Commit cbecf57f authored by Johann's avatar Johann

Release v1.5.0

Javan Whistling Duck release.

Change-Id: If44c9ca16a8188b68759325fbacc771365cb4af8
parent bc98bf65
Adrian Grange <agrange@google.com>
Alex Converse <aconverse@google.com> <alex.converse@gmail.com>
Adrian Grange <agrange@google.com> <agrange@agrange-macbookpro.roam.corp.google.com>
Aℓex Converse <aconverse@google.com>
Aℓex Converse <aconverse@google.com> <alex.converse@gmail.com>
Alexis Ballier <aballier@gentoo.org> <alexis.ballier@gmail.com>
Alpha Lam <hclam@google.com> <hclam@chromium.org>
Deb Mukherjee <debargha@google.com>
Erik Niemeyer <erik.a.niemeyer@intel.com> <erik.a.niemeyer@gmail.com>
Guillaume Martres <gmartres@google.com> <smarter3@gmail.com>
Hangyu Kuang <hkuang@google.com>
Hangyu Kuang <hkuang@google.com> <hkuang@hkuang-macbookpro.roam.corp.google.com>
Hui Su <huisu@google.com>
Jacky Chen <jackychen@google.com>
Jim Bankoski <jimbankoski@google.com>
Johann Koenig <johannkoenig@google.com>
Johann Koenig <johannkoenig@google.com> <johann.koenig@duck.com>
Johann Koenig <johannkoenig@google.com> <johannkoenig@dhcp-172-19-7-52.mtv.corp.google.com>
Johann Koenig <johannkoenig@google.com> <johann.koenig@gmail.com>
John Koleszar <jkoleszar@google.com>
Joshua Litt <joshualitt@google.com> <joshualitt@chromium.org>
Marco Paniconi <marpan@google.com>
......@@ -17,10 +24,13 @@ Pascal Massimino <pascal.massimino@gmail.com>
Paul Wilkins <paulwilkins@google.com>
Ralph Giles <giles@xiph.org> <giles@entropywave.com>
Ralph Giles <giles@xiph.org> <giles@mozilla.com>
Ronald S. Bultje <rsbultje@gmail.com> <rbultje@google.com>
Sami Pietilä <samipietila@google.com>
Tamar Levy <tamar.levy@intel.com>
Tamar Levy <tamar.levy@intel.com> <levytamar82@gmail.com>
Tero Rintaluoma <teror@google.com> <tero.rintaluoma@on2.com>
Timothy B. Terriberry <tterribe@xiph.org> Tim Terriberry <tterriberry@mozilla.com>
Tom Finegan <tomfinegan@google.com>
Tom Finegan <tomfinegan@google.com> <tomfinegan@chromium.org>
Yaowu Xu <yaowu@google.com> <yaowu@xuyaowu.com>
Yaowu Xu <yaowu@google.com> <yaowu@YAOWU2-W.ad.corp.google.com>
......@@ -5,9 +5,9 @@ Aaron Watry <awatry@gmail.com>
Abo Talib Mahfoodh <ab.mahfoodh@gmail.com>
Adam Xu <adam@xuyaowu.com>
Adrian Grange <agrange@google.com>
Aℓex Converse <aconverse@google.com>
Ahmad Sharif <asharif@google.com>
Alexander Voronov <avoronov@graphics.cs.msu.ru>
Alex Converse <aconverse@google.com>
Alexis Ballier <aballier@gentoo.org>
Alok Ahuja <waveletcoeff@gmail.com>
Alpha Lam <hclam@google.com>
......@@ -16,8 +16,10 @@ Ami Fischman <fischman@chromium.org>
Andoni Morales Alastruey <ylatuya@gmail.com>
Andres Mejia <mcitadel@gmail.com>
Andrew Russell <anrussell@google.com>
Angie Chiang <angiebird@google.com>
Aron Rosenberg <arosenberg@logitech.com>
Attila Nagy <attilanagy@google.com>
Brion Vibber <bvibber@wikimedia.org>
changjun.yang <changjun.yang@intel.com>
