diff --git a/common/aidl/Android.bp b/common/aidl/Android.bp
new file mode 100644
index 0000000..f55e799
--- /dev/null
+++ b/common/aidl/Android.bp
@@ -0,0 +1,21 @@
+aidl_interface {
+    name: "android.hardware.common",
+    host_supported: true,
+    vendor_available: true,
+    vndk: {
+        enabled: true,
+        support_system_process: true,
+    },
+    srcs: [
+        "android/hardware/common/*.aidl",
+    ],
+    stability: "vintf",
+    backend: {
+        java: {
+            enabled: false,
+        },
+        cpp: {
+            enabled: false,
+        },
+    },
+}
diff --git a/common/aidl/aidl_api/android.hardware.common/current/android/hardware/common/NativeHandle.aidl b/common/aidl/aidl_api/android.hardware.common/current/android/hardware/common/NativeHandle.aidl
new file mode 100644
index 0000000..f37b7d5
--- /dev/null
+++ b/common/aidl/aidl_api/android.hardware.common/current/android/hardware/common/NativeHandle.aidl
@@ -0,0 +1,23 @@
+///////////////////////////////////////////////////////////////////////////////
+// THIS FILE IS IMMUTABLE. DO NOT EDIT IN ANY CASE.                          //
+///////////////////////////////////////////////////////////////////////////////
+
+// This file is a snapshot of an AIDL interface (or parcelable). Do not try to
+// edit this file. It looks like you are doing that because you have modified
+// an AIDL interface in a backward-incompatible way, e.g., deleting a function
+// from an interface or a field from a parcelable and it broke the build. That
+// breakage is intended.
+//
+// You must not make a backward incompatible changes to the AIDL files built
+// with the aidl_interface module type with versions property set. The module
+// type is used to build AIDL files in a way that they can be used across
+// independently updatable components of the system. If a device is shipped
+// with such a backward incompatible change, it has a high risk of breaking
+// later when a module using the interface is updated, e.g., Mainline modules.
+
+package android.hardware.common;
+@VintfStability
+parcelable NativeHandle {
+  ParcelFileDescriptor[] fds;
+  int[] ints;
+}
diff --git a/common/aidl/android/hardware/common/NativeHandle.aidl b/common/aidl/android/hardware/common/NativeHandle.aidl
new file mode 100644
index 0000000..2c250a2
--- /dev/null
+++ b/common/aidl/android/hardware/common/NativeHandle.aidl
@@ -0,0 +1,26 @@
+/*
+ * 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.common;
+
+/**
+ * Representation of a native handle.
+ */
+@VintfStability
+parcelable NativeHandle {
+    ParcelFileDescriptor[] fds;
+    int[] ints;
+}
diff --git a/graphics/common/aidl/Android.bp b/graphics/common/aidl/Android.bp
new file mode 100644
index 0000000..601cabc
--- /dev/null
+++ b/graphics/common/aidl/Android.bp
@@ -0,0 +1,24 @@
+aidl_interface {
+    name: "android.hardware.graphics.common",
+    host_supported: true,
+    vendor_available: true,
+    vndk: {
+        enabled: true,
+        support_system_process: true,
+    },
+    srcs: [
+        "android/hardware/graphics/common/*.aidl",
+    ],
+    stability: "vintf",
+    imports: [
+        "android.hardware.common",
+    ],
+    backend: {
+        java: {
+            enabled: false,
+        },
+        cpp: {
+            enabled: false,
+        },
+    },
+}
diff --git a/graphics/common/aidl/aidl_api/android.hardware.graphics.common/current/android/hardware/graphics/common/BufferUsage.aidl b/graphics/common/aidl/aidl_api/android.hardware.graphics.common/current/android/hardware/graphics/common/BufferUsage.aidl
new file mode 100644
index 0000000..58eefc4
--- /dev/null
+++ b/graphics/common/aidl/aidl_api/android.hardware.graphics.common/current/android/hardware/graphics/common/BufferUsage.aidl
@@ -0,0 +1,47 @@
+///////////////////////////////////////////////////////////////////////////////
+// THIS FILE IS IMMUTABLE. DO NOT EDIT IN ANY CASE.                          //
+///////////////////////////////////////////////////////////////////////////////
+
+// This file is a snapshot of an AIDL interface (or parcelable). Do not try to
+// edit this file. It looks like you are doing that because you have modified
+// an AIDL interface in a backward-incompatible way, e.g., deleting a function
+// from an interface or a field from a parcelable and it broke the build. That
+// breakage is intended.
+//
+// You must not make a backward incompatible changes to the AIDL files built
+// with the aidl_interface module type with versions property set. The module
+// type is used to build AIDL files in a way that they can be used across
+// independently updatable components of the system. If a device is shipped
+// with such a backward incompatible change, it has a high risk of breaking
+// later when a module using the interface is updated, e.g., Mainline modules.
