Magic-Mirror/gecko-dev
1
0
Fork 0
mirror of https://github.com/mozilla/gecko-dev.git synced 2024-10-08 19:04:45 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Bug 608166 - Add ARM assembly optimizations for libtheora r=chris.double,tterribe,khuey a=b-f

Browse Source
This commit is contained in:
Timothy B. Terriberry 2010-11-08 09:47:34 +02:00
parent dc49da79ab
commit bed266bd26
62 changed files with 8124 additions and 2975 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
media/libtheora/AUTHORS
View File

@ -16,6 +16,9 @@ Nils Pipenbrinck
Monty
- MMX optimized functions
David Schleef
- C64x port
Aaron Colwell
Thomas Vander Stichele
Jan Gerber
@ -45,5 +48,7 @@ Arc Riley
Rodolphe Ortalo
- Bug fixes
Robin Watts
- ARM code optimisations
and other Xiph.org contributors

24
media/libtheora/CHANGES
View File

@ -1,6 +1,26 @@
libteora 1.2.0alpha1 (2010 September 23)
- New 'ptalarbvorm' encoder with better rate/distortion optimization
- New th_encode_ctl option for copying configuration from an existing
setup header, useful for splicing streams.
- Returns TH_DUPFRAME in more cases.
- Add ARM optimizations
- Add TI C64x+ DSP optimizations
- Other performance improvements
- Rename speedlevel 2 to 3 and provide a new speedlevel 2
- Various minor bug fixes
libtheora 1.1.2 (unreleased snapshot)
- no changes recorded
- Fix Huffman table decoding with OC_HUFF_SLUSH is set to 0
- Fix a frame size bug in player_example
- Add support for passing a buffer the size of the picture
region, rather than a full padded frame to th_encode_ycbcr_in()
as was possible with the legacy pre-1.0 API.
- 4:4:4 support in player_example using software yuv->rgb
- Better rgb->yuv conversion in png2theora
- Clean up warnings and local variables
- Build and documentation fixes
libtheora 1.1.1 (2009 October 1)
@ -128,7 +148,7 @@ libtheora 1.0beta1 (2007 September 22)
- Granulepos scheme modified to match other codecs. This bumps
the bitstream revision to 3.2.1. Bitstreams marked 3.2.0 are
handled correctly by this decoder. Older decoders will show
a one frame sync error in the less noticable direction.
a one frame sync error in the less noticeable direction.
libtheora 1.0alpha8 (2007 September 18)

14
media/libtheora/README
View File

@ -1,5 +1,5 @@
-------------------------------------------------------------------------
The Xiph.org Foundation's libtheora 1.1
The Xiph.org Foundation's libtheora 1.2
-------------------------------------------------------------------------
*** What is Theora?
@ -12,10 +12,13 @@ while allow it a longer useful lifetime as an competitive codec.
The 1.0 release decoder supported all the new features, but the
encoder is nearly identical to the VP3 code.
The 1.1 release features a completely rewritten encoder, offering
The 1.1 release featured a completely rewritten encoder, offering
better performance and compression, and making more complete use
of the format's feature set. Files produced by both encoders can
be decoded by either release.
of the format's feature set.
The 1.2 release features significant additional improvements in
compression and performance. Files produced by newer encoders can
be decoded by earlier releases.
*** Where is Theora?
@ -41,6 +44,7 @@ Requirements summary:
as above,
libvorbis and libvorbisenc 1.0.1 or newer.
(libvorbis 1.3.1 or newer for 5.1 audio)
For creating a source distribution package:
@ -66,7 +70,7 @@ Windows build support is included in the win32 directory.
Project files for Apple XCode are included in the macosx directory.
There is also an experimental scons build.
There is also a more limited scons build.
*** How do I use the sample encoder?

4
media/libtheora/README_MOZILLA
View File

@ -2,6 +2,4 @@ The source from this directory was copied from the theora subversion trunk
using the update.sh script. The changes made were those applied by update.sh,
the addition/update of Makefile.in files for the Mozilla build system.
The subversion revision used was r16712.
bug559343.patch: Silence Coverity warning.
The subversion revision used was r17578.

22
media/libtheora/bug559343.patch
View File

@ -1,22 +0,0 @@
diff --git a/media/libtheora/lib/state.c b/media/libtheora/lib/state.c
--- a/media/libtheora/lib/state.c
+++ b/media/libtheora/lib/state.c
@@ -87,17 +87,17 @@ static void oc_sb_create_plane_mapping(o
int quadi;
int i;
/*Figure out how many rows of blocks in this super block lie within the
image.*/
jmax=_hfrags-x;
if(jmax>4)jmax=4;
else if(jmax<=0)break;
/*By default, set all fragment indices to -1.*/
- memset(_sb_maps[sbi][0],0xFF,sizeof(_sb_maps[sbi]));
+ memset(_sb_maps[sbi],0xFF,sizeof(_sb_maps[sbi]));
/*Fill in the fragment map for this super block.*/
xfrag=yfrag+x;
for(i=0;i<imax;i++){
int j;
for(j=0;j<jmax;j++){
_sb_maps[sbi][SB_MAP[i][j][0]][SB_MAP[i][j][1]]=xfrag+j;
}
xfrag+=_hfrags;

90
media/libtheora/include/theora/config.h
View File

@ -1,90 +0,0 @@
/* config.h. Generated from config.h.in by configure. */
/* config.h.in. Generated from configure.ac by autoheader. */
/* libcairo is available for visual debugging output */
/* #undef HAVE_CAIRO */
/* Define to 1 if you have the <dlfcn.h> header file. */
#define HAVE_DLFCN_H 1
/* Define to 1 if you have the <inttypes.h> header file. */
#define HAVE_INTTYPES_H 1
/* Define to 1 if you have the <machine/soundcard.h> header file. */
/* #undef HAVE_MACHINE_SOUNDCARD_H */
/* Define to 1 if you have the <memory.h> header file. */
#define HAVE_MEMORY_H 1
/* Define to 1 if you have the <soundcard.h> header file. */
/* #undef HAVE_SOUNDCARD_H */
/* Define to 1 if you have the <stdint.h> header file. */
#define HAVE_STDINT_H 1
/* Define to 1 if you have the <stdlib.h> header file. */
#define HAVE_STDLIB_H 1
/* Define to 1 if you have the <strings.h> header file. */
#define HAVE_STRINGS_H 1
/* Define to 1 if you have the <string.h> header file. */
#define HAVE_STRING_H 1
/* Define to 1 if you have the <sys/soundcard.h> header file. */
/* #undef HAVE_SYS_SOUNDCARD_H */
/* Define to 1 if you have the <sys/stat.h> header file. */
#define HAVE_SYS_STAT_H 1
/* Define to 1 if you have the <sys/types.h> header file. */
#define HAVE_SYS_TYPES_H 1
/* Define to 1 if you have the <unistd.h> header file. */
#define HAVE_UNISTD_H 1
/* Define to the sub-directory in which libtool stores uninstalled libraries.
*/
#define LT_OBJDIR ".libs/"
/* Define to 1 if your C compiler doesn't accept -c and -o together. */
/* #undef NO_MINUS_C_MINUS_O */
/* make use of x86_64 asm optimization */
/* #undef OC_X86_64_ASM */
/* make use of x86 asm optimization */
/**/
/* Name of package */
#define PACKAGE "libtheora"
/* Define to the address where bug reports for this package should be sent. */
#define PACKAGE_BUGREPORT ""
/* Define to the full name of this package. */
#define PACKAGE_NAME "libtheora"
/* Define to the full name and version of this package. */
#define PACKAGE_STRING "libtheora 1.1.1+svn"
/* Define to the one symbol short name of this package. */
#define PACKAGE_TARNAME "libtheora"
/* Define to the home page for this package. */
#define PACKAGE_URL ""
/* Define to the version of this package. */
#define PACKAGE_VERSION "1.1.1+svn"
/* Define to 1 if you have the ANSI C header files. */
#define STDC_HEADERS 1
/* Define to exclude encode support from the build */
/* #undef THEORA_DISABLE_ENCODE */
/* Define to exclude floating point code from the build */
/* #undef THEORA_DISABLE_FLOAT */
/* Version number of package */
#define VERSION "1.1.1+svn"

2
media/libtheora/include/theora/theora.h
View File

@ -179,7 +179,7 @@ typedef enum {
OC_PF_420, /**< Chroma subsampling by 2 in each direction (4:2:0) */
OC_PF_RSVD, /**< Reserved value */
OC_PF_422, /**< Horizonatal chroma subsampling by 2 (4:2:2) */
OC_PF_444, /**< No chroma subsampling at all (4:4:4) */
OC_PF_444 /**< No chroma subsampling at all (4:4:4) */
} theora_pixelformat;
/**

3
media/libtheora/include/theora/theoradec.h
View File

@ -283,7 +283,8 @@ extern int th_decode_ctl(th_dec_ctx *_dec,int _req,void *_buf,
* \retval 0 Success.
* A new decoded frame can be retrieved by calling
* th_decode_ycbcr_out().
* \retval TH_DUPFRAME The packet represented a dropped (0-byte) frame.
* \retval TH_DUPFRAME The packet represented a dropped frame (either a
* 0-byte frame or an INTER frame with no coded blocks).
* The player can skip the call to th_decode_ycbcr_out(),
* as the contents of the decoded frame buffer have not
* changed.

88
media/libtheora/include/theora/theoraenc.h
View File

@ -43,7 +43,7 @@ extern "C" {
* <tt>NULL</tt> may be specified to revert to the default tables.
*
* \param[in] _buf <tt>#th_huff_code[#TH_NHUFFMAN_TABLES][#TH_NDCT_TOKENS]</tt>
* \retval TH_EFAULT \a _enc_ctx is <tt>NULL</tt>.
* \retval TH_EFAULT \a _enc is <tt>NULL</tt>.
* \retval TH_EINVAL Encoding has already begun or one or more of the given
* tables is not full or prefix-free, \a _buf is
* <tt>NULL</tt> and \a _buf_sz is not zero, or \a _buf is
@ -57,7 +57,7 @@ extern "C" {
* <tt>NULL</tt> may be specified to revert to the default parameters.
*
* \param[in] _buf #th_quant_info
* \retval TH_EFAULT \a _enc_ctx is <tt>NULL</tt>.
* \retval TH_EFAULT \a _enc is <tt>NULL</tt>.
* \retval TH_EINVAL Encoding has already begun, \a _buf is
* <tt>NULL</tt> and \a _buf_sz is not zero,
* or \a _buf is non-<tt>NULL</tt> and
@ -73,7 +73,7 @@ extern "C" {
* \param[in] _buf <tt>ogg_uint32_t</tt>: The maximum distance between key
* frames.
* \param[out] _buf <tt>ogg_uint32_t</tt>: The actual maximum distance set.
* \retval TH_EFAULT \a _enc_ctx or \a _buf is <tt>NULL</tt>.
* \retval TH_EFAULT \a _enc or \a _buf is <tt>NULL</tt>.
* \retval TH_EINVAL \a _buf_sz is not <tt>sizeof(ogg_uint32_t)</tt>.
* \retval TH_EIMPL Not supported by this implementation.*/
#define TH_ENCCTL_SET_KEYFRAME_FREQUENCY_FORCE (4)
@ -101,7 +101,7 @@ extern "C" {
* 4:2:0, the picture region is smaller than the full frame,
* or if encoding has begun, preventing the quantization
* tables and codebooks from being set.
* \retval TH_EFAULT \a _enc_ctx or \a _buf is <tt>NULL</tt>.
* \retval TH_EFAULT \a _enc or \a _buf is <tt>NULL</tt>.
* \retval TH_EINVAL \a _buf_sz is not <tt>sizeof(int)</tt>.
* \retval TH_EIMPL Not supported by this implementation.*/
#define TH_ENCCTL_SET_VP3_COMPATIBLE (10)
@ -114,7 +114,7 @@ extern "C" {
* the current encoding mode (VBR vs. constant quality, etc.).
*
* \param[out] _buf <tt>int</tt>: The maximum encoding speed level.
* \retval TH_EFAULT \a _enc_ctx or \a _buf is <tt>NULL</tt>.
* \retval TH_EFAULT \a _enc or \a _buf is <tt>NULL</tt>.
* \retval TH_EINVAL \a _buf_sz is not <tt>sizeof(int)</tt>.
* \retval TH_EIMPL Not supported by this implementation in the current
* encoding mode.*/
@ -124,7 +124,7 @@ extern "C" {
*
* \param[in] _buf <tt>int</tt>: The new encoding speed level.
* 0 is slowest, larger values use less CPU.
* \retval TH_EFAULT \a _enc_ctx or \a _buf is <tt>NULL</tt>.
* \retval TH_EFAULT \a _enc or \a _buf is <tt>NULL</tt>.
* \retval TH_EINVAL \a _buf_sz is not <tt>sizeof(int)</tt>, or the
* encoding speed level is out of bounds.
* The maximum encoding speed level may be
@ -142,7 +142,7 @@ extern "C" {
*
* \param[out] _buf <tt>int</tt>: The current encoding speed level.
* 0 is slowest, larger values use less CPU.
* \retval TH_EFAULT \a _enc_ctx or \a _buf is <tt>NULL</tt>.
* \retval TH_EFAULT \a _enc or \a _buf is <tt>NULL</tt>.
* \retval TH_EINVAL \a _buf_sz is not <tt>sizeof(int)</tt>.
* \retval TH_EIMPL Not supported by this implementation in the current
* encoding mode.*/
@ -162,7 +162,7 @@ extern "C" {
*
* \param[in] _buf <tt>int</tt>: The number of duplicates to produce.
* If this is negative or zero, no duplicates will be produced.
* \retval TH_EFAULT \a _enc_ctx or \a _buf is <tt>NULL</tt>.
* \retval TH_EFAULT \a _enc or \a _buf is <tt>NULL</tt>.
* \retval TH_EINVAL \a _buf_sz is not <tt>sizeof(int)</tt>, or the
* number of duplicates is greater than or equal to the
* maximum keyframe interval.
@ -187,7 +187,7 @@ extern "C" {
* use.
* - #TH_RATECTL_CAP_UNDERFLOW: Don't try to make up shortfalls
* later.
* \retval TH_EFAULT \a _enc_ctx or \a _buf is <tt>NULL</tt>.
* \retval TH_EFAULT \a _enc or \a _buf is <tt>NULL</tt>.
* \retval TH_EINVAL \a _buf_sz is not <tt>sizeof(int)</tt> or rate control
* is not enabled.
* \retval TH_EIMPL Not supported by this implementation in the current
@ -211,7 +211,7 @@ extern "C" {
* \param[in] _buf <tt>int</tt>: Requested size of the reservoir measured in
* frames.
* \param[out] _buf <tt>int</tt>: The actual size of the reservoir set.
* \retval TH_EFAULT \a _enc_ctx or \a _buf is <tt>NULL</tt>.
* \retval TH_EFAULT \a _enc or \a _buf is <tt>NULL</tt>.
* \retval TH_EINVAL \a _buf_sz is not <tt>sizeof(int)</tt>, or rate control
* is not enabled. The buffer has an implementation
* defined minimum and maximum size and the value in _buf
@ -243,7 +243,7 @@ extern "C" {
* application.
* \retval >=0 The number of bytes of metric data available in the
* returned buffer.
* \retval TH_EFAULT \a _enc_ctx or \a _buf is <tt>NULL</tt>.
* \retval TH_EFAULT \a _enc or \a _buf is <tt>NULL</tt>.
* \retval TH_EINVAL \a _buf_sz is not <tt>sizeof(char *)</tt>, no target
* bitrate has been set, or the first call was made after
* the first frame was submitted for encoding.
@ -283,7 +283,7 @@ extern "C" {
* of bytes consumed.
* \retval >0 The number of bytes of metric data required/consumed.
* \retval 0 No more data is required before the next frame.
* \retval TH_EFAULT \a _enc_ctx is <tt>NULL</tt>.
* \retval TH_EFAULT \a _enc is <tt>NULL</tt>.
* \retval TH_EINVAL No target bitrate has been set, or the first call was
* made after the first frame was submitted for
* encoding.
@ -306,7 +306,7 @@ extern "C" {
* \param[in] _buf <tt>int</tt>: The new target quality, in the range 0...63,
* inclusive.
* \retval 0 Success.
* \retval TH_EFAULT \a _enc_ctx or \a _buf is <tt>NULL</tt>.
* \retval TH_EFAULT \a _enc or \a _buf is <tt>NULL</tt>.
* \retval TH_EINVAL A target bitrate has already been specified, or the
* quality index was not in the range 0...63.
* \retval TH_EIMPL Not supported by this implementation.*/
@ -328,10 +328,50 @@ extern "C" {
*
* \param[in] _buf <tt>long</tt>: The new target bitrate, in bits per second.
* \retval 0 Success.
* \retval TH_EFAULT \a _enc_ctx or \a _buf is <tt>NULL</tt>.
* \retval TH_EFAULT \a _enc or \a _buf is <tt>NULL</tt>.
* \retval TH_EINVAL The target bitrate was not positive.
* \retval TH_EIMPL Not supported by this implementation.*/
#define TH_ENCCTL_SET_BITRATE (30)
/**Sets the configuration to be compatible with that from the given setup
* header.
* This sets the Huffman codebooks and quantization parameters to match those
* found in the given setup header.
* This guarantees that packets encoded by this encoder will be decodable using
* a decoder configured with the passed-in setup header.
* It does <em>not</em> guarantee that th_encode_flushheader() will produce a
* bit-identical setup header, only that they will be compatible.
* If you need a bit-identical setup header, then use the one you passed into
* this command, and not the one returned by th_encode_flushheader().
*
* This also does <em>not</em> enable or disable VP3 compatibility; that is not
* signaled in the setup header (or anywhere else in the encoded stream), and
* is controlled independently by the #TH_ENCCTL_SET_VP3_COMPATIBLE function.
* If you wish to enable VP3 compatibility mode <em>and</em> want the codebooks
* and quantization parameters to match the given setup header, you should
* enable VP3 compatibility before invoking this command, otherwise the
* codebooks and quantization parameters will be reset to the VP3 defaults.
*
* The current encoder does not support Huffman codebooks which do not contain
* codewords for all 32 tokens.
* Such codebooks are legal, according to the specification, but cannot be
* configured with this function.
*
* \param[in] _buf <tt>unsigned char[]</tt>: The encoded setup header to copy
* the configuration from.
* This should be the original,
* undecoded setup header packet,
* and <em>not</em> a #th_setup_info
* structure filled in by
* th_decode_headerin().
* \retval TH_EFAULT \a _enc or \a _buf is <tt>NULL</tt>.
* \retval TH_EINVAL Encoding has already begun, so the codebooks and
* quantization parameters cannot be changed, or the
* data in the setup header was not supported by this
* encoder.
* \retval TH_EBADHEADER \a _buf did not contain a valid setup header packet.
* \retval TH_ENOTFORMAT \a _buf did not contain a Theora header at all.
* \retval TH_EIMPL Not supported by this implementation.*/
#define TH_ENCCTL_SET_COMPAT_CONFIG (32)
/*@}*/
@ -441,11 +481,25 @@ extern int th_encode_flushheader(th_enc_ctx *_enc,
/**Submits an uncompressed frame to the encoder.
* \param _enc A #th_enc_ctx handle.
* \param _ycbcr A buffer of Y'CbCr data to encode.
* If the width and height of the buffer matches the frame size
* the encoder was initialized with, the encoder will only
* reference the portion inside the picture region.
* Any data outside this region will be ignored, and need not map
* to a valid address.
* Alternatively, you can pass a buffer equal to the size of the
* picture region, if this is less than the full frame size.
* When using subsampled chroma planes, odd picture sizes or odd
* picture offsets may require an unexpected chroma plane size,
* and their use is generally discouraged, as they will not be
* well-supported by players and other media frameworks.
* See Section 4.4 of
* <a href="http://www.theora.org/doc/Theora.pdf">the Theora
* specification</a> for details if you wish to use them anyway.
* \retval 0 Success.
* \retval TH_EFAULT \a _enc or \a _ycbcr is <tt>NULL</tt>.
* \retval TH_EINVAL The buffer size does not match the frame size the encoder
* was initialized with, or encoding has already
* completed.*/
* \retval TH_EINVAL The buffer size matches neither the frame size nor the
* picture size the encoder was initialized with, or
* encoding has already completed.*/
extern int th_encode_ycbcr_in(th_enc_ctx *_enc,th_ycbcr_buffer _ycbcr);
/**Retrieves encoded video data packets.
* This should be called repeatedly after each frame is submitted to flush any

128
media/libtheora/lib/Makefile.in
View File

@ -34,15 +34,15 @@
#
# ***** END LICENSE BLOCK *****
DEPTH = ../../..
topsrcdir = @top_srcdir@
srcdir = @srcdir@
DEPTH = ../../..
topsrcdir = @top_srcdir@
srcdir = @srcdir@
include $(DEPTH)/config/autoconf.mk
MODULE = theora
LIBRARY_NAME = theora
FORCE_STATIC_LIB= 1
MODULE = theora
LIBRARY_NAME = theora
FORCE_STATIC_LIB = 1
# The encoder is currently not included.
DEFINES += -DTHEORA_DISABLE_ENCODE
@ -50,51 +50,103 @@ DEFINES += -DTHEORA_DISABLE_ENCODE
ifeq ($(findstring 86,$(OS_TEST)), 86)
ifneq ($(OS_ARCH),SunOS)
ifneq ($(OS_ARCH)$(OS_TEST),WINNTx86_64)
DEFINES += -DOC_X86_ASM -DUSE_ASM
DEFINES += -DOC_X86_ASM
ifeq (64,$(findstring 64,$(OS_TEST)))
DEFINES += -DOC_X86_64_ASM
endif
endif
endif
endif
VPATH := $(srcdir)
VPATH := $(srcdir)
CSRCS = \
apiwrapper.c \
bitpack.c \
decapiwrapper.c \
decinfo.c \
decode.c \
dequant.c \
encoder_disabled.c \
fragment.c \
huffdec.c \
idct.c \
info.c \
internal.c \
quant.c \
state.c \
$(NULL)
CSRCS = \
apiwrapper.c \
bitpack.c \
decapiwrapper.c \
decinfo.c \
decode.c \
dequant.c \
fragment.c \
huffdec.c \
idct.c \
info.c \
internal.c \
quant.c \
state.c \
$(NULL)
ifeq ($(findstring 86,$(OS_TEST)), 86)
ifdef _MSC_VER
ifneq (64,$(findstring 64,$(OS_TEST)))
VPATH += $(srcdir)/x86_vc
VPATH += $(srcdir)/x86_vc
CSRCS += \
mmxidct.c \
mmxfrag.c \
mmxstate.c \
x86state.c \
$(NULL)
CSRCS += \
mmxidct.c \
mmxfrag.c \
mmxstate.c \
x86state.c \
x86cpu.c \
$(NULL)
endif
else
VPATH += $(srcdir)/x86
VPATH += $(srcdir)/x86
CSRCS += \
mmxidct.c \
mmxfrag.c \
mmxstate.c \
sse2idct.c \
x86state.c \
x86cpu.c \
$(NULL)
endif
endif
ifdef GNU_AS
ifeq ($(findstring arm,$(OS_TEST)), arm)
VPATH += $(srcdir)/arm
CSRCS += \
armcpu.c \
armstate.c \
$(NULL)
DEFINES += -DOC_ARM_ASM -DOC_ARM_ASM_EDSP -DOC_ARM_ASM_MEDIA -DOC_ARM_ASM_NEON
# The Android NDK doesn't pre-define anything to indicate the OS it's on, so
# do it for them.
ifeq ($(OS_TARGET),Android)
DEFINES += -D__linux__
endif
THEORA_ASFILES = \
armbits.s \
armfrag.s \
armidct.s \
armloop.s \
armopts.s \
$(NULL)
ASFILES = $(patsubst %.s,%-gnu.$(ASM_SUFFIX),$(THEORA_ASFILES))
# These flags are a lie; they're just used to enable the requisite
# opcodes; actual arch detection is done at runtime.
ASFLAGS = -march=armv7-a -mfpu=neon
armfrag-gnu.$(ASM_SUFFIX): armopts-gnu.S
armidct-gnu.$(ASM_SUFFIX): armopts-gnu.S
armloop-gnu.$(ASM_SUFFIX): armopts-gnu.S
# armopts needs a specific rule, because arm2gnu.pl will always add the .S
# suffix when translating the files that include it.
armopts-gnu.S: armopts.s
$(PERL) $(srcdir)/arm/arm2gnu.pl < $< > $@
# For all others, we can use an implicit rule with the configured $(ASM_SUFFIX).
%-gnu.$(ASM_SUFFIX): %.s
$(PERL) $(srcdir)/arm/arm2gnu.pl < $< > $@
CSRCS += \
mmxidct.c \
mmxfrag.c \
mmxstate.c \
x86state.c \
$(NULL)
endif
endif

2
media/libtheora/lib/apiwrapper.h
View File

@ -21,7 +21,7 @@
# include <theora/theora.h>
# include "theora/theoradec.h"
# include "theora/theoraenc.h"
# include "internal.h"
# include "state.h"
typedef struct th_api_wrapper th_api_wrapper;
typedef struct th_api_info th_api_info;

