gecko/content/media/ogg/nsOggCodecState.cpp

521 lines
16 KiB
C++

/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
/* vim:set ts=2 sw=2 sts=2 et cindent: */
/* ***** BEGIN LICENSE BLOCK *****
* Version: MPL 1.1/GPL 2.0/LGPL 2.1
*
* The contents of this file are subject to the Mozilla Public License Version
* 1.1 (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.mozilla.org/MPL/
*
* Software distributed under the License is distributed on an "AS IS" basis,
* WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
* for the specific language governing rights and limitations under the
* License.
*
* The Original Code is Mozilla code.
*
* The Initial Developer of the Original Code is the Mozilla Corporation.
* Portions created by the Initial Developer are Copyright (C) 2010
* the Initial Developer. All Rights Reserved.
*
* Contributor(s):
* Chris Double <chris.double@double.co.nz>
* Chris Pearce <chris@pearce.org.nz>
*
* Alternatively, the contents of this file may be used under the terms of
* either the GNU General Public License Version 2 or later (the "GPL"), or
* the GNU Lesser General Public License Version 2.1 or later (the "LGPL"),
* in which case the provisions of the GPL or the LGPL are applicable instead
* of those above. If you wish to allow use of your version of this file only
* under the terms of either the GPL or the LGPL, and not to allow others to
* use your version of this file under the terms of the MPL, indicate your
* decision by deleting the provisions above and replace them with the notice
* and other provisions required by the GPL or the LGPL. If you do not delete
* the provisions above, a recipient may use your version of this file under
* the terms of any one of the MPL, the GPL or the LGPL.
*
* ***** END LICENSE BLOCK ***** */
#include "nsDebug.h"
#include "nsOggCodecState.h"
#include "nsOggDecoder.h"
#include <string.h>
#include "nsTraceRefcnt.h"
#include "VideoUtils.h"
/*
The maximum height and width of the video. Used for
sanitizing the memory allocation of the RGB buffer.
The maximum resolution we anticipate encountering in the
wild is 2160p - 3840x2160 pixels.
*/
#define MAX_VIDEO_WIDTH 4000
#define MAX_VIDEO_HEIGHT 3000
// Adds two 64bit numbers, retuns PR_TRUE if addition succeeded, or PR_FALSE
// if addition would result in an overflow.
static PRBool AddOverflow(PRInt64 a, PRInt64 b, PRInt64& aResult);
// 64 bit integer multiplication with overflow checking. Returns PR_TRUE
// if the multiplication was successful, or PR_FALSE if the operation resulted
// in an integer overflow.
static PRBool MulOverflow(PRInt64 a, PRInt64 b, PRInt64& aResult);
static PRBool MulOverflow32(PRUint32 a, PRUint32 b, PRUint32& aResult)
{
// 32 bit integer multiplication with overflow checking. Returns PR_TRUE
// if the multiplication was successful, or PR_FALSE if the operation resulted
// in an integer overflow.
PRUint64 a64 = a;
PRUint64 b64 = b;
PRUint64 r64 = a64 * b64;
if (r64 > PR_UINT32_MAX)
return PR_FALSE;
aResult = static_cast<PRUint32>(r64);
return PR_TRUE;
}
nsOggCodecState*
nsOggCodecState::Create(ogg_page* aPage)
{
nsAutoPtr<nsOggCodecState> codecState;
if (aPage->body_len > 6 && memcmp(aPage->body+1, "theora", 6) == 0) {
codecState = new nsTheoraState(aPage);
} else if (aPage->body_len > 6 && memcmp(aPage->body+1, "vorbis", 6) == 0) {
codecState = new nsVorbisState(aPage);
} else if (aPage->body_len > 8 && memcmp(aPage->body, "fishead\0", 8) == 0) {
codecState = new nsSkeletonState(aPage);
} else {
codecState = new nsOggCodecState(aPage);
}
return codecState->nsOggCodecState::Init() ? codecState.forget() : nsnull;
}
nsOggCodecState::nsOggCodecState(ogg_page* aBosPage) :
mPacketCount(0),
mSerial(ogg_page_serialno(aBosPage)),
mActive(PR_FALSE),
mDoneReadingHeaders(PR_FALSE)
{
MOZ_COUNT_CTOR(nsOggCodecState);
memset(&mState, 0, sizeof(ogg_stream_state));
}
nsOggCodecState::~nsOggCodecState() {
MOZ_COUNT_DTOR(nsOggCodecState);
int ret = ogg_stream_clear(&mState);
NS_ASSERTION(ret == 0, "ogg_stream_clear failed");
}
nsresult nsOggCodecState::Reset() {
if (ogg_stream_reset(&mState) != 0) {
return NS_ERROR_FAILURE;
}
mBuffer.Erase();
return NS_OK;
}
PRBool nsOggCodecState::Init() {