+
+package android.hardware.graphics.common;
+@Backing(type="long") @VintfStability
+enum BufferUsage {
+  CPU_READ_MASK = 15,
+  CPU_READ_NEVER = 0,
+  CPU_READ_RARELY = 2,
+  CPU_READ_OFTEN = 3,
+  CPU_WRITE_MASK = 240,
+  CPU_WRITE_NEVER = 0,
+  CPU_WRITE_RARELY = 32,
+  CPU_WRITE_OFTEN = 48,
+  GPU_TEXTURE = 256,
+  GPU_RENDER_TARGET = 512,
+  COMPOSER_OVERLAY = 2048,
+  COMPOSER_CLIENT_TARGET = 4096,
+  PROTECTED = 16384,
+  COMPOSER_CURSOR = 32768,
+  VIDEO_ENCODER = 65536,
+  CAMERA_OUTPUT = 131072,
+  CAMERA_INPUT = 262144,
+  RENDERSCRIPT = 1048576,
+  VIDEO_DECODER = 4194304,
+  SENSOR_DIRECT_DATA = 8388608,
+  GPU_CUBE_MAP = 33554432,
+  GPU_MIPMAP_COMPLETE = 67108864,
+  HW_IMAGE_ENCODER = 134217728,
+  GPU_DATA_BUFFER = 16777216,
+  VENDOR_MASK = -268435456,
+  VENDOR_MASK_HI = -281474976710656,
+}
diff --git a/graphics/common/aidl/aidl_api/android.hardware.graphics.common/current/android/hardware/graphics/common/HardwareBuffer.aidl b/graphics/common/aidl/aidl_api/android.hardware.graphics.common/current/android/hardware/graphics/common/HardwareBuffer.aidl
new file mode 100644
index 0000000..72222e1
--- /dev/null
+++ b/graphics/common/aidl/aidl_api/android.hardware.graphics.common/current/android/hardware/graphics/common/HardwareBuffer.aidl
@@ -0,0 +1,23 @@
+///////////////////////////////////////////////////////////////////////////////
+// THIS FILE IS IMMUTABLE. DO NOT EDIT IN ANY CASE.                          //
+///////////////////////////////////////////////////////////////////////////////
+
+// This file is a snapshot of an AIDL interface (or parcelable). Do not try to
+// edit this file. It looks like you are doing that because you have modified
+// an AIDL interface in a backward-incompatible way, e.g., deleting a function
+// from an interface or a field from a parcelable and it broke the build. That
+// breakage is intended.
+//
+// You must not make a backward incompatible changes to the AIDL files built
+// with the aidl_interface module type with versions property set. The module
+// type is used to build AIDL files in a way that they can be used across
+// independently updatable components of the system. If a device is shipped
+// with such a backward incompatible change, it has a high risk of breaking
+// later when a module using the interface is updated, e.g., Mainline modules.
+
+package android.hardware.graphics.common;
+@VintfStability
+parcelable HardwareBuffer {
+  android.hardware.graphics.common.HardwareBufferDescription description;
+  android.hardware.common.NativeHandle handle;
+}
diff --git a/graphics/common/aidl/aidl_api/android.hardware.graphics.common/current/android/hardware/graphics/common/HardwareBufferDescription.aidl b/graphics/common/aidl/aidl_api/android.hardware.graphics.common/current/android/hardware/graphics/common/HardwareBufferDescription.aidl
new file mode 100644
index 0000000..8b12169
--- /dev/null
+++ b/graphics/common/aidl/aidl_api/android.hardware.graphics.common/current/android/hardware/graphics/common/HardwareBufferDescription.aidl
@@ -0,0 +1,27 @@
+///////////////////////////////////////////////////////////////////////////////
+// THIS FILE IS IMMUTABLE. DO NOT EDIT IN ANY CASE.                          //
+///////////////////////////////////////////////////////////////////////////////
+
+// This file is a snapshot of an AIDL interface (or parcelable). Do not try to
+// edit this file. It looks like you are doing that because you have modified
+// an AIDL interface in a backward-incompatible way, e.g., deleting a function
+// from an interface or a field from a parcelable and it broke the build. That
+// breakage is intended.
+//
+// You must not make a backward incompatible changes to the AIDL files built
+// with the aidl_interface module type with versions property set. The module
+// type is used to build AIDL files in a way that they can be used across
+// independently updatable components of the system. If a device is shipped
+// with such a backward incompatible change, it has a high risk of breaking
+// later when a module using the interface is updated, e.g., Mainline modules.