271
media/libtheora/lib/arm/arm2gnu.pl Executable file
View File

@ -0,0 +1,271 @@
#!/usr/bin/perl
my $bigend; # little/big endian
eval 'exec /usr/local/bin/perl -S $0 ${1+"$@"}'
if $running_under_some_shell;
while ($ARGV[0] =~ /^-/) {
$_ = shift;
last if /^--/;
if (/^-n/) {
$nflag++;
next;
}
die "I don't recognize this switch: $_\\n";
}
$printit++ unless $nflag;
$\ = "\n"; # automatically add newline on print
$n=0;
$thumb = 0; # ARM mode by default, not Thumb.
LINE:
while (<>) {
# For ADRLs we need to add a new line after the substituted one.
$addPadding = 0;
# First, we do not dare to touch *anything* inside double quotes, do we?
# Second, if you want a dollar character in the string,
# insert two of them -- that's how ARM C and assembler treat strings.
s/^([A-Za-z_]\w*)[ \t]+DCB[ \t]*\"/$1: .ascii \"/ && do { s/\$\$/\$/g; next };
s/\bDCB\b[ \t]*\"/.ascii \"/ && do { s/\$\$/\$/g; next };
s/^(\S+)\s+RN\s+(\S+)/$1 .req r$2/ && do { s/\$\$/\$/g; next };
# If there's nothing on a line but a comment, don't try to apply any further
# substitutions (this is a cheap hack to avoid mucking up the license header)
s/^([ \t]*);/$1@/ && do { s/\$\$/\$/g; next };
# If substituted -- leave immediately !
s/@/,:/;
s/;/@/;
while ( /@.*'/ ) {
s/(@.*)'/$1/g;
}
s/\{FALSE\}/0/g;
s/\{TRUE\}/1/g;
s/\{(\w\w\w\w+)\}/$1/g;
s/\bINCLUDE[ \t]*([^ \t\n]+)/.include \"$1\"/;
s/\bGET[ \t]*([^ \t\n]+)/.include \"${ my $x=$1; $x =~ s|\.s|-gnu.S|; \$x }\"/;
s/\bIMPORT\b/.extern/;
s/\bEXPORT\b/.global/;
s/^(\s+)\[/$1IF/;
s/^(\s+)\|/$1ELSE/;
s/^(\s+)\]/$1ENDIF/;
s/IF *:DEF:/ .ifdef/;
s/IF *:LNOT: *:DEF:/ .ifndef/;
s/ELSE/ .else/;
s/ENDIF/ .endif/;
if( /\bIF\b/ ) {
s/\bIF\b/ .if/;
s/=/==/;
}
if ( $n == 2) {
s/\$/\\/g;
}
if ($n == 1) {
s/\$//g;
s/label//g;
$n = 2;
}
if ( /MACRO/ ) {
s/MACRO *\n/.macro/;
$n=1;
}
if ( /\bMEND\b/ ) {
s/\bMEND\b/.endm/;
$n=0;
}
# ".rdata" doesn't work in 'as' version 2.13.2, as it is ".rodata" there.
#
if ( /\bAREA\b/ ) {
s/^(.+)CODE(.+)READONLY(.*)/ .text/;
s/^(.+)DATA(.+)READONLY(.*)/ .section .rdata\n .align 2/;
s/^(.+)\|\|\.data\|\|(.+)/ .data\n .align 2/;
s/^(.+)\|\|\.bss\|\|(.+)/ .bss/;
}
s/\|\|\.constdata\$(\d+)\|\|/.L_CONST$1/; # ||.constdata$3||
s/\|\|\.bss\$(\d+)\|\|/.L_BSS$1/; # ||.bss$2||
s/\|\|\.data\$(\d+)\|\|/.L_DATA$1/; # ||.data$2||
s/\|\|([a-zA-Z0-9_]+)\@([a-zA-Z0-9_]+)\|\|/@ $&/;
s/^(\s+)\%(\s)/ .space $1/;
s/\|(.+)\.(\d+)\|/\.$1_$2/; # |L80.123| -> .L80_123
s/\bCODE32\b/.code 32/ && do {$thumb = 0};
s/\bCODE16\b/.code 16/ && do {$thumb = 1};
if (/\bPROC\b/)
{
print " .thumb_func" if ($thumb);
s/\bPROC\b/@ $&/;
}
s/^(\s*)(S|Q|SH|U|UQ|UH)ASX\b/$1$2ADDSUBX/;
s/^(\s*)(S|Q|SH|U|UQ|UH)SAX\b/$1$2SUBADDX/;
s/\bENDP\b/@ $&/;
s/\bSUBT\b/@ $&/;
s/\bDATA\b/@ $&/; # DATA directive is deprecated -- Asm guide, p.7-25
s/\bKEEP\b/@ $&/;
s/\bEXPORTAS\b/@ $&/;
s/\|\|(.)+\bEQU\b/@ $&/;
s/\|\|([\w\$]+)\|\|/$1/;
s/\bENTRY\b/@ $&/;
s/\bASSERT\b/@ $&/;
s/\bGBLL\b/@ $&/;
s/\bGBLA\b/@ $&/;
s/^\W+OPT\b/@ $&/;
s/:OR:/|/g;
s/:SHL:/<</g;
s/:SHR:/>>/g;
s/:AND:/&/g;
s/:LAND:/&&/g;
s/CPSR/cpsr/;
s/SPSR/spsr/;
s/ALIGN$/.balign 4/;
s/ALIGN\s+([0-9x]+)$/.balign $1/;
s/psr_cxsf/psr_all/;
s/LTORG/.ltorg/;
s/^([A-Za-z_]\w*)[ \t]+EQU/ .set $1,/;
s/^([A-Za-z_]\w*)[ \t]+SETL/ .set $1,/;
s/^([A-Za-z_]\w*)[ \t]+SETA/ .set $1,/;
s/^([A-Za-z_]\w*)[ \t]+\*/ .set $1,/;
# {PC} + 0xdeadfeed --> . + 0xdeadfeed
s/\{PC\} \+/ \. +/;
# Single hex constant on the line !
#
# >>> NOTE <<<
# Double-precision floats in gcc are always mixed-endian, which means
# bytes in two words are little-endian, but words are big-endian.
# So, 0x0000deadfeed0000 would be stored as 0x0000dead at low address
# and 0xfeed0000 at high address.
#
s/\bDCFD\b[ \t]+0x([a-fA-F0-9]{8})([a-fA-F0-9]{8})/.long 0x$1, 0x$2/;
# Only decimal constants on the line, no hex !
s/\bDCFD\b[ \t]+([0-9\.\-]+)/.double $1/;
# Single hex constant on the line !
# s/\bDCFS\b[ \t]+0x([a-f0-9]{8})([a-f0-9]{8})/.long 0x$1, 0x$2/;
# Only decimal constants on the line, no hex !
# s/\bDCFS\b[ \t]+([0-9\.\-]+)/.double $1/;
s/\bDCFS[ \t]+0x/.word 0x/;
s/\bDCFS\b/.float/;
s/^([A-Za-z_]\w*)[ \t]+DCD/$1 .word/;
s/\bDCD\b/.word/;
s/^([A-Za-z_]\w*)[ \t]+DCW/$1 .short/;
s/\bDCW\b/.short/;
s/^([A-Za-z_]\w*)[ \t]+DCB/$1 .byte/;
s/\bDCB\b/.byte/;
s/^([A-Za-z_]\w*)[ \t]+\%/.comm $1,/;
s/^[A-Za-z_\.]\w+/$&:/;
s/^(\d+)/$1:/;
s/\%(\d+)/$1b_or_f/;
s/\%[Bb](\d+)/$1b/;
s/\%[Ff](\d+)/$1f/;
s/\%[Ff][Tt](\d+)/$1f/;
s/&([\dA-Fa-f]+)/0x$1/;
if ( /\b2_[01]+\b/ ) {
s/\b2_([01]+)\b/conv$1&&&&/g;
while ( /[01][01][01][01]&&&&/ ) {
s/0000&&&&/&&&&0/g;
s/0001&&&&/&&&&1/g;
s/0010&&&&/&&&&2/g;
s/0011&&&&/&&&&3/g;
s/0100&&&&/&&&&4/g;
s/0101&&&&/&&&&5/g;
s/0110&&&&/&&&&6/g;
s/0111&&&&/&&&&7/g;
s/1000&&&&/&&&&8/g;
s/1001&&&&/&&&&9/g;
s/1010&&&&/&&&&A/g;
s/1011&&&&/&&&&B/g;
s/1100&&&&/&&&&C/g;
s/1101&&&&/&&&&D/g;
s/1110&&&&/&&&&E/g;
s/1111&&&&/&&&&F/g;
}
s/000&&&&/&&&&0/g;
s/001&&&&/&&&&1/g;
s/010&&&&/&&&&2/g;
s/011&&&&/&&&&3/g;
s/100&&&&/&&&&4/g;
s/101&&&&/&&&&5/g;
s/110&&&&/&&&&6/g;
s/111&&&&/&&&&7/g;
s/00&&&&/&&&&0/g;
s/01&&&&/&&&&1/g;
s/10&&&&/&&&&2/g;
s/11&&&&/&&&&3/g;
s/0&&&&/&&&&0/g;
s/1&&&&/&&&&1/g;
s/conv&&&&/0x/g;
}
if ( /commandline/)
{
if( /-bigend/)
{
$bigend=1;
}
}
if ( /\bDCDU\b/ )
{
my $cmd=$_;
my $value;
my $w1;
my $w2;
my $w3;
my $w4;
s/\s+DCDU\b/@ $&/;
$cmd =~ /\bDCDU\b\s+0x(\d+)/;
$value = $1;
$value =~ /(\w\w)(\w\w)(\w\w)(\w\w)/;
$w1 = $1;
$w2 = $2;
$w3 = $3;
$w4 = $4;
if( $bigend ne "")
{
# big endian
print " .byte 0x".$w1;
print " .byte 0x".$w2;
print " .byte 0x".$w3;
print " .byte 0x".$w4;
}
else
{
# little endian
print " .byte 0x".$w4;
print " .byte 0x".$w3;
print " .byte 0x".$w2;
print " .byte 0x".$w1;
}
}
if ( /\badrl\b/i )
{
s/\badrl\s+(\w+)\s*,\s*(\w+)/ldr $1,=$2/i;
$addPadding = 1;
}
s/\bEND\b/@ END/;
} continue {
printf ("%s", $_) if $printit;
if ($addPadding != 0)
{
printf (" mov r0,r0\n");
$addPadding = 0;
}
}

32
media/libtheora/lib/arm/armbits.h Normal file
View File

@ -0,0 +1,32 @@
/********************************************************************
* *
* THIS FILE IS PART OF THE OggTheora SOFTWARE CODEC SOURCE CODE. *
* USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS *
* GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
* IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING. *
* *
* THE Theora SOURCE CODE IS COPYRIGHT (C) 2002-2010 *
* by the Xiph.Org Foundation and contributors http://www.xiph.org/ *
* *
********************************************************************
function:
last mod: $Id: x86int.h 17344 2010-07-21 01:42:18Z tterribe $
********************************************************************/
#if !defined(_arm_armbits_H)
# define _arm_armbits_H (1)
# include "../bitpack.h"
# include "armcpu.h"
# if defined(OC_ARM_ASM)
# define oc_pack_read oc_pack_read_arm
# define oc_pack_read1 oc_pack_read1_arm
# define oc_huff_token_decode oc_huff_token_decode_arm
# endif
long oc_pack_read_arm(oc_pack_buf *_b,int _bits);
int oc_pack_read1_arm(oc_pack_buf *_b);
int oc_huff_token_decode_arm(oc_pack_buf *_b,const ogg_int16_t *_tree);
#endif

230
media/libtheora/lib/arm/armbits.s Normal file
View File

@ -0,0 +1,230 @@
;********************************************************************
;* *
;* THIS FILE IS PART OF THE OggTheora SOFTWARE CODEC SOURCE CODE. *
;* USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS *
;* GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
;* IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING. *
;* *
;* THE Theora SOURCE CODE IS COPYRIGHT (C) 2002-2010 *
;* by the Xiph.Org Foundation and contributors http://www.xiph.org/ *
;* *
;********************************************************************
;
; function:
; last mod: $Id: armbits.s 17481 2010-10-03 22:49:42Z tterribe $
;
;********************************************************************
AREA |.text|, CODE, READONLY
EXPORT oc_pack_read_arm
EXPORT oc_pack_read1_arm
EXPORT oc_huff_token_decode_arm
oc_pack_read1_arm PROC
; r0 = oc_pack_buf *_b
ADD r12,r0,#8
LDMIA r12,{r2,r3} ; r2 = window
; Stall... ; r3 = available
; Stall...
SUBS r3,r3,#1 ; r3 = available-1, available<1 => LT
BLT oc_pack_read1_refill
MOV r0,r2,LSR #31 ; r0 = window>>31
MOV r2,r2,LSL #1 ; r2 = window<<=1
STMIA r12,{r2,r3} ; window = r2
; available = r3
MOV PC,r14
ENDP
oc_pack_read_arm PROC
; r0 = oc_pack_buf *_b
; r1 = int _bits
ADD r12,r0,#8
LDMIA r12,{r2,r3} ; r2 = window
; Stall... ; r3 = available
; Stall...
SUBS r3,r3,r1 ; r3 = available-_bits, available<_bits => LT
BLT oc_pack_read_refill
RSB r0,r1,#32 ; r0 = 32-_bits
MOV r0,r2,LSR r0 ; r0 = window>>32-_bits
MOV r2,r2,LSL r1 ; r2 = window<<=_bits
STMIA r12,{r2,r3} ; window = r2
; available = r3
MOV PC,r14
; We need to refill window.
oc_pack_read1_refill
MOV r1,#1
oc_pack_read_refill
STMFD r13!,{r10,r11,r14}
LDMIA r0,{r10,r11} ; r10 = stop
; r11 = ptr
RSB r0,r1,#32 ; r0 = 32-_bits
RSB r3,r3,r0 ; r3 = 32-available
; We can use unsigned compares for both the pointers and for available
; (allowing us to chain condition codes) because available will never be
; larger than 32 (or we wouldn't be here), and thus 32-available will never be
; negative.
CMP r10,r11 ; ptr<stop => HI
CMPHI r3,#7 ; available<=24 => HI
LDRHIB r14,[r11],#1 ; r14 = *ptr++
SUBHI r3,#8 ; available += 8
; (HI) Stall...
ORRHI r2,r14,LSL r3 ; r2 = window|=r14<<32-available
CMPHI r10,r11 ; ptr<stop => HI
CMPHI r3,#7 ; available<=24 => HI
LDRHIB r14,[r11],#1 ; r14 = *ptr++
SUBHI r3,#8 ; available += 8
; (HI) Stall...
ORRHI r2,r14,LSL r3 ; r2 = window|=r14<<32-available
CMPHI r10,r11 ; ptr<stop => HI
CMPHI r3,#7 ; available<=24 => HI
LDRHIB r14,[r11],#1 ; r14 = *ptr++
SUBHI r3,#8 ; available += 8
; (HI) Stall...
ORRHI r2,r14,LSL r3 ; r2 = window|=r14<<32-available
CMPHI r10,r11 ; ptr<stop => HI
CMPHI r3,#7 ; available<=24 => HI
LDRHIB r14,[r11],#1 ; r14 = *ptr++
SUBHI r3,#8 ; available += 8
; (HI) Stall...
ORRHI r2,r14,LSL r3 ; r2 = window|=r14<<32-available
SUBS r3,r0,r3 ; r3 = available-=_bits, available<bits => GT
BLT oc_pack_read_refill_last
MOV r0,r2,LSR r0 ; r0 = window>>32-_bits
MOV r2,r2,LSL r1 ; r2 = window<<=_bits
STR r11,[r12,#-4] ; ptr = r11
STMIA r12,{r2,r3} ; window = r2
; available = r3
LDMFD r13!,{r10,r11,PC}
; Either we wanted to read more than 24 bits and didn't have enough room to
; stuff the last byte into the window, or we hit the end of the packet.
oc_pack_read_refill_last
CMP r11,r10 ; ptr<stop => LO
; If we didn't hit the end of the packet, then pull enough of the next byte to
; to fill up the window.
LDRLOB r14,[r11] ; (LO) r14 = *ptr
; Otherwise, set the EOF flag and pretend we have lots of available bits.
MOVHS r14,#1 ; (HS) r14 = 1
ADDLO r10,r3,r1 ; (LO) r10 = available
STRHS r14,[r12,#8] ; (HS) eof = 1
ANDLO r10,r10,#7 ; (LO) r10 = available&7
MOVHS r3,#1<<30 ; (HS) available = OC_LOTS_OF_BITS
ORRLO r2,r14,LSL r10 ; (LO) r2 = window|=*ptr>>(available&7)
MOV r0,r2,LSR r0 ; r0 = window>>32-_bits
MOV r2,r2,LSL r1 ; r2 = window<<=_bits
STR r11,[r12,#-4] ; ptr = r11
STMIA r12,{r2,r3} ; window = r2
; available = r3
LDMFD r13!,{r10,r11,PC}
ENDP
oc_huff_token_decode_arm PROC
; r0 = oc_pack_buf *_b
; r1 = const ogg_int16_t *_tree
STMFD r13!,{r4,r5,r10,r14}
LDRSH r10,[r1] ; r10 = n=_tree[0]
LDMIA r0,{r2-r5} ; r2 = stop
; Stall... ; r3 = ptr
; Stall... ; r4 = window
; r5 = available
CMP r10,r5 ; n>available => GT
BGT oc_huff_token_decode_refill0
RSB r14,r10,#32 ; r14 = 32-n
MOV r14,r4,LSR r14 ; r14 = bits=window>>32-n
ADD r14,r1,r14,LSL #1 ; r14 = _tree+bits
LDRSH r12,[r14,#2] ; r12 = node=_tree[1+bits]
; Stall...
; Stall...
RSBS r14,r12,#0 ; r14 = -node, node>0 => MI
BMI oc_huff_token_decode_continue
MOV r10,r14,LSR #8 ; r10 = n=node>>8
MOV r4,r4,LSL r10 ; r4 = window<<=n
SUB r5,r10 ; r5 = available-=n
STMIB r0,{r3-r5} ; ptr = r3
; window = r4
; available = r5
AND r0,r14,#255 ; r0 = node&255
LDMFD r13!,{r4,r5,r10,pc}
; The first tree node wasn't enough to reach a leaf, read another
oc_huff_token_decode_continue
ADD r12,r1,r12,LSL #1 ; r12 = _tree+node
MOV r4,r4,LSL r10 ; r4 = window<<=n
SUB r5,r5,r10 ; r5 = available-=n
LDRSH r10,[r12],#2 ; r10 = n=_tree[node]
; Stall... ; r12 = _tree+node+1
; Stall...
CMP r10,r5 ; n>available => GT
BGT oc_huff_token_decode_refill
RSB r14,r10,#32 ; r14 = 32-n
MOV r14,r4,LSR r14 ; r14 = bits=window>>32-n
ADD r12,r12,r14 ;
LDRSH r12,[r12,r14] ; r12 = node=_tree[node+1+bits]
; Stall...
; Stall...
RSBS r14,r12,#0 ; r14 = -node, node>0 => MI
BMI oc_huff_token_decode_continue
MOV r10,r14,LSR #8 ; r10 = n=node>>8
MOV r4,r4,LSL r10 ; r4 = window<<=n
SUB r5,r10 ; r5 = available-=n
STMIB r0,{r3-r5} ; ptr = r3
; window = r4
; available = r5
AND r0,r14,#255 ; r0 = node&255
LDMFD r13!,{r4,r5,r10,pc}
oc_huff_token_decode_refill0
ADD r12,r1,#2 ; r12 = _tree+1
oc_huff_token_decode_refill
; We can't possibly need more than 15 bits, so available must be <= 15.
; Therefore we can load at least two bytes without checking it.
CMP r2,r3 ; ptr<stop => HI
LDRHIB r14,[r3],#1 ; r14 = *ptr++
RSBHI r5,r5,#24 ; (HI) available = 32-(available+=8)
RSBLS r5,r5,#32 ; (LS) r5 = 32-available
ORRHI r4,r14,LSL r5 ; r4 = window|=r14<<32-available
CMPHI r2,r3 ; ptr<stop => HI
LDRHIB r14,[r3],#1 ; r14 = *ptr++
SUBHI r5,#8 ; available += 8
; (HI) Stall...
ORRHI r4,r14,LSL r5 ; r4 = window|=r14<<32-available
; We can use unsigned compares for both the pointers and for available
; (allowing us to chain condition codes) because available will never be
; larger than 32 (or we wouldn't be here), and thus 32-available will never be
; negative.
CMPHI r2,r3 ; ptr<stop => HI
CMPHI r5,#7 ; available<=24 => HI
LDRHIB r14,[r3],#1 ; r14 = *ptr++
SUBHI r5,#8 ; available += 8
; (HI) Stall...
ORRHI r4,r14,LSL r5 ; r4 = window|=r14<<32-available
CMP r2,r3 ; ptr<stop => HI
MOVLS r5,#-1<<30 ; (LS) available = OC_LOTS_OF_BITS+32
CMPHI r5,#7 ; (HI) available<=24 => HI
LDRHIB r14,[r3],#1 ; (HI) r14 = *ptr++
SUBHI r5,#8 ; (HI) available += 8
; (HI) Stall...
ORRHI r4,r14,LSL r5 ; (HI) r4 = window|=r14<<32-available
RSB r14,r10,#32 ; r14 = 32-n
MOV r14,r4,LSR r14 ; r14 = bits=window>>32-n
ADD r12,r12,r14 ;
LDRSH r12,[r12,r14] ; r12 = node=_tree[node+1+bits]
RSB r5,r5,#32 ; r5 = available
; Stall...
RSBS r14,r12,#0 ; r14 = -node, node>0 => MI
BMI oc_huff_token_decode_continue
MOV r10,r14,LSR #8 ; r10 = n=node>>8
MOV r4,r4,LSL r10 ; r4 = window<<=n
SUB r5,r10 ; r5 = available-=n
STMIB r0,{r3-r5} ; ptr = r3
; window = r4
; available = r5
AND r0,r14,#255 ; r0 = node&255
LDMFD r13!,{r4,r5,r10,pc}
ENDP
END

116
media/libtheora/lib/arm/armcpu.c Normal file
View File

@ -0,0 +1,116 @@
/********************************************************************
* *
* THIS FILE IS PART OF THE OggTheora SOFTWARE CODEC SOURCE CODE. *
* USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS *
* GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
* IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING. *
* *
* THE Theora SOURCE CODE IS COPYRIGHT (C) 2002-2010 *
* by the Xiph.Org Foundation and contributors http://www.xiph.org/ *
* *
********************************************************************
CPU capability detection for ARM processors.
function:
last mod: $Id: cpu.c 17344 2010-07-21 01:42:18Z tterribe $
********************************************************************/
#include "armcpu.h"
#if !defined(OC_ARM_ASM)|| \
!defined(OC_ARM_ASM_EDSP)&&!defined(OC_ARM_ASM_ARMV6)&& \
!defined(OC_ARM_ASM_NEON)
ogg_uint32_t oc_cpu_flags_get(void){
return 0;
}
#elif defined(_MSC_VER)
/*For GetExceptionCode() and EXCEPTION_ILLEGAL_INSTRUCTION.*/
# define WIN32_LEAN_AND_MEAN
# define WIN32_EXTRA_LEAN
# include <windows.h>
ogg_uint32_t oc_cpu_flags_get(void){
ogg_uint32_t flags;
flags=0;
/*MSVC has no inline __asm support for ARM, but it does let you __emit
instructions via their assembled hex code.
All of these instructions should be essentially nops.*/
# if defined(OC_ARM_ASM_EDSP)
__try{
/*PLD [r13]*/
__emit(0xF5DDF000);
flags|=OC_CPU_ARM_EDSP;
}
__except(GetExceptionCode()==EXCEPTION_ILLEGAL_INSTRUCTION){
/*Ignore exception.*/
}
# if defined(OC_ARM_ASM_MEDIA)
__try{
/*SHADD8 r3,r3,r3*/
__emit(0xE6333F93);
flags|=OC_CPU_ARM_MEDIA;
}
__except(GetExceptionCode()==EXCEPTION_ILLEGAL_INSTRUCTION){
/*Ignore exception.*/
}
# if defined(OC_ARM_ASM_NEON)
__try{
/*VORR q0,q0,q0*/
__emit(0xF2200150);
flags|=OC_CPU_ARM_NEON;
}
__except(GetExceptionCode()==EXCEPTION_ILLEGAL_INSTRUCTION){
/*Ignore exception.*/
}
# endif
# endif
# endif
return flags;
}
#elif defined(__linux__)
# include <stdio.h>
# include <stdlib.h>
# include <string.h>
ogg_uint32_t oc_cpu_flags_get(void){
ogg_uint32_t flags;
FILE *fin;
flags=0;
/*Reading /proc/self/auxv would be easier, but that doesn't work reliably on
Android.
This also means that detection will fail in Scratchbox.*/
fin=fopen("/proc/cpuinfo","r");
if(fin!=NULL){
/*512 should be enough for anybody (it's even enough for all the flags that
x86 has accumulated... so far).*/
char buf[512];
while(fgets(buf,511,fin)!=NULL){
if(memcmp(buf,"Features",8)==0){
char *p;
p=strstr(buf," edsp");
if(p!=NULL&&(p[5]==' '||p[5]=='\n'))flags|=OC_CPU_ARM_EDSP;
p=strstr(buf," neon");
if(p!=NULL&&(p[5]==' '||p[5]=='\n'))flags|=OC_CPU_ARM_NEON;
}
if(memcmp(buf,"CPU architecture:",17)==0){
int version;
version=atoi(buf+17);
if(version>=6)flags|=OC_CPU_ARM_MEDIA;
}
}
fclose(fin);
}
return flags;
}
#else
/*The feature registers which can tell us what the processor supports are
accessible in priveleged modes only, so we can't have a general user-space
detection method like on x86.*/
# error "Configured to use ARM asm but no CPU detection method available for " \
"your platform. Reconfigure with --disable-asm (or send patches)."
#endif

29
media/libtheora/lib/arm/armcpu.h Normal file
View File

@ -0,0 +1,29 @@
/********************************************************************
* *
* THIS FILE IS PART OF THE OggTheora SOFTWARE CODEC SOURCE CODE. *
* USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS *
* GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
* IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING. *
* *
* THE Theora SOURCE CODE IS COPYRIGHT (C) 2002-2010 *
* by the Xiph.Org Foundation and contributors http://www.xiph.org/ *
* *
********************************************************************
function:
last mod: $Id: cpu.h 17344 2010-07-21 01:42:18Z tterribe $
********************************************************************/
#if !defined(_arm_armcpu_H)
# define _arm_armcpu_H (1)
#include "../internal.h"
/*"Parallel instructions" from ARM v6 and above.*/
#define OC_CPU_ARM_MEDIA (1<<24)
/*Flags chosen to match arch/arm/include/asm/hwcap.h in the Linux kernel.*/
#define OC_CPU_ARM_EDSP (1<<7)
#define OC_CPU_ARM_NEON (1<<12)
ogg_uint32_t oc_cpu_flags_get(void);
#endif