int ret = ogg_stream_init(&mState, mSerial);
return ret == 0;
}
void nsPageQueue::Append(ogg_page* aPage) {
ogg_page* p = new ogg_page();
p->header_len = aPage->header_len;
p->body_len = aPage->body_len;
p->header = new unsigned char[p->header_len + p->body_len];
p->body = p->header + p->header_len;
memcpy(p->header, aPage->header, p->header_len);
memcpy(p->body, aPage->body, p->body_len);
nsDeque::Push(p);
}
PRBool nsOggCodecState::PageInFromBuffer() {
if (mBuffer.IsEmpty())
return PR_FALSE;
ogg_page *p = mBuffer.PeekFront();
int ret = ogg_stream_pagein(&mState, p);
NS_ENSURE_TRUE(ret == 0, PR_FALSE);
mBuffer.PopFront();
delete p->header;
delete p;
return PR_TRUE;
}
nsTheoraState::nsTheoraState(ogg_page* aBosPage) :
nsOggCodecState(aBosPage),
mSetup(0),
mCtx(0),
mFrameDuration(0),
mPixelAspectRatio(0)
{
MOZ_COUNT_CTOR(nsTheoraState);
th_info_init(&mInfo);
th_comment_init(&mComment);
}
nsTheoraState::~nsTheoraState() {
MOZ_COUNT_DTOR(nsTheoraState);
th_setup_free(mSetup);
th_decode_free(mCtx);
th_comment_clear(&mComment);
th_info_clear(&mInfo);
}
PRBool nsTheoraState::Init() {
if (!mActive)
return PR_FALSE;
mCtx = th_decode_alloc(&mInfo, mSetup);
if (mCtx == NULL) {
return mActive = PR_FALSE;
}
PRInt64 n = mInfo.fps_numerator;
PRInt64 d = mInfo.fps_denominator;
PRInt64 f;
if (!MulOverflow(1000, d, f)) {
return mActive = PR_FALSE;
}
f /= n;
if (f > PR_UINT32_MAX) {
return mActive = PR_FALSE;
}
mFrameDuration = static_cast<PRUint32>(f);
n = mInfo.aspect_numerator;
d = mInfo.aspect_denominator;
mPixelAspectRatio = (n == 0 || d == 0) ?
1.0f : static_cast<float>(n) / static_cast<float>(d);
// Ensure the frame isn't larger than our prescribed maximum.
PRUint32 pixels;
if (!MulOverflow32(mInfo.pic_width, mInfo.pic_height, pixels) ||
pixels > MAX_VIDEO_WIDTH * MAX_VIDEO_HEIGHT ||
pixels == 0)
{
return mActive = PR_FALSE;
}
return PR_TRUE;
}
PRBool
nsTheoraState::DecodeHeader(ogg_packet* aPacket)
{
mPacketCount++;
int ret = th_decode_headerin(&mInfo,
&mComment,
&mSetup,
aPacket);
// We must determine when we've read the last header packet.
// th_decode_headerin() does not tell us when it's read the last header, so
// we must keep track of the headers externally.
//
// There are 3 header packets, the Identification, Comment, and Setup
// headers, which must be in that order. If they're out of order, the file
// is invalid. If we've successfully read a header, and it's the setup
// header, then we're done reading headers. The first byte of each packet
// determines it's type as follows:
// 0x80 -> Identification header
// 0x81 -> Comment header
// 0x82 -> Setup header
// See http://www.theora.org/doc/Theora.pdf Chapter 6, "Bitstream Headers",
// for more details of the Ogg/Theora containment scheme.
PRBool isSetupHeader = aPacket->bytes > 0 && aPacket->packet[0] == 0x82;
if (ret < 0 || mPacketCount > 3) {
// We've received an error, or the first three packets weren't valid
// header packets, assume bad input, and don't activate the bitstream.
mDoneReadingHeaders = PR_TRUE;
} else if (ret > 0 && isSetupHeader && mPacketCount == 3) {
// Successfully read the three header packets.
mDoneReadingHeaders = PR_TRUE;
mActive = PR_TRUE;
}
return mDoneReadingHeaders;
}
PRInt64
nsTheoraState::Time(PRInt64 granulepos) {
if (granulepos < 0 || !mActive || mInfo.fps_numerator == 0) {
return -1;
}
PRInt64 t = 0;
PRInt64 frameno = th_granule_frame(mCtx, granulepos);
if (!AddOverflow(frameno, 1, t))
return -1;
if (!MulOverflow(t, 1000, t))
return -1;
if (!MulOverflow(t, mInfo.fps_denominator, t))
return -1;
return t / mInfo.fps_numerator;
}
PRInt64 nsTheoraState::StartTime(PRInt64 granulepos) {
if (granulepos < 0 || !mActive || mInfo.fps_numerator == 0) {
return -1;
}
PRInt64 t = 0;
PRInt64 frameno = th_granule_frame(mCtx, granulepos);
if (!MulOverflow(frameno, 1000, t))
return -1;
if (!MulOverflow(t, mInfo.fps_denominator, t))
return -1;
return t / mInfo.fps_numerator;
}
PRInt64
nsTheoraState::MaxKeyframeOffset()
{
// Determine the maximum time in milliseconds by which a key frame could
// offset for the theora bitstream. Theora granulepos encode time as:
// ((key_frame_number << granule_shift) + frame_offset).