Charles 'Buck' Krasic <ckrasic@google.com>
chm <chm@rock-chips.com>
......@@ -27,6 +29,7 @@ Deb Mukherjee <debargha@google.com>
Dim Temp <dimtemp0@gmail.com>
Dmitry Kovalev <dkovalev@google.com>
Dragan Mrdjan <dmrdjan@mips.com>
Ed Baker <edward.baker@intel.com>
Ehsan Akhgari <ehsan.akhgari@gmail.com>
Erik Niemeyer <erik.a.niemeyer@intel.com>
Fabio Pedretti <fabio.ped@libero.it>
......@@ -34,6 +37,8 @@ Frank Galligan <fgalligan@google.com>
Fredrik Söderquist <fs@opera.com>
Fritz Koenig <frkoenig@google.com>
Gaute Strokkenes <gaute.strokkenes@broadcom.com>
Geza Lore <gezalore@gmail.com>
Ghislain MARY <ghislainmary2@gmail.com>
Giuseppe Scrivano <gscrivano@gnu.org>
Gordana Cmiljanovic <gordana.cmiljanovic@imgtec.com>
Guillaume Martres <gmartres@google.com>
......@@ -44,7 +49,7 @@ Henrik Lundin <hlundin@google.com>
Hui Su <huisu@google.com>
Ivan Maltz <ivanmaltz@google.com>
Jacek Caban <cjacek@gmail.com>
JackyChen <jackychen@google.com>
Jacky Chen <jackychen@google.com>
James Berry <jamesberry@google.com>
James Yu <james.yu@linaro.org>
James Zern <jzern@google.com>
......@@ -60,9 +65,11 @@ Jingning Han <jingning@google.com>
Joey Parrish <joeyparrish@google.com>
Johann Koenig <johannkoenig@google.com>
John Koleszar <jkoleszar@google.com>
Johnny Klonaris <google@jawknee.com>
John Stark <jhnstrk@gmail.com>
Joshua Bleecher Snyder <josh@treelinelabs.com>
Joshua Litt <joshualitt@google.com>
Julia Robson <juliamrobson@gmail.com>
Justin Clift <justin@salasaga.org>
Justin Lebar <justin.lebar@gmail.com>
KO Myung-Hun <komh@chollian.net>
......@@ -82,6 +89,7 @@ Mike Hommey <mhommey@mozilla.com>
Mikhal Shemer <mikhal@google.com>
Minghai Shang <minghai@google.com>
Morton Jonuschat <yabawock@gmail.com>
Nico Weber <thakis@chromium.org>
Parag Salasakar <img.mips1@gmail.com>
Pascal Massimino <pascal.massimino@gmail.com>
Patrik Westin <patrik.westin@gmail.com>
......@@ -96,7 +104,7 @@ Rafael Ávila de Espíndola <rafael.espindola@gmail.com>
Rafaël Carré <funman@videolan.org>
Ralph Giles <giles@xiph.org>
Rob Bradford <rob@linux.intel.com>
Ronald S. Bultje <rbultje@google.com>
Ronald S. Bultje <rsbultje@gmail.com>
Rui Ueyama <ruiu@google.com>
Sami Pietilä <samipietila@google.com>
Scott Graham <scottmg@chromium.org>
......@@ -104,6 +112,7 @@ Scott LaVarnway <slavarnway@google.com>
Sean McGovern <gseanmcg@gmail.com>
Sergey Ulanov <sergeyu@chromium.org>
Shimon Doodkin <helpmepro1@gmail.com>
Shunyao Li <shunyaoli@google.com>
Stefan Holmer <holmer@google.com>
Suman Sunkara <sunkaras@google.com>
Taekhyun Kim <takim@nvidia.com>
......
xxxx-yy-zz v1.4.0 "Changes for next release"
vpxenc is changed to use VP9 by default.
Encoder controls added for 1 pass SVC.
Decoder control to toggle on/off loopfilter.
2015-11-09 v1.5.0 "Javan Whistling Duck"
This release improves upon the VP9 encoder and speeds up the encoding and
decoding processes.