656
media/libtheora/lib/arm/armfrag.s Normal file
View File

@ -0,0 +1,656 @@
;********************************************************************
;* *
;* THIS FILE IS PART OF THE OggTheora SOFTWARE CODEC SOURCE CODE. *
;* USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS *
;* GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
;* IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING. *
;* *
;* THE Theora SOURCE CODE IS COPYRIGHT (C) 2002-2010 *
;* by the Xiph.Org Foundation and contributors http://www.xiph.org/ *
;* *
;********************************************************************
; Original implementation:
; Copyright (C) 2009 Robin Watts for Pinknoise Productions Ltd
; last mod: $Id: armfrag.s 17481 2010-10-03 22:49:42Z tterribe $
;********************************************************************
AREA |.text|, CODE, READONLY
GET armopts.s
; Vanilla ARM v4 versions
EXPORT oc_frag_copy_list_arm
EXPORT oc_frag_recon_intra_arm
EXPORT oc_frag_recon_inter_arm
EXPORT oc_frag_recon_inter2_arm
oc_frag_copy_list_arm PROC
; r0 = _dst_frame
; r1 = _src_frame
; r2 = _ystride
; r3 = _fragis
; <> = _nfragis
; <> = _frag_buf_offs
LDR r12,[r13] ; r12 = _nfragis
STMFD r13!,{r4-r6,r11,r14}
SUBS r12, r12, #1
LDR r4,[r3],#4 ; r4 = _fragis[fragii]
LDRGE r14,[r13,#4*6] ; r14 = _frag_buf_offs
BLT ofcl_arm_end
SUB r2, r2, #4
ofcl_arm_lp
LDR r11,[r14,r4,LSL #2] ; r11 = _frag_buf_offs[_fragis[fragii]]
SUBS r12, r12, #1
; Stall (on XScale)
ADD r4, r1, r11 ; r4 = _src_frame+frag_buf_off
LDR r6, [r4], #4
ADD r11,r0, r11 ; r11 = _dst_frame+frag_buf_off
LDR r5, [r4], r2
STR r6, [r11],#4
LDR r6, [r4], #4
STR r5, [r11],r2
LDR r5, [r4], r2
STR r6, [r11],#4
LDR r6, [r4], #4
STR r5, [r11],r2
LDR r5, [r4], r2
STR r6, [r11],#4
LDR r6, [r4], #4
STR r5, [r11],r2
LDR r5, [r4], r2
STR r6, [r11],#4
LDR r6, [r4], #4
STR r5, [r11],r2
LDR r5, [r4], r2
STR r6, [r11],#4
LDR r6, [r4], #4
STR r5, [r11],r2
LDR r5, [r4], r2
STR r6, [r11],#4
LDR r6, [r4], #4
STR r5, [r11],r2
LDR r5, [r4], r2
STR r6, [r11],#4
LDR r6, [r4], #4
STR r5, [r11],r2
LDR r5, [r4]
LDRGE r4,[r3],#4 ; r4 = _fragis[fragii]
STR r6, [r11],#4
STR r5, [r11]
BGE ofcl_arm_lp
ofcl_arm_end
LDMFD r13!,{r4-r6,r11,PC}
oc_frag_recon_intra_arm
; r0 = unsigned char *_dst
; r1 = int _ystride
; r2 = const ogg_int16_t _residue[64]
STMFD r13!,{r4,r5,r14}
MOV r14,#8
MOV r5, #255
SUB r1, r1, #7
ofrintra_lp_arm
LDRSH r3, [r2], #2
LDRSH r4, [r2], #2
LDRSH r12,[r2], #2
ADDS r3, r3, #128
CMPGT r5, r3
EORLT r3, r5, r3, ASR #32
STRB r3, [r0], #1
ADDS r4, r4, #128
CMPGT r5, r4
EORLT r4, r5, r4, ASR #32
LDRSH r3, [r2], #2
STRB r4, [r0], #1
ADDS r12,r12,#128
CMPGT r5, r12
EORLT r12,r5, r12,ASR #32
LDRSH r4, [r2], #2
STRB r12,[r0], #1
ADDS r3, r3, #128
CMPGT r5, r3
EORLT r3, r5, r3, ASR #32
LDRSH r12,[r2], #2
STRB r3, [r0], #1
ADDS r4, r4, #128
CMPGT r5, r4
EORLT r4, r5, r4, ASR #32
LDRSH r3, [r2], #2
STRB r4, [r0], #1
ADDS r12,r12,#128
CMPGT r5, r12
EORLT r12,r5, r12,ASR #32
LDRSH r4, [r2], #2
STRB r12,[r0], #1
ADDS r3, r3, #128
CMPGT r5, r3
EORLT r3, r5, r3, ASR #32
STRB r3, [r0], #1
ADDS r4, r4, #128
CMPGT r5, r4
EORLT r4, r5, r4, ASR #32
STRB r4, [r0], r1
SUBS r14,r14,#1
BGT ofrintra_lp_arm
LDMFD r13!,{r4,r5,PC}
ENDP
oc_frag_recon_inter_arm PROC
; r0 = unsigned char *dst
; r1 = const unsigned char *src
; r2 = int ystride
; r3 = const ogg_int16_t residue[64]
STMFD r13!,{r5,r9-r11,r14}
MOV r9, #8
MOV r5, #255
SUB r2, r2, #7
ofrinter_lp_arm
LDRSH r12,[r3], #2
LDRB r14,[r1], #1
LDRSH r11,[r3], #2
LDRB r10,[r1], #1
ADDS r12,r12,r14
CMPGT r5, r12
EORLT r12,r5, r12,ASR #32
STRB r12,[r0], #1
ADDS r11,r11,r10
CMPGT r5, r11
LDRSH r12,[r3], #2
LDRB r14,[r1], #1
EORLT r11,r5, r11,ASR #32
STRB r11,[r0], #1
ADDS r12,r12,r14
CMPGT r5, r12
LDRSH r11,[r3], #2
LDRB r10,[r1], #1
EORLT r12,r5, r12,ASR #32
STRB r12,[r0], #1
ADDS r11,r11,r10
CMPGT r5, r11
LDRSH r12,[r3], #2
LDRB r14,[r1], #1
EORLT r11,r5, r11,ASR #32
STRB r11,[r0], #1
ADDS r12,r12,r14
CMPGT r5, r12
LDRSH r11,[r3], #2
LDRB r10,[r1], #1
EORLT r12,r5, r12,ASR #32
STRB r12,[r0], #1
ADDS r11,r11,r10
CMPGT r5, r11
LDRSH r12,[r3], #2
LDRB r14,[r1], #1
EORLT r11,r5, r11,ASR #32
STRB r11,[r0], #1
ADDS r12,r12,r14
CMPGT r5, r12
LDRSH r11,[r3], #2
LDRB r10,[r1], r2
EORLT r12,r5, r12,ASR #32
STRB r12,[r0], #1
ADDS r11,r11,r10
CMPGT r5, r11
EORLT r11,r5, r11,ASR #32
STRB r11,[r0], r2
SUBS r9, r9, #1
BGT ofrinter_lp_arm
LDMFD r13!,{r5,r9-r11,PC}
ENDP
oc_frag_recon_inter2_arm PROC
; r0 = unsigned char *dst
; r1 = const unsigned char *src1
; r2 = const unsigned char *src2
; r3 = int ystride
LDR r12,[r13]
; r12= const ogg_int16_t residue[64]
STMFD r13!,{r4-r8,r14}
MOV r14,#8
MOV r8, #255
SUB r3, r3, #7
ofrinter2_lp_arm
LDRB r5, [r1], #1
LDRB r6, [r2], #1
LDRSH r4, [r12],#2
LDRB r7, [r1], #1
ADD r5, r5, r6
ADDS r5, r4, r5, LSR #1
CMPGT r8, r5
LDRB r6, [r2], #1
LDRSH r4, [r12],#2
EORLT r5, r8, r5, ASR #32
STRB r5, [r0], #1
ADD r7, r7, r6
ADDS r7, r4, r7, LSR #1
CMPGT r8, r7
LDRB r5, [r1], #1
LDRB r6, [r2], #1
LDRSH r4, [r12],#2
EORLT r7, r8, r7, ASR #32
STRB r7, [r0], #1
ADD r5, r5, r6
ADDS r5, r4, r5, LSR #1
CMPGT r8, r5
LDRB r7, [r1], #1
LDRB r6, [r2], #1
LDRSH r4, [r12],#2
EORLT r5, r8, r5, ASR #32
STRB r5, [r0], #1
ADD r7, r7, r6
ADDS r7, r4, r7, LSR #1
CMPGT r8, r7
LDRB r5, [r1], #1
LDRB r6, [r2], #1
LDRSH r4, [r12],#2
EORLT r7, r8, r7, ASR #32
STRB r7, [r0], #1
ADD r5, r5, r6
ADDS r5, r4, r5, LSR #1
CMPGT r8, r5
LDRB r7, [r1], #1
LDRB r6, [r2], #1
LDRSH r4, [r12],#2
EORLT r5, r8, r5, ASR #32
STRB r5, [r0], #1
ADD r7, r7, r6
ADDS r7, r4, r7, LSR #1
CMPGT r8, r7
LDRB r5, [r1], #1
LDRB r6, [r2], #1
LDRSH r4, [r12],#2
EORLT r7, r8, r7, ASR #32
STRB r7, [r0], #1
ADD r5, r5, r6
ADDS r5, r4, r5, LSR #1
CMPGT r8, r5
LDRB r7, [r1], r3
LDRB r6, [r2], r3
LDRSH r4, [r12],#2
EORLT r5, r8, r5, ASR #32
STRB r5, [r0], #1
ADD r7, r7, r6
ADDS r7, r4, r7, LSR #1
CMPGT r8, r7
EORLT r7, r8, r7, ASR #32
STRB r7, [r0], r3
SUBS r14,r14,#1
BGT ofrinter2_lp_arm
LDMFD r13!,{r4-r8,PC}
ENDP
[ OC_ARM_ASM_EDSP
EXPORT oc_frag_copy_list_edsp
oc_frag_copy_list_edsp PROC
; r0 = _dst_frame
; r1 = _src_frame
; r2 = _ystride
; r3 = _fragis
; <> = _nfragis
; <> = _frag_buf_offs
LDR r12,[r13] ; r12 = _nfragis
STMFD r13!,{r4-r11,r14}
SUBS r12, r12, #1
LDRGE r5, [r3],#4 ; r5 = _fragis[fragii]
LDRGE r14,[r13,#4*10] ; r14 = _frag_buf_offs
BLT ofcl_edsp_end
ofcl_edsp_lp
MOV r4, r1
LDR r5, [r14,r5, LSL #2] ; r5 = _frag_buf_offs[_fragis[fragii]]
SUBS r12, r12, #1
; Stall (on XScale)
LDRD r6, [r4, r5]! ; r4 = _src_frame+frag_buf_off
LDRD r8, [r4, r2]!
; Stall
STRD r6, [r5, r0]! ; r5 = _dst_frame+frag_buf_off
STRD r8, [r5, r2]!
; Stall
LDRD r6, [r4, r2]! ; On Xscale at least, doing 3 consecutive
LDRD r8, [r4, r2]! ; loads causes a stall, but that's no worse
LDRD r10,[r4, r2]! ; than us only doing 2, and having to do
; another pair of LDRD/STRD later on.
; Stall
STRD r6, [r5, r2]!
STRD r8, [r5, r2]!
STRD r10,[r5, r2]!
LDRD r6, [r4, r2]!
LDRD r8, [r4, r2]!
LDRD r10,[r4, r2]!
STRD r6, [r5, r2]!
STRD r8, [r5, r2]!
STRD r10,[r5, r2]!
LDRGE r5, [r3],#4 ; r5 = _fragis[fragii]
BGE ofcl_edsp_lp
ofcl_edsp_end
LDMFD r13!,{r4-r11,PC}
ENDP
]
[ OC_ARM_ASM_MEDIA
EXPORT oc_frag_recon_intra_v6
EXPORT oc_frag_recon_inter_v6
EXPORT oc_frag_recon_inter2_v6
oc_frag_recon_intra_v6 PROC
; r0 = unsigned char *_dst
; r1 = int _ystride
; r2 = const ogg_int16_t _residue[64]
STMFD r13!,{r4-r6,r14}
MOV r14,#8
MOV r12,r2
LDR r6, =0x00800080
ofrintra_v6_lp
LDRD r2, [r12],#8 ; r2 = 11110000 r3 = 33332222
LDRD r4, [r12],#8 ; r4 = 55554444 r5 = 77776666
SUBS r14,r14,#1
QADD16 r2, r2, r6
QADD16 r3, r3, r6
QADD16 r4, r4, r6
QADD16 r5, r5, r6
USAT16 r2, #8, r2 ; r2 = __11__00
USAT16 r3, #8, r3 ; r3 = __33__22
USAT16 r4, #8, r4 ; r4 = __55__44
USAT16 r5, #8, r5 ; r5 = __77__66
ORR r2, r2, r2, LSR #8 ; r2 = __111100
ORR r3, r3, r3, LSR #8 ; r3 = __333322
ORR r4, r4, r4, LSR #8 ; r4 = __555544
ORR r5, r5, r5, LSR #8 ; r5 = __777766
PKHBT r2, r2, r3, LSL #16 ; r2 = 33221100
PKHBT r3, r4, r5, LSL #16 ; r3 = 77665544
STRD r2, [r0], r1
BGT ofrintra_v6_lp
LDMFD r13!,{r4-r6,PC}
ENDP
oc_frag_recon_inter_v6 PROC
; r0 = unsigned char *_dst
; r1 = const unsigned char *_src
; r2 = int _ystride
; r3 = const ogg_int16_t _residue[64]
STMFD r13!,{r4-r7,r14}
MOV r14,#8
ofrinter_v6_lp
LDRD r6, [r3], #8 ; r6 = 11110000 r7 = 33332222
SUBS r14,r14,#1
[ OC_ARM_CAN_UNALIGN_LDRD
LDRD r4, [r1], r2 ; Unaligned ; r4 = 33221100 r5 = 77665544
|
LDR r5, [r1, #4]
LDR r4, [r1], r2
]
PKHBT r12,r6, r7, LSL #16 ; r12= 22220000
PKHTB r7, r7, r6, ASR #16 ; r7 = 33331111
UXTB16 r6,r4 ; r6 = __22__00
UXTB16 r4,r4, ROR #8 ; r4 = __33__11
QADD16 r12,r12,r6 ; r12= xx22xx00
QADD16 r4, r7, r4 ; r4 = xx33xx11
LDRD r6, [r3], #8 ; r6 = 55554444 r7 = 77776666
USAT16 r4, #8, r4 ; r4 = __33__11
USAT16 r12,#8,r12 ; r12= __22__00
ORR r4, r12,r4, LSL #8 ; r4 = 33221100
PKHBT r12,r6, r7, LSL #16 ; r12= 66664444
PKHTB r7, r7, r6, ASR #16 ; r7 = 77775555
UXTB16 r6,r5 ; r6 = __66__44
UXTB16 r5,r5, ROR #8 ; r5 = __77__55
QADD16 r12,r12,r6 ; r12= xx66xx44
QADD16 r5, r7, r5 ; r5 = xx77xx55
USAT16 r12,#8, r12 ; r12= __66__44
USAT16 r5, #8, r5 ; r4 = __77__55
ORR r5, r12,r5, LSL #8 ; r5 = 33221100
STRD r4, [r0], r2
BGT ofrinter_v6_lp
LDMFD r13!,{r4-r7,PC}
ENDP
oc_frag_recon_inter2_v6 PROC
; r0 = unsigned char *_dst
; r1 = const unsigned char *_src1
; r2 = const unsigned char *_src2
; r3 = int _ystride
LDR r12,[r13]
; r12= const ogg_int16_t _residue[64]
STMFD r13!,{r4-r9,r14}
MOV r14,#8
ofrinter2_v6_lp
LDRD r6, [r12,#8] ; r6 = 55554444 r7 = 77776666
SUBS r14,r14,#1
LDR r4, [r1, #4] ; Unaligned ; r4 = src1[1] = 77665544
LDR r5, [r2, #4] ; Unaligned ; r5 = src2[1] = 77665544
PKHBT r8, r6, r7, LSL #16 ; r8 = 66664444
PKHTB r9, r7, r6, ASR #16 ; r9 = 77775555
UHADD8 r4, r4, r5 ; r4 = (src1[7,6,5,4] + src2[7,6,5,4])>>1
UXTB16 r5, r4 ; r5 = __66__44
UXTB16 r4, r4, ROR #8 ; r4 = __77__55
QADD16 r8, r8, r5 ; r8 = xx66xx44
QADD16 r9, r9, r4 ; r9 = xx77xx55
LDRD r6,[r12],#16 ; r6 = 33332222 r7 = 11110000
USAT16 r8, #8, r8 ; r8 = __66__44
LDR r4, [r1], r3 ; Unaligned ; r4 = src1[0] = 33221100
USAT16 r9, #8, r9 ; r9 = __77__55
LDR r5, [r2], r3 ; Unaligned ; r5 = src2[0] = 33221100
ORR r9, r8, r9, LSL #8 ; r9 = 77665544
PKHBT r8, r6, r7, LSL #16 ; r8 = 22220000
UHADD8 r4, r4, r5 ; r4 = (src1[3,2,1,0] + src2[3,2,1,0])>>1
PKHTB r7, r7, r6, ASR #16 ; r7 = 33331111
UXTB16 r5, r4 ; r5 = __22__00
UXTB16 r4, r4, ROR #8 ; r4 = __33__11
QADD16 r8, r8, r5 ; r8 = xx22xx00
QADD16 r7, r7, r4 ; r7 = xx33xx11
USAT16 r8, #8, r8 ; r8 = __22__00
USAT16 r7, #8, r7 ; r7 = __33__11
ORR r8, r8, r7, LSL #8 ; r8 = 33221100
STRD r8, [r0], r3
BGT ofrinter2_v6_lp
LDMFD r13!,{r4-r9,PC}
ENDP
]
[ OC_ARM_ASM_NEON
EXPORT oc_frag_copy_list_neon
EXPORT oc_frag_recon_intra_neon
EXPORT oc_frag_recon_inter_neon
EXPORT oc_frag_recon_inter2_neon
oc_frag_copy_list_neon PROC
; r0 = _dst_frame
; r1 = _src_frame
; r2 = _ystride
; r3 = _fragis
; <> = _nfragis
; <> = _frag_buf_offs
LDR r12,[r13] ; r12 = _nfragis
STMFD r13!,{r4-r7,r14}
CMP r12, #1
LDRGE r6, [r3] ; r6 = _fragis[fragii]
LDRGE r14,[r13,#4*6] ; r14 = _frag_buf_offs
BLT ofcl_neon_end
; Stall (2 on Xscale)
LDR r6, [r14,r6, LSL #2] ; r6 = _frag_buf_offs[_fragis[fragii]]
; Stall (on XScale)
MOV r7, r6 ; Guarantee PLD points somewhere valid.
ofcl_neon_lp
ADD r4, r1, r6
VLD1.64 {D0}, [r4@64], r2
ADD r5, r0, r6
VLD1.64 {D1}, [r4@64], r2
SUBS r12, r12, #1
VLD1.64 {D2}, [r4@64], r2
LDRGT r6, [r3,#4]! ; r6 = _fragis[fragii]
VLD1.64 {D3}, [r4@64], r2
LDRGT r6, [r14,r6, LSL #2] ; r6 = _frag_buf_offs[_fragis[fragii]]
VLD1.64 {D4}, [r4@64], r2
ADDGT r7, r1, r6
VLD1.64 {D5}, [r4@64], r2
PLD [r7]
VLD1.64 {D6}, [r4@64], r2
PLD [r7, r2]
VLD1.64 {D7}, [r4@64]
PLD [r7, r2, LSL #1]
VST1.64 {D0}, [r5@64], r2
ADDGT r7, r7, r2, LSL #2
VST1.64 {D1}, [r5@64], r2
PLD [r7, -r2]
VST1.64 {D2}, [r5@64], r2
PLD [r7]
VST1.64 {D3}, [r5@64], r2
PLD [r7, r2]
VST1.64 {D4}, [r5@64], r2
PLD [r7, r2, LSL #1]
VST1.64 {D5}, [r5@64], r2
ADDGT r7, r7, r2, LSL #2
VST1.64 {D6}, [r5@64], r2
PLD [r7, -r2]
VST1.64 {D7}, [r5@64]
BGT ofcl_neon_lp
ofcl_neon_end
LDMFD r13!,{r4-r7,PC}
ENDP
oc_frag_recon_intra_neon PROC
; r0 = unsigned char *_dst
; r1 = int _ystride
; r2 = const ogg_int16_t _residue[64]
MOV r3, #128
VDUP.S16 Q0, r3
VLDMIA r2, {D16-D31} ; D16= 3333222211110000 etc ; 9(8) cycles
VQADD.S16 Q8, Q8, Q0
VQADD.S16 Q9, Q9, Q0
VQADD.S16 Q10,Q10,Q0
VQADD.S16 Q11,Q11,Q0
VQADD.S16 Q12,Q12,Q0
VQADD.S16 Q13,Q13,Q0
VQADD.S16 Q14,Q14,Q0
VQADD.S16 Q15,Q15,Q0
VQMOVUN.S16 D16,Q8 ; D16= 7766554433221100 ; 1 cycle
VQMOVUN.S16 D17,Q9 ; D17= FFEEDDCCBBAA9988 ; 1 cycle
VQMOVUN.S16 D18,Q10 ; D18= NNMMLLKKJJIIHHGG ; 1 cycle
VST1.64 {D16},[r0@64], r1
VQMOVUN.S16 D19,Q11 ; D19= VVUUTTSSRRQQPPOO ; 1 cycle
VST1.64 {D17},[r0@64], r1
VQMOVUN.S16 D20,Q12 ; D20= ddccbbaaZZYYXXWW ; 1 cycle
VST1.64 {D18},[r0@64], r1
VQMOVUN.S16 D21,Q13 ; D21= llkkjjiihhggffee ; 1 cycle
VST1.64 {D19},[r0@64], r1
VQMOVUN.S16 D22,Q14 ; D22= ttssrrqqppoonnmm ; 1 cycle
VST1.64 {D20},[r0@64], r1
VQMOVUN.S16 D23,Q15 ; D23= !!@@zzyyxxwwvvuu ; 1 cycle
VST1.64 {D21},[r0@64], r1
VST1.64 {D22},[r0@64], r1
VST1.64 {D23},[r0@64], r1
MOV PC,R14
ENDP
oc_frag_recon_inter_neon PROC
; r0 = unsigned char *_dst
; r1 = const unsigned char *_src
; r2 = int _ystride
; r3 = const ogg_int16_t _residue[64]
VLDMIA r3, {D16-D31} ; D16= 3333222211110000 etc ; 9(8) cycles
VLD1.64 {D0}, [r1], r2
VLD1.64 {D2}, [r1], r2
VMOVL.U8 Q0, D0 ; Q0 = __77__66__55__44__33__22__11__00
VLD1.64 {D4}, [r1], r2
VMOVL.U8 Q1, D2 ; etc
VLD1.64 {D6}, [r1], r2
VMOVL.U8 Q2, D4
VMOVL.U8 Q3, D6
VQADD.S16 Q8, Q8, Q0
VLD1.64 {D0}, [r1], r2
VQADD.S16 Q9, Q9, Q1
VLD1.64 {D2}, [r1], r2
VQADD.S16 Q10,Q10,Q2
VLD1.64 {D4}, [r1], r2
VQADD.S16 Q11,Q11,Q3
VLD1.64 {D6}, [r1], r2
VMOVL.U8 Q0, D0
VMOVL.U8 Q1, D2
VMOVL.U8 Q2, D4
VMOVL.U8 Q3, D6
VQADD.S16 Q12,Q12,Q0
VQADD.S16 Q13,Q13,Q1
VQADD.S16 Q14,Q14,Q2
VQADD.S16 Q15,Q15,Q3
VQMOVUN.S16 D16,Q8
VQMOVUN.S16 D17,Q9
VQMOVUN.S16 D18,Q10
VST1.64 {D16},[r0@64], r2
VQMOVUN.S16 D19,Q11
VST1.64 {D17},[r0@64], r2
VQMOVUN.S16 D20,Q12
VST1.64 {D18},[r0@64], r2
VQMOVUN.S16 D21,Q13
VST1.64 {D19},[r0@64], r2
VQMOVUN.S16 D22,Q14
VST1.64 {D20},[r0@64], r2
VQMOVUN.S16 D23,Q15
VST1.64 {D21},[r0@64], r2
VST1.64 {D22},[r0@64], r2
VST1.64 {D23},[r0@64], r2
MOV PC,R14
ENDP
oc_frag_recon_inter2_neon PROC
; r0 = unsigned char *_dst
; r1 = const unsigned char *_src1
; r2 = const unsigned char *_src2
; r3 = int _ystride
LDR r12,[r13]
; r12= const ogg_int16_t _residue[64]
VLDMIA r12,{D16-D31}
VLD1.64 {D0}, [r1], r3
VLD1.64 {D4}, [r2], r3
VLD1.64 {D1}, [r1], r3
VLD1.64 {D5}, [r2], r3
VHADD.U8 Q2, Q0, Q2 ; Q2 = FFEEDDCCBBAA99887766554433221100
VLD1.64 {D2}, [r1], r3
VLD1.64 {D6}, [r2], r3
VMOVL.U8 Q0, D4 ; Q0 = __77__66__55__44__33__22__11__00
VLD1.64 {D3}, [r1], r3
VMOVL.U8 Q2, D5 ; etc
VLD1.64 {D7}, [r2], r3
VHADD.U8 Q3, Q1, Q3
VQADD.S16 Q8, Q8, Q0
VQADD.S16 Q9, Q9, Q2
VLD1.64 {D0}, [r1], r3
VMOVL.U8 Q1, D6
VLD1.64 {D4}, [r2], r3
VMOVL.U8 Q3, D7
VLD1.64 {D1}, [r1], r3
VQADD.S16 Q10,Q10,Q1
VLD1.64 {D5}, [r2], r3
VQADD.S16 Q11,Q11,Q3
VLD1.64 {D2}, [r1], r3
VHADD.U8 Q2, Q0, Q2
VLD1.64 {D6}, [r2], r3
VLD1.64 {D3}, [r1], r3
VMOVL.U8 Q0, D4
VLD1.64 {D7}, [r2], r3
VMOVL.U8 Q2, D5
VHADD.U8 Q3, Q1, Q3
VQADD.S16 Q12,Q12,Q0
VQADD.S16 Q13,Q13,Q2
VMOVL.U8 Q1, D6
VMOVL.U8 Q3, D7
VQADD.S16 Q14,Q14,Q1
VQADD.S16 Q15,Q15,Q3
VQMOVUN.S16 D16,Q8
VQMOVUN.S16 D17,Q9
VQMOVUN.S16 D18,Q10
VST1.64 {D16},[r0@64], r3
VQMOVUN.S16 D19,Q11
VST1.64 {D17},[r0@64], r3
VQMOVUN.S16 D20,Q12
VST1.64 {D18},[r0@64], r3
VQMOVUN.S16 D21,Q13
VST1.64 {D19},[r0@64], r3
VQMOVUN.S16 D22,Q14
VST1.64 {D20},[r0@64], r3
VQMOVUN.S16 D23,Q15
VST1.64 {D21},[r0@64], r3
VST1.64 {D22},[r0@64], r3
VST1.64 {D23},[r0@64], r3
MOV PC,R14
ENDP
]
END

1908
media/libtheora/lib/arm/armidct.s Normal file
View File

File diff suppressed because it is too large Load Diff

126
media/libtheora/lib/arm/armint.h Normal file
View File

@ -0,0 +1,126 @@
/********************************************************************
* *
* THIS FILE IS PART OF THE OggTheora SOFTWARE CODEC SOURCE CODE. *
* USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS *
* GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
* IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING. *
* *
* THE Theora SOURCE CODE IS COPYRIGHT (C) 2002-2010 *
* by the Xiph.Org Foundation and contributors http://www.xiph.org/ *
* *
********************************************************************
function:
last mod: $Id: x86int.h 17344 2010-07-21 01:42:18Z tterribe $
********************************************************************/
#if !defined(_arm_armint_H)
# define _arm_armint_H (1)
# include "../internal.h"
# if defined(OC_ARM_ASM)
# if defined(__ARMEB__)
# error "Big-endian configurations are not supported by the ARM asm. " \
"Reconfigure with --disable-asm or undefine OC_ARM_ASM."
# endif
# define oc_state_accel_init oc_state_accel_init_arm
/*This function is implemented entirely in asm, so it's helpful to pull out all
of the things that depend on structure offsets.
We reuse the function pointer with the wrong prototype, though.*/
# define oc_state_loop_filter_frag_rows(_state,_bv,_refi,_pli, \
_fragy0,_fragy_end) \
((oc_loop_filter_frag_rows_arm_func) \
(_state)->opt_vtable.state_loop_filter_frag_rows)( \
(_state)->ref_frame_data[(_refi)],(_state)->ref_ystride[(_pli)], \
(_bv), \
(_state)->frags, \
(_state)->fplanes[(_pli)].froffset \
+(_fragy0)*(ptrdiff_t)(_state)->fplanes[(_pli)].nhfrags, \
(_state)->fplanes[(_pli)].froffset \
+(_fragy_end)*(ptrdiff_t)(_state)->fplanes[(_pli)].nhfrags, \
(_state)->fplanes[(_pli)].froffset, \
(_state)->fplanes[(_pli)].froffset+(_state)->fplanes[(_pli)].nfrags, \
(_state)->frag_buf_offs, \
(_state)->fplanes[(_pli)].nhfrags)
/*For everything else the default vtable macros are fine.*/
# define OC_STATE_USE_VTABLE (1)
# endif
# include "../state.h"
# include "armcpu.h"
# if defined(OC_ARM_ASM)
typedef void (*oc_loop_filter_frag_rows_arm_func)(
unsigned char *_ref_frame_data,int _ystride,signed char _bv[256],
const oc_fragment *_frags,ptrdiff_t _fragi0,ptrdiff_t _fragi0_end,
ptrdiff_t _fragi_top,ptrdiff_t _fragi_bot,
const ptrdiff_t *_frag_buf_offs,int _nhfrags);
void oc_state_accel_init_arm(oc_theora_state *_state);
void oc_frag_copy_list_arm(unsigned char *_dst_frame,
const unsigned char *_src_frame,int _ystride,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,const ptrdiff_t *_frag_buf_offs);
void oc_frag_recon_intra_arm(unsigned char *_dst,int _ystride,
const ogg_int16_t *_residue);
void oc_frag_recon_inter_arm(unsigned char *_dst,const unsigned char *_src,
int _ystride,const ogg_int16_t *_residue);
void oc_frag_recon_inter2_arm(unsigned char *_dst,const unsigned char *_src1,
const unsigned char *_src2,int _ystride,const ogg_int16_t *_residue);
void oc_idct8x8_1_arm(ogg_int16_t _y[64],ogg_uint16_t _dc);
void oc_idct8x8_arm(ogg_int16_t _y[64],ogg_int16_t _x[64],int _last_zzi);
void oc_state_frag_recon_arm(const oc_theora_state *_state,ptrdiff_t _fragi,
int _pli,ogg_int16_t _dct_coeffs[128],int _last_zzi,ogg_uint16_t _dc_quant);
void oc_loop_filter_frag_rows_arm(unsigned char *_ref_frame_data,
int _ystride,signed char *_bv,const oc_fragment *_frags,ptrdiff_t _fragi0,
ptrdiff_t _fragi0_end,ptrdiff_t _fragi_top,ptrdiff_t _fragi_bot,
const ptrdiff_t *_frag_buf_offs,int _nhfrags);
# if defined(OC_ARM_ASM_EDSP)
void oc_frag_copy_list_edsp(unsigned char *_dst_frame,
const unsigned char *_src_frame,int _ystride,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,const ptrdiff_t *_frag_buf_offs);
# if defined(OC_ARM_ASM_MEDIA)
void oc_frag_recon_intra_v6(unsigned char *_dst,int _ystride,
const ogg_int16_t *_residue);
void oc_frag_recon_inter_v6(unsigned char *_dst,const unsigned char *_src,
int _ystride,const ogg_int16_t *_residue);
void oc_frag_recon_inter2_v6(unsigned char *_dst,const unsigned char *_src1,
const unsigned char *_src2,int _ystride,const ogg_int16_t *_residue);
void oc_idct8x8_1_v6(ogg_int16_t _y[64],ogg_uint16_t _dc);
void oc_idct8x8_v6(ogg_int16_t _y[64],ogg_int16_t _x[64],int _last_zzi);
void oc_state_frag_recon_v6(const oc_theora_state *_state,ptrdiff_t _fragi,
int _pli,ogg_int16_t _dct_coeffs[128],int _last_zzi,ogg_uint16_t _dc_quant);
void oc_loop_filter_init_v6(signed char *_bv,int _flimit);
void oc_loop_filter_frag_rows_v6(unsigned char *_ref_frame_data,
int _ystride,signed char *_bv,const oc_fragment *_frags,ptrdiff_t _fragi0,
ptrdiff_t _fragi0_end,ptrdiff_t _fragi_top,ptrdiff_t _fragi_bot,
const ptrdiff_t *_frag_buf_offs,int _nhfrags);
# if defined(OC_ARM_ASM_NEON)
void oc_frag_copy_list_neon(unsigned char *_dst_frame,
const unsigned char *_src_frame,int _ystride,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,const ptrdiff_t *_frag_buf_offs);
void oc_frag_recon_intra_neon(unsigned char *_dst,int _ystride,
const ogg_int16_t *_residue);
void oc_frag_recon_inter_neon(unsigned char *_dst,const unsigned char *_src,
int _ystride,const ogg_int16_t *_residue);
void oc_frag_recon_inter2_neon(unsigned char *_dst,const unsigned char *_src1,
const unsigned char *_src2,int _ystride,const ogg_int16_t *_residue);
void oc_idct8x8_1_neon(ogg_int16_t _y[64],ogg_uint16_t _dc);
void oc_idct8x8_neon(ogg_int16_t _y[64],ogg_int16_t _x[64],int _last_zzi);
void oc_state_frag_recon_neon(const oc_theora_state *_state,ptrdiff_t _fragi,
int _pli,ogg_int16_t _dct_coeffs[128],int _last_zzi,ogg_uint16_t _dc_quant);
void oc_loop_filter_init_neon(signed char *_bv,int _flimit);
void oc_loop_filter_frag_rows_neon(unsigned char *_ref_frame_data,
int _ystride,signed char *_bv,const oc_fragment *_frags,ptrdiff_t _fragi0,
ptrdiff_t _fragi0_end,ptrdiff_t _fragi_top,ptrdiff_t _fragi_bot,
const ptrdiff_t *_frag_buf_offs,int _nhfrags);
# endif
# endif
# endif
# endif
#endif