+
+package android.hardware.graphics.common;
+@VintfStability
+parcelable HardwareBufferDescription {
+  int width;
+  int height;
+  int layers;
+  android.hardware.graphics.common.PixelFormat format;
+  android.hardware.graphics.common.BufferUsage usage;
+  int stride;
+}
diff --git a/graphics/common/aidl/aidl_api/android.hardware.graphics.common/current/android/hardware/graphics/common/PixelFormat.aidl b/graphics/common/aidl/aidl_api/android.hardware.graphics.common/current/android/hardware/graphics/common/PixelFormat.aidl
new file mode 100644
index 0000000..e5f0470
--- /dev/null
+++ b/graphics/common/aidl/aidl_api/android.hardware.graphics.common/current/android/hardware/graphics/common/PixelFormat.aidl
@@ -0,0 +1,50 @@
+///////////////////////////////////////////////////////////////////////////////
+// THIS FILE IS IMMUTABLE. DO NOT EDIT IN ANY CASE.                          //
+///////////////////////////////////////////////////////////////////////////////
+
+// This file is a snapshot of an AIDL interface (or parcelable). Do not try to
+// edit this file. It looks like you are doing that because you have modified
+// an AIDL interface in a backward-incompatible way, e.g., deleting a function
+// from an interface or a field from a parcelable and it broke the build. That
+// breakage is intended.
+//
+// You must not make a backward incompatible changes to the AIDL files built
+// with the aidl_interface module type with versions property set. The module
+// type is used to build AIDL files in a way that they can be used across
+// independently updatable components of the system. If a device is shipped
+// with such a backward incompatible change, it has a high risk of breaking
+// later when a module using the interface is updated, e.g., Mainline modules.
+
+package android.hardware.graphics.common;
+@Backing(type="int") @VintfStability
+enum PixelFormat {
+  UNSPECIFIED = 0,
+  RGBA_8888 = 1,
+  RGBX_8888 = 2,
+  RGB_888 = 3,
+  RGB_565 = 4,
+  BGRA_8888 = 5,
+  YCBCR_422_SP = 16,
+  YCRCB_420_SP = 17,
+  YCBCR_422_I = 20,
+  RGBA_FP16 = 22,
+  RAW16 = 32,
+  BLOB = 33,
+  IMPLEMENTATION_DEFINED = 34,
+  YCBCR_420_888 = 35,
+  RAW_OPAQUE = 36,
+  RAW10 = 37,
+  RAW12 = 38,
+  RGBA_1010102 = 43,
+  Y8 = 538982489,
+  Y16 = 540422489,
+  YV12 = 842094169,
+  DEPTH_16 = 48,
+  DEPTH_24 = 49,
+  DEPTH_24_STENCIL_8 = 50,
+  DEPTH_32F = 51,
+  DEPTH_32F_STENCIL_8 = 52,
+  STENCIL_8 = 53,
+  YCBCR_P010 = 54,
+  HSV_888 = 55,
+}
diff --git a/graphics/common/aidl/android/hardware/graphics/common/BufferUsage.aidl b/graphics/common/aidl/android/hardware/graphics/common/BufferUsage.aidl
new file mode 100644
index 0000000..d978f46
--- /dev/null
+++ b/graphics/common/aidl/android/hardware/graphics/common/BufferUsage.aidl
@@ -0,0 +1,114 @@
+/*
+ * 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.graphics.common;
+
+/**
+ * Buffer usage definitions.
+ */
+@VintfStability
+@Backing(type="long")
+enum BufferUsage {
+    /** bit 0-3 is an enum */
+    CPU_READ_MASK                      = 0xf,
+    /** buffer is never read by CPU */
+    CPU_READ_NEVER                     = 0,
+    /** buffer is rarely read by CPU */
+    CPU_READ_RARELY                    = 2,
+    /** buffer is often read by CPU */
+    CPU_READ_OFTEN                     = 3,
+
+    /** bit 4-7 is an enum */
+    CPU_WRITE_MASK                     = 0xf << 4,
+    /** buffer is never written by CPU */
+    CPU_WRITE_NEVER                    = 0 << 4,
+    /** buffer is rarely written by CPU */
+    CPU_WRITE_RARELY                   = 2 << 4,
+    /** buffer is often written by CPU */
+    CPU_WRITE_OFTEN                    = 3 << 4,
+
+    /** buffer is used as a GPU texture */
+    GPU_TEXTURE                        = 1 << 8,
+
+    /** buffer is used as a GPU render target */
+    GPU_RENDER_TARGET                  = 1 << 9,
+
+    /** bit 10 must be zero */
+
+    /** buffer is used as a composer HAL overlay layer */
+    COMPOSER_OVERLAY                   = 1 << 11,
+    /** buffer is used as a composer HAL client target */
+    COMPOSER_CLIENT_TARGET             = 1 << 12,
+
+    /** bit 13 must be zero */
+
+    /**
+     * Buffer is allocated with hardware-level protection against copying the
+     * contents (or information derived from the contents) into unprotected
+     * memory.