- Upgrading:
This release is ABI incompatible with 1.4.0. It drops deprecated VP8
controls and adds a variety of VP9 controls for testing.
The vpxenc utility now prefers VP9 by default.
- Enhancements:
Faster VP9 encoding and decoding
Smaller library size by combining functions used by VP8 and VP9
- Bug Fixes:
A variety of fuzzing issues
2015-04-03 v1.4.0 "Indian Runner Duck"
This release includes significant improvements to the VP9 codec.
......
......@@ -260,7 +260,7 @@ OBJS-yes += $(LIBVPX_OBJS)
LIBS-$(if yes,$(CONFIG_STATIC)) += $(BUILD_PFX)libvpx.a $(BUILD_PFX)libvpx_g.a
$(BUILD_PFX)libvpx_g.a: $(LIBVPX_OBJS)
SO_VERSION_MAJOR := 2
SO_VERSION_MAJOR := 3
SO_VERSION_MINOR := 0
SO_VERSION_PATCH := 0
ifeq ($(filter darwin%,$(TGT_OS)),$(TGT_OS))
......
/*
* Copyright (c) 2010 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include "./vpx_config.h"
#include "vpx_mem/vpx_mem.h"
#include "vp10/common/alloccommon.h"
#include "vp10/common/blockd.h"
#include "vp10/common/entropymode.h"
#include "vp10/common/entropymv.h"
#include "vp10/common/onyxc_int.h"
void vp10_set_mb_mi(VP10_COMMON *cm, int width, int height) {
const int aligned_width = ALIGN_POWER_OF_TWO(width, MI_SIZE_LOG2);
const int aligned_height = ALIGN_POWER_OF_TWO(height, MI_SIZE_LOG2);
cm->mi_cols = aligned_width >> MI_SIZE_LOG2;
cm->mi_rows = aligned_height >> MI_SIZE_LOG2;
cm->mi_stride = calc_mi_size(cm->mi_cols);
cm->mb_cols = (cm->mi_cols + 1) >> 1;
cm->mb_rows = (cm->mi_rows + 1) >> 1;
cm->MBs = cm->mb_rows * cm->mb_cols;
}
static int alloc_seg_map(VP10_COMMON *cm, int seg_map_size) {
int i;
for (i = 0; i < NUM_PING_PONG_BUFFERS; ++i) {
cm->seg_map_array[i] = (uint8_t *)vpx_calloc(seg_map_size, 1);
if (cm->seg_map_array[i] == NULL)
return 1;
}
cm->seg_map_alloc_size = seg_map_size;
// Init the index.
cm->seg_map_idx = 0;
cm->prev_seg_map_idx = 1;
cm->current_frame_seg_map = cm->seg_map_array[cm->seg_map_idx];
if (!cm->frame_parallel_decode)
cm->last_frame_seg_map = cm->seg_map_array[cm->prev_seg_map_idx];
return 0;
}
static void free_seg_map(VP10_COMMON *cm) {
int i;
for (i = 0; i < NUM_PING_PONG_BUFFERS; ++i) {
vpx_free(cm->seg_map_array[i]);
cm->seg_map_array[i] = NULL;
}
cm->current_frame_seg_map = NULL;
if (!cm->frame_parallel_decode) {
cm->last_frame_seg_map = NULL;
}
}
void vp10_free_ref_frame_buffers(BufferPool *pool) {
int i;
for (i = 0; i < FRAME_BUFFERS; ++i) {
if (pool->frame_bufs[i].ref_count > 0 &&
pool->frame_bufs[i].raw_frame_buffer.data != NULL) {
pool->release_fb_cb(pool->cb_priv, &pool->frame_bufs[i].raw_frame_buffer);
pool->frame_bufs[i].ref_count = 0;
}
vpx_free(pool->frame_bufs[i].mvs);
pool->frame_bufs[i].mvs = NULL;
vpx_free_frame_buffer(&pool->frame_bufs[i].buf);
}
}
void vp10_free_postproc_buffers(VP10_COMMON *cm) {
#if CONFIG_VP9_POSTPROC
vpx_free_frame_buffer(&cm->post_proc_buffer);
vpx_free_frame_buffer(&cm->post_proc_buffer_int);
#else
(void)cm;
#endif
}
void vp10_free_context_buffers(VP10_COMMON *cm) {
cm->free_mi(cm);
free_seg_map(cm);
vpx_free(cm->above_context);
cm->above_context = NULL;
vpx_free(cm->above_seg_context);
cm->above_seg_context = NULL;
}
int vp10_alloc_context_buffers(VP10_COMMON *cm, int width, int height) {
int new_mi_size;
vp10_set_mb_mi(cm, width, height);
new_mi_size = cm->mi_stride * calc_mi_size(cm->mi_rows);
if (cm->mi_alloc_size < new_mi_size) {
cm->free_mi(cm);
if (cm->alloc_mi(cm, new_mi_size))
goto fail;
}
if (cm->seg_map_alloc_size < cm->mi_rows * cm->mi_cols) {
// Create the segmentation map structure and set to 0.