676
media/libtheora/lib/arm/armloop.s Normal file
View File

@ -0,0 +1,676 @@
;********************************************************************
;* *
;* THIS FILE IS PART OF THE OggTheora SOFTWARE CODEC SOURCE CODE. *
;* USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS *
;* GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
;* IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING. *
;* *
;* THE Theora SOURCE CODE IS COPYRIGHT (C) 2002-2010 *
;* by the Xiph.Org Foundation and contributors http://www.xiph.org/ *
;* *
;********************************************************************
; Original implementation:
; Copyright (C) 2009 Robin Watts for Pinknoise Productions Ltd
; last mod: $Id: armloop.s 17481 2010-10-03 22:49:42Z tterribe $
;********************************************************************
AREA |.text|, CODE, READONLY
GET armopts.s
EXPORT oc_loop_filter_frag_rows_arm
; Which bit this is depends on the order of packing within a bitfield.
; Hopefully that doesn't change among any of the relevant compilers.
OC_FRAG_CODED_FLAG * 1
; Vanilla ARM v4 version
loop_filter_h_arm PROC
; r0 = unsigned char *_pix
; r1 = int _ystride
; r2 = int *_bv
; preserves r0-r3
STMFD r13!,{r3-r6,r14}
MOV r14,#8
MOV r6, #255
lfh_arm_lp
LDRB r3, [r0, #-2] ; r3 = _pix[0]
LDRB r12,[r0, #1] ; r12= _pix[3]
LDRB r4, [r0, #-1] ; r4 = _pix[1]
LDRB r5, [r0] ; r5 = _pix[2]
SUB r3, r3, r12 ; r3 = _pix[0]-_pix[3]+4
ADD r3, r3, #4
SUB r12,r5, r4 ; r12= _pix[2]-_pix[1]
ADD r12,r12,r12,LSL #1 ; r12= 3*(_pix[2]-_pix[1])
ADD r12,r12,r3 ; r12= _pix[0]-_pix[3]+3*(_pix[2]-_pix[1])+4
MOV r12,r12,ASR #3
LDRSB r12,[r2, r12]
; Stall (2 on Xscale)
ADDS r4, r4, r12
CMPGT r6, r4
EORLT r4, r6, r4, ASR #32
SUBS r5, r5, r12
CMPGT r6, r5
EORLT r5, r6, r5, ASR #32
STRB r4, [r0, #-1]
STRB r5, [r0], r1
SUBS r14,r14,#1
BGT lfh_arm_lp
SUB r0, r0, r1, LSL #3
LDMFD r13!,{r3-r6,PC}
ENDP
loop_filter_v_arm PROC
; r0 = unsigned char *_pix
; r1 = int _ystride
; r2 = int *_bv
; preserves r0-r3
STMFD r13!,{r3-r6,r14}
MOV r14,#8
MOV r6, #255
lfv_arm_lp
LDRB r3, [r0, -r1, LSL #1] ; r3 = _pix[0]
LDRB r12,[r0, r1] ; r12= _pix[3]
LDRB r4, [r0, -r1] ; r4 = _pix[1]
LDRB r5, [r0] ; r5 = _pix[2]
SUB r3, r3, r12 ; r3 = _pix[0]-_pix[3]+4
ADD r3, r3, #4
SUB r12,r5, r4 ; r12= _pix[2]-_pix[1]
ADD r12,r12,r12,LSL #1 ; r12= 3*(_pix[2]-_pix[1])
ADD r12,r12,r3 ; r12= _pix[0]-_pix[3]+3*(_pix[2]-_pix[1])+4
MOV r12,r12,ASR #3
LDRSB r12,[r2, r12]
; Stall (2 on Xscale)
ADDS r4, r4, r12
CMPGT r6, r4
EORLT r4, r6, r4, ASR #32
SUBS r5, r5, r12
CMPGT r6, r5
EORLT r5, r6, r5, ASR #32
STRB r4, [r0, -r1]
STRB r5, [r0], #1
SUBS r14,r14,#1
BGT lfv_arm_lp
SUB r0, r0, #8
LDMFD r13!,{r3-r6,PC}
ENDP
oc_loop_filter_frag_rows_arm PROC
; r0 = _ref_frame_data
; r1 = _ystride
; r2 = _bv
; r3 = _frags
; r4 = _fragi0
; r5 = _fragi0_end
; r6 = _fragi_top
; r7 = _fragi_bot
; r8 = _frag_buf_offs
; r9 = _nhfrags
MOV r12,r13
STMFD r13!,{r0,r4-r11,r14}
LDMFD r12,{r4-r9}
ADD r2, r2, #127 ; _bv += 127
CMP r4, r5 ; if(_fragi0>=_fragi0_end)
BGE oslffri_arm_end ; bail
SUBS r9, r9, #1 ; r9 = _nhfrags-1 if (r9<=0)
BLE oslffri_arm_end ; bail
ADD r3, r3, r4, LSL #2 ; r3 = &_frags[fragi]
ADD r8, r8, r4, LSL #2 ; r8 = &_frag_buf_offs[fragi]
SUB r7, r7, r9 ; _fragi_bot -= _nhfrags;
oslffri_arm_lp1
MOV r10,r4 ; r10= fragi = _fragi0
ADD r11,r4, r9 ; r11= fragi_end-1=fragi+_nhfrags-1
oslffri_arm_lp2
LDR r14,[r3], #4 ; r14= _frags[fragi] _frags++
LDR r0, [r13] ; r0 = _ref_frame_data
LDR r12,[r8], #4 ; r12= _frag_buf_offs[fragi] _frag_buf_offs++
TST r14,#OC_FRAG_CODED_FLAG
BEQ oslffri_arm_uncoded
CMP r10,r4 ; if (fragi>_fragi0)
ADD r0, r0, r12 ; r0 = _ref_frame_data + _frag_buf_offs[fragi]
BLGT loop_filter_h_arm
CMP r4, r6 ; if (_fragi0>_fragi_top)
BLGT loop_filter_v_arm
CMP r10,r11 ; if(fragi+1<fragi_end)===(fragi<fragi_end-1)
LDRLT r12,[r3] ; r12 = _frags[fragi+1]
ADD r0, r0, #8
ADD r10,r10,#1 ; r10 = fragi+1;
ANDLT r12,r12,#OC_FRAG_CODED_FLAG
CMPLT r12,#OC_FRAG_CODED_FLAG ; && _frags[fragi+1].coded==0
BLLT loop_filter_h_arm
CMP r10,r7 ; if (fragi<_fragi_bot)
LDRLT r12,[r3, r9, LSL #2] ; r12 = _frags[fragi+1+_nhfrags-1]
SUB r0, r0, #8
ADD r0, r0, r1, LSL #3
ANDLT r12,r12,#OC_FRAG_CODED_FLAG
CMPLT r12,#OC_FRAG_CODED_FLAG
BLLT loop_filter_v_arm
CMP r10,r11 ; while(fragi<=fragi_end-1)
BLE oslffri_arm_lp2
MOV r4, r10 ; r4 = fragi0 += _nhfrags
CMP r4, r5
BLT oslffri_arm_lp1
oslffri_arm_end
LDMFD r13!,{r0,r4-r11,PC}
oslffri_arm_uncoded
ADD r10,r10,#1
CMP r10,r11
BLE oslffri_arm_lp2
MOV r4, r10 ; r4 = _fragi0 += _nhfrags
CMP r4, r5
BLT oslffri_arm_lp1
LDMFD r13!,{r0,r4-r11,PC}
ENDP
[ OC_ARM_ASM_MEDIA
EXPORT oc_loop_filter_init_v6
EXPORT oc_loop_filter_frag_rows_v6
oc_loop_filter_init_v6 PROC
; r0 = _bv
; r1 = _flimit (=L from the spec)
MVN r1, r1, LSL #1 ; r1 = <0xFFFFFF|255-2*L>
AND r1, r1, #255 ; r1 = ll=r1&0xFF
ORR r1, r1, r1, LSL #8 ; r1 = <ll|ll>
PKHBT r1, r1, r1, LSL #16 ; r1 = <ll|ll|ll|ll>
STR r1, [r0]
MOV PC,r14
ENDP
; We could use the same strategy as the v filter below, but that would require
; 40 instructions to load the data and transpose it into columns and another
; 32 to write out the results at the end, plus the 52 instructions to do the
; filtering itself.
; This is slightly less, and less code, even assuming we could have shared the
; 52 instructions in the middle with the other function.
; It executes slightly fewer instructions than the ARMv6 approach David Conrad
; proposed for FFmpeg, but not by much:
; http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/2010-February/083141.html
; His is a lot less code, though, because it only does two rows at once instead
; of four.
loop_filter_h_v6 PROC
; r0 = unsigned char *_pix
; r1 = int _ystride
; r2 = int _ll
; preserves r0-r3
STMFD r13!,{r4-r11,r14}
LDR r12,=0x10003
BL loop_filter_h_core_v6
ADD r0, r0, r1, LSL #2
BL loop_filter_h_core_v6
SUB r0, r0, r1, LSL #2
LDMFD r13!,{r4-r11,PC}
ENDP
loop_filter_h_core_v6 PROC
; r0 = unsigned char *_pix
; r1 = int _ystride
; r2 = int _ll
; r12= 0x10003
; Preserves r0-r3, r12; Clobbers r4-r11.
LDR r4,[r0, #-2]! ; r4 = <p3|p2|p1|p0>
; Single issue
LDR r5,[r0, r1]! ; r5 = <q3|q2|q1|q0>
UXTB16 r6, r4, ROR #16 ; r6 = <p0|p2>
UXTB16 r4, r4, ROR #8 ; r4 = <p3|p1>
UXTB16 r7, r5, ROR #16 ; r7 = <q0|q2>
UXTB16 r5, r5, ROR #8 ; r5 = <q3|q1>
PKHBT r8, r4, r5, LSL #16 ; r8 = <__|q1|__|p1>
PKHBT r9, r6, r7, LSL #16 ; r9 = <__|q2|__|p2>
SSUB16 r6, r4, r6 ; r6 = <p3-p0|p1-p2>
SMLAD r6, r6, r12,r12 ; r6 = <????|(p3-p0)+3*(p1-p2)+3>
SSUB16 r7, r5, r7 ; r7 = <q3-q0|q1-q2>
SMLAD r7, r7, r12,r12 ; r7 = <????|(q0-q3)+3*(q2-q1)+4>
LDR r4,[r0, r1]! ; r4 = <r3|r2|r1|r0>
MOV r6, r6, ASR #3 ; r6 = <??????|(p3-p0)+3*(p1-p2)+3>>3>
LDR r5,[r0, r1]! ; r5 = <s3|s2|s1|s0>
PKHBT r11,r6, r7, LSL #13 ; r11= <??|-R_q|??|-R_p>
UXTB16 r6, r4, ROR #16 ; r6 = <r0|r2>
UXTB16 r11,r11 ; r11= <__|-R_q|__|-R_p>
UXTB16 r4, r4, ROR #8 ; r4 = <r3|r1>
UXTB16 r7, r5, ROR #16 ; r7 = <s0|s2>
PKHBT r10,r6, r7, LSL #16 ; r10= <__|s2|__|r2>
SSUB16 r6, r4, r6 ; r6 = <r3-r0|r1-r2>
UXTB16 r5, r5, ROR #8 ; r5 = <s3|s1>
SMLAD r6, r6, r12,r12 ; r6 = <????|(r3-r0)+3*(r2-r1)+3>
SSUB16 r7, r5, r7 ; r7 = <r3-r0|r1-r2>
SMLAD r7, r7, r12,r12 ; r7 = <????|(s0-s3)+3*(s2-s1)+4>
ORR r9, r9, r10, LSL #8 ; r9 = <s2|q2|r2|p2>
MOV r6, r6, ASR #3 ; r6 = <??????|(r0-r3)+3*(r2-r1)+4>>3>
PKHBT r10,r4, r5, LSL #16 ; r10= <__|s1|__|r1>
PKHBT r6, r6, r7, LSL #13 ; r6 = <??|-R_s|??|-R_r>
ORR r8, r8, r10, LSL #8 ; r8 = <s1|q1|r1|p1>
UXTB16 r6, r6 ; r6 = <__|-R_s|__|-R_r>
MOV r10,#0
ORR r6, r11,r6, LSL #8 ; r6 = <-R_s|-R_q|-R_r|-R_p>
; Single issue
; There's no min, max or abs instruction.
; SSUB8 and SEL will work for abs, and we can do all the rest with
; unsigned saturated adds, which means the GE flags are still all
; set when we're done computing lflim(abs(R_i),L).
; This allows us to both add and subtract, and split the results by
; the original sign of R_i.
SSUB8 r7, r10,r6
; Single issue
SEL r7, r7, r6 ; r7 = abs(R_i)
; Single issue
UQADD8 r4, r7, r2 ; r4 = 255-max(2*L-abs(R_i),0)
; Single issue
UQADD8 r7, r7, r4
; Single issue
UQSUB8 r7, r7, r4 ; r7 = min(abs(R_i),max(2*L-abs(R_i),0))
; Single issue
UQSUB8 r4, r8, r7
UQADD8 r5, r9, r7
UQADD8 r8, r8, r7
UQSUB8 r9, r9, r7
SEL r8, r8, r4 ; r8 = p1+lflim(R_i,L)
SEL r9, r9, r5 ; r9 = p2-lflim(R_i,L)
MOV r5, r9, LSR #24 ; r5 = s2
STRB r5, [r0,#2]!
MOV r4, r8, LSR #24 ; r4 = s1
STRB r4, [r0,#-1]
MOV r5, r9, LSR #8 ; r5 = r2
STRB r5, [r0,-r1]!
MOV r4, r8, LSR #8 ; r4 = r1
STRB r4, [r0,#-1]
MOV r5, r9, LSR #16 ; r5 = q2
STRB r5, [r0,-r1]!
MOV r4, r8, LSR #16 ; r4 = q1
STRB r4, [r0,#-1]
; Single issue
STRB r9, [r0,-r1]!
; Single issue
STRB r8, [r0,#-1]
MOV PC,r14
ENDP
; This uses the same strategy as the MMXEXT version for x86, except that UHADD8
; computes (a+b>>1) instead of (a+b+1>>1) like PAVGB.
; This works just as well, with the following procedure for computing the
; filter value, f:
; u = ~UHADD8(p1,~p2);
; v = UHADD8(~p1,p2);
; m = v-u;
; a = m^UHADD8(m^p0,m^~p3);
; f = UHADD8(UHADD8(a,u1),v1);
; where f = 127+R, with R in [-127,128] defined as in the spec.
; This is exactly the same amount of arithmetic as the version that uses PAVGB
; as the basic operator.
; It executes about 2/3 the number of instructions of David Conrad's approach,
; but requires more code, because it does all eight columns at once, instead
; of four at a time.
loop_filter_v_v6 PROC
; r0 = unsigned char *_pix
; r1 = int _ystride
; r2 = int _ll
; preserves r0-r11
STMFD r13!,{r4-r11,r14}
LDRD r6, [r0, -r1]! ; r7, r6 = <p5|p1>
LDRD r4, [r0, -r1] ; r5, r4 = <p4|p0>
LDRD r8, [r0, r1]! ; r9, r8 = <p6|p2>
MVN r14,r6 ; r14= ~p1
LDRD r10,[r0, r1] ; r11,r10= <p7|p3>
; Filter the first four columns.
MVN r12,r8 ; r12= ~p2
UHADD8 r14,r14,r8 ; r14= v1=~p1+p2>>1
UHADD8 r12,r12,r6 ; r12= p1+~p2>>1
MVN r10, r10 ; r10=~p3
MVN r12,r12 ; r12= u1=~p1+p2+1>>1
SSUB8 r14,r14,r12 ; r14= m1=v1-u1
; Single issue
EOR r4, r4, r14 ; r4 = m1^p0
EOR r10,r10,r14 ; r10= m1^~p3
UHADD8 r4, r4, r10 ; r4 = (m1^p0)+(m1^~p3)>>1
; Single issue
EOR r4, r4, r14 ; r4 = a1=m1^((m1^p0)+(m1^~p3)>>1)
SADD8 r14,r14,r12 ; r14= v1=m1+u1
UHADD8 r4, r4, r12 ; r4 = a1+u1>>1
MVN r12,r9 ; r12= ~p6
UHADD8 r4, r4, r14 ; r4 = f1=(a1+u1>>1)+v1>>1
; Filter the second four columns.
MVN r14,r7 ; r14= ~p5
UHADD8 r12,r12,r7 ; r12= p5+~p6>>1
UHADD8 r14,r14,r9 ; r14= v2=~p5+p6>>1
MVN r12,r12 ; r12= u2=~p5+p6+1>>1
MVN r11,r11 ; r11=~p7
SSUB8 r10,r14,r12 ; r10= m2=v2-u2
; Single issue
EOR r5, r5, r10 ; r5 = m2^p4
EOR r11,r11,r10 ; r11= m2^~p7
UHADD8 r5, r5, r11 ; r5 = (m2^p4)+(m2^~p7)>>1
; Single issue
EOR r5, r5, r10 ; r5 = a2=m2^((m2^p4)+(m2^~p7)>>1)
; Single issue
UHADD8 r5, r5, r12 ; r5 = a2+u2>>1
LDR r12,=0x7F7F7F7F ; r12 = {127}x4
UHADD8 r5, r5, r14 ; r5 = f2=(a2+u2>>1)+v2>>1
; Now split f[i] by sign.
; There's no min or max instruction.
; We could use SSUB8 and SEL, but this is just as many instructions and
; dual issues more (for v7 without NEON).
UQSUB8 r10,r4, r12 ; r10= R_i>0?R_i:0
UQSUB8 r4, r12,r4 ; r4 = R_i<0?-R_i:0
UQADD8 r11,r10,r2 ; r11= 255-max(2*L-abs(R_i<0),0)
UQADD8 r14,r4, r2 ; r14= 255-max(2*L-abs(R_i>0),0)
UQADD8 r10,r10,r11
UQADD8 r4, r4, r14
UQSUB8 r10,r10,r11 ; r10= min(abs(R_i<0),max(2*L-abs(R_i<0),0))
UQSUB8 r4, r4, r14 ; r4 = min(abs(R_i>0),max(2*L-abs(R_i>0),0))
UQSUB8 r11,r5, r12 ; r11= R_i>0?R_i:0
UQADD8 r6, r6, r10
UQSUB8 r8, r8, r10
UQSUB8 r5, r12,r5 ; r5 = R_i<0?-R_i:0
UQSUB8 r6, r6, r4 ; r6 = p1+lflim(R_i,L)
UQADD8 r8, r8, r4 ; r8 = p2-lflim(R_i,L)
UQADD8 r10,r11,r2 ; r10= 255-max(2*L-abs(R_i<0),0)
UQADD8 r14,r5, r2 ; r14= 255-max(2*L-abs(R_i>0),0)
UQADD8 r11,r11,r10
UQADD8 r5, r5, r14
UQSUB8 r11,r11,r10 ; r11= min(abs(R_i<0),max(2*L-abs(R_i<0),0))
UQSUB8 r5, r5, r14 ; r5 = min(abs(R_i>0),max(2*L-abs(R_i>0),0))
UQADD8 r7, r7, r11
UQSUB8 r9, r9, r11
UQSUB8 r7, r7, r5 ; r7 = p5+lflim(R_i,L)
STRD r6, [r0, -r1] ; [p5:p1] = [r7: r6]
UQADD8 r9, r9, r5 ; r9 = p6-lflim(R_i,L)
STRD r8, [r0] ; [p6:p2] = [r9: r8]
LDMFD r13!,{r4-r11,PC}
ENDP
oc_loop_filter_frag_rows_v6 PROC
; r0 = _ref_frame_data
; r1 = _ystride
; r2 = _bv
; r3 = _frags
; r4 = _fragi0
; r5 = _fragi0_end
; r6 = _fragi_top
; r7 = _fragi_bot
; r8 = _frag_buf_offs
; r9 = _nhfrags
MOV r12,r13
STMFD r13!,{r0,r4-r11,r14}
LDMFD r12,{r4-r9}
LDR r2, [r2] ; ll = *(int *)_bv
CMP r4, r5 ; if(_fragi0>=_fragi0_end)
BGE oslffri_v6_end ; bail
SUBS r9, r9, #1 ; r9 = _nhfrags-1 if (r9<=0)
BLE oslffri_v6_end ; bail
ADD r3, r3, r4, LSL #2 ; r3 = &_frags[fragi]
ADD r8, r8, r4, LSL #2 ; r8 = &_frag_buf_offs[fragi]
SUB r7, r7, r9 ; _fragi_bot -= _nhfrags;
oslffri_v6_lp1
MOV r10,r4 ; r10= fragi = _fragi0
ADD r11,r4, r9 ; r11= fragi_end-1=fragi+_nhfrags-1
oslffri_v6_lp2
LDR r14,[r3], #4 ; r14= _frags[fragi] _frags++
LDR r0, [r13] ; r0 = _ref_frame_data
LDR r12,[r8], #4 ; r12= _frag_buf_offs[fragi] _frag_buf_offs++
TST r14,#OC_FRAG_CODED_FLAG
BEQ oslffri_v6_uncoded
CMP r10,r4 ; if (fragi>_fragi0)
ADD r0, r0, r12 ; r0 = _ref_frame_data + _frag_buf_offs[fragi]
BLGT loop_filter_h_v6
CMP r4, r6 ; if (fragi0>_fragi_top)
BLGT loop_filter_v_v6
CMP r10,r11 ; if(fragi+1<fragi_end)===(fragi<fragi_end-1)
LDRLT r12,[r3] ; r12 = _frags[fragi+1]
ADD r0, r0, #8
ADD r10,r10,#1 ; r10 = fragi+1;
ANDLT r12,r12,#OC_FRAG_CODED_FLAG
CMPLT r12,#OC_FRAG_CODED_FLAG ; && _frags[fragi+1].coded==0
BLLT loop_filter_h_v6
CMP r10,r7 ; if (fragi<_fragi_bot)
LDRLT r12,[r3, r9, LSL #2] ; r12 = _frags[fragi+1+_nhfrags-1]
SUB r0, r0, #8
ADD r0, r0, r1, LSL #3
ANDLT r12,r12,#OC_FRAG_CODED_FLAG
CMPLT r12,#OC_FRAG_CODED_FLAG
BLLT loop_filter_v_v6
CMP r10,r11 ; while(fragi<=fragi_end-1)
BLE oslffri_v6_lp2
MOV r4, r10 ; r4 = fragi0 += nhfrags
CMP r4, r5
BLT oslffri_v6_lp1
oslffri_v6_end
LDMFD r13!,{r0,r4-r11,PC}
oslffri_v6_uncoded
ADD r10,r10,#1
CMP r10,r11
BLE oslffri_v6_lp2
MOV r4, r10 ; r4 = fragi0 += nhfrags
CMP r4, r5
BLT oslffri_v6_lp1
LDMFD r13!,{r0,r4-r11,PC}
ENDP
]
[ OC_ARM_ASM_NEON
EXPORT oc_loop_filter_init_neon
EXPORT oc_loop_filter_frag_rows_neon
oc_loop_filter_init_neon PROC
; r0 = _bv
; r1 = _flimit (=L from the spec)
MOV r1, r1, LSL #1 ; r1 = 2*L
VDUP.S16 Q15, r1 ; Q15= 2L in U16s
VST1.64 {D30,D31}, [r0@128]
MOV PC,r14
ENDP
loop_filter_h_neon PROC
; r0 = unsigned char *_pix
; r1 = int _ystride
; r2 = int *_bv
; preserves r0-r3
; We assume Q15= 2*L in U16s
; My best guesses at cycle counts (and latency)--vvv
SUB r12,r0, #2
; Doing a 2-element structure load saves doing two VTRN's below, at the
; cost of using two more slower single-lane loads vs. the faster
; all-lane loads.
; It's less code this way, though, and benches a hair faster, but it
; leaves D2 and D4 swapped.
VLD2.16 {D0[],D2[]}, [r12], r1 ; D0 = ____________1100 2,1
; D2 = ____________3322
VLD2.16 {D4[],D6[]}, [r12], r1 ; D4 = ____________5544 2,1
; D6 = ____________7766
VLD2.16 {D0[1],D2[1]},[r12], r1 ; D0 = ________99881100 3,1
; D2 = ________BBAA3322
VLD2.16 {D4[1],D6[1]},[r12], r1 ; D4 = ________DDCC5544 3,1
; D6 = ________FFEE7766
VLD2.16 {D0[2],D2[2]},[r12], r1 ; D0 = ____GGHH99881100 3,1
; D2 = ____JJIIBBAA3322
VLD2.16 {D4[2],D6[2]},[r12], r1 ; D4 = ____KKLLDDCC5544 3,1
; D6 = ____NNMMFFEE7766
VLD2.16 {D0[3],D2[3]},[r12], r1 ; D0 = PPOOGGHH99881100 3,1
; D2 = RRQQJJIIBBAA3322
VLD2.16 {D4[3],D6[3]},[r12], r1 ; D4 = TTSSKKLLDDCC5544 3,1
; D6 = VVUUNNMMFFEE7766
VTRN.8 D0, D4 ; D0 = SSOOKKGGCC884400 D4 = TTPPLLHHDD995511 1,1
VTRN.8 D2, D6 ; D2 = UUQQMMIIEEAA6622 D6 = VVRRNNJJFFBB7733 1,1
VSUBL.U8 Q0, D0, D6 ; Q0 = 00 - 33 in S16s 1,3
VSUBL.U8 Q8, D2, D4 ; Q8 = 22 - 11 in S16s 1,3
ADD r12,r0, #8
VADD.S16 Q0, Q0, Q8 ; 1,3
PLD [r12]
VADD.S16 Q0, Q0, Q8 ; 1,3
PLD [r12,r1]
VADD.S16 Q0, Q0, Q8 ; Q0 = [0-3]+3*[2-1] 1,3
PLD [r12,r1, LSL #1]
VRSHR.S16 Q0, Q0, #3 ; Q0 = f = ([0-3]+3*[2-1]+4)>>3 1,4
ADD r12,r12,r1, LSL #2
; We want to do
; f = CLAMP(MIN(-2L-f,0), f, MAX(2L-f,0))
; = ((f >= 0) ? MIN( f ,MAX(2L- f ,0)) : MAX( f , MIN(-2L- f ,0)))
; = ((f >= 0) ? MIN(|f|,MAX(2L-|f|,0)) : MAX(-|f|, MIN(-2L+|f|,0)))
; = ((f >= 0) ? MIN(|f|,MAX(2L-|f|,0)) :-MIN( |f|,-MIN(-2L+|f|,0)))
; = ((f >= 0) ? MIN(|f|,MAX(2L-|f|,0)) :-MIN( |f|, MAX( 2L-|f|,0)))
; So we've reduced the left and right hand terms to be the same, except
; for a negation.
; Stall x3
VABS.S16 Q9, Q0 ; Q9 = |f| in U16s 1,4
PLD [r12,-r1]
VSHR.S16 Q0, Q0, #15 ; Q0 = -1 or 0 according to sign 1,3
PLD [r12]
VQSUB.U16 Q10,Q15,Q9 ; Q10= MAX(2L-|f|,0) in U16s 1,4
PLD [r12,r1]
VMOVL.U8 Q1, D2 ; Q2 = __UU__QQ__MM__II__EE__AA__66__22 2,3
PLD [r12,r1,LSL #1]
VMIN.U16 Q9, Q10,Q9 ; Q9 = MIN(|f|,MAX(2L-|f|)) 1,4
ADD r12,r12,r1, LSL #2
; Now we need to correct for the sign of f.
; For negative elements of Q0, we want to subtract the appropriate
; element of Q9. For positive elements we want to add them. No NEON
; instruction exists to do this, so we need to negate the negative
; elements, and we can then just add them. a-b = a-(1+!b) = a-1+!b
VADD.S16 Q9, Q9, Q0 ; 1,3
PLD [r12,-r1]
VEOR.S16 Q9, Q9, Q0 ; Q9 = real value of f 1,3
; Bah. No VRSBW.U8
; Stall (just 1 as Q9 not needed to second pipeline stage. I think.)
VADDW.U8 Q2, Q9, D4 ; Q1 = xxTTxxPPxxLLxxHHxxDDxx99xx55xx11 1,3
VSUB.S16 Q1, Q1, Q9 ; Q2 = xxUUxxQQxxMMxxIIxxEExxAAxx66xx22 1,3
VQMOVUN.S16 D4, Q2 ; D4 = TTPPLLHHDD995511 1,1
VQMOVUN.S16 D2, Q1 ; D2 = UUQQMMIIEEAA6622 1,1
SUB r12,r0, #1
VTRN.8 D4, D2 ; D4 = QQPPIIHHAA992211 D2 = MMLLEEDD6655 1,1
VST1.16 {D4[0]}, [r12], r1
VST1.16 {D2[0]}, [r12], r1
VST1.16 {D4[1]}, [r12], r1
VST1.16 {D2[1]}, [r12], r1
VST1.16 {D4[2]}, [r12], r1
VST1.16 {D2[2]}, [r12], r1
VST1.16 {D4[3]}, [r12], r1
VST1.16 {D2[3]}, [r12], r1
MOV PC,r14
ENDP
loop_filter_v_neon PROC
; r0 = unsigned char *_pix
; r1 = int _ystride
; r2 = int *_bv
; preserves r0-r3
; We assume Q15= 2*L in U16s
; My best guesses at cycle counts (and latency)--vvv
SUB r12,r0, r1, LSL #1
VLD1.64 {D0}, [r12@64], r1 ; D0 = SSOOKKGGCC884400 2,1
VLD1.64 {D2}, [r12@64], r1 ; D2 = TTPPLLHHDD995511 2,1
VLD1.64 {D4}, [r12@64], r1 ; D4 = UUQQMMIIEEAA6622 2,1
VLD1.64 {D6}, [r12@64] ; D6 = VVRRNNJJFFBB7733 2,1
VSUBL.U8 Q8, D4, D2 ; Q8 = 22 - 11 in S16s 1,3
VSUBL.U8 Q0, D0, D6 ; Q0 = 00 - 33 in S16s 1,3
ADD r12, #8
VADD.S16 Q0, Q0, Q8 ; 1,3
PLD [r12]
VADD.S16 Q0, Q0, Q8 ; 1,3
PLD [r12,r1]
VADD.S16 Q0, Q0, Q8 ; Q0 = [0-3]+3*[2-1] 1,3
SUB r12, r0, r1
VRSHR.S16 Q0, Q0, #3 ; Q0 = f = ([0-3]+3*[2-1]+4)>>3 1,4
; We want to do
; f = CLAMP(MIN(-2L-f,0), f, MAX(2L-f,0))
; = ((f >= 0) ? MIN( f ,MAX(2L- f ,0)) : MAX( f , MIN(-2L- f ,0)))
; = ((f >= 0) ? MIN(|f|,MAX(2L-|f|,0)) : MAX(-|f|, MIN(-2L+|f|,0)))
; = ((f >= 0) ? MIN(|f|,MAX(2L-|f|,0)) :-MIN( |f|,-MIN(-2L+|f|,0)))
; = ((f >= 0) ? MIN(|f|,MAX(2L-|f|,0)) :-MIN( |f|, MAX( 2L-|f|,0)))
; So we've reduced the left and right hand terms to be the same, except
; for a negation.
; Stall x3
VABS.S16 Q9, Q0 ; Q9 = |f| in U16s 1,4
VSHR.S16 Q0, Q0, #15 ; Q0 = -1 or 0 according to sign 1,3
; Stall x2
VQSUB.U16 Q10,Q15,Q9 ; Q10= MAX(2L-|f|,0) in U16s 1,4
VMOVL.U8 Q2, D4 ; Q2 = __UU__QQ__MM__II__EE__AA__66__22 2,3
; Stall x2
VMIN.U16 Q9, Q10,Q9 ; Q9 = MIN(|f|,MAX(2L-|f|)) 1,4
; Now we need to correct for the sign of f.
; For negative elements of Q0, we want to subtract the appropriate
; element of Q9. For positive elements we want to add them. No NEON
; instruction exists to do this, so we need to negate the negative
; elements, and we can then just add them. a-b = a-(1+!b) = a-1+!b
; Stall x3
VADD.S16 Q9, Q9, Q0 ; 1,3
; Stall x2
VEOR.S16 Q9, Q9, Q0 ; Q9 = real value of f 1,3
; Bah. No VRSBW.U8
; Stall (just 1 as Q9 not needed to second pipeline stage. I think.)
VADDW.U8 Q1, Q9, D2 ; Q1 = xxTTxxPPxxLLxxHHxxDDxx99xx55xx11 1,3
VSUB.S16 Q2, Q2, Q9 ; Q2 = xxUUxxQQxxMMxxIIxxEExxAAxx66xx22 1,3
VQMOVUN.S16 D2, Q1 ; D2 = TTPPLLHHDD995511 1,1
VQMOVUN.S16 D4, Q2 ; D4 = UUQQMMIIEEAA6622 1,1
VST1.64 {D2}, [r12@64], r1
VST1.64 {D4}, [r12@64], r1
MOV PC,r14
ENDP
oc_loop_filter_frag_rows_neon PROC
; r0 = _ref_frame_data
; r1 = _ystride
; r2 = _bv
; r3 = _frags
; r4 = _fragi0
; r5 = _fragi0_end
; r6 = _fragi_top
; r7 = _fragi_bot
; r8 = _frag_buf_offs
; r9 = _nhfrags
MOV r12,r13
STMFD r13!,{r0,r4-r11,r14}
LDMFD r12,{r4-r9}
CMP r4, r5 ; if(_fragi0>=_fragi0_end)
BGE oslffri_neon_end; bail
SUBS r9, r9, #1 ; r9 = _nhfrags-1 if (r9<=0)
BLE oslffri_neon_end ; bail
VLD1.64 {D30,D31}, [r2@128] ; Q15= 2L in U16s
ADD r3, r3, r4, LSL #2 ; r3 = &_frags[fragi]
ADD r8, r8, r4, LSL #2 ; r8 = &_frag_buf_offs[fragi]
SUB r7, r7, r9 ; _fragi_bot -= _nhfrags;
oslffri_neon_lp1
MOV r10,r4 ; r10= fragi = _fragi0
ADD r11,r4, r9 ; r11= fragi_end-1=fragi+_nhfrags-1
oslffri_neon_lp2
LDR r14,[r3], #4 ; r14= _frags[fragi] _frags++
LDR r0, [r13] ; r0 = _ref_frame_data
LDR r12,[r8], #4 ; r12= _frag_buf_offs[fragi] _frag_buf_offs++
TST r14,#OC_FRAG_CODED_FLAG
BEQ oslffri_neon_uncoded
CMP r10,r4 ; if (fragi>_fragi0)
ADD r0, r0, r12 ; r0 = _ref_frame_data + _frag_buf_offs[fragi]
BLGT loop_filter_h_neon
CMP r4, r6 ; if (_fragi0>_fragi_top)
BLGT loop_filter_v_neon
CMP r10,r11 ; if(fragi+1<fragi_end)===(fragi<fragi_end-1)
LDRLT r12,[r3] ; r12 = _frags[fragi+1]
ADD r0, r0, #8
ADD r10,r10,#1 ; r10 = fragi+1;
ANDLT r12,r12,#OC_FRAG_CODED_FLAG
CMPLT r12,#OC_FRAG_CODED_FLAG ; && _frags[fragi+1].coded==0
BLLT loop_filter_h_neon
CMP r10,r7 ; if (fragi<_fragi_bot)
LDRLT r12,[r3, r9, LSL #2] ; r12 = _frags[fragi+1+_nhfrags-1]
SUB r0, r0, #8
ADD r0, r0, r1, LSL #3
ANDLT r12,r12,#OC_FRAG_CODED_FLAG
CMPLT r12,#OC_FRAG_CODED_FLAG
BLLT loop_filter_v_neon
CMP r10,r11 ; while(fragi<=fragi_end-1)
BLE oslffri_neon_lp2
MOV r4, r10 ; r4 = _fragi0 += _nhfrags
CMP r4, r5
BLT oslffri_neon_lp1
oslffri_neon_end
LDMFD r13!,{r0,r4-r11,PC}
oslffri_neon_uncoded
ADD r10,r10,#1
CMP r10,r11
BLE oslffri_neon_lp2
MOV r4, r10 ; r4 = _fragi0 += _nhfrags
CMP r4, r5
BLT oslffri_neon_lp1
LDMFD r13!,{r0,r4-r11,PC}
ENDP
]
END

