
+     * @param strength The intensity of haptic event to trigger.
+     * @param callback A callback used to inform Frameworks of state change, if supported.
+     * @return status Whether the effect was successfully performed or not. Must
+     *     return Status::UNSUPPORTED_OPERATION if the effect is not supported.
+     * @return lengthMs The length of time the event is expected to take in
+     *     milliseconds. This doesn't need to be perfectly accurate, but should be a reasonable
+     *     approximation. Should be a positive, non-zero value if the returned status is Status::OK,
+     *     and set to 0 otherwise.
+     */
+    perform_1_4(Effect effect, EffectStrength strength, IVibratorCallback callback)
+        generates (Status status, uint32_t lengthMs);
+};
diff --git a/vibrator/1.4/IVibratorCallback.hal b/vibrator/1.4/IVibratorCallback.hal
new file mode 100644
index 0000000..76281bc
--- /dev/null
+++ b/vibrator/1.4/IVibratorCallback.hal
@@ -0,0 +1,21 @@
+/*
+ * Copyright (C) 2019 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package android.hardware.vibrator@1.4;
+
+interface IVibratorCallback {
+    oneway onComplete();
+};
diff --git a/vibrator/1.4/types.hal b/vibrator/1.4/types.hal
new file mode 100644
index 0000000..acc49b1
--- /dev/null
+++ b/vibrator/1.4/types.hal
@@ -0,0 +1,22 @@
+/*
+ * Copyright (C) 2019 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package android.hardware.vibrator@1.4;
+
+enum Capabilities : uint32_t {
+    ON_COMPLETION_CALLBACK = 1 << 0,
+    PERFORM_COMPLETION_CALLBACK = 1 << 1,
+};