free_seg_map(cm);
if (alloc_seg_map(cm, cm->mi_rows * cm->mi_cols))
goto fail;
}
if (cm->above_context_alloc_cols < cm->mi_cols) {
vpx_free(cm->above_context);
cm->above_context = (ENTROPY_CONTEXT *)vpx_calloc(
2 * mi_cols_aligned_to_sb(cm->mi_cols) * MAX_MB_PLANE,
sizeof(*cm->above_context));
if (!cm->above_context) goto fail;
vpx_free(cm->above_seg_context);
cm->above_seg_context = (PARTITION_CONTEXT *)vpx_calloc(
mi_cols_aligned_to_sb(cm->mi_cols), sizeof(*cm->above_seg_context));
if (!cm->above_seg_context) goto fail;
cm->above_context_alloc_cols = cm->mi_cols;
}
return 0;
fail:
vp10_free_context_buffers(cm);
return 1;
}
void vp10_remove_common(VP10_COMMON *cm) {
vp10_free_context_buffers(cm);
vpx_free(cm->fc);
cm->fc = NULL;
vpx_free(cm->frame_contexts);
cm->frame_contexts = NULL;
}
void vp10_init_context_buffers(VP10_COMMON *cm) {
cm->setup_mi(cm);
if (cm->last_frame_seg_map && !cm->frame_parallel_decode)
memset(cm->last_frame_seg_map, 0, cm->mi_rows * cm->mi_cols);
}
void vp10_swap_current_and_last_seg_map(VP10_COMMON *cm) {
// Swap indices.
const int tmp = cm->seg_map_idx;
cm->seg_map_idx = cm->prev_seg_map_idx;
cm->prev_seg_map_idx = tmp;
cm->current_frame_seg_map = cm->seg_map_array[cm->seg_map_idx];
cm->last_frame_seg_map = cm->seg_map_array[cm->prev_seg_map_idx];
}
/*
* Copyright (c) 2010 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#ifndef VP10_COMMON_ALLOCCOMMON_H_
#define VP10_COMMON_ALLOCCOMMON_H_
#define INVALID_IDX -1 // Invalid buffer index.