// Therefore the maximum possible time by which any frame could be offset
// from a keyframe is the duration of (1 << granule_shift) - 1) frames.
PRInt64 frameDuration;
PRInt64 keyframeDiff;
PRInt64 shift = mInfo.keyframe_granule_shift;
// Max number of frames keyframe could possibly be offset.
keyframeDiff = (1 << shift) - 1;
// Length of frame in ms.
PRInt64 d = 0; // d will be 0 if multiplication overflows.
MulOverflow(1000, mInfo.fps_denominator, d);
frameDuration = d / mInfo.fps_numerator;
// Total time in ms keyframe can be offset from any given frame.
return frameDuration * keyframeDiff;
}
nsresult nsVorbisState::Reset()
{
nsresult res = NS_OK;
if (mActive && vorbis_synthesis_restart(&mDsp) != 0) {
res = NS_ERROR_FAILURE;
}
if (NS_FAILED(nsOggCodecState::Reset())) {
return NS_ERROR_FAILURE;
}
return res;
}
nsVorbisState::nsVorbisState(ogg_page* aBosPage) :
nsOggCodecState(aBosPage)
{
MOZ_COUNT_CTOR(nsVorbisState);
vorbis_info_init(&mInfo);
vorbis_comment_init(&mComment);
memset(&mDsp, 0, sizeof(vorbis_dsp_state));
memset(&mBlock, 0, sizeof(vorbis_block));
}
nsVorbisState::~nsVorbisState() {
MOZ_COUNT_DTOR(nsVorbisState);
vorbis_block_clear(&mBlock);
vorbis_dsp_clear(&mDsp);
vorbis_info_clear(&mInfo);
vorbis_comment_clear(&mComment);
}
PRBool nsVorbisState::DecodeHeader(ogg_packet* aPacket) {
mPacketCount++;
int ret = vorbis_synthesis_headerin(&mInfo,
&mComment,
aPacket);
// We must determine when we've read the last header packet.
// vorbis_synthesis_headerin() does not tell us when it's read the last
// header, so we must keep track of the headers externally.
//
// There are 3 header packets, the Identification, Comment, and Setup
// headers, which must be in that order. If they're out of order, the file
// is invalid. If we've successfully read a header, and it's the setup
// header, then we're done reading headers. The first byte of each packet
// determines it's type as follows:
// 0x1 -> Identification header
// 0x3 -> Comment header
// 0x5 -> Setup header
// For more details of the Vorbis/Ogg containment scheme, see the Vorbis I
// Specification, Chapter 4, Codec Setup and Packet Decode:
// http://www.xiph.org/vorbis/doc/Vorbis_I_spec.html#x1-580004
PRBool isSetupHeader = aPacket->bytes > 0 && aPacket->packet[0] == 0x5;
if (ret < 0 || mPacketCount > 3) {
// We've received an error, or the first three packets weren't valid
// header packets, assume bad input, and don't activate the bitstream.
mDoneReadingHeaders = PR_TRUE;
} else if (ret == 0 && isSetupHeader && mPacketCount == 3) {
// Successfully read the three header packets, activate the bitstream.
mDoneReadingHeaders = PR_TRUE;
mActive = PR_TRUE;
}
return mDoneReadingHeaders;
}
PRBool nsVorbisState::Init()
{
if (!mActive)
return PR_FALSE;
int ret = vorbis_synthesis_init(&mDsp, &mInfo);
if (ret != 0) {
NS_WARNING("vorbis_synthesis_init() failed initializing vorbis bitstream");
return mActive = PR_FALSE;
}
ret = vorbis_block_init(&mDsp, &mBlock);
if (ret != 0) {
NS_WARNING("vorbis_block_init() failed initializing vorbis bitstream");
if (mActive) {
vorbis_dsp_clear(&mDsp);
}
return mActive = PR_FALSE;
}
return PR_TRUE;
}
PRInt64 nsVorbisState::Time(PRInt64 granulepos) {
if (granulepos == -1 || !mActive || mDsp.vi->rate == 0) {
return -1;
}
PRInt64 t = 0;
MulOverflow(1000, granulepos, t);
return t / mDsp.vi->rate;
}
nsSkeletonState::nsSkeletonState(ogg_page* aBosPage)
: nsOggCodecState(aBosPage)
{
MOZ_COUNT_CTOR(nsSkeletonState);
}
nsSkeletonState::~nsSkeletonState()
{
MOZ_COUNT_DTOR(nsSkeletonState);
}
PRBool nsSkeletonState::DecodeHeader(ogg_packet* aPacket)
{
if (aPacket->e_o_s) {
mActive = PR_TRUE;
mDoneReadingHeaders = PR_TRUE;
}
return mDoneReadingHeaders;
}
// Adds two 64bit numbers, retuns PR_TRUE if addition succeeded, or PR_FALSE
// if addition would result in an overflow.