39
media/libtheora/lib/arm/armopts.s Normal file
View File

@ -0,0 +1,39 @@
;********************************************************************
;* *
;* THIS FILE IS PART OF THE OggTheora SOFTWARE CODEC SOURCE CODE. *
;* USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS *
;* GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
;* IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING. *
;* *
;* THE Theora SOURCE CODE IS COPYRIGHT (C) 2002-2010 *
;* by the Xiph.Org Foundation and contributors http://www.xiph.org/ *
;* *
;********************************************************************
; Original implementation:
; Copyright (C) 2009 Robin Watts for Pinknoise Productions Ltd
; last mod: $Id: armopts.s.in 17430 2010-09-22 21:54:09Z tterribe $
;********************************************************************
; Set the following to 1 if we have EDSP instructions
; (LDRD/STRD, etc., ARMv5E and later).
OC_ARM_ASM_EDSP * 1
; Set the following to 1 if we have ARMv6 media instructions.
OC_ARM_ASM_MEDIA * 1
; Set the following to 1 if we have NEON (some ARMv7)
OC_ARM_ASM_NEON * 1
; Set the following to 1 if LDR/STR can work on unaligned addresses
; This is assumed to be true for ARMv6 and later code
OC_ARM_CAN_UNALIGN * 0
; Large unaligned loads and stores are often configured to cause an exception.
; They cause an 8 cycle stall when they cross a 128-bit (load) or 64-bit (store)
; boundary, so it's usually a bad idea to use them anyway if they can be
; avoided.
; Set the following to 1 if LDRD/STRD can work on unaligned addresses
OC_ARM_CAN_UNALIGN_LDRD * 0
END

219
media/libtheora/lib/arm/armstate.c Normal file
View File

@ -0,0 +1,219 @@
/********************************************************************
* *
* THIS FILE IS PART OF THE OggTheora SOFTWARE CODEC SOURCE CODE. *
* USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS *
* GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
* IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING. *
* *
* THE Theora SOURCE CODE IS COPYRIGHT (C) 2002-2010 *
* by the Xiph.Org Foundation and contributors http://www.xiph.org/ *
* *
********************************************************************
function:
last mod: $Id: x86state.c 17344 2010-07-21 01:42:18Z tterribe $
********************************************************************/
#include "armint.h"
#if defined(OC_ARM_ASM)
# if defined(OC_ARM_ASM_NEON)
/*This table has been modified from OC_FZIG_ZAG by baking an 8x8 transpose into
the destination.*/
static const unsigned char OC_FZIG_ZAG_NEON[128]={
0, 8, 1, 2, 9,16,24,17,
10, 3, 4,11,18,25,32,40,
33,26,19,12, 5, 6,13,20,
27,34,41,48,56,49,42,35,
28,21,14, 7,15,22,29,36,
43,50,57,58,51,44,37,30,
23,31,38,45,52,59,60,53,
46,39,47,54,61,62,55,63,
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64
};
# endif
void oc_state_accel_init_arm(oc_theora_state *_state){
oc_state_accel_init_c(_state);
_state->cpu_flags=oc_cpu_flags_get();
# if defined(OC_STATE_USE_VTABLE)
_state->opt_vtable.frag_copy_list=oc_frag_copy_list_arm;
_state->opt_vtable.frag_recon_intra=oc_frag_recon_intra_arm;
_state->opt_vtable.frag_recon_inter=oc_frag_recon_inter_arm;
_state->opt_vtable.frag_recon_inter2=oc_frag_recon_inter2_arm;
_state->opt_vtable.idct8x8=oc_idct8x8_arm;
_state->opt_vtable.state_frag_recon=oc_state_frag_recon_arm;
/*Note: We _must_ set this function pointer, because the macro in armint.h
calls it with different arguments, so the C version will segfault.*/
_state->opt_vtable.state_loop_filter_frag_rows=
(oc_state_loop_filter_frag_rows_func)oc_loop_filter_frag_rows_arm;
# endif
# if defined(OC_ARM_ASM_EDSP)
if(_state->cpu_flags&OC_CPU_ARM_EDSP){
# if defined(OC_STATE_USE_VTABLE)
_state->opt_vtable.frag_copy_list=oc_frag_copy_list_edsp;
# endif
}
# if defined(OC_ARM_ASM_MEDIA)
if(_state->cpu_flags&OC_CPU_ARM_MEDIA){
# if defined(OC_STATE_USE_VTABLE)
_state->opt_vtable.frag_recon_intra=oc_frag_recon_intra_v6;
_state->opt_vtable.frag_recon_inter=oc_frag_recon_inter_v6;
_state->opt_vtable.frag_recon_inter2=oc_frag_recon_inter2_v6;
_state->opt_vtable.idct8x8=oc_idct8x8_v6;
_state->opt_vtable.state_frag_recon=oc_state_frag_recon_v6;
_state->opt_vtable.loop_filter_init=oc_loop_filter_init_v6;
_state->opt_vtable.state_loop_filter_frag_rows=
(oc_state_loop_filter_frag_rows_func)oc_loop_filter_frag_rows_v6;
# endif
}
# if defined(OC_ARM_ASM_NEON)
if(_state->cpu_flags&OC_CPU_ARM_NEON){
# if defined(OC_STATE_USE_VTABLE)
_state->opt_vtable.frag_copy_list=oc_frag_copy_list_neon;
_state->opt_vtable.frag_recon_intra=oc_frag_recon_intra_neon;
_state->opt_vtable.frag_recon_inter=oc_frag_recon_inter_neon;
_state->opt_vtable.frag_recon_inter2=oc_frag_recon_inter2_neon;
_state->opt_vtable.state_frag_recon=oc_state_frag_recon_neon;
_state->opt_vtable.loop_filter_init=oc_loop_filter_init_neon;
_state->opt_vtable.state_loop_filter_frag_rows=
(oc_state_loop_filter_frag_rows_func)oc_loop_filter_frag_rows_neon;
_state->opt_vtable.idct8x8=oc_idct8x8_neon;
# endif
_state->opt_data.dct_fzig_zag=OC_FZIG_ZAG_NEON;
}
# endif
# endif
# endif
}
void oc_state_frag_recon_arm(const oc_theora_state *_state,ptrdiff_t _fragi,
int _pli,ogg_int16_t _dct_coeffs[128],int _last_zzi,ogg_uint16_t _dc_quant){
unsigned char *dst;
ptrdiff_t frag_buf_off;
int ystride;
int refi;
/*Apply the inverse transform.*/
/*Special case only having a DC component.*/
if(_last_zzi<2){
ogg_uint16_t p;
/*We round this dequant product (and not any of the others) because there's
no iDCT rounding.*/
p=(ogg_uint16_t)(_dct_coeffs[0]*(ogg_int32_t)_dc_quant+15>>5);
oc_idct8x8_1_arm(_dct_coeffs+64,p);
}
else{
/*First, dequantize the DC coefficient.*/
_dct_coeffs[0]=(ogg_int16_t)(_dct_coeffs[0]*(int)_dc_quant);
oc_idct8x8_arm(_dct_coeffs+64,_dct_coeffs,_last_zzi);
}
/*Fill in the target buffer.*/
frag_buf_off=_state->frag_buf_offs[_fragi];
refi=_state->frags[_fragi].refi;
ystride=_state->ref_ystride[_pli];
dst=_state->ref_frame_data[OC_FRAME_SELF]+frag_buf_off;
if(refi==OC_FRAME_SELF)oc_frag_recon_intra_arm(dst,ystride,_dct_coeffs+64);
else{
const unsigned char *ref;
int mvoffsets[2];
ref=_state->ref_frame_data[refi]+frag_buf_off;
if(oc_state_get_mv_offsets(_state,mvoffsets,_pli,
_state->frag_mvs[_fragi])>1){
oc_frag_recon_inter2_arm(dst,ref+mvoffsets[0],ref+mvoffsets[1],ystride,
_dct_coeffs+64);
}
else oc_frag_recon_inter_arm(dst,ref+mvoffsets[0],ystride,_dct_coeffs+64);
}
}
# if defined(OC_ARM_ASM_MEDIA)
void oc_state_frag_recon_v6(const oc_theora_state *_state,ptrdiff_t _fragi,
int _pli,ogg_int16_t _dct_coeffs[128],int _last_zzi,ogg_uint16_t _dc_quant){
unsigned char *dst;
ptrdiff_t frag_buf_off;
int ystride;
int refi;
/*Apply the inverse transform.*/
/*Special case only having a DC component.*/
if(_last_zzi<2){
ogg_uint16_t p;
/*We round this dequant product (and not any of the others) because there's
no iDCT rounding.*/
p=(ogg_uint16_t)(_dct_coeffs[0]*(ogg_int32_t)_dc_quant+15>>5);
oc_idct8x8_1_v6(_dct_coeffs+64,p);
}
else{
/*First, dequantize the DC coefficient.*/
_dct_coeffs[0]=(ogg_int16_t)(_dct_coeffs[0]*(int)_dc_quant);
oc_idct8x8_v6(_dct_coeffs+64,_dct_coeffs,_last_zzi);
}
/*Fill in the target buffer.*/
frag_buf_off=_state->frag_buf_offs[_fragi];
refi=_state->frags[_fragi].refi;
ystride=_state->ref_ystride[_pli];
dst=_state->ref_frame_data[OC_FRAME_SELF]+frag_buf_off;
if(refi==OC_FRAME_SELF)oc_frag_recon_intra_v6(dst,ystride,_dct_coeffs+64);
else{
const unsigned char *ref;
int mvoffsets[2];
ref=_state->ref_frame_data[refi]+frag_buf_off;
if(oc_state_get_mv_offsets(_state,mvoffsets,_pli,
_state->frag_mvs[_fragi])>1){
oc_frag_recon_inter2_v6(dst,ref+mvoffsets[0],ref+mvoffsets[1],ystride,
_dct_coeffs+64);
}
else oc_frag_recon_inter_v6(dst,ref+mvoffsets[0],ystride,_dct_coeffs+64);
}
}
# if defined(OC_ARM_ASM_NEON)
void oc_state_frag_recon_neon(const oc_theora_state *_state,ptrdiff_t _fragi,
int _pli,ogg_int16_t _dct_coeffs[128],int _last_zzi,ogg_uint16_t _dc_quant){
unsigned char *dst;
ptrdiff_t frag_buf_off;
int ystride;
int refi;
/*Apply the inverse transform.*/
/*Special case only having a DC component.*/
if(_last_zzi<2){
ogg_uint16_t p;
/*We round this dequant product (and not any of the others) because there's
no iDCT rounding.*/
p=(ogg_uint16_t)(_dct_coeffs[0]*(ogg_int32_t)_dc_quant+15>>5);
oc_idct8x8_1_neon(_dct_coeffs+64,p);
}
else{
/*First, dequantize the DC coefficient.*/
_dct_coeffs[0]=(ogg_int16_t)(_dct_coeffs[0]*(int)_dc_quant);
oc_idct8x8_neon(_dct_coeffs+64,_dct_coeffs,_last_zzi);
}
/*Fill in the target buffer.*/
frag_buf_off=_state->frag_buf_offs[_fragi];
refi=_state->frags[_fragi].refi;
ystride=_state->ref_ystride[_pli];
dst=_state->ref_frame_data[OC_FRAME_SELF]+frag_buf_off;
if(refi==OC_FRAME_SELF)oc_frag_recon_intra_neon(dst,ystride,_dct_coeffs+64);
else{
const unsigned char *ref;
int mvoffsets[2];
ref=_state->ref_frame_data[refi]+frag_buf_off;
if(oc_state_get_mv_offsets(_state,mvoffsets,_pli,
_state->frag_mvs[_fragi])>1){
oc_frag_recon_inter2_neon(dst,ref+mvoffsets[0],ref+mvoffsets[1],ystride,
_dct_coeffs+64);
}
else oc_frag_recon_inter_neon(dst,ref+mvoffsets[0],ystride,_dct_coeffs+64);
}
}
# endif
# endif
#endif

23
media/libtheora/lib/bitpack.c
View File

@ -11,7 +11,7 @@
********************************************************************
function: packing variable sized words into an octet stream
last mod: $Id: bitpack.c 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: bitpack.c 17410 2010-09-21 21:53:48Z tterribe $
********************************************************************/
#include <string.h>
@ -32,15 +32,18 @@ static oc_pb_window oc_pack_refill(oc_pack_buf *_b,int _bits){
const unsigned char *stop;
oc_pb_window window;
int available;
unsigned shift;
stop=_b->stop;
ptr=_b->ptr;
window=_b->window;
available=_b->bits;
ptr=_b->ptr;
stop=_b->stop;
while(available<=OC_PB_WINDOW_SIZE-8&&ptr<stop){
available+=8;
window|=(oc_pb_window)*ptr++<<OC_PB_WINDOW_SIZE-available;
shift=OC_PB_WINDOW_SIZE-available;
while(7<shift&&ptr<stop){
shift-=8;
window|=(oc_pb_window)*ptr++<<shift;
}
_b->ptr=ptr;
available=OC_PB_WINDOW_SIZE-shift;
if(_bits>available){
if(ptr>=stop){
_b->eof=1;
@ -67,7 +70,7 @@ void oc_pack_adv1(oc_pack_buf *_b){
}
/*Here we assume that 0<=_bits&&_bits<=32.*/
long oc_pack_read(oc_pack_buf *_b,int _bits){
long oc_pack_read_c(oc_pack_buf *_b,int _bits){
oc_pb_window window;
int available;
long result;
@ -82,12 +85,12 @@ long oc_pack_read(oc_pack_buf *_b,int _bits){
available-=_bits;
window<<=1;
window<<=_bits-1;
_b->bits=available;
_b->window=window;
_b->bits=available;
return result;
}
int oc_pack_read1(oc_pack_buf *_b){
int oc_pack_read1_c(oc_pack_buf *_b){
oc_pb_window window;
int available;
int result;
@ -100,8 +103,8 @@ int oc_pack_read1(oc_pack_buf *_b){
result=window>>OC_PB_WINDOW_SIZE-1;
available--;
window<<=1;
_b->bits=available;
_b->window=window;
_b->bits=available;
return result;
}

27
media/libtheora/lib/bitpack.h
View File

@ -16,15 +16,32 @@
********************************************************************/
#if !defined(_bitpack_H)
# define _bitpack_H (1)
# include <stddef.h>
# include <limits.h>
# include "internal.h"
typedef unsigned long oc_pb_window;
typedef size_t oc_pb_window;
typedef struct oc_pack_buf oc_pack_buf;
/*Custom bitpacker implementations.*/
# if defined(OC_ARM_ASM)
# include "arm/armbits.h"
# endif
# if !defined(oc_pack_read)
# define oc_pack_read oc_pack_read_c
# endif
# if !defined(oc_pack_read1)
# define oc_pack_read1 oc_pack_read1_c
# endif
# if !defined(oc_huff_token_decode)
# define oc_huff_token_decode oc_huff_token_decode_c
# endif
# define OC_PB_WINDOW_SIZE ((int)sizeof(oc_pb_window)*CHAR_BIT)
/*This is meant to be a large, positive constant that can still be efficiently
loaded as an immediate (on platforms like ARM, for example).
@ -34,9 +51,9 @@ typedef struct oc_pack_buf oc_pack_buf;
struct oc_pack_buf{
oc_pb_window window;
const unsigned char *ptr;
const unsigned char *stop;
const unsigned char *ptr;
oc_pb_window window;
int bits;
int eof;
};
@ -45,8 +62,8 @@ void oc_pack_readinit(oc_pack_buf *_b,unsigned char *_buf,long _bytes);
int oc_pack_look1(oc_pack_buf *_b);
void oc_pack_adv1(oc_pack_buf *_b);
/*Here we assume 0<=_bits&&_bits<=32.*/
long oc_pack_read(oc_pack_buf *_b,int _bits);
int oc_pack_read1(oc_pack_buf *_b);
long oc_pack_read_c(oc_pack_buf *_b,int _bits);
int oc_pack_read1_c(oc_pack_buf *_b);
/* returns -1 for read beyond EOF, or the number of whole bytes available */
long oc_pack_bytes_left(oc_pack_buf *_b);

32
media/libtheora/lib/config.h
View File

@ -32,7 +32,7 @@
#define HAVE_STRING_H 1
/* Define to 1 if you have the <sys/soundcard.h> header file. */
/* #undef HAVE_SYS_SOUNDCARD_H */
#define HAVE_SYS_SOUNDCARD_H 1
/* Define to 1 if you have the <sys/stat.h> header file. */
#define HAVE_SYS_STAT_H 1
@ -43,18 +43,29 @@
/* Define to 1 if you have the <unistd.h> header file. */
#define HAVE_UNISTD_H 1
/* Define to the sub-directory in which libtool stores uninstalled libraries.
*/
#define LT_OBJDIR ".libs/"
/* Define to 1 if your C compiler doesn't accept -c and -o together. */
/* #undef NO_MINUS_C_MINUS_O */
/* make use of arm asm optimization */
/* Define if assembler supports EDSP instructions */
/* Define if assembler supports ARMv6 media instructions */
/* Define if compiler supports NEON instructions */
/* make use of c64x+ asm optimization */
/* #undef OC_C64X_ASM */
/* make use of x86_64 asm optimization */
/* #undef OC_X86_64_ASM */
/* make use of x86 asm optimization */
/**/
/* #undef OC_X86_ASM */
/* Name of package */
#define PACKAGE "libtheora"
@ -66,16 +77,13 @@
#define PACKAGE_NAME "libtheora"
/* Define to the full name and version of this package. */
#define PACKAGE_STRING "libtheora 1.1.1+svn"
#define PACKAGE_STRING "libtheora 1.2.0alpha1+svn"
/* Define to the one symbol short name of this package. */
#define PACKAGE_TARNAME "libtheora"
/* Define to the home page for this package. */
#define PACKAGE_URL ""
/* Define to the version of this package. */
#define PACKAGE_VERSION "1.1.1+svn"
#define PACKAGE_VERSION "1.2.0alpha1+svn"
/* Define to 1 if you have the ANSI C header files. */
#define STDC_HEADERS 1
@ -87,4 +95,4 @@
/* #undef THEORA_DISABLE_FLOAT */
/* Version number of package */
#define VERSION "1.1.1+svn"
#define VERSION "1.2.0alpha1+svn"

6
media/libtheora/lib/decinfo.c
View File

@ -11,7 +11,7 @@
********************************************************************
function:
last mod: $Id: decinfo.c 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: decinfo.c 17276 2010-06-05 05:57:05Z tterribe $
********************************************************************/
@ -128,6 +128,10 @@ static int oc_comment_unpack(oc_pack_buf *_opb,th_comment *_tc){
_tc->comments*sizeof(_tc->comment_lengths[0]));
_tc->user_comments=(char **)_ogg_malloc(
_tc->comments*sizeof(_tc->user_comments[0]));
if(_tc->comment_lengths==NULL||_tc->user_comments==NULL){
_tc->comments=0;
return TH_EFAULT;
}
for(i=0;i<_tc->comments;i++){
len=oc_unpack_length(_opb);
if(len<0||len>oc_pack_bytes_left(_opb)){