#ifdef __cplusplus
extern "C" {
#endif
struct VP10Common;
struct BufferPool;
void vp10_remove_common(struct VP10Common *cm);
int vp10_alloc_context_buffers(struct VP10Common *cm, int width, int height);
void vp10_init_context_buffers(struct VP10Common *cm);
void vp10_free_context_buffers(struct VP10Common *cm);
void vp10_free_ref_frame_buffers(struct BufferPool *pool);
void vp10_free_postproc_buffers(struct VP10Common *cm);
int vp10_alloc_state_buffers(struct VP10Common *cm, int width, int height);
void vp10_free_state_buffers(struct VP10Common *cm);
void vp10_set_mb_mi(struct VP10Common *cm, int width, int height);
void vp10_swap_current_and_last_seg_map(struct VP10Common *cm);
#ifdef __cplusplus
} // extern "C"
#endif
#endif // VP10_COMMON_ALLOCCOMMON_H_
/*
* Copyright (c) 2014 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include <arm_neon.h>
#include <assert.h>
#include "./vp10_rtcd.h"
#include "./vpx_config.h"
#include "vp10/common/common.h"
static int16_t sinpi_1_9 = 0x14a3;
static int16_t sinpi_2_9 = 0x26c9;
static int16_t sinpi_3_9 = 0x3441;
static int16_t sinpi_4_9 = 0x3b6c;
static int16_t cospi_8_64 = 0x3b21;
static int16_t cospi_16_64 = 0x2d41;
static int16_t cospi_24_64 = 0x187e;
static INLINE void TRANSPOSE4X4(
int16x8_t *q8s16,
int16x8_t *q9s16) {
int32x4_t q8s32, q9s32;
int16x4x2_t d0x2s16, d1x2s16;
int32x4x2_t q0x2s32;
d0x2s16 = vtrn_s16(vget_low_s16(*q8s16), vget_high_s16(*q8s16));
d1x2s16 = vtrn_s16(vget_low_s16(*q9s16), vget_high_s16(*q9s16));
q8s32 = vreinterpretq_s32_s16(vcombine_s16(d0x2s16.val[0], d0x2s16.val[1]));
q9s32 = vreinterpretq_s32_s16(vcombine_s16(d1x2s16.val[0], d1x2s16.val[1]));
q0x2s32 = vtrnq_s32(q8s32, q9s32);
*q8s16 = vreinterpretq_s16_s32(q0x2s32.val[0]);
*q9s16 = vreinterpretq_s16_s32(q0x2s32.val[1]);
return;
}
static INLINE void GENERATE_COSINE_CONSTANTS(
int16x4_t *d0s16,
int16x4_t *d1s16,
int16x4_t *d2s16) {
*d0s16 = vdup_n_s16(cospi_8_64);
*d1s16 = vdup_n_s16(cospi_16_64);
*d2s16 = vdup_n_s16(cospi_24_64);
return;
}
static INLINE void GENERATE_SINE_CONSTANTS(
int16x4_t *d3s16,
int16x4_t *d4s16,
int16x4_t *d5s16,
int16x8_t *q3s16) {
*d3s16 = vdup_n_s16(sinpi_1_9);
*d4s16 = vdup_n_s16(sinpi_2_9);
*q3s16 = vdupq_n_s16(sinpi_3_9);
*d5s16 = vdup_n_s16(sinpi_4_9);
return;
}
static INLINE void IDCT4x4_1D(
int16x4_t *d0s16,
int16x4_t *d1s16,
int16x4_t *d2s16,
int16x8_t *q8s16,
int16x8_t *q9s16) {
int16x4_t d16s16, d17s16, d18s16, d19s16, d23s16, d24s16;
int16x4_t d26s16, d27s16, d28s16, d29s16;
int32x4_t q10s32, q13s32, q14s32, q15s32;
int16x8_t q13s16, q14s16;
d16s16 = vget_low_s16(*q8s16);
d17s16 = vget_high_s16(*q8s16);
d18s16 = vget_low_s16(*q9s16);
d19s16 = vget_high_s16(*q9s16);
d23s16 = vadd_s16(d16s16, d18s16);
d24s16 = vsub_s16(d16s16, d18s16);