static PRBool AddOverflow(PRInt64 a, PRInt64 b, PRInt64& aResult) {
if (b < 1) {
if (PR_INT64_MIN - b <= a) {
aResult = a + b;
return PR_TRUE;
}
} else if (PR_INT64_MAX - b >= a) {
aResult = a + b;
return PR_TRUE;
}
return PR_FALSE;
}
// 64 bit integer multiplication with overflow checking. Returns PR_TRUE
// if the multiplication was successful, or PR_FALSE if the operation resulted
// in an integer overflow.
static PRBool MulOverflow(PRInt64 a, PRInt64 b, PRInt64& aResult) {
// We break a multiplication a * b into of sign_a * sign_b * abs(a) * abs(b)
//
// This is equivalent to:
//
// (sign_a * sign_b) * ((a_hi * 2^32) + a_lo) * ((b_hi * 2^32) + b_lo)
//
// Which is equivalent to:
//
// (sign_a * sign_b) *
// ((a_hi * b_hi << 64) +
// (a_hi * b_lo << 32) + (a_lo * b_hi << 32) +
// a_lo * b_lo)
//
// So to check if a*b overflows, we must check each sub part of the above
// sum.
//
// Note: -1 * PR_INT64_MIN == PR_INT64_MIN ; we can't negate PR_INT64_MIN!
// Note: Shift of negative numbers is undefined.
//
// Figure out the sign after multiplication. Then we can just work with
// unsigned numbers.
PRInt64 sign = (!(a < 0) == !(b < 0)) ? 1 : -1;
PRInt64 abs_a = (a < 0) ? -a : a;
PRInt64 abs_b = (b < 0) ? -b : b;
if (abs_a < 0) {
NS_ASSERTION(a == PR_INT64_MIN, "How else can this happen?");
if (b == 0 || b == 1) {
aResult = a * b;
return PR_TRUE;
} else {
return PR_FALSE;
}
}
if (abs_b < 0) {
NS_ASSERTION(b == PR_INT64_MIN, "How else can this happen?");
if (a == 0 || a == 1) {
aResult = a * b;
return PR_TRUE;
} else {
return PR_FALSE;
}
}
NS_ASSERTION(abs_a >= 0 && abs_b >= 0, "abs values must be non-negative");
PRInt64 a_hi = abs_a >> 32;
PRInt64 a_lo = abs_a & 0xFFFFFFFF;
PRInt64 b_hi = abs_b >> 32;
PRInt64 b_lo = abs_b & 0xFFFFFFFF;
NS_ASSERTION((a_hi<<32) + a_lo == abs_a, "Partition must be correct");
NS_ASSERTION((b_hi<<32) + b_lo == abs_b, "Partition must be correct");
// In the sub-equation (a_hi * b_hi << 64), if a_hi or b_hi
// are non-zero, this will overflow as it's shifted by 64.
// Abort if this overflows.
if (a_hi != 0 && b_hi != 0) {
return PR_FALSE;
}
// We can now assume that either a_hi or b_hi is 0.
NS_ASSERTION(a_hi == 0 || b_hi == 0, "One of these must be 0");
// Next we calculate:
// (a_hi * b_lo << 32) + (a_lo * b_hi << 32)
// We can factor this as:
// (a_hi * b_lo + a_lo * b_hi) << 32
PRInt64 q = a_hi * b_lo + a_lo * b_hi;
if (q > PR_INT32_MAX) {
// q will overflow when we shift by 32; abort.
return PR_FALSE;
}
q <<= 32;
// Both a_lo and b_lo are less than INT32_MAX, so can't overflow.
PRUint64 lo = a_lo * b_lo;
if (lo > PR_INT64_MAX) {
return PR_FALSE;
}
// Add the final result. We must check for overflow during addition.
if (!AddOverflow(q, static_cast<PRInt64>(lo), aResult)) {
return PR_FALSE;
}
aResult *= sign;
NS_ASSERTION(a * b == aResult, "We didn't overflow, but result is wrong!");
return PR_TRUE;
}