+     */
+    PROTECTED                          = 1 << 14,
+
+    /** buffer is used as a hwcomposer HAL cursor layer */
+    COMPOSER_CURSOR                    = 1 << 15,
+
+    /** buffer is used as a video encoder input */
+    VIDEO_ENCODER                      = 1 << 16,
+
+    /** buffer is used as a camera HAL output */
+    CAMERA_OUTPUT                      = 1 << 17,
+
+    /** buffer is used as a camera HAL input */
+    CAMERA_INPUT                       = 1 << 18,
+
+    /** bit 19 must be zero */
+
+    /** buffer is used as a renderscript allocation */
+    RENDERSCRIPT                       = 1 << 20,
+
+    /** bit 21 must be zero */
+
+    /** buffer is used as a video decoder output */
+    VIDEO_DECODER                      = 1 << 22,
+
+    /** buffer is used as a sensor direct report output */
+    SENSOR_DIRECT_DATA                 = 1 << 23,
+
+    /** buffer is used as a cube map texture */
+    GPU_CUBE_MAP                       = 1 << 25,
+
+    /** buffer contains a complete mipmap hierarchy */
+    GPU_MIPMAP_COMPLETE                = 1 << 26,
+
+    /**
+     * Buffer is used as input for HEIC encoder.
+     */
+    HW_IMAGE_ENCODER                   = 1 << 27,
+
+    /**
+     * buffer is used as as an OpenGL shader storage or uniform
+     * buffer object
+     */
+    GPU_DATA_BUFFER                    = 1 << 24,
+
+    /** bits 25-27 must be zero and are reserved for future versions */
+    /** bits 28-31 are reserved for vendor extensions */
+    VENDOR_MASK                        = 0xf << 28,
+
+    /** bits 32-47 must be zero and are reserved for future versions */
+    /** bits 48-63 are reserved for vendor extensions */
+    VENDOR_MASK_HI                     = (1L * 0xffff) << 48,
+}
diff --git a/graphics/common/aidl/android/hardware/graphics/common/HardwareBuffer.aidl b/graphics/common/aidl/android/hardware/graphics/common/HardwareBuffer.aidl
new file mode 100644
index 0000000..5a22c0f
--- /dev/null
+++ b/graphics/common/aidl/android/hardware/graphics/common/HardwareBuffer.aidl
@@ -0,0 +1,32 @@
+/*
+ * 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.graphics.common;
+
+import android.hardware.common.NativeHandle;
+import android.hardware.graphics.common.HardwareBufferDescription;
+
+/**
+ * Stable AIDL counterpart of AHardwareBuffer.
+ *
+ * @note This is different from the public HardwareBuffer.
+ * @sa +ndk libnativewindow#AHardwareBuffer
+ */
+@VintfStability
+parcelable HardwareBuffer {
+    HardwareBufferDescription description;
+    NativeHandle handle;
+}
diff --git a/graphics/common/aidl/android/hardware/graphics/common/HardwareBufferDescription.aidl b/graphics/common/aidl/android/hardware/graphics/common/HardwareBufferDescription.aidl
new file mode 100644
index 0000000..e1e3492
--- /dev/null
+++ b/graphics/common/aidl/android/hardware/graphics/common/HardwareBufferDescription.aidl
@@ -0,0 +1,35 @@
+/*
+ * 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.graphics.common;
+
+import android.hardware.graphics.common.BufferUsage;
+import android.hardware.graphics.common.PixelFormat;
+
+/**
+ * Stable AIDL counterpart of AHardwareBuffer_Desc.
+ *
+ * @sa +ndk libnativewindow#AHardwareBuffer_Desc
+ */
+@VintfStability
+parcelable HardwareBufferDescription {
+    int width;
+    int height;
+    int layers;
+    PixelFormat format;
+    BufferUsage usage;
+    int stride;
+}
diff --git a/graphics/common/aidl/android/hardware/graphics/common/PixelFormat.aidl b/graphics/common/aidl/android/hardware/graphics/common/PixelFormat.aidl
new file mode 100644
index 0000000..4942462
--- /dev/null
+++ b/graphics/common/aidl/android/hardware/graphics/common/PixelFormat.aidl
@@ -0,0 +1,516 @@
+/*
+ * 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.graphics.common;
+
+/**
+ * Pixel formats for graphics buffers.
+ */
+@VintfStability
+@Backing(type="int")
+enum PixelFormat {
+    /**
+     * This value may be used in an operation where the format is optional.
+     */
+    UNSPECIFIED                        = 0,
+    /**
+     * 32-bit format that has 8-bit R, G, B, and A components, in that order,
+     * from the lowest memory address to the highest memory address.