q15s32 = vmull_s16(d17s16, *d2s16);
q10s32 = vmull_s16(d17s16, *d0s16);
q13s32 = vmull_s16(d23s16, *d1s16);
q14s32 = vmull_s16(d24s16, *d1s16);
q15s32 = vmlsl_s16(q15s32, d19s16, *d0s16);
q10s32 = vmlal_s16(q10s32, d19s16, *d2s16);
d26s16 = vqrshrn_n_s32(q13s32, 14);
d27s16 = vqrshrn_n_s32(q14s32, 14);
d29s16 = vqrshrn_n_s32(q15s32, 14);
d28s16 = vqrshrn_n_s32(q10s32, 14);
q13s16 = vcombine_s16(d26s16, d27s16);
q14s16 = vcombine_s16(d28s16, d29s16);
*q8s16 = vaddq_s16(q13s16, q14s16);
*q9s16 = vsubq_s16(q13s16, q14s16);
*q9s16 = vcombine_s16(vget_high_s16(*q9s16),
vget_low_s16(*q9s16)); // vswp
return;
}
static INLINE void IADST4x4_1D(
int16x4_t *d3s16,
int16x4_t *d4s16,
int16x4_t *d5s16,
int16x8_t *q3s16,
int16x8_t *q8s16,
int16x8_t *q9s16) {
int16x4_t d6s16, d16s16, d17s16, d18s16, d19s16;
int32x4_t q8s32, q9s32, q10s32, q11s32, q12s32, q13s32, q14s32, q15s32;
d6s16 = vget_low_s16(*q3s16);
d16s16 = vget_low_s16(*q8s16);
d17s16 = vget_high_s16(*q8s16);
d18s16 = vget_low_s16(*q9s16);
d19s16 = vget_high_s16(*q9s16);
q10s32 = vmull_s16(*d3s16, d16s16);
q11s32 = vmull_s16(*d4s16, d16s16);
q12s32 = vmull_s16(d6s16, d17s16);
q13s32 = vmull_s16(*d5s16, d18s16);
q14s32 = vmull_s16(*d3s16, d18s16);
q15s32 = vmovl_s16(d16s16);
q15s32 = vaddw_s16(q15s32, d19s16);
q8s32 = vmull_s16(*d4s16, d19s16);
q15s32 = vsubw_s16(q15s32, d18s16);
q9s32 = vmull_s16(*d5s16, d19s16);
q10s32 = vaddq_s32(q10s32, q13s32);
q10s32 = vaddq_s32(q10s32, q8s32);
q11s32 = vsubq_s32(q11s32, q14s32);
q8s32 = vdupq_n_s32(sinpi_3_9);
q11s32 = vsubq_s32(q11s32, q9s32);
q15s32 = vmulq_s32(q15s32, q8s32);
q13s32 = vaddq_s32(q10s32, q12s32);
q10s32 = vaddq_s32(q10s32, q11s32);
q14s32 = vaddq_s32(q11s32, q12s32);
q10s32 = vsubq_s32(q10s32, q12s32);
d16s16 = vqrshrn_n_s32(q13s32, 14);
d17s16 = vqrshrn_n_s32(q14s32, 14);
d18s16 = vqrshrn_n_s32(q15s32, 14);
d19s16 = vqrshrn_n_s32(q10s32, 14);
*q8s16 = vcombine_s16(d16s16, d17s16);
*q9s16 = vcombine_s16(d18s16, d19s16);
return;
}
void vp10_iht4x4_16_add_neon(const tran_low_t *input, uint8_t *dest,
int dest_stride, int tx_type) {
uint8x8_t d26u8, d27u8;
int16x4_t d0s16, d1s16, d2s16, d3s16, d4s16, d5s16;
uint32x2_t d26u32, d27u32;
int16x8_t q3s16, q8s16, q9s16;
uint16x8_t q8u16, q9u16;
d26u32 = d27u32 = vdup_n_u32(0);
q8s16 = vld1q_s16(input);
q9s16 = vld1q_s16(input + 8);
TRANSPOSE4X4(&q8s16, &q9s16);
switch (tx_type) {
case 0: // idct_idct is not supported. Fall back to C
vp10_iht4x4_16_add_c(input, dest, dest_stride, tx_type);
return;
break;
case 1: // iadst_idct
// generate constants
GENERATE_COSINE_CONSTANTS(&d0s16, &d1s16, &d2s16);
GENERATE_SINE_CONSTANTS(&d3s16, &d4s16, &d5s16, &q3s16);
// first transform rows
IDCT4x4_1D(&d0s16, &d1s16, &d2s16, &q8s16, &q9s16);