153
media/libtheora/lib/decint.h
View File

@ -11,7 +11,7 @@
********************************************************************
function:
last mod: $Id: decint.h 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: decint.h 17457 2010-09-24 02:05:49Z tterribe $
********************************************************************/
@ -19,15 +19,39 @@
#if !defined(_decint_H)
# define _decint_H (1)
# include "theora/theoradec.h"
# include "internal.h"
# include "state.h"
# include "bitpack.h"
typedef struct th_setup_info oc_setup_info;
typedef struct th_dec_ctx oc_dec_ctx;
# include "huffdec.h"
# include "dequant.h"
typedef struct th_setup_info oc_setup_info;
typedef struct oc_dec_opt_vtable oc_dec_opt_vtable;
typedef struct oc_dec_pipeline_state oc_dec_pipeline_state;
typedef struct th_dec_ctx oc_dec_ctx;
/*Decoder-specific accelerated functions.*/
# if defined(OC_C64X_ASM)
# include "c64x/c64xdec.h"
# endif
# if !defined(oc_dec_accel_init)
# define oc_dec_accel_init oc_dec_accel_init_c
# endif
# if defined(OC_DEC_USE_VTABLE)
# if !defined(oc_dec_dc_unpredict_mcu_plane)
# define oc_dec_dc_unpredict_mcu_plane(_dec,_pipe,_pli) \
((*(_dec)->opt_vtable.dc_unpredict_mcu_plane)(_dec,_pipe,_pli))
# endif
# else
# if !defined(oc_dec_dc_unpredict_mcu_plane)
# define oc_dec_dc_unpredict_mcu_plane oc_dec_dc_unpredict_mcu_plane_c
# endif
# endif
/*Constants for the packet-in state machine specific to the decoder.*/
/*Next packet to read: Data packet.*/
@ -37,71 +61,126 @@ typedef struct th_dec_ctx oc_dec_ctx;
struct th_setup_info{
/*The Huffman codes.*/
oc_huff_node *huff_tables[TH_NHUFFMAN_TABLES];
ogg_int16_t *huff_tables[TH_NHUFFMAN_TABLES];
/*The quantization parameters.*/
th_quant_info qinfo;
};
/*Decoder specific functions with accelerated variants.*/
struct oc_dec_opt_vtable{
void (*dc_unpredict_mcu_plane)(oc_dec_ctx *_dec,
oc_dec_pipeline_state *_pipe,int _pli);
};
struct oc_dec_pipeline_state{
/*Decoded DCT coefficients.
These are placed here instead of on the stack so that they can persist
between blocks, which makes clearing them back to zero much faster when
only a few non-zero coefficients were decoded.
It requires at least 65 elements because the zig-zag index array uses the
65th element as a dumping ground for out-of-range indices to protect us
from buffer overflow.
We make it fully twice as large so that the second half can serve as the
reconstruction buffer, which saves passing another parameter to all the
acceleration functios.
It also solves problems with 16-byte alignment for NEON on ARM.
gcc (as of 4.2.1) only seems to be able to give stack variables 8-byte
alignment, and silently produces incorrect results if you ask for 16.
Finally, keeping it off the stack means there's less likely to be a data
hazard beween the NEON co-processor and the regular ARM core, which avoids
unnecessary stalls.*/
OC_ALIGN16(ogg_int16_t dct_coeffs[128]);
OC_ALIGN16(signed char bounding_values[256]);
ptrdiff_t ti[3][64];
ptrdiff_t ebi[3][64];
ptrdiff_t eob_runs[3][64];
const ptrdiff_t *coded_fragis[3];
const ptrdiff_t *uncoded_fragis[3];
ptrdiff_t ncoded_fragis[3];
ptrdiff_t nuncoded_fragis[3];
const ogg_uint16_t *dequant[3][3][2];
int fragy0[3];
int fragy_end[3];
int pred_last[3][4];
int mcu_nvfrags;
int loop_filter;
int pp_level;
};
struct th_dec_ctx{
/*Shared encoder/decoder state.*/
oc_theora_state state;
oc_theora_state state;
/*Whether or not packets are ready to be emitted.
This takes on negative values while there are remaining header packets to
be emitted, reaches 0 when the codec is ready for input, and goes to 1
when a frame has been processed and a data packet is ready.*/
int packet_state;
int packet_state;
/*Buffer in which to assemble packets.*/
oc_pack_buf opb;
oc_pack_buf opb;
/*Huffman decode trees.*/
oc_huff_node *huff_tables[TH_NHUFFMAN_TABLES];
ogg_int16_t *huff_tables[TH_NHUFFMAN_TABLES];
/*The index of the first token in each plane for each coefficient.*/
ptrdiff_t ti0[3][64];
ptrdiff_t ti0[3][64];
/*The number of outstanding EOB runs at the start of each coefficient in each
plane.*/
ptrdiff_t eob_runs[3][64];
ptrdiff_t eob_runs[3][64];
/*The DCT token lists.*/
unsigned char *dct_tokens;
unsigned char *dct_tokens;
/*The extra bits associated with DCT tokens.*/
unsigned char *extra_bits;
unsigned char *extra_bits;
/*The number of dct tokens unpacked so far.*/
int dct_tokens_count;
int dct_tokens_count;
/*The out-of-loop post-processing level.*/
int pp_level;
int pp_level;
/*The DC scale used for out-of-loop deblocking.*/
int pp_dc_scale[64];
int pp_dc_scale[64];
/*The sharpen modifier used for out-of-loop deringing.*/
int pp_sharp_mod[64];
int pp_sharp_mod[64];
/*The DC quantization index of each block.*/
unsigned char *dc_qis;
unsigned char *dc_qis;
/*The variance of each block.*/
int *variances;
int *variances;
/*The storage for the post-processed frame buffer.*/
unsigned char *pp_frame_data;
unsigned char *pp_frame_data;
/*Whether or not the post-processsed frame buffer has space for chroma.*/
int pp_frame_state;
int pp_frame_state;
/*The buffer used for the post-processed frame.
Note that this is _not_ guaranteed to have the same strides and offsets as
the reference frame buffers.*/
th_ycbcr_buffer pp_frame_buf;
th_ycbcr_buffer pp_frame_buf;
/*The striped decode callback function.*/
th_stripe_callback stripe_cb;
th_stripe_callback stripe_cb;
oc_dec_pipeline_state pipe;
# if defined(OC_DEC_USE_VTABLE)
/*Table for decoder acceleration functions.*/
oc_dec_opt_vtable opt_vtable;
# endif
# if defined(HAVE_CAIRO)
/*Output metrics for debugging.*/
int telemetry;
int telemetry_mbmode;
int telemetry_mv;
int telemetry_qi;
int telemetry_bits;
int telemetry_frame_bytes;
int telemetry_coding_bytes;
int telemetry_mode_bytes;
int telemetry_mv_bytes;
int telemetry_qi_bytes;
int telemetry_dc_bytes;
unsigned char *telemetry_frame_data;
int telemetry;
int telemetry_mbmode;
int telemetry_mv;
int telemetry_qi;
int telemetry_bits;
int telemetry_frame_bytes;
int telemetry_coding_bytes;
int telemetry_mode_bytes;
int telemetry_mv_bytes;
int telemetry_qi_bytes;
int telemetry_dc_bytes;
unsigned char *telemetry_frame_data;
# endif
};
/*Default pure-C implementations of decoder-specific accelerated functions.*/
void oc_dec_accel_init_c(oc_dec_ctx *_dec);
void oc_dec_dc_unpredict_mcu_plane_c(oc_dec_ctx *_dec,
oc_dec_pipeline_state *_pipe,int _pli);
#endif

727
media/libtheora/lib/decode.c
View File

File diff suppressed because it is too large Load Diff

493
media/libtheora/lib/encint.h
View File

@ -1,493 +0,0 @@
/********************************************************************
* *
* THIS FILE IS PART OF THE OggTheora SOFTWARE CODEC SOURCE CODE. *
* USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS *
* GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
* IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING. *
* *
* THE Theora SOURCE CODE IS COPYRIGHT (C) 2002-2009 *
* by the Xiph.Org Foundation http://www.xiph.org/ *
* *
********************************************************************
function:
last mod: $Id: encint.h 16503 2009-08-22 18:14:02Z giles $
********************************************************************/
#if !defined(_encint_H)
# define _encint_H (1)
# if defined(HAVE_CONFIG_H)
# include "config.h"
# endif
# include "theora/theoraenc.h"
# include "internal.h"
# include "ocintrin.h"
# include "mathops.h"
# include "enquant.h"
# include "huffenc.h"
/*# define OC_COLLECT_METRICS*/
typedef oc_mv oc_mv2[2];
typedef struct oc_enc_opt_vtable oc_enc_opt_vtable;
typedef struct oc_mb_enc_info oc_mb_enc_info;
typedef struct oc_mode_scheme_chooser oc_mode_scheme_chooser;
typedef struct oc_iir_filter oc_iir_filter;
typedef struct oc_frame_metrics oc_frame_metrics;
typedef struct oc_rc_state oc_rc_state;
typedef struct th_enc_ctx oc_enc_ctx;
typedef struct oc_token_checkpoint oc_token_checkpoint;
/*Constants for the packet-out state machine specific to the encoder.*/
/*Next packet to emit: Data packet, but none are ready yet.*/
#define OC_PACKET_EMPTY (0)
/*Next packet to emit: Data packet, and one is ready.*/
#define OC_PACKET_READY (1)
/*All features enabled.*/
#define OC_SP_LEVEL_SLOW (0)
/*Enable early skip.*/
#define OC_SP_LEVEL_EARLY_SKIP (1)
/*Disable motion compensation.*/
#define OC_SP_LEVEL_NOMC (2)
/*Maximum valid speed level.*/
#define OC_SP_LEVEL_MAX (2)
/*The bits used for each of the MB mode codebooks.*/
extern const unsigned char OC_MODE_BITS[2][OC_NMODES];
/*The bits used for each of the MV codebooks.*/
extern const unsigned char OC_MV_BITS[2][64];
/*The minimum value that can be stored in a SB run for each codeword.
The last entry is the upper bound on the length of a single SB run.*/
extern const ogg_uint16_t OC_SB_RUN_VAL_MIN[8];
/*The bits used for each SB run codeword.*/
extern const unsigned char OC_SB_RUN_CODE_NBITS[7];
/*The bits used for each block run length (starting with 1).*/
extern const unsigned char OC_BLOCK_RUN_CODE_NBITS[30];
/*Encoder specific functions with accelerated variants.*/
struct oc_enc_opt_vtable{
unsigned (*frag_sad)(const unsigned char *_src,
const unsigned char *_ref,int _ystride);
unsigned (*frag_sad_thresh)(const unsigned char *_src,
const unsigned char *_ref,int _ystride,unsigned _thresh);
unsigned (*frag_sad2_thresh)(const unsigned char *_src,
const unsigned char *_ref1,const unsigned char *_ref2,int _ystride,
unsigned _thresh);
unsigned (*frag_satd_thresh)(const unsigned char *_src,
const unsigned char *_ref,int _ystride,unsigned _thresh);
unsigned (*frag_satd2_thresh)(const unsigned char *_src,
const unsigned char *_ref1,const unsigned char *_ref2,int _ystride,
unsigned _thresh);
unsigned (*frag_intra_satd)(const unsigned char *_src,int _ystride);
void (*frag_sub)(ogg_int16_t _diff[64],const unsigned char *_src,
const unsigned char *_ref,int _ystride);
void (*frag_sub_128)(ogg_int16_t _diff[64],
const unsigned char *_src,int _ystride);
void (*frag_copy2)(unsigned char *_dst,
const unsigned char *_src1,const unsigned char *_src2,int _ystride);
void (*frag_recon_intra)(unsigned char *_dst,int _ystride,
const ogg_int16_t _residue[64]);
void (*frag_recon_inter)(unsigned char *_dst,
const unsigned char *_src,int _ystride,const ogg_int16_t _residue[64]);
void (*fdct8x8)(ogg_int16_t _y[64],const ogg_int16_t _x[64]);
};
void oc_enc_vtable_init(oc_enc_ctx *_enc);
/*Encoder-specific macroblock information.*/
struct oc_mb_enc_info{
/*Neighboring macro blocks that have MVs available from the current frame.*/
unsigned cneighbors[4];
/*Neighboring macro blocks to use for MVs from the previous frame.*/
unsigned pneighbors[4];
/*The number of current-frame neighbors.*/
unsigned char ncneighbors;
/*The number of previous-frame neighbors.*/
unsigned char npneighbors;
/*Flags indicating which MB modes have been refined.*/
unsigned char refined;
/*Motion vectors for a macro block for the current frame and the
previous two frames.
Each is a set of 2 vectors against OC_FRAME_GOLD and OC_FRAME_PREV, which
can be used to estimate constant velocity and constant acceleration
predictors.
Uninitialized MVs are (0,0).*/
oc_mv2 analysis_mv[3];
/*Current unrefined analysis MVs.*/
oc_mv unref_mv[2];
/*Unrefined block MVs.*/
oc_mv block_mv[4];
/*Refined block MVs.*/
oc_mv ref_mv[4];
/*Minimum motion estimation error from the analysis stage.*/
ogg_uint16_t error[2];
/*MB error for half-pel refinement for each frame type.*/
unsigned satd[2];
/*Block error for half-pel refinement.*/
unsigned block_satd[4];
};
/*State machine to estimate the opportunity cost of coding a MB mode.*/
struct oc_mode_scheme_chooser{
/*Pointers to the a list containing the index of each mode in the mode
alphabet used by each scheme.
The first entry points to the dynamic scheme0_ranks, while the remaining 7
point to the constant entries stored in OC_MODE_SCHEMES.*/
const unsigned char *mode_ranks[8];
/*The ranks for each mode when coded with scheme 0.
These are optimized so that the more frequent modes have lower ranks.*/
unsigned char scheme0_ranks[OC_NMODES];
/*The list of modes, sorted in descending order of frequency, that
corresponds to the ranks above.*/
unsigned char scheme0_list[OC_NMODES];
/*The number of times each mode has been chosen so far.*/
int mode_counts[OC_NMODES];
/*The list of mode coding schemes, sorted in ascending order of bit cost.*/
unsigned char scheme_list[8];
/*The number of bits used by each mode coding scheme.*/
ptrdiff_t scheme_bits[8];
};
void oc_mode_scheme_chooser_init(oc_mode_scheme_chooser *_chooser);
/*A 2nd order low-pass Bessel follower.
We use this for rate control because it has fast reaction time, but is
critically damped.*/
struct oc_iir_filter{
ogg_int32_t c[2];
ogg_int64_t g;
ogg_int32_t x[2];
ogg_int32_t y[2];
};
/*The 2-pass metrics associated with a single frame.*/
struct oc_frame_metrics{
/*The log base 2 of the scale factor for this frame in Q24 format.*/
ogg_int32_t log_scale;
/*The number of application-requested duplicates of this frame.*/
unsigned dup_count:31;
/*The frame type from pass 1.*/
unsigned frame_type:1;
};
/*Rate control state information.*/
struct oc_rc_state{
/*The target average bits per frame.*/
ogg_int64_t bits_per_frame;
/*The current buffer fullness (bits available to be used).*/
ogg_int64_t fullness;
/*The target buffer fullness.
This is where we'd like to be by the last keyframe the appears in the next
buf_delay frames.*/
ogg_int64_t target;
/*The maximum buffer fullness (total size of the buffer).*/
ogg_int64_t max;
/*The log of the number of pixels in a frame in Q57 format.*/
ogg_int64_t log_npixels;
/*The exponent used in the rate model in Q8 format.*/
unsigned exp[2];
/*The number of frames to distribute the buffer usage over.*/
int buf_delay;
/*The total drop count from the previous frame.
This includes duplicates explicitly requested via the
TH_ENCCTL_SET_DUP_COUNT API as well as frames we chose to drop ourselves.*/
ogg_uint32_t prev_drop_count;
/*The log of an estimated scale factor used to obtain the real framerate, for
VFR sources or, e.g., 12 fps content doubled to 24 fps, etc.*/
ogg_int64_t log_drop_scale;
/*The log of estimated scale factor for the rate model in Q57 format.*/
ogg_int64_t log_scale[2];
/*The log of the target quantizer level in Q57 format.*/
ogg_int64_t log_qtarget;
/*Will we drop frames to meet bitrate target?*/
unsigned char drop_frames;
/*Do we respect the maximum buffer fullness?*/
unsigned char cap_overflow;
/*Can the reservoir go negative?*/
unsigned char cap_underflow;
/*Second-order lowpass filters to track scale and VFR.*/
oc_iir_filter scalefilter[2];
int inter_count;
int inter_delay;
int inter_delay_target;
oc_iir_filter vfrfilter;
/*Two-pass mode state.
0 => 1-pass encoding.
1 => 1st pass of 2-pass encoding.
2 => 2nd pass of 2-pass encoding.*/
int twopass;
/*Buffer for current frame metrics.*/
unsigned char twopass_buffer[48];
/*The number of bytes in the frame metrics buffer.
When 2-pass encoding is enabled, this is set to 0 after each frame is
submitted, and must be non-zero before the next frame will be accepted.*/
int twopass_buffer_bytes;
int twopass_buffer_fill;
/*Whether or not to force the next frame to be a keyframe.*/
unsigned char twopass_force_kf;
/*The metrics for the previous frame.*/
oc_frame_metrics prev_metrics;
/*The metrics for the current frame.*/
oc_frame_metrics cur_metrics;
/*The buffered metrics for future frames.*/
oc_frame_metrics *frame_metrics;
int nframe_metrics;
int cframe_metrics;
/*The index of the current frame in the circular metric buffer.*/
int frame_metrics_head;
/*The frame count of each type (keyframes, delta frames, and dup frames);
32 bits limits us to 2.268 years at 60 fps.*/
ogg_uint32_t frames_total[3];
/*The number of frames of each type yet to be processed.*/
ogg_uint32_t frames_left[3];
/*The sum of the scale values for each frame type.*/
ogg_int64_t scale_sum[2];
/*The start of the window over which the current scale sums are taken.*/
int scale_window0;
/*The end of the window over which the current scale sums are taken.*/
int scale_window_end;
/*The frame count of each type in the current 2-pass window; this does not
include dup frames.*/
int nframes[3];
/*The total accumulated estimation bias.*/
ogg_int64_t rate_bias;
};
void oc_rc_state_init(oc_rc_state *_rc,oc_enc_ctx *_enc);
void oc_rc_state_clear(oc_rc_state *_rc);
void oc_enc_rc_resize(oc_enc_ctx *_enc);
int oc_enc_select_qi(oc_enc_ctx *_enc,int _qti,int _clamp);
void oc_enc_calc_lambda(oc_enc_ctx *_enc,int _frame_type);
int oc_enc_update_rc_state(oc_enc_ctx *_enc,
long _bits,int _qti,int _qi,int _trial,int _droppable);
int oc_enc_rc_2pass_out(oc_enc_ctx *_enc,unsigned char **_buf);
int oc_enc_rc_2pass_in(oc_enc_ctx *_enc,unsigned char *_buf,size_t _bytes);
/*The internal encoder state.*/
struct th_enc_ctx{
/*Shared encoder/decoder state.*/
oc_theora_state state;
/*Buffer in which to assemble packets.*/
oggpack_buffer opb;
/*Encoder-specific macroblock information.*/
oc_mb_enc_info *mb_info;
/*DC coefficients after prediction.*/
ogg_int16_t *frag_dc;
/*The list of coded macro blocks, in coded order.*/
unsigned *coded_mbis;
/*The number of coded macro blocks.*/
size_t ncoded_mbis;
/*Whether or not packets are ready to be emitted.
This takes on negative values while there are remaining header packets to
be emitted, reaches 0 when the codec is ready for input, and becomes
positive when a frame has been processed and data packets are ready.*/
int packet_state;
/*The maximum distance between keyframes.*/
ogg_uint32_t keyframe_frequency_force;
/*The number of duplicates to produce for the next frame.*/
ogg_uint32_t dup_count;
/*The number of duplicates remaining to be emitted for the current frame.*/
ogg_uint32_t nqueued_dups;
/*The number of duplicates emitted for the last frame.*/
ogg_uint32_t prev_dup_count;
/*The current speed level.*/
int sp_level;
/*Whether or not VP3 compatibility mode has been enabled.*/
unsigned char vp3_compatible;
/*Whether or not any INTER frames have been coded.*/
unsigned char coded_inter_frame;
/*Whether or not previous frame was dropped.*/
unsigned char prevframe_dropped;
/*Stores most recently chosen Huffman tables for each frame type, DC and AC
coefficients, and luma and chroma tokens.
The actual Huffman table used for a given coefficient depends not only on
the choice made here, but also its index in the zig-zag ordering.*/
unsigned char huff_idxs[2][2][2];
/*Current count of bits used by each MV coding mode.*/
size_t mv_bits[2];
/*The mode scheme chooser for estimating mode coding costs.*/
oc_mode_scheme_chooser chooser;
/*The number of vertical super blocks in an MCU.*/
int mcu_nvsbs;
/*The SSD error for skipping each fragment in the current MCU.*/
unsigned *mcu_skip_ssd;
/*The DCT token lists for each coefficient and each plane.*/
unsigned char **dct_tokens[3];
/*The extra bits associated with each DCT token.*/
ogg_uint16_t **extra_bits[3];
/*The number of DCT tokens for each coefficient for each plane.*/
ptrdiff_t ndct_tokens[3][64];
/*Pending EOB runs for each coefficient for each plane.*/
ogg_uint16_t eob_run[3][64];
/*The offset of the first DCT token for each coefficient for each plane.*/
unsigned char dct_token_offs[3][64];
/*The last DC coefficient for each plane and reference frame.*/
int dc_pred_last[3][3];
#if defined(OC_COLLECT_METRICS)
/*Fragment SATD statistics for MB mode estimation metrics.*/
unsigned *frag_satd;
/*Fragment SSD statistics for MB mode estimation metrics.*/
unsigned *frag_ssd;
#endif
/*The R-D optimization parameter.*/
int lambda;
/*The huffman tables in use.*/
th_huff_code huff_codes[TH_NHUFFMAN_TABLES][TH_NDCT_TOKENS];
/*The quantization parameters in use.*/
th_quant_info qinfo;
oc_iquant *enquant_tables[64][3][2];
oc_iquant_table enquant_table_data[64][3][2];
/*An "average" quantizer for each quantizer type (INTRA or INTER) and qi
value.
This is used to paramterize the rate control decisions.
They are kept in the log domain to simplify later processing.
Keep in mind these are DCT domain quantizers, and so are scaled by an
additional factor of 4 from the pixel domain.*/
ogg_int64_t log_qavg[2][64];
/*The buffer state used to drive rate control.*/
oc_rc_state rc;
/*Table for encoder acceleration functions.*/
oc_enc_opt_vtable opt_vtable;
};
void oc_enc_analyze_intra(oc_enc_ctx *_enc,int _recode);
int oc_enc_analyze_inter(oc_enc_ctx *_enc,int _allow_keyframe,int _recode);
#if defined(OC_COLLECT_METRICS)
void oc_enc_mode_metrics_collect(oc_enc_ctx *_enc);
void oc_enc_mode_metrics_dump(oc_enc_ctx *_enc);
#endif
/*Perform fullpel motion search for a single MB against both reference frames.*/
void oc_mcenc_search(oc_enc_ctx *_enc,int _mbi);
/*Refine a MB MV for one frame.*/
void oc_mcenc_refine1mv(oc_enc_ctx *_enc,int _mbi,int _frame);
/*Refine the block MVs.*/
void oc_mcenc_refine4mv(oc_enc_ctx *_enc,int _mbi);
/*Used to rollback a tokenlog transaction when we retroactively decide to skip
a fragment.
A checkpoint is taken right before each token is added.*/
struct oc_token_checkpoint{
/*The color plane the token was added to.*/
unsigned char pli;
/*The zig-zag index the token was added to.*/
unsigned char zzi;
/*The outstanding EOB run count before the token was added.*/
ogg_uint16_t eob_run;
/*The token count before the token was added.*/
ptrdiff_t ndct_tokens;
};
void oc_enc_tokenize_start(oc_enc_ctx *_enc);
int oc_enc_tokenize_ac(oc_enc_ctx *_enc,int _pli,ptrdiff_t _fragi,
ogg_int16_t *_qdct,const ogg_uint16_t *_dequant,const ogg_int16_t *_dct,
int _zzi,oc_token_checkpoint **_stack,int _acmin);
void oc_enc_tokenlog_rollback(oc_enc_ctx *_enc,
const oc_token_checkpoint *_stack,int _n);
void oc_enc_pred_dc_frag_rows(oc_enc_ctx *_enc,
int _pli,int _fragy0,int _frag_yend);
void oc_enc_tokenize_dc_frag_list(oc_enc_ctx *_enc,int _pli,
const ptrdiff_t *_coded_fragis,ptrdiff_t _ncoded_fragis,
int _prev_ndct_tokens1,int _prev_eob_run1);
void oc_enc_tokenize_finish(oc_enc_ctx *_enc);
/*Utility routine to encode one of the header packets.*/
int oc_state_flushheader(oc_theora_state *_state,int *_packet_state,
oggpack_buffer *_opb,const th_quant_info *_qinfo,
const th_huff_code _codes[TH_NHUFFMAN_TABLES][TH_NDCT_TOKENS],
const char *_vendor,th_comment *_tc,ogg_packet *_op);
/*Encoder-specific accelerated functions.*/
void oc_enc_frag_sub(const oc_enc_ctx *_enc,ogg_int16_t _diff[64],
const unsigned char *_src,const unsigned char *_ref,int _ystride);
void oc_enc_frag_sub_128(const oc_enc_ctx *_enc,ogg_int16_t _diff[64],
const unsigned char *_src,int _ystride);
unsigned oc_enc_frag_sad(const oc_enc_ctx *_enc,const unsigned char *_src,
const unsigned char *_ref,int _ystride);
unsigned oc_enc_frag_sad_thresh(const oc_enc_ctx *_enc,
const unsigned char *_src,const unsigned char *_ref,int _ystride,
unsigned _thresh);
unsigned oc_enc_frag_sad2_thresh(const oc_enc_ctx *_enc,
const unsigned char *_src,const unsigned char *_ref1,
const unsigned char *_ref2,int _ystride,unsigned _thresh);
unsigned oc_enc_frag_satd_thresh(const oc_enc_ctx *_enc,
const unsigned char *_src,const unsigned char *_ref,int _ystride,
unsigned _thresh);
unsigned oc_enc_frag_satd2_thresh(const oc_enc_ctx *_enc,
const unsigned char *_src,const unsigned char *_ref1,
const unsigned char *_ref2,int _ystride,unsigned _thresh);
unsigned oc_enc_frag_intra_satd(const oc_enc_ctx *_enc,
const unsigned char *_src,int _ystride);
void oc_enc_frag_copy2(const oc_enc_ctx *_enc,unsigned char *_dst,
const unsigned char *_src1,const unsigned char *_src2,int _ystride);
void oc_enc_frag_recon_intra(const oc_enc_ctx *_enc,
unsigned char *_dst,int _ystride,const ogg_int16_t _residue[64]);
void oc_enc_frag_recon_inter(const oc_enc_ctx *_enc,unsigned char *_dst,
const unsigned char *_src,int _ystride,const ogg_int16_t _residue[64]);
void oc_enc_fdct8x8(const oc_enc_ctx *_enc,ogg_int16_t _y[64],
const ogg_int16_t _x[64]);
/*Default pure-C implementations.*/
void oc_enc_vtable_init_c(oc_enc_ctx *_enc);
void oc_enc_frag_sub_c(ogg_int16_t _diff[64],
const unsigned char *_src,const unsigned char *_ref,int _ystride);
void oc_enc_frag_sub_128_c(ogg_int16_t _diff[64],
const unsigned char *_src,int _ystride);
void oc_enc_frag_copy2_c(unsigned char *_dst,
const unsigned char *_src1,const unsigned char *_src2,int _ystride);
unsigned oc_enc_frag_sad_c(const unsigned char *_src,
const unsigned char *_ref,int _ystride);
unsigned oc_enc_frag_sad_thresh_c(const unsigned char *_src,
const unsigned char *_ref,int _ystride,unsigned _thresh);
unsigned oc_enc_frag_sad2_thresh_c(const unsigned char *_src,
const unsigned char *_ref1,const unsigned char *_ref2,int _ystride,
unsigned _thresh);
unsigned oc_enc_frag_satd_thresh_c(const unsigned char *_src,
const unsigned char *_ref,int _ystride,unsigned _thresh);
unsigned oc_enc_frag_satd2_thresh_c(const unsigned char *_src,
const unsigned char *_ref1,const unsigned char *_ref2,int _ystride,
unsigned _thresh);
unsigned oc_enc_frag_intra_satd_c(const unsigned char *_src,int _ystride);
void oc_enc_fdct8x8_c(ogg_int16_t _y[64],const ogg_int16_t _x[64]);
#endif

67
media/libtheora/lib/encoder_disabled.c
View File

@ -1,67 +0,0 @@
/********************************************************************
* *
* THIS FILE IS PART OF THE OggTheora SOFTWARE CODEC SOURCE CODE. *
* USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS *
* GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
* IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING. *
* *
* THE Theora SOURCE CODE IS COPYRIGHT (C) 2002-2009 *
* by the Xiph.Org Foundation http://www.xiph.org/ *
* *
********************************************************************
function:
last mod: $Id: encoder_disabled.c 16503 2009-08-22 18:14:02Z giles $
********************************************************************/
#include "apiwrapper.h"
#include "encint.h"
th_enc_ctx *th_encode_alloc(const th_info *_info){
return NULL;
}
void th_encode_free(th_enc_ctx *_enc){}
int th_encode_ctl(th_enc_ctx *_enc,int _req,void *_buf,size_t _buf_sz){
return OC_DISABLED;
}
int th_encode_flushheader(th_enc_ctx *_enc,th_comment *_tc,ogg_packet *_op){
return OC_DISABLED;
}
int th_encode_ycbcr_in(th_enc_ctx *_enc,th_ycbcr_buffer _img){
return OC_DISABLED;
}
int th_encode_packetout(th_enc_ctx *_enc,int _last_p,ogg_packet *_op){
return OC_DISABLED;
}
int theora_encode_init(theora_state *_te,theora_info *_ci){
return OC_DISABLED;
}
int theora_encode_YUVin(theora_state *_te,yuv_buffer *_yuv){
return OC_DISABLED;
}
int theora_encode_packetout(theora_state *_te,int _last_p,ogg_packet *_op){
return OC_DISABLED;
}
int theora_encode_header(theora_state *_te,ogg_packet *_op){
return OC_DISABLED;
}
int theora_encode_comment(theora_comment *_tc,ogg_packet *_op){
return OC_DISABLED;
}
int theora_encode_tables(theora_state *_te,ogg_packet *_op){
return OC_DISABLED;
}

27
media/libtheora/lib/enquant.h
View File

@ -1,27 +0,0 @@
#if !defined(_enquant_H)
# define _enquant_H (1)
# include "quant.h"
typedef struct oc_iquant oc_iquant;
#define OC_QUANT_MAX_LOG (OC_Q57(OC_STATIC_ILOG_32(OC_QUANT_MAX)-1))
/*Used to compute x/d via ((x*m>>16)+x>>l)+(x<0))
(i.e., one 16x16->16 mul, 2 shifts, and 2 adds).
This is not an approximation; for 16-bit x and d, it is exact.*/
struct oc_iquant{
ogg_int16_t m;
ogg_int16_t l;
};
typedef oc_iquant oc_iquant_table[64];
void oc_quant_params_pack(oggpack_buffer *_opb,const th_quant_info *_qinfo);
void oc_enquant_tables_init(ogg_uint16_t *_dequant[64][3][2],
oc_iquant *_enquant[64][3][2],const th_quant_info *_qinfo);
void oc_enquant_qavg_init(ogg_int64_t _log_qavg[2][64],
ogg_uint16_t *_dequant[64][3][2],int _pixel_fmt);
#endif