+     *
+     * The component values are unsigned normalized to the range [0, 1], whose
+     * interpretation is defined by the dataspace.
+     */
+    RGBA_8888                          = 0x1,
+
+    /**
+     * 32-bit format that has 8-bit R, G, B, and unused components, in that
+     * order, from the lowest memory address to the highest memory address.
+     *
+     * The component values are unsigned normalized to the range [0, 1], whose
+     * interpretation is defined by the dataspace.
+     */
+    RGBX_8888                          = 0x2,
+
+    /**
+     * 24-bit format that has 8-bit R, G, and B components, in that order,
+     * from the lowest memory address to the highest memory address.
+     *
+     * The component values are unsigned normalized to the range [0, 1], whose
+     * interpretation is defined by the dataspace.
+     */
+    RGB_888                            = 0x3,
+
+    /**
+     * 16-bit packed format that has 5-bit R, 6-bit G, and 5-bit B components,
+     * in that order, from the most-sigfinicant bits to the least-significant
+     * bits.
+     *
+     * The component values are unsigned normalized to the range [0, 1], whose
+     * interpretation is defined by the dataspace.
+     */
+    RGB_565                            = 0x4,
+
+    /**
+     * 32-bit format that has 8-bit B, G, R, and A components, in that order,
+     * from the lowest memory address to the highest memory address.
+     *
+     * The component values are unsigned normalized to the range [0, 1], whose
+     * interpretation is defined by the dataspace.
+     */
+    BGRA_8888                          = 0x5,
+
+    /**
+     * Legacy formats deprecated in favor of YCBCR_420_888.
+     */
+    YCBCR_422_SP                       = 0x10,  // NV16
+    YCRCB_420_SP                       = 0x11,  // NV21
+    YCBCR_422_I                        = 0x14,  // YUY2
+
+    /**
+     * 64-bit format that has 16-bit R, G, B, and A components, in that order,
+     * from the lowest memory address to the highest memory address.
+     *
+     * The component values are signed floats, whose interpretation is defined
+     * by the dataspace.
+     */
+    RGBA_FP16                          = 0x16,
+
+    /**
+     * RAW16 is a single-channel, 16-bit, little endian format, typically
+     * representing raw Bayer-pattern images from an image sensor, with minimal
+     * processing.
+     *
+     * The exact pixel layout of the data in the buffer is sensor-dependent, and
+     * needs to be queried from the camera device.
+     *
+     * Generally, not all 16 bits are used; more common values are 10 or 12
+     * bits. If not all bits are used, the lower-order bits are filled first.
+     * All parameters to interpret the raw data (black and white points,
+     * color space, etc) must be queried from the camera device.
+     *
+     * This format assumes
+     * - an even width
+     * - an even height
+     * - a horizontal stride multiple of 16 pixels
+     * - a vertical stride equal to the height
+     * - strides are specified in pixels, not in bytes
+     *
+     *   size = stride * height * 2
+     *
+     * This format must be accepted by the allocator when used with the
+     * following usage flags:
+     *
+     *    - BufferUsage::CAMERA_*
+     *    - BufferUsage::CPU_*
+     *    - BufferUsage::RENDERSCRIPT
+     *
+     * The mapping of the dataspace to buffer contents for RAW16 is as
+     * follows:
+     *
+     *  Dataspace value               | Buffer contents
+     * -------------------------------+-----------------------------------------
+     *  Dataspace::ARBITRARY          | Raw image sensor data, layout is as
+     *                                | defined above.
+     *  Dataspace::DEPTH              | Unprocessed implementation-dependent raw
+     *                                | depth measurements, opaque with 16 bit
+     *                                | samples.
+     *  Other                         | Unsupported
+     */
+    RAW16                              = 0x20,
+
+    /**
+     * BLOB is used to carry task-specific data which does not have a standard
+     * image structure. The details of the format are left to the two
+     * endpoints.
+     *
+     * A typical use case is for transporting JPEG-compressed images from the
+     * Camera HAL to the framework or to applications.
+     *
+     * Buffers of this format must have a height of 1, and width equal to their
+     * size in bytes.
+     *
+     * The mapping of the dataspace to buffer contents for BLOB is as
+     * follows:
+     *
+     *  Dataspace value               | Buffer contents
+     * -------------------------------+-----------------------------------------
+     *  Dataspace::JFIF               | An encoded JPEG image
+     *  Dataspace::DEPTH              | An android_depth_points buffer
+     *  Dataspace::SENSOR             | Sensor event data.
+     *  Other                         | Unsupported
+     */
+    BLOB                               = 0x21,
+
+    /**
+     * A format indicating that the choice of format is entirely up to the
+     * allocator.
+     *
+     * The allocator should examine the usage bits passed in when allocating a
+     * buffer with this format, and it should derive the pixel format from
+     * those usage flags. This format must never be used with any of the
+     * BufferUsage::CPU_* usage flags.