// transpose the matrix
TRANSPOSE4X4(&q8s16, &q9s16);
// then transform columns
IADST4x4_1D(&d3s16, &d4s16, &d5s16, &q3s16, &q8s16, &q9s16);
break;
case 2: // idct_iadst
// generate constantsyy
GENERATE_COSINE_CONSTANTS(&d0s16, &d1s16, &d2s16);
GENERATE_SINE_CONSTANTS(&d3s16, &d4s16, &d5s16, &q3s16);
// first transform rows
IADST4x4_1D(&d3s16, &d4s16, &d5s16, &q3s16, &q8s16, &q9s16);
// transpose the matrix
TRANSPOSE4X4(&q8s16, &q9s16);
// then transform columns
IDCT4x4_1D(&d0s16, &d1s16, &d2s16, &q8s16, &q9s16);
break;
case 3: // iadst_iadst
// generate constants
GENERATE_SINE_CONSTANTS(&d3s16, &d4s16, &d5s16, &q3s16);
// first transform rows
IADST4x4_1D(&d3s16, &d4s16, &d5s16, &q3s16, &q8s16, &q9s16);
// transpose the matrix
TRANSPOSE4X4(&q8s16, &q9s16);
// then transform columns
IADST4x4_1D(&d3s16, &d4s16, &d5s16, &q3s16, &q8s16, &q9s16);
break;
default: // iadst_idct
assert(0);
break;
}
q8s16 = vrshrq_n_s16(q8s16, 4);
q9s16 = vrshrq_n_s16(q9s16, 4);
d26u32 = vld1_lane_u32((const uint32_t *)dest, d26u32, 0);
dest += dest_stride;
d26u32 = vld1_lane_u32((const uint32_t *)dest, d26u32, 1);
dest += dest_stride;
d27u32 = vld1_lane_u32((const uint32_t *)dest, d27u32, 0);
dest += dest_stride;
d27u32 = vld1_lane_u32((const uint32_t *)dest, d27u32, 1);
q8u16 = vaddw_u8(vreinterpretq_u16_s16(q8s16), vreinterpret_u8_u32(d26u32));
q9u16 = vaddw_u8(vreinterpretq_u16_s16(q9s16), vreinterpret_u8_u32(d27u32));
d26u8 = vqmovun_s16(vreinterpretq_s16_u16(q8u16));
d27u8 = vqmovun_s16(vreinterpretq_s16_u16(q9u16));
vst1_lane_u32((uint32_t *)dest, vreinterpret_u32_u8(d27u8), 1);
dest -= dest_stride;
vst1_lane_u32((uint32_t *)dest, vreinterpret_u32_u8(d27u8), 0);
dest -= dest_stride;
vst1_lane_u32((uint32_t *)dest, vreinterpret_u32_u8(d26u8), 1);
dest -= dest_stride;
vst1_lane_u32((uint32_t *)dest, vreinterpret_u32_u8(d26u8), 0);
return;
}
This diff is collapsed.
/*
* Copyright (c) 2014 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include "vp10/common/blockd.h"
PREDICTION_MODE vp10_left_block_mode(const MODE_INFO *cur_mi,
const MODE_INFO *left_mi, int b) {
if (b == 0 || b == 2) {
if (!left_mi || is_inter_block(&left_mi->mbmi))
return DC_PRED;
return get_y_mode(left_mi, b + 1);
} else {
assert(b == 1 || b == 3);
return cur_mi->bmi[b - 1].as_mode;
}
}
PREDICTION_MODE vp10_above_block_mode(const MODE_INFO *cur_mi,
const MODE_INFO *above_mi, int b) {
if (b == 0 || b == 1) {
if (!above_mi || is_inter_block(&above_mi->mbmi))
return DC_PRED;
return get_y_mode(above_mi, b + 2);
} else {
assert(b == 2 || b == 3);
return cur_mi->bmi[b - 2].as_mode;
}
}
void vp10_foreach_transformed_block_in_plane(
const MACROBLOCKD *const xd, BLOCK_SIZE bsize, int plane,
foreach_transformed_block_visitor visit, void *arg) {