43
media/libtheora/lib/fragment.c
View File

@ -11,17 +11,12 @@
********************************************************************
function:
last mod: $Id: fragment.c 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: fragment.c 17410 2010-09-21 21:53:48Z tterribe $
********************************************************************/
#include <string.h>
#include "internal.h"
void oc_frag_copy(const oc_theora_state *_state,unsigned char *_dst,
const unsigned char *_src,int _ystride){
(*_state->opt_vtable.frag_copy)(_dst,_src,_ystride);
}
void oc_frag_copy_c(unsigned char *_dst,const unsigned char *_src,int _ystride){
int i;
for(i=8;i-->0;){
@ -31,9 +26,24 @@ void oc_frag_copy_c(unsigned char *_dst,const unsigned char *_src,int _ystride){
}
}
void oc_frag_recon_intra(const oc_theora_state *_state,unsigned char *_dst,
int _ystride,const ogg_int16_t _residue[64]){
_state->opt_vtable.frag_recon_intra(_dst,_ystride,_residue);
/*Copies the fragments specified by the lists of fragment indices from one
frame to another.
_dst_frame: The reference frame to copy to.
_src_frame: The reference frame to copy from.
_ystride: The row stride of the reference frames.
_fragis: A pointer to a list of fragment indices.
_nfragis: The number of fragment indices to copy.
_frag_buf_offs: The offsets of fragments in the reference frames.*/
void oc_frag_copy_list_c(unsigned char *_dst_frame,
const unsigned char *_src_frame,int _ystride,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,const ptrdiff_t *_frag_buf_offs){
ptrdiff_t fragii;
for(fragii=0;fragii<_nfragis;fragii++){
ptrdiff_t frag_buf_off;
frag_buf_off=_frag_buf_offs[_fragis[fragii]];
oc_frag_copy_c(_dst_frame+frag_buf_off,
_src_frame+frag_buf_off,_ystride);
}
}
void oc_frag_recon_intra_c(unsigned char *_dst,int _ystride,
@ -46,11 +56,6 @@ void oc_frag_recon_intra_c(unsigned char *_dst,int _ystride,
}
}
void oc_frag_recon_inter(const oc_theora_state *_state,unsigned char *_dst,
const unsigned char *_src,int _ystride,const ogg_int16_t _residue[64]){
_state->opt_vtable.frag_recon_inter(_dst,_src,_ystride,_residue);
}
void oc_frag_recon_inter_c(unsigned char *_dst,
const unsigned char *_src,int _ystride,const ogg_int16_t _residue[64]){
int i;
@ -62,12 +67,6 @@ void oc_frag_recon_inter_c(unsigned char *_dst,
}
}
void oc_frag_recon_inter2(const oc_theora_state *_state,unsigned char *_dst,
const unsigned char *_src1,const unsigned char *_src2,int _ystride,
const ogg_int16_t _residue[64]){
_state->opt_vtable.frag_recon_inter2(_dst,_src1,_src2,_ystride,_residue);
}
void oc_frag_recon_inter2_c(unsigned char *_dst,const unsigned char *_src1,
const unsigned char *_src2,int _ystride,const ogg_int16_t _residue[64]){
int i;
@ -80,8 +79,4 @@ void oc_frag_recon_inter2_c(unsigned char *_dst,const unsigned char *_src1,
}
}
void oc_restore_fpu(const oc_theora_state *_state){
_state->opt_vtable.restore_fpu();
}
void oc_restore_fpu_c(void){}

712
media/libtheora/lib/huffdec.c
View File

@ -11,7 +11,7 @@
********************************************************************
function:
last mod: $Id: huffdec.c 16702 2009-11-15 00:40:55Z tterribe $
last mod: $Id: huffdec.c 17577 2010-10-29 04:00:07Z tterribe $
********************************************************************/
@ -22,14 +22,60 @@
#include "decint.h"
/*The ANSI offsetof macro is broken on some platforms (e.g., older DECs).*/
#define _ogg_offsetof(_type,_field)\
((size_t)((char *)&((_type *)0)->_field-(char *)0))
/*The number of internal tokens associated with each of the spec tokens.*/
static const unsigned char OC_DCT_TOKEN_MAP_ENTRIES[TH_NDCT_TOKENS]={
1,1,1,4,8,1,1,8,1,1,1,1,1,2,2,2,2,4,8,2,2,2,4,2,2,2,2,2,8,2,4,8
};
/*Instead of storing every branching in the tree, subtrees can be collapsed
into one node, with a table of size 1<<nbits pointing directly to its
descedents nbits levels down.
This allows more than one bit to be read at a time, and avoids following all
the intermediate branches with next to no increased code complexity once
the collapsed tree has been built.
We do _not_ require that a subtree be complete to be collapsed, but instead
store duplicate pointers in the table, and record the actual depth of the
node below its parent.
This tells us the number of bits to advance the stream after reaching it.
This turns out to be equivalent to the method described in \cite{Hash95},
without the requirement that codewords be sorted by length.
If the codewords were sorted by length (so-called ``canonical-codes''), they
could be decoded much faster via either Lindell and Moffat's approach or
Hashemian's Condensed Huffman Code approach, the latter of which has an
extremely small memory footprint.
We can't use Choueka et al.'s finite state machine approach, which is
extremely fast, because we can't allow multiple symbols to be output at a
time; the codebook can and does change between symbols.
It also has very large memory requirements, which impairs cache coherency.
We store the tree packed in an array of 16-bit integers (words).
Each node consists of a single word, followed consecutively by two or more
indices of its children.
Let n be the value of this first word.
This is the number of bits that need to be read to traverse the node, and
must be positive.
1<<n entries follow in the array, each an index to a child node.
If the child is positive, then it is the index of another internal node in
the table.
If the child is negative or zero, then it is a leaf node.
These are stored directly in the child pointer to save space, since they only
require a single word.
If a leaf node would have been encountered before reading n bits, then it is
duplicated the necessary number of times in this table.
Leaf nodes pack both a token value and their actual depth in the tree.
The token in the leaf node is (-leaf&255).
The number of bits that need to be consumed to reach the leaf, starting from
the current node, is (-leaf>>8).
@ARTICLE{Hash95,
author="Reza Hashemian",
title="Memory Efficient and High-Speed Search {Huffman} Coding",
journal="{IEEE} Transactions on Communications",
volume=43,
number=10,
pages="2576--2581",
month=Oct,
year=1995
}*/
/*The map from external spec-defined tokens to internal tokens.
This is constructed so that any extra bits read with the original token value
@ -99,391 +145,371 @@ static const unsigned char OC_DCT_TOKEN_MAP[TH_NDCT_TOKENS]={
40
};
/*These three functions are really part of the bitpack.c module, but
they are only used here.
Declaring local static versions so they can be inlined saves considerable
function call overhead.*/
static oc_pb_window oc_pack_refill(oc_pack_buf *_b,int _bits){
const unsigned char *ptr;
const unsigned char *stop;
oc_pb_window window;
int available;
window=_b->window;
available=_b->bits;
ptr=_b->ptr;
stop=_b->stop;
/*This version of _refill() doesn't bother setting eof because we won't
check for it after we've started decoding DCT tokens.*/
if(ptr>=stop)available=OC_LOTS_OF_BITS;
while(available<=OC_PB_WINDOW_SIZE-8){
available+=8;
window|=(oc_pb_window)*ptr++<<OC_PB_WINDOW_SIZE-available;
if(ptr>=stop)available=OC_LOTS_OF_BITS;
}
_b->ptr=ptr;
if(_bits>available)window|=*ptr>>(available&7);
_b->bits=available;
return window;
}
/*The log base 2 of number of internal tokens associated with each of the spec
tokens (i.e., how many of the extra bits are folded into the token value).
Increasing the maximum value beyond 3 will enlarge the amount of stack
required for tree construction.*/
static const unsigned char OC_DCT_TOKEN_MAP_LOG_NENTRIES[TH_NDCT_TOKENS]={
0,0,0,2,3,0,0,3,0,0,0,0,0,1,1,1,1,2,3,1,1,1,2,1,1,1,1,1,3,1,2,3
};
/*Read in bits without advancing the bit pointer.
Here we assume 0<=_bits&&_bits<=32.*/
static long oc_pack_look(oc_pack_buf *_b,int _bits){
oc_pb_window window;
int available;
long result;
window=_b->window;
available=_b->bits;
if(_bits==0)return 0;
if(_bits>available)_b->window=window=oc_pack_refill(_b,_bits);
result=window>>OC_PB_WINDOW_SIZE-_bits;
return result;
}
/*Advance the bit pointer.*/
static void oc_pack_adv(oc_pack_buf *_b,int _bits){
/*We ignore the special cases for _bits==0 and _bits==32 here, since they are
never used actually used.
OC_HUFF_SLUSH (defined below) would have to be at least 27 to actually read
32 bits in a single go, and would require a 32 GB lookup table (assuming
8 byte pointers, since 4 byte pointers couldn't fit such a table).*/
_b->window<<=_bits;
_b->bits-=_bits;
}
/*The log_2 of the size of a lookup table is allowed to grow to relative to
the number of unique nodes it contains.
E.g., if OC_HUFF_SLUSH is 2, then at most 75% of the space in the tree is
wasted (each node will have an amortized cost of at most 20 bytes when using
4-byte pointers).
/*The size a lookup table is allowed to grow to relative to the number of
unique nodes it contains.
E.g., if OC_HUFF_SLUSH is 4, then at most 75% of the space in the tree is
wasted (1/4 of the space must be used).
Larger numbers can decode tokens with fewer read operations, while smaller
numbers may save more space (requiring as little as 8 bytes amortized per
node, though there will be more nodes).
numbers may save more space.
With a sample file:
32233473 read calls are required when no tree collapsing is done (100.0%).
19269269 read calls are required when OC_HUFF_SLUSH is 0 (59.8%).
11144969 read calls are required when OC_HUFF_SLUSH is 1 (34.6%).
10538563 read calls are required when OC_HUFF_SLUSH is 2 (32.7%).
10192578 read calls are required when OC_HUFF_SLUSH is 3 (31.6%).
Since a value of 1 gets us the vast majority of the speed-up with only a
small amount of wasted memory, this is what we use.*/
#define OC_HUFF_SLUSH (1)
19269269 read calls are required when OC_HUFF_SLUSH is 1 (59.8%).
11144969 read calls are required when OC_HUFF_SLUSH is 2 (34.6%).
10538563 read calls are required when OC_HUFF_SLUSH is 4 (32.7%).
10192578 read calls are required when OC_HUFF_SLUSH is 8 (31.6%).
Since a value of 2 gets us the vast majority of the speed-up with only a
small amount of wasted memory, this is what we use.
This value must be less than 128, or you could create a tree with more than
32767 entries, which would overflow the 16-bit words used to index it.*/
#define OC_HUFF_SLUSH (2)
/*The root of the tree is on the fast path, and a larger value here is more
beneficial than elsewhere in the tree.
7 appears to give the best performance, trading off between increased use of
the single-read fast path and cache footprint for the tables, though
obviously this will depend on your cache size.
Using 7 here, the VP3 tables are about twice as large compared to using 2.*/
#define OC_ROOT_HUFF_SLUSH (7)
/*Determines the size in bytes of a Huffman tree node that represents a
/*Unpacks a Huffman codebook.
_opb: The buffer to unpack from.
_tokens: Stores a list of internal tokens, in the order they were found in
the codebook, and the lengths of their corresponding codewords.
This is enough to completely define the codebook, while minimizing
stack usage and avoiding temporary allocations (for platforms
where free() is a no-op).
Return: The number of internal tokens in the codebook, or a negative value
on error.*/
int oc_huff_tree_unpack(oc_pack_buf *_opb,unsigned char _tokens[256][2]){
ogg_uint32_t code;
int len;
int ntokens;
int nleaves;
code=0;
len=ntokens=nleaves=0;
for(;;){
long bits;
bits=oc_pack_read1(_opb);
/*Only process nodes so long as there's more bits in the buffer.*/
if(oc_pack_bytes_left(_opb)<0)return TH_EBADHEADER;
/*Read an internal node:*/
if(!bits){
len++;
/*Don't allow codewords longer than 32 bits.*/
if(len>32)return TH_EBADHEADER;
}
/*Read a leaf node:*/
else{
ogg_uint32_t code_bit;
int neb;
int nentries;
int token;
/*Don't allow more than 32 spec-tokens per codebook.*/
if(++nleaves>32)return TH_EBADHEADER;
bits=oc_pack_read(_opb,OC_NDCT_TOKEN_BITS);
neb=OC_DCT_TOKEN_MAP_LOG_NENTRIES[bits];
token=OC_DCT_TOKEN_MAP[bits];
nentries=1<<neb;
while(nentries-->0){
_tokens[ntokens][0]=(unsigned char)token++;
_tokens[ntokens][1]=(unsigned char)(len+neb);
ntokens++;
}
code_bit=0x80000000U>>len-1;
while(len>0&&(code&code_bit)){
code^=code_bit;
code_bit<<=1;
len--;
}
if(len<=0)break;
code|=code_bit;
}
}
return ntokens;
}
/*Count how many tokens would be required to fill a subtree at depth _depth.
_tokens: A list of internal tokens, in the order they are found in the
codebook, and the lengths of their corresponding codewords.
_depth: The depth of the desired node in the corresponding tree structure.
Return: The number of tokens that belong to that subtree.*/
static int oc_huff_subtree_tokens(unsigned char _tokens[][2],int _depth){
ogg_uint32_t code;
int ti;
code=0;
ti=0;
do{
if(_tokens[ti][1]-_depth<32)code+=0x80000000U>>_tokens[ti++][1]-_depth;
else{
/*Because of the expanded internal tokens, we can have codewords as long
as 35 bits.
A single recursion here is enough to advance past them.*/
code++;
ti+=oc_huff_subtree_tokens(_tokens+ti,_depth+31);
}
}
while(code<0x80000000U);
return ti;
}
/*Compute the number of bits to use for a collapsed tree node at the given
depth.
_tokens: A list of internal tokens, in the order they are found in the
codebook, and the lengths of their corresponding codewords.
_ntokens: The number of tokens corresponding to this tree node.
_depth: The depth of this tree node.
Return: The number of bits to use for a collapsed tree node rooted here.
This is always at least one, even if this was a leaf node.*/
static int oc_huff_tree_collapse_depth(unsigned char _tokens[][2],
int _ntokens,int _depth){
int got_leaves;
int loccupancy;
int occupancy;
int slush;
int nbits;
int best_nbits;
slush=_depth>0?OC_HUFF_SLUSH:OC_ROOT_HUFF_SLUSH;
/*It's legal to have a tree with just a single node, which requires no bits
to decode and always returns the same token.
However, no encoder actually does this (yet).
To avoid a special case in oc_huff_token_decode(), we force the number of
lookahead bits to be at least one.
This will produce a tree that looks ahead one bit and then advances the
stream zero bits.*/
nbits=1;
occupancy=2;
got_leaves=1;
do{
int ti;
if(got_leaves)best_nbits=nbits;
nbits++;
got_leaves=0;
loccupancy=occupancy;
for(occupancy=ti=0;ti<_ntokens;occupancy++){
if(_tokens[ti][1]<_depth+nbits)ti++;
else if(_tokens[ti][1]==_depth+nbits){
got_leaves=1;
ti++;
}
else ti+=oc_huff_subtree_tokens(_tokens+ti,_depth+nbits);
}
}
while(occupancy>loccupancy&&occupancy*slush>=1<<nbits);
return best_nbits;
}
/*Determines the size in words of a Huffman tree node that represents a
subtree of depth _nbits.
_nbits: The depth of the subtree.
If this is 0, the node is a leaf node.
Otherwise 1<<_nbits pointers are allocated for children.
Return: The number of bytes required to store the node.*/
This must be greater than zero.
Return: The number of words required to store the node.*/
static size_t oc_huff_node_size(int _nbits){
size_t size;
size=_ogg_offsetof(oc_huff_node,nodes);
if(_nbits>0)size+=sizeof(oc_huff_node *)*(1<<_nbits);
return size;
return 1+(1<<_nbits);
}
static oc_huff_node *oc_huff_node_init(char **_storage,size_t _size,int _nbits){
oc_huff_node *ret;
ret=(oc_huff_node *)*_storage;
ret->nbits=(unsigned char)_nbits;
(*_storage)+=_size;
return ret;
}
/*Determines the size in bytes of a Huffman tree.
_nbits: The depth of the subtree.
If this is 0, the node is a leaf node.
Otherwise storage for 1<<_nbits pointers are added for children.
Return: The number of bytes required to store the tree.*/
static size_t oc_huff_tree_size(const oc_huff_node *_node){
size_t size;
size=oc_huff_node_size(_node->nbits);
if(_node->nbits){
int nchildren;
int i;
nchildren=1<<_node->nbits;
for(i=0;i<nchildren;i+=1<<_node->nbits-_node->nodes[i]->depth){
size+=oc_huff_tree_size(_node->nodes[i]);
}
}
return size;
}
/*Unpacks a sub-tree from the given buffer.
_opb: The buffer to unpack from.
_binodes: The nodes to store the sub-tree in.
_nbinodes: The number of nodes available for the sub-tree.
Return: 0 on success, or a negative value on error.*/
static int oc_huff_tree_unpack(oc_pack_buf *_opb,
oc_huff_node *_binodes,int _nbinodes){
oc_huff_node *binode;
long bits;
int nused;
if(_nbinodes<1)return TH_EBADHEADER;
binode=_binodes;
nused=0;
bits=oc_pack_read1(_opb);
if(oc_pack_bytes_left(_opb)<0)return TH_EBADHEADER;
/*Read an internal node:*/
if(!bits){
int ret;
nused++;
binode->nbits=1;
binode->depth=1;
binode->nodes[0]=_binodes+nused;
ret=oc_huff_tree_unpack(_opb,_binodes+nused,_nbinodes-nused);
if(ret>=0){
nused+=ret;
binode->nodes[1]=_binodes+nused;
ret=oc_huff_tree_unpack(_opb,_binodes+nused,_nbinodes-nused);
}
if(ret<0)return ret;
nused+=ret;
}
/*Read a leaf node:*/
else{
int ntokens;
int token;
int i;
bits=oc_pack_read(_opb,OC_NDCT_TOKEN_BITS);
if(oc_pack_bytes_left(_opb)<0)return TH_EBADHEADER;
/*Find out how many internal tokens we translate this external token into.*/
ntokens=OC_DCT_TOKEN_MAP_ENTRIES[bits];
if(_nbinodes<2*ntokens-1)return TH_EBADHEADER;
/*Fill in a complete binary tree pointing to the internal tokens.*/
for(i=1;i<ntokens;i<<=1){
int j;
binode=_binodes+nused;
nused+=i;
for(j=0;j<i;j++){
binode[j].nbits=1;
binode[j].depth=1;
binode[j].nodes[0]=_binodes+nused+2*j;
binode[j].nodes[1]=_binodes+nused+2*j+1;
/*Produces a collapsed-tree representation of the given token list.
_tree: The storage for the collapsed Huffman tree.
This may be NULL to compute the required storage size instead of
constructing the tree.
_tokens: A list of internal tokens, in the order they are found in the
codebook, and the lengths of their corresponding codewords.
_ntokens: The number of tokens corresponding to this tree node.
Return: The number of words required to store the tree.*/
static size_t oc_huff_tree_collapse(ogg_int16_t *_tree,
unsigned char _tokens[][2],int _ntokens){
ogg_int16_t node[34];
unsigned char depth[34];
unsigned char last[34];
size_t ntree;
int ti;
int l;
depth[0]=0;
last[0]=(unsigned char)(_ntokens-1);
ntree=0;
ti=0;
l=0;
do{
int nbits;
nbits=oc_huff_tree_collapse_depth(_tokens+ti,last[l]+1-ti,depth[l]);
node[l]=(ogg_int16_t)ntree;
ntree+=oc_huff_node_size(nbits);
if(_tree!=NULL)_tree[node[l]++]=(ogg_int16_t)nbits;
do{
while(ti<=last[l]&&_tokens[ti][1]<=depth[l]+nbits){
if(_tree!=NULL){
ogg_int16_t leaf;
int nentries;
nentries=1<<depth[l]+nbits-_tokens[ti][1];
leaf=(ogg_int16_t)-(_tokens[ti][1]-depth[l]<<8|_tokens[ti][0]);
while(nentries-->0)_tree[node[l]++]=leaf;
}
ti++;
}
if(ti<=last[l]){
/*We need to recurse*/
depth[l+1]=(unsigned char)(depth[l]+nbits);
if(_tree!=NULL)_tree[node[l]++]=(ogg_int16_t)ntree;
l++;
last[l]=
(unsigned char)(ti+oc_huff_subtree_tokens(_tokens+ti,depth[l])-1);
break;
}
/*Pop back up a level of recursion.*/
else if(l-->0)nbits=depth[l+1]-depth[l];
}
/*And now the leaf nodes with those tokens.*/
token=OC_DCT_TOKEN_MAP[bits];
for(i=0;i<ntokens;i++){
binode=_binodes+nused++;
binode->nbits=0;
binode->depth=1;
binode->token=token+i;
}
while(l>=0);
}
return nused;
}
/*Finds the depth of shortest branch of the given sub-tree.
The tree must be binary.
_binode: The root of the given sub-tree.
_binode->nbits must be 0 or 1.
Return: The smallest depth of a leaf node in this sub-tree.
0 indicates this sub-tree is a leaf node.*/
static int oc_huff_tree_mindepth(oc_huff_node *_binode){
int depth0;
int depth1;
if(_binode->nbits==0)return 0;
depth0=oc_huff_tree_mindepth(_binode->nodes[0]);
depth1=oc_huff_tree_mindepth(_binode->nodes[1]);
return OC_MINI(depth0,depth1)+1;
}
/*Finds the number of internal nodes at a given depth, plus the number of
leaves at that depth or shallower.
The tree must be binary.
_binode: The root of the given sub-tree.
_binode->nbits must be 0 or 1.
Return: The number of entries that would be contained in a jump table of the
given depth.*/
static int oc_huff_tree_occupancy(oc_huff_node *_binode,int _depth){
if(_binode->nbits==0||_depth<=0)return 1;
else{
return oc_huff_tree_occupancy(_binode->nodes[0],_depth-1)+
oc_huff_tree_occupancy(_binode->nodes[1],_depth-1);
}
}
/*Makes a copy of the given Huffman tree.
_node: The Huffman tree to copy.
Return: The copy of the Huffman tree.*/
static oc_huff_node *oc_huff_tree_copy(const oc_huff_node *_node,
char **_storage){
oc_huff_node *ret;
ret=oc_huff_node_init(_storage,oc_huff_node_size(_node->nbits),_node->nbits);
ret->depth=_node->depth;
if(_node->nbits){
int nchildren;
int i;
int inext;
nchildren=1<<_node->nbits;
for(i=0;i<nchildren;){
ret->nodes[i]=oc_huff_tree_copy(_node->nodes[i],_storage);
inext=i+(1<<_node->nbits-ret->nodes[i]->depth);
while(++i<inext)ret->nodes[i]=ret->nodes[i-1];
}
}
else ret->token=_node->token;
return ret;
}
static size_t oc_huff_tree_collapse_size(oc_huff_node *_binode,int _depth){
size_t size;
int mindepth;
int depth;
int loccupancy;
int occupancy;
if(_binode->nbits!=0&&_depth>0){
return oc_huff_tree_collapse_size(_binode->nodes[0],_depth-1)+
oc_huff_tree_collapse_size(_binode->nodes[1],_depth-1);
}
depth=mindepth=oc_huff_tree_mindepth(_binode);
occupancy=1<<mindepth;
do{
loccupancy=occupancy;
occupancy=oc_huff_tree_occupancy(_binode,++depth);
}
while(occupancy>loccupancy&&occupancy>=1<<OC_MAXI(depth-OC_HUFF_SLUSH,0));
depth--;
size=oc_huff_node_size(depth);
if(depth>0){
size+=oc_huff_tree_collapse_size(_binode->nodes[0],depth-1);
size+=oc_huff_tree_collapse_size(_binode->nodes[1],depth-1);
}
return size;
}
static oc_huff_node *oc_huff_tree_collapse(oc_huff_node *_binode,
char **_storage);
/*Fills the given nodes table with all the children in the sub-tree at the
given depth.
The nodes in the sub-tree with a depth less than that stored in the table
are freed.
The sub-tree must be binary and complete up until the given depth.
_nodes: The nodes table to fill.
_binode: The root of the sub-tree to fill it with.
_binode->nbits must be 0 or 1.
_level: The current level in the table.
0 indicates that the current node should be stored, regardless of
whether it is a leaf node or an internal node.
_depth: The depth of the nodes to fill the table with, relative to their
parent.*/
static void oc_huff_node_fill(oc_huff_node **_nodes,
oc_huff_node *_binode,int _level,int _depth,char **_storage){
if(_level<=0||_binode->nbits==0){
int i;
_binode->depth=(unsigned char)(_depth-_level);
_nodes[0]=oc_huff_tree_collapse(_binode,_storage);
for(i=1;i<1<<_level;i++)_nodes[i]=_nodes[0];
}
else{
_level--;
oc_huff_node_fill(_nodes,_binode->nodes[0],_level,_depth,_storage);
_nodes+=1<<_level;
oc_huff_node_fill(_nodes,_binode->nodes[1],_level,_depth,_storage);
}
}
/*Finds the largest complete sub-tree rooted at the current node and collapses
it into a single node.
This procedure is then applied recursively to all the children of that node.
_binode: The root of the sub-tree to collapse.
_binode->nbits must be 0 or 1.
Return: The new root of the collapsed sub-tree.*/
static oc_huff_node *oc_huff_tree_collapse(oc_huff_node *_binode,
char **_storage){
oc_huff_node *root;
size_t size;
int mindepth;
int depth;
int loccupancy;
int occupancy;
depth=mindepth=oc_huff_tree_mindepth(_binode);
occupancy=1<<mindepth;
do{
loccupancy=occupancy;
occupancy=oc_huff_tree_occupancy(_binode,++depth);
}
while(occupancy>loccupancy&&occupancy>=1<<OC_MAXI(depth-OC_HUFF_SLUSH,0));
depth--;
if(depth<=0)return oc_huff_tree_copy(_binode,_storage);
size=oc_huff_node_size(depth);
root=oc_huff_node_init(_storage,size,depth);
root->depth=_binode->depth;
oc_huff_node_fill(root->nodes,_binode,depth,depth,_storage);
return root;
while(l>=0);
return ntree;
}
/*Unpacks a set of Huffman trees, and reduces them to a collapsed
representation.
_opb: The buffer to unpack the trees from.
_nodes: The table to fill with the Huffman trees.
Return: 0 on success, or a negative value on error.*/
Return: 0 on success, or a negative value on error.
The caller is responsible for cleaning up any partially initialized
_nodes on failure.*/
int oc_huff_trees_unpack(oc_pack_buf *_opb,
oc_huff_node *_nodes[TH_NHUFFMAN_TABLES]){
ogg_int16_t *_nodes[TH_NHUFFMAN_TABLES]){
int i;
for(i=0;i<TH_NHUFFMAN_TABLES;i++){
oc_huff_node nodes[511];
char *storage;
size_t size;
int ret;
unsigned char tokens[256][2];
int ntokens;
ogg_int16_t *tree;
size_t size;
/*Unpack the full tree into a temporary buffer.*/
ret=oc_huff_tree_unpack(_opb,nodes,sizeof(nodes)/sizeof(*nodes));
if(ret<0)return ret;
/*Figure out how big the collapsed tree will be.*/
size=oc_huff_tree_collapse_size(nodes,0);
storage=(char *)_ogg_calloc(1,size);
if(storage==NULL)return TH_EFAULT;
/*And collapse it.*/
_nodes[i]=oc_huff_tree_collapse(nodes,&storage);
ntokens=oc_huff_tree_unpack(_opb,tokens);
if(ntokens<0)return ntokens;
/*Figure out how big the collapsed tree will be and allocate space for it.*/
size=oc_huff_tree_collapse(NULL,tokens,ntokens);
/*This should never happen; if it does it means you set OC_HUFF_SLUSH or
OC_ROOT_HUFF_SLUSH too large.*/
if(size>32767)return TH_EIMPL;
tree=(ogg_int16_t *)_ogg_malloc(size*sizeof(*tree));
if(tree==NULL)return TH_EFAULT;
/*Construct the collapsed the tree.*/
oc_huff_tree_collapse(tree,tokens,ntokens);
_nodes[i]=tree;
}
return 0;
}
/*Determines the size in words of a Huffman subtree.
_tree: The complete Huffman tree.
_node: The index of the root of the desired subtree.
Return: The number of words required to store the tree.*/
static size_t oc_huff_tree_size(const ogg_int16_t *_tree,int _node){
size_t size;
int nchildren;
int n;
int i;
n=_tree[_node];
size=oc_huff_node_size(n);
nchildren=1<<n;
i=0;
do{
int child;
child=_tree[_node+i+1];
if(child<=0)i+=1<<n-(-child>>8);
else{
size+=oc_huff_tree_size(_tree,child);
i++;
}
}
while(i<nchildren);
return size;
}
/*Makes a copy of the given set of Huffman trees.
_dst: The array to store the copy in.
_src: The array of trees to copy.*/
int oc_huff_trees_copy(oc_huff_node *_dst[TH_NHUFFMAN_TABLES],
const oc_huff_node *const _src[TH_NHUFFMAN_TABLES]){
int oc_huff_trees_copy(ogg_int16_t *_dst[TH_NHUFFMAN_TABLES],
const ogg_int16_t *const _src[TH_NHUFFMAN_TABLES]){
int total;
int i;
total=0;
for(i=0;i<TH_NHUFFMAN_TABLES;i++){
size_t size;
char *storage;
size=oc_huff_tree_size(_src[i]);
storage=(char *)_ogg_calloc(1,size);
if(storage==NULL){
size_t size;
size=oc_huff_tree_size(_src[i],0);
total+=size;
_dst[i]=(ogg_int16_t *)_ogg_malloc(size*sizeof(*_dst[i]));
if(_dst[i]==NULL){
while(i-->0)_ogg_free(_dst[i]);
return TH_EFAULT;
}
_dst[i]=oc_huff_tree_copy(_src[i],&storage);
memcpy(_dst[i],_src[i],size*sizeof(*_dst[i]));
}
return 0;
}
/*Frees the memory used by a set of Huffman trees.
_nodes: The array of trees to free.*/
void oc_huff_trees_clear(oc_huff_node *_nodes[TH_NHUFFMAN_TABLES]){
void oc_huff_trees_clear(ogg_int16_t *_nodes[TH_NHUFFMAN_TABLES]){
int i;
for(i=0;i<TH_NHUFFMAN_TABLES;i++)_ogg_free(_nodes[i]);
}
/*Unpacks a single token using the given Huffman tree.
_opb: The buffer to unpack the token from.
_node: The tree to unpack the token with.
Return: The token value.*/
int oc_huff_token_decode(oc_pack_buf *_opb,const oc_huff_node *_node){
long bits;
while(_node->nbits!=0){
bits=oc_pack_look(_opb,_node->nbits);
_node=_node->nodes[bits];
oc_pack_adv(_opb,_node->depth);
int oc_huff_token_decode_c(oc_pack_buf *_opb,const ogg_int16_t *_tree){
const unsigned char *ptr;
const unsigned char *stop;
oc_pb_window window;
int available;
long bits;
int node;
int n;
ptr=_opb->ptr;
window=_opb->window;
stop=_opb->stop;
available=_opb->bits;
node=0;
for(;;){
n=_tree[node];
if(n>available){
unsigned shift;
shift=OC_PB_WINDOW_SIZE-available;
do{
/*We don't bother setting eof because we won't check for it after we've
started decoding DCT tokens.*/
if(ptr>=stop){
shift=(unsigned)-OC_LOTS_OF_BITS;
break;
}
shift-=8;
window|=(oc_pb_window)*ptr++<<shift;
}
while(shift>=8);
/*Note: We never request more than 24 bits, so there's no need to fill in
the last partial byte here.*/
available=OC_PB_WINDOW_SIZE-shift;
}
bits=window>>OC_PB_WINDOW_SIZE-n;
node=_tree[node+1+bits];
if(node<=0)break;
window<<=n;
available-=n;
}
return _node->token;
node=-node;
n=node>>8;
window<<=n;
available-=n;
_opb->ptr=ptr;
_opb->window=window;
_opb->bits=available;
return node&255;
}