+     *
+     * Even when the internally chosen format has an alpha component, the
+     * clients must assume the alpha vlaue to be 1.0.
+     *
+     * The interpretation of the component values is defined by the dataspace.
+     */
+    IMPLEMENTATION_DEFINED             = 0x22,
+
+    /**
+     * This format allows platforms to use an efficient YCbCr/YCrCb 4:2:0
+     * buffer layout, while still describing the general format in a
+     * layout-independent manner. While called YCbCr, it can be used to
+     * describe formats with either chromatic ordering, as well as
+     * whole planar or semiplanar layouts.
+     *
+     * This format must be accepted by the allocator when BufferUsage::CPU_*
+     * are set.
+     *
+     * Buffers with this format must be locked with IMapper::lockYCbCr.
+     * Locking with IMapper::lock must return an error.
+     *
+     * The interpretation of the component values is defined by the dataspace.
+     */
+    YCBCR_420_888                      = 0x23,
+
+    /**
+     * RAW_OPAQUE is a format for unprocessed raw image buffers coming from an
+     * image sensor. The actual structure of buffers of this format is
+     * implementation-dependent.
+     *
+     * This format must be accepted by the allocator when used with the
+     * following usage flags:
+     *
+     *    - BufferUsage::CAMERA_*
+     *    - BufferUsage::CPU_*
+     *    - BufferUsage::RENDERSCRIPT
+     *
+     * The mapping of the dataspace to buffer contents for RAW_OPAQUE is as
+     * follows:
+     *
+     *  Dataspace value               | Buffer contents
+     * -------------------------------+-----------------------------------------
+     *  Dataspace::ARBITRARY          | Raw image sensor data.
+     *  Other                         | Unsupported
+     */
+    RAW_OPAQUE                         = 0x24,
+
+    /**
+     * RAW10 is a single-channel, 10-bit per pixel, densely packed in each row,
+     * unprocessed format, usually representing raw Bayer-pattern images coming from
+     * an image sensor.
+     *
+     * In an image buffer with this format, starting from the first pixel of each
+     * row, each 4 consecutive pixels are packed into 5 bytes (40 bits). Each one
+     * of the first 4 bytes contains the top 8 bits of each pixel, The fifth byte
+     * contains the 2 least significant bits of the 4 pixels, the exact layout data
+     * for each 4 consecutive pixels is illustrated below (Pi[j] stands for the jth
+     * bit of the ith pixel):
+     *
+     *          bit 7                                     bit 0
+     *          =====|=====|=====|=====|=====|=====|=====|=====|
+     * Byte 0: |P0[9]|P0[8]|P0[7]|P0[6]|P0[5]|P0[4]|P0[3]|P0[2]|
+     *         |-----|-----|-----|-----|-----|-----|-----|-----|
+     * Byte 1: |P1[9]|P1[8]|P1[7]|P1[6]|P1[5]|P1[4]|P1[3]|P1[2]|
+     *         |-----|-----|-----|-----|-----|-----|-----|-----|
+     * Byte 2: |P2[9]|P2[8]|P2[7]|P2[6]|P2[5]|P2[4]|P2[3]|P2[2]|
+     *         |-----|-----|-----|-----|-----|-----|-----|-----|
+     * Byte 3: |P3[9]|P3[8]|P3[7]|P3[6]|P3[5]|P3[4]|P3[3]|P3[2]|
+     *         |-----|-----|-----|-----|-----|-----|-----|-----|
+     * Byte 4: |P3[1]|P3[0]|P2[1]|P2[0]|P1[1]|P1[0]|P0[1]|P0[0]|
+     *          ===============================================
+     *
+     * This format assumes
+     * - a width multiple of 4 pixels
+     * - an even height
+     * - a vertical stride equal to the height
+     * - strides are specified in bytes, not in pixels
+     *
+     *   size = stride * height
+     *
+     * When stride is equal to width * (10 / 8), there will be no padding bytes at
+     * the end of each row, the entire image data is densely packed. When stride is
+     * larger than width * (10 / 8), padding bytes will be present at the end of each
+     * row (including the last row).
+     *
+     * This format must be accepted by the allocator when used with the
+     * following usage flags:
+     *
+     *    - BufferUsage::CAMERA_*
+     *    - BufferUsage::CPU_*
+     *    - BufferUsage::RENDERSCRIPT
+     *
+     * The mapping of the dataspace to buffer contents for RAW10 is as
+     * follows:
+     *
+     *  Dataspace value               | Buffer contents
+     * -------------------------------+-----------------------------------------
+     *  Dataspace::ARBITRARY          | Raw image sensor data.
+     *  Other                         | Unsupported
+     */
+    RAW10                              = 0x25,
+
+    /**
+     * RAW12 is a single-channel, 12-bit per pixel, densely packed in each row,
+     * unprocessed format, usually representing raw Bayer-pattern images coming from
+     * an image sensor.
+     *
+     * In an image buffer with this format, starting from the first pixel of each
+     * row, each two consecutive pixels are packed into 3 bytes (24 bits). The first
+     * and second byte contains the top 8 bits of first and second pixel. The third
+     * byte contains the 4 least significant bits of the two pixels, the exact layout
+     * data for each two consecutive pixels is illustrated below (Pi[j] stands for
+     * the jth bit of the ith pixel):
+     *
+     *           bit 7                                            bit 0
+     *          ======|======|======|======|======|======|======|======|
+     * Byte 0: |P0[11]|P0[10]|P0[ 9]|P0[ 8]|P0[ 7]|P0[ 6]|P0[ 5]|P0[ 4]|
+     *         |------|------|------|------|------|------|------|------|
+     * Byte 1: |P1[11]|P1[10]|P1[ 9]|P1[ 8]|P1[ 7]|P1[ 6]|P1[ 5]|P1[ 4]|
+     *         |------|------|------|------|------|------|------|------|
+     * Byte 2: |P1[ 3]|P1[ 2]|P1[ 1]|P1[ 0]|P0[ 3]|P0[ 2]|P0[ 1]|P0[ 0]|
+     *          =======================================================
+     *
+     * This format assumes:
+     * - a width multiple of 4 pixels
+     * - an even height
+     * - a vertical stride equal to the height
+     * - strides are specified in bytes, not in pixels
+     *
+     *   size = stride * height
+     *
+     * When stride is equal to width * (12 / 8), there will be no padding bytes at
+     * the end of each row, the entire image data is densely packed. When stride is
+     * larger than width * (12 / 8), padding bytes will be present at the end of
+     * each row (including the last row).
+     *
+     * This format must be accepted by the allocator when used with the
+     * following usage flags:
+     *
+     *    - BufferUsage::CAMERA_*
+     *    - BufferUsage::CPU_*
+     *    - BufferUsage::RENDERSCRIPT
+     *
+     * The mapping of the dataspace to buffer contents for RAW12 is as
+     * follows:
+     *
+     *  Dataspace value               | Buffer contents
+     * -------------------------------+-----------------------------------------
+     *  Dataspace::ARBITRARY          | Raw image sensor data.
+     *  Other                         | Unsupported
+     */
+    RAW12                              = 0x26,
+
+    /** 0x27 to 0x2A are reserved for flexible formats */
+
+    /**
+     * 32-bit packed format that has 2-bit A, 10-bit B, G, and R components,
+     * in that order, from the most-sigfinicant bits to the least-significant
+     * bits.
+     *
+     * The component values are unsigned normalized to the range [0, 1], whose
+     * interpretation is defined by the dataspace.
+     */
+    RGBA_1010102                       = 0x2B,
+
+    /**
+     * 0x100 - 0x1FF
+     *
+     * This range is reserved for vendor extensions. Formats in this range
+     * must support BufferUsage::GPU_TEXTURE. Clients must assume they do not
+     * have an alpha component.
+     */
+
+    /**
+     * Y8 is a YUV planar format comprised of a WxH Y plane, with each pixel
+     * being represented by 8 bits. It is equivalent to just the Y plane from
+     * YV12.
+     *
+     * This format assumes
+     * - an even width
+     * - an even height
+     * - a horizontal stride multiple of 16 pixels
+     * - a vertical stride equal to the height
+     *
+     *   size = stride * height
+     *
+     * This format must be accepted by the allocator when used with the
+     * following usage flags:
+     *
+     *    - BufferUsage::CAMERA_*
+     *    - BufferUsage::CPU_*
+     *
+     * The component values are unsigned normalized to the range [0, 1], whose
+     * interpretation is defined by the dataspace.
+     */
+    Y8                                 = 0x20203859,
+
+    /**
+     * Y16 is a YUV planar format comprised of a WxH Y plane, with each pixel
+     * being represented by 16 bits. It is just like Y8, but has double the
+     * bits per pixel (little endian).
+     *
+     * This format assumes
+     * - an even width
+     * - an even height
+     * - a horizontal stride multiple of 16 pixels
+     * - a vertical stride equal to the height
+     * - strides are specified in pixels, not in bytes
+     *
+     *   size = stride * height * 2
+     *
+     * This format must be accepted by the allocator when used with the
+     * following usage flags:
+     *
+     *    - BufferUsage::CAMERA_*
+     *    - BufferUsage::CPU_*
+     *
+     * The component values are unsigned normalized to the range [0, 1], whose
+     * interpretation is defined by the dataspace. When the dataspace is
+     * Dataspace::DEPTH, each pixel is a distance value measured by a depth
+     * camera, plus an associated confidence value.