72
media/libtheora/lib/huffdec.h
View File

@ -11,7 +11,7 @@
********************************************************************
function:
last mod: $Id: huffdec.h 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: huffdec.h 17410 2010-09-21 21:53:48Z tterribe $
********************************************************************/
@ -22,71 +22,11 @@
typedef struct oc_huff_node oc_huff_node;
/*A node in the Huffman tree.
Instead of storing every branching in the tree, subtrees can be collapsed
into one node, with a table of size 1<<nbits pointing directly to its
descedents nbits levels down.
This allows more than one bit to be read at a time, and avoids following all
the intermediate branches with next to no increased code complexity once
the collapsed tree has been built.
We do _not_ require that a subtree be complete to be collapsed, but instead
store duplicate pointers in the table, and record the actual depth of the
node below its parent.
This tells us the number of bits to advance the stream after reaching it.
This turns out to be equivalent to the method described in \cite{Hash95},
without the requirement that codewords be sorted by length.
If the codewords were sorted by length (so-called ``canonical-codes''), they
could be decoded much faster via either Lindell and Moffat's approach or
Hashemian's Condensed Huffman Code approach, the latter of which has an
extremely small memory footprint.
We can't use Choueka et al.'s finite state machine approach, which is
extremely fast, because we can't allow multiple symbols to be output at a
time; the codebook can and does change between symbols.
It also has very large memory requirements, which impairs cache coherency.
@ARTICLE{Hash95,
author="Reza Hashemian",
title="Memory Efficient and High-Speed Search {Huffman} Coding",
journal="{IEEE} Transactions on Communications",
volume=43,
number=10,
pages="2576--2581",
month=Oct,
year=1995
}*/
struct oc_huff_node{
/*The number of bits of the code needed to descend through this node.
0 indicates a leaf node.
Otherwise there are 1<<nbits nodes in the nodes table, which can be
indexed by reading nbits bits from the stream.*/
unsigned char nbits;
/*The value of a token stored in a leaf node.
The value in non-leaf nodes is undefined.*/
unsigned char token;
/*The depth of the current node, relative to its parent in the collapsed
tree.
This can be less than its parent's nbits value, in which case there are
1<<nbits-depth copies of this node in the table, and the bitstream should
only be advanced depth bits after reaching this node.*/
unsigned char depth;
/*The table of child nodes.
The ACTUAL size of this array is 1<<nbits, despite what the declaration
below claims.
The exception is that for leaf nodes the size is 0.*/
oc_huff_node *nodes[2];
};
int oc_huff_trees_unpack(oc_pack_buf *_opb,
oc_huff_node *_nodes[TH_NHUFFMAN_TABLES]);
int oc_huff_trees_copy(oc_huff_node *_dst[TH_NHUFFMAN_TABLES],
const oc_huff_node *const _src[TH_NHUFFMAN_TABLES]);
void oc_huff_trees_clear(oc_huff_node *_nodes[TH_NHUFFMAN_TABLES]);
int oc_huff_token_decode(oc_pack_buf *_opb,const oc_huff_node *_node);
ogg_int16_t *_nodes[TH_NHUFFMAN_TABLES]);
int oc_huff_trees_copy(ogg_int16_t *_dst[TH_NHUFFMAN_TABLES],
const ogg_int16_t *const _src[TH_NHUFFMAN_TABLES]);
void oc_huff_trees_clear(ogg_int16_t *_nodes[TH_NHUFFMAN_TABLES]);
int oc_huff_token_decode_c(oc_pack_buf *_opb,const ogg_int16_t *_node);
#endif

19
media/libtheora/lib/huffenc.h
View File

@ -1,19 +0,0 @@
#if !defined(_huffenc_H)
# define _huffenc_H (1)
# include "huffman.h"
typedef th_huff_code th_huff_table[TH_NDCT_TOKENS];
extern const th_huff_code
TH_VP31_HUFF_CODES[TH_NHUFFMAN_TABLES][TH_NDCT_TOKENS];
int oc_huff_codes_pack(oggpack_buffer *_opb,
const th_huff_code _codes[TH_NHUFFMAN_TABLES][TH_NDCT_TOKENS]);
#endif

58
media/libtheora/lib/idct.c
View File

@ -11,7 +11,7 @@
********************************************************************
function:
last mod: $Id: idct.c 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: idct.c 17410 2010-09-21 21:53:48Z tterribe $
********************************************************************/
@ -231,18 +231,18 @@ static void idct8_1(ogg_int16_t *_y,const ogg_int16_t _x[1]){
_y: The buffer to store the result in.
This may be the same as _x.
_x: The input coefficients.*/
static void oc_idct8x8_3(ogg_int16_t _y[64],const ogg_int16_t _x[64]){
const ogg_int16_t *in;
ogg_int16_t *end;
ogg_int16_t *out;
ogg_int16_t w[64];
static void oc_idct8x8_3(ogg_int16_t _y[64],ogg_int16_t _x[64]){
ogg_int16_t w[64];
int i;
/*Transform rows of x into columns of w.*/
idct8_2(w,_x);
idct8_1(w+1,_x+8);
/*Transform rows of w into columns of y.*/
for(in=w,out=_y,end=out+8;out<end;in+=8,out++)idct8_2(out,in);
for(i=0;i<8;i++)idct8_2(_y+i,w+i*8);
/*Adjust for the scale factor.*/
for(out=_y,end=out+64;out<end;out++)*out=(ogg_int16_t)(*out+8>>4);
for(i=0;i<64;i++)_y[i]=(ogg_int16_t)(_y[i]+8>>4);
/*Clear input data for next block (decoder only).*/
if(_x!=_y)_x[0]=_x[1]=_x[8]=0;
}
/*Performs an inverse 8x8 Type-II DCT transform.
@ -260,20 +260,20 @@ static void oc_idct8x8_3(ogg_int16_t _y[64],const ogg_int16_t _x[64]){
_y: The buffer to store the result in.
This may be the same as _x.
_x: The input coefficients.*/
static void oc_idct8x8_10(ogg_int16_t _y[64],const ogg_int16_t _x[64]){
const ogg_int16_t *in;
ogg_int16_t *end;
ogg_int16_t *out;
ogg_int16_t w[64];
static void oc_idct8x8_10(ogg_int16_t _y[64],ogg_int16_t _x[64]){
ogg_int16_t w[64];
int i;
/*Transform rows of x into columns of w.*/
idct8_4(w,_x);
idct8_3(w+1,_x+8);
idct8_2(w+2,_x+16);
idct8_1(w+3,_x+24);
/*Transform rows of w into columns of y.*/
for(in=w,out=_y,end=out+8;out<end;in+=8,out++)idct8_4(out,in);
for(i=0;i<8;i++)idct8_4(_y+i,w+i*8);
/*Adjust for the scale factor.*/
for(out=_y,end=out+64;out<end;out++)*out=(ogg_int16_t)(*out+8>>4);
for(i=0;i<64;i++)_y[i]=(ogg_int16_t)(_y[i]+8>>4);
/*Clear input data for next block (decoder only).*/
if(_x!=_y)_x[0]=_x[1]=_x[2]=_x[3]=_x[8]=_x[9]=_x[10]=_x[16]=_x[17]=_x[24]=0;
}
/*Performs an inverse 8x8 Type-II DCT transform.
@ -282,28 +282,22 @@ static void oc_idct8x8_10(ogg_int16_t _y[64],const ogg_int16_t _x[64]){
_y: The buffer to store the result in.
This may be the same as _x.
_x: The input coefficients.*/
static void oc_idct8x8_slow(ogg_int16_t _y[64],const ogg_int16_t _x[64]){
const ogg_int16_t *in;
ogg_int16_t *end;
ogg_int16_t *out;
ogg_int16_t w[64];
static void oc_idct8x8_slow(ogg_int16_t _y[64],ogg_int16_t _x[64]){
ogg_int16_t w[64];
int i;
/*Transform rows of x into columns of w.*/
for(in=_x,out=w,end=out+8;out<end;in+=8,out++)idct8(out,in);
for(i=0;i<8;i++)idct8(w+i,_x+i*8);
/*Transform rows of w into columns of y.*/
for(in=w,out=_y,end=out+8;out<end;in+=8,out++)idct8(out,in);
for(i=0;i<8;i++)idct8(_y+i,w+i*8);
/*Adjust for the scale factor.*/
for(out=_y,end=out+64;out<end;out++)*out=(ogg_int16_t)(*out+8>>4);
}
void oc_idct8x8(const oc_theora_state *_state,ogg_int16_t _y[64],
int _last_zzi){
(*_state->opt_vtable.idct8x8)(_y,_last_zzi);
for(i=0;i<64;i++)_y[i]=(ogg_int16_t)(_y[i]+8>>4);
if(_x!=_y)for(i=0;i<64;i++)_x[i]=0;
}
/*Performs an inverse 8x8 Type-II DCT transform.
The input is assumed to be scaled by a factor of 4 relative to orthonormal
version of the transform.*/
void oc_idct8x8_c(ogg_int16_t _y[64],int _last_zzi){
void oc_idct8x8_c(ogg_int16_t _y[64],ogg_int16_t _x[64],int _last_zzi){
/*_last_zzi is subtly different from an actual count of the number of
coefficients we decoded for this block.
It contains the value of zzi BEFORE the final token in the block was
@ -329,7 +323,7 @@ void oc_idct8x8_c(ogg_int16_t _y[64],int _last_zzi){
gets.
Needless to say we inherited this approach from VP3.*/
/*Then perform the iDCT.*/
if(_last_zzi<3)oc_idct8x8_3(_y,_y);
else if(_last_zzi<10)oc_idct8x8_10(_y,_y);
else oc_idct8x8_slow(_y,_y);
if(_last_zzi<=3)oc_idct8x8_3(_y,_x);
else if(_last_zzi<=10)oc_idct8x8_10(_y,_x);
else oc_idct8x8_slow(_y,_x);
}

90
media/libtheora/lib/internal.c
View File

@ -11,7 +11,7 @@
********************************************************************
function:
last mod: $Id: internal.c 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: internal.c 17506 2010-10-13 02:52:41Z tterribe $
********************************************************************/
@ -97,79 +97,29 @@ int oc_ilog(unsigned _v){
/*The function used to fill in the chroma plane motion vectors for a macro
block when 4 different motion vectors are specified in the luma plane.
This version is for use with chroma decimated in the X and Y directions
(4:2:0).
_cbmvs: The chroma block-level motion vectors to fill in.
_lbmvs: The luma block-level motion vectors.*/
static void oc_set_chroma_mvs00(oc_mv _cbmvs[4],const oc_mv _lbmvs[4]){
int dx;
int dy;
dx=_lbmvs[0][0]+_lbmvs[1][0]+_lbmvs[2][0]+_lbmvs[3][0];
dy=_lbmvs[0][1]+_lbmvs[1][1]+_lbmvs[2][1]+_lbmvs[3][1];
_cbmvs[0][0]=(signed char)OC_DIV_ROUND_POW2(dx,2,2);
_cbmvs[0][1]=(signed char)OC_DIV_ROUND_POW2(dy,2,2);
void *oc_aligned_malloc(size_t _sz,size_t _align){
unsigned char *p;
if(_align-1>UCHAR_MAX||(_align&_align-1)||_sz>~(size_t)0-_align)return NULL;
p=(unsigned char *)_ogg_malloc(_sz+_align);
if(p!=NULL){
int offs;
offs=((p-(unsigned char *)0)-1&_align-1);
p[offs]=offs;
p+=offs+1;
}
return p;
}
/*The function used to fill in the chroma plane motion vectors for a macro
block when 4 different motion vectors are specified in the luma plane.
This version is for use with chroma decimated in the Y direction.
_cbmvs: The chroma block-level motion vectors to fill in.
_lbmvs: The luma block-level motion vectors.*/
static void oc_set_chroma_mvs01(oc_mv _cbmvs[4],const oc_mv _lbmvs[4]){
int dx;
int dy;
dx=_lbmvs[0][0]+_lbmvs[2][0];
dy=_lbmvs[0][1]+_lbmvs[2][1];
_cbmvs[0][0]=(signed char)OC_DIV_ROUND_POW2(dx,1,1);
_cbmvs[0][1]=(signed char)OC_DIV_ROUND_POW2(dy,1,1);
dx=_lbmvs[1][0]+_lbmvs[3][0];
dy=_lbmvs[1][1]+_lbmvs[3][1];
_cbmvs[1][0]=(signed char)OC_DIV_ROUND_POW2(dx,1,1);
_cbmvs[1][1]=(signed char)OC_DIV_ROUND_POW2(dy,1,1);
void oc_aligned_free(void *_ptr){
unsigned char *p;
p=(unsigned char *)_ptr;
if(p!=NULL){
int offs;
offs=*--p;
_ogg_free(p-offs);
}
}
/*The function used to fill in the chroma plane motion vectors for a macro
block when 4 different motion vectors are specified in the luma plane.
This version is for use with chroma decimated in the X direction (4:2:2).
_cbmvs: The chroma block-level motion vectors to fill in.
_lbmvs: The luma block-level motion vectors.*/
static void oc_set_chroma_mvs10(oc_mv _cbmvs[4],const oc_mv _lbmvs[4]){
int dx;
int dy;
dx=_lbmvs[0][0]+_lbmvs[1][0];
dy=_lbmvs[0][1]+_lbmvs[1][1];
_cbmvs[0][0]=(signed char)OC_DIV_ROUND_POW2(dx,1,1);
_cbmvs[0][1]=(signed char)OC_DIV_ROUND_POW2(dy,1,1);
dx=_lbmvs[2][0]+_lbmvs[3][0];
dy=_lbmvs[2][1]+_lbmvs[3][1];
_cbmvs[2][0]=(signed char)OC_DIV_ROUND_POW2(dx,1,1);
_cbmvs[2][1]=(signed char)OC_DIV_ROUND_POW2(dy,1,1);
}
/*The function used to fill in the chroma plane motion vectors for a macro
block when 4 different motion vectors are specified in the luma plane.
This version is for use with no chroma decimation (4:4:4).
_cbmvs: The chroma block-level motion vectors to fill in.
_lmbmv: The luma macro-block level motion vector to fill in for use in
prediction.
_lbmvs: The luma block-level motion vectors.*/
static void oc_set_chroma_mvs11(oc_mv _cbmvs[4],const oc_mv _lbmvs[4]){
memcpy(_cbmvs,_lbmvs,4*sizeof(_lbmvs[0]));
}
/*A table of functions used to fill in the chroma plane motion vectors for a
macro block when 4 different motion vectors are specified in the luma
plane.*/
const oc_set_chroma_mvs_func OC_SET_CHROMA_MVS_TABLE[TH_PF_NFORMATS]={
(oc_set_chroma_mvs_func)oc_set_chroma_mvs00,
(oc_set_chroma_mvs_func)oc_set_chroma_mvs01,
(oc_set_chroma_mvs_func)oc_set_chroma_mvs10,
(oc_set_chroma_mvs_func)oc_set_chroma_mvs11
};
void **oc_malloc_2d(size_t _height,size_t _width,size_t _sz){
size_t rowsz;

445
media/libtheora/lib/internal.h
View File

@ -11,7 +11,7 @@
********************************************************************
function:
last mod: $Id: internal.h 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: internal.h 17578 2010-10-29 04:21:26Z tterribe $
********************************************************************/
#if !defined(_internal_H)
@ -19,10 +19,20 @@
# include <stdlib.h>
# include <limits.h>
# if defined(HAVE_CONFIG_H)
# include <config.h>
# include "config.h"
# endif
# include "theora/codec.h"
# include "theora/theora.h"
# include "ocintrin.h"
# if !defined(__GNUC_PREREQ)
# if defined(__GNUC__)&&defined(__GNUC_MINOR__)
# define __GNUC_PREREQ(_maj,_min) \
((__GNUC__<<16)+__GNUC_MINOR__>=((_maj)<<16)+(_min))
# else
# define __GNUC_PREREQ(_maj,_min) 0
# endif
# endif
# if defined(_MSC_VER)
/*Disable missing EMMS warnings.*/
@ -31,24 +41,25 @@
# pragma warning(disable:4554)
# endif
/*You, too, gcc.*/
# if defined(__GNUC_PREREQ)
# if __GNUC_PREREQ(4,2)
# pragma GCC diagnostic ignored "-Wparentheses"
# endif
# if __GNUC_PREREQ(4,2)
# pragma GCC diagnostic ignored "-Wparentheses"
# endif
# include "ocintrin.h"
# include "huffman.h"
# include "quant.h"
/*Some assembly constructs require aligned operands.*/
# if defined(OC_X86_ASM)
/*Some assembly constructs require aligned operands.
The following macros are _only_ intended for structure member declarations.
Although they will sometimes work on stack variables, gcc will often silently
ignore them.
A separate set of macros could be made for manual stack alignment, but we
don't actually require it anywhere.*/
# if defined(OC_X86_ASM)||defined(OC_ARM_ASM)
# if defined(__GNUC__)
# define OC_ALIGN8(expr) expr __attribute__((aligned(8)))
# define OC_ALIGN16(expr) expr __attribute__((aligned(16)))
# elif defined(_MSC_VER)
# define OC_ALIGN8(expr) __declspec (align(8)) expr
# define OC_ALIGN16(expr) __declspec (align(16)) expr
# else
# error "Alignment macros required for this platform."
# endif
# endif
# if !defined(OC_ALIGN8)
@ -60,19 +71,8 @@
typedef struct oc_sb_flags oc_sb_flags;
typedef struct oc_border_info oc_border_info;
typedef struct oc_fragment oc_fragment;
typedef struct oc_fragment_plane oc_fragment_plane;
typedef struct oc_base_opt_vtable oc_base_opt_vtable;
typedef struct oc_base_opt_data oc_base_opt_data;
typedef struct oc_state_dispatch_vtable oc_state_dispatch_vtable;
typedef struct oc_theora_state oc_theora_state;
/*This library's version.*/
# define OC_VENDOR_STRING "Xiph.Org libtheora 1.1 20090822 (Thusnelda)"
# define OC_VENDOR_STRING "Xiph.Org libtheora 1.2.0alpha 20100924 (Ptalarbvorm)"
/*Theora bitstream version.*/
# define TH_VERSION_MAJOR (3)
@ -83,315 +83,6 @@ typedef struct oc_theora_state oc_theora_state;
((_info)->version_minor>(_min)||(_info)->version_minor==(_min)&& \
(_info)->version_subminor>=(_sub)))
/*A keyframe.*/
#define OC_INTRA_FRAME (0)
/*A predicted frame.*/
#define OC_INTER_FRAME (1)
/*A frame of unknown type (frame type decision has not yet been made).*/
#define OC_UNKWN_FRAME (-1)
/*The amount of padding to add to the reconstructed frame buffers on all
sides.
This is used to allow unrestricted motion vectors without special casing.
This must be a multiple of 2.*/
#define OC_UMV_PADDING (16)
/*Frame classification indices.*/
/*The previous golden frame.*/
#define OC_FRAME_GOLD (0)
/*The previous frame.*/
#define OC_FRAME_PREV (1)
/*The current frame.*/
#define OC_FRAME_SELF (2)
/*The input or output buffer.*/
#define OC_FRAME_IO (3)
/*Macroblock modes.*/
/*Macro block is invalid: It is never coded.*/
#define OC_MODE_INVALID (-1)
/*Encoded difference from the same macro block in the previous frame.*/
#define OC_MODE_INTER_NOMV (0)
/*Encoded with no motion compensated prediction.*/
#define OC_MODE_INTRA (1)
/*Encoded difference from the previous frame offset by the given motion
vector.*/
#define OC_MODE_INTER_MV (2)
/*Encoded difference from the previous frame offset by the last coded motion
vector.*/
#define OC_MODE_INTER_MV_LAST (3)
/*Encoded difference from the previous frame offset by the second to last
coded motion vector.*/
#define OC_MODE_INTER_MV_LAST2 (4)
/*Encoded difference from the same macro block in the previous golden
frame.*/
#define OC_MODE_GOLDEN_NOMV (5)
/*Encoded difference from the previous golden frame offset by the given motion
vector.*/
#define OC_MODE_GOLDEN_MV (6)
/*Encoded difference from the previous frame offset by the individual motion
vectors given for each block.*/
#define OC_MODE_INTER_MV_FOUR (7)
/*The number of (coded) modes.*/
#define OC_NMODES (8)
/*Determines the reference frame used for a given MB mode.*/
#define OC_FRAME_FOR_MODE(_x) \
OC_UNIBBLE_TABLE32(OC_FRAME_PREV,OC_FRAME_SELF,OC_FRAME_PREV,OC_FRAME_PREV, \
OC_FRAME_PREV,OC_FRAME_GOLD,OC_FRAME_GOLD,OC_FRAME_PREV,(_x))
/*Constants for the packet state machine common between encoder and decoder.*/
/*Next packet to emit/read: Codec info header.*/
#define OC_PACKET_INFO_HDR (-3)
/*Next packet to emit/read: Comment header.*/
#define OC_PACKET_COMMENT_HDR (-2)
/*Next packet to emit/read: Codec setup header.*/
#define OC_PACKET_SETUP_HDR (-1)
/*No more packets to emit/read.*/
#define OC_PACKET_DONE (INT_MAX)
/*Super blocks are 32x32 segments of pixels in a single color plane indexed
in image order.
Internally, super blocks are broken up into four quadrants, each of which
contains a 2x2 pattern of blocks, each of which is an 8x8 block of pixels.
Quadrants, and the blocks within them, are indexed in a special order called
a "Hilbert curve" within the super block.
In order to differentiate between the Hilbert-curve indexing strategy and
the regular image order indexing strategy, blocks indexed in image order
are called "fragments".
Fragments are indexed in image order, left to right, then bottom to top,
from Y' plane to Cb plane to Cr plane.
The co-located fragments in all image planes corresponding to the location
of a single quadrant of a luma plane super block form a macro block.
Thus there is only a single set of macro blocks for all planes, each of which
contains between 6 and 12 fragments, depending on the pixel format.
Therefore macro block information is kept in a separate set of arrays from
super blocks to avoid unused space in the other planes.
The lists are indexed in super block order.
That is, the macro block corresponding to the macro block mbi in (luma plane)
super block sbi is at index (sbi<<2|mbi).
Thus the number of macro blocks in each dimension is always twice the number
of super blocks, even when only an odd number fall inside the coded frame.
These "extra" macro blocks are just an artifact of our internal data layout,
and not part of the coded stream; they are flagged with a negative MB mode.*/
/*A single quadrant of the map from a super block to fragment numbers.*/
typedef ptrdiff_t oc_sb_map_quad[4];
/*A map from a super block to fragment numbers.*/
typedef oc_sb_map_quad oc_sb_map[4];
/*A single plane of the map from a macro block to fragment numbers.*/
typedef ptrdiff_t oc_mb_map_plane[4];
/*A map from a macro block to fragment numbers.*/
typedef oc_mb_map_plane oc_mb_map[3];
/*A motion vector.*/
typedef signed char oc_mv[2];
/*Super block information.*/
struct oc_sb_flags{
unsigned char coded_fully:1;
unsigned char coded_partially:1;
unsigned char quad_valid:4;
};
/*Information about a fragment which intersects the border of the displayable
region.
This marks which pixels belong to the displayable region.*/
struct oc_border_info{
/*A bit mask marking which pixels are in the displayable region.
Pixel (x,y) corresponds to bit (y<<3|x).*/
ogg_int64_t mask;
/*The number of pixels in the displayable region.
This is always positive, and always less than 64.*/
int npixels;
};
/*Fragment information.*/
struct oc_fragment{
/*A flag indicating whether or not this fragment is coded.*/
unsigned coded:1;
/*A flag indicating that this entire fragment lies outside the displayable
region of the frame.
Note the contrast with an invalid macro block, which is outside the coded
frame, not just the displayable one.
There are no fragments outside the coded frame by construction.*/
unsigned invalid:1;
/*The index of the quality index used for this fragment's AC coefficients.*/
unsigned qii:6;
/*The mode of the macroblock this fragment belongs to.*/
unsigned mb_mode:3;
/*The index of the associated border information for fragments which lie
partially outside the displayable region.
For fragments completely inside or outside this region, this is -1.
Note that the C standard requires an explicit signed keyword for bitfield
types, since some compilers may treat them as unsigned without it.*/
signed int borderi:5;
/*The prediction-corrected DC component.
Note that the C standard requires an explicit signed keyword for bitfield
types, since some compilers may treat them as unsigned without it.*/
signed int dc:16;
};
/*A description of each fragment plane.*/
struct oc_fragment_plane{
/*The number of fragments in the horizontal direction.*/
int nhfrags;
/*The number of fragments in the vertical direction.*/
int nvfrags;
/*The offset of the first fragment in the plane.*/
ptrdiff_t froffset;
/*The total number of fragments in the plane.*/
ptrdiff_t nfrags;
/*The number of super blocks in the horizontal direction.*/
unsigned nhsbs;
/*The number of super blocks in the vertical direction.*/
unsigned nvsbs;
/*The offset of the first super block in the plane.*/
unsigned sboffset;
/*The total number of super blocks in the plane.*/
unsigned nsbs;
};
/*The shared (encoder and decoder) functions that have accelerated variants.*/
struct oc_base_opt_vtable{
void (*frag_copy)(unsigned char *_dst,
const unsigned char *_src,int _ystride);
void (*frag_recon_intra)(unsigned char *_dst,int _ystride,
const ogg_int16_t _residue[64]);
void (*frag_recon_inter)(unsigned char *_dst,
const unsigned char *_src,int _ystride,const ogg_int16_t _residue[64]);
void (*frag_recon_inter2)(unsigned char *_dst,const unsigned char *_src1,
const unsigned char *_src2,int _ystride,const ogg_int16_t _residue[64]);
void (*idct8x8)(ogg_int16_t _y[64],int _last_zzi);
void (*state_frag_recon)(const oc_theora_state *_state,ptrdiff_t _fragi,
int _pli,ogg_int16_t _dct_coeffs[64],int _last_zzi,ogg_uint16_t _dc_quant);
void (*state_frag_copy_list)(const oc_theora_state *_state,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,
int _dst_frame,int _src_frame,int _pli);
void (*state_loop_filter_frag_rows)(const oc_theora_state *_state,
int _bv[256],int _refi,int _pli,int _fragy0,int _fragy_end);
void (*restore_fpu)(void);
};
/*The shared (encoder and decoder) tables that vary according to which variants
of the above functions are used.*/
struct oc_base_opt_data{
const unsigned char *dct_fzig_zag;
};
/*State information common to both the encoder and decoder.*/
struct oc_theora_state{
/*The stream information.*/
th_info info;
/*Table for shared accelerated functions.*/
oc_base_opt_vtable opt_vtable;
/*Table for shared data used by accelerated functions.*/
oc_base_opt_data opt_data;
/*CPU flags to detect the presence of extended instruction sets.*/
ogg_uint32_t cpu_flags;
/*The fragment plane descriptions.*/
oc_fragment_plane fplanes[