+     */
+    Y16                                = 0x20363159,
+
+    /**
+     * YV12 is a 4:2:0 YCrCb planar format comprised of a WxH Y plane followed
+     * by (W/2) x (H/2) Cr and Cb planes.
+     *
+     * This format assumes
+     * - an even width
+     * - an even height
+     * - a horizontal stride multiple of 16 pixels
+     * - a vertical stride equal to the height
+     *
+     *   y_size = stride * height
+     *   c_stride = ALIGN(stride/2, 16)
+     *   c_size = c_stride * height/2
+     *   size = y_size + c_size * 2
+     *   cr_offset = y_size
+     *   cb_offset = y_size + c_size
+     *
+     * This range is reserved for vendor extensions. Formats in this range
+     * must support BufferUsage::GPU_TEXTURE. Clients must assume they do not
+     * have an alpha component.
+     *
+     * This format must be accepted by the allocator when used with the
+     * following usage flags:
+     *
+     *    - BufferUsage::CAMERA_*
+     *    - BufferUsage::CPU_*
+     *    - BufferUsage::GPU_TEXTURE
+     *
+     * The component values are unsigned normalized to the range [0, 1], whose
+     * interpretation is defined by the dataspace.
+     */
+    YV12                               = 0x32315659, // YCrCb 4:2:0 Planar
+
+    /**
+     * 16-bit format that has a single 16-bit depth component.
+     *
+     * The component values are unsigned normalized to the range [0, 1], whose
+     * interpretation is defined by the dataspace.
+     */
+    DEPTH_16                           = 0x30,
+
+    /**
+     * 32-bit format that has a single 24-bit depth component and, optionally,
+     * 8 bits that are unused.
+     *
+     * The component values are unsigned normalized to the range [0, 1], whose
+     * interpretation is defined by the dataspace.
+     */
+    DEPTH_24                           = 0x31,
+
+    /**
+     * 32-bit format that has a 24-bit depth component and an 8-bit stencil
+     * component packed into 32-bits.
+     *
+     * The depth component values are unsigned normalized to the range [0, 1],
+     * whose interpretation is defined by the dataspace. The stencil values are
+     * unsigned integers, whose interpretation is defined by the dataspace.
+     */
+    DEPTH_24_STENCIL_8                 = 0x32,
+
+    /**
+     * 32-bit format that has a single 32-bit depth component.
+     *
+     * The component values are signed floats, whose interpretation is defined
+     * by the dataspace.
+     */
+    DEPTH_32F                          = 0x33,
+
+    /**
+     * Two-component format that has a 32-bit depth component, an 8-bit stencil
+     * component, and optionally 24-bits unused.
+     *
+     * The depth component values are signed floats, whose interpretation is
+     * defined by the dataspace. The stencil bits are unsigned integers, whose
+     * interpretation is defined by the dataspace.
+     */
+    DEPTH_32F_STENCIL_8                = 0x34,
+
+    /**
+     * 8-bit format that has a single 8-bit stencil component.
+     *
+     * The component values are unsigned integers, whose interpretation is
+     * defined by the dataspace.
+     */
+    STENCIL_8                          = 0x35,
+
+    /**
+     * P010 is a 4:2:0 YCbCr semiplanar format comprised of a WxH Y plane
+     * followed immediately by a Wx(H/2) CbCr plane. Each sample is
+     * represented by a 16-bit little-endian value, with the lower 6 bits set
+     * to zero.
+     *
+     * This format assumes
+     * - an even height
+     * - a vertical stride equal to the height
+     *
+     *   stride_in_bytes = stride * 2
+     *   y_size = stride_in_bytes * height
+     *   cbcr_size = stride_in_bytes * (height / 2)
+     *   cb_offset = y_size
+     *   cr_offset = cb_offset + 2
+     *
+     * This format must be accepted by the allocator when used with the
+     * following usage flags:
+     *
+     *    - BufferUsage::VIDEO_*
+     *    - BufferUsage::CPU_*
+     *    - BufferUsage::GPU_TEXTURE
+     *
+     * The component values are unsigned normalized to the range [0, 1], whose
+     * interpretation is defined by the dataspace.
+     *
+     * This format is appropriate for 10bit video content.
+     *
+     * Buffers with this format must be locked with IMapper::lockYCbCr
+     * or with IMapper::lock.
+     */
+    YCBCR_P010                         = 0x36,
+
+    /**
+     * 24-bit format that has 8-bit H, S, and V components, in that order,
+     * from the lowest memory address to the highest memory address.
+     *
+     * The component values are unsigned normalized to the range [0, 1], whose
+     * interpretation is defined by the dataspace.
+    */
+    HSV_888                            = 0x37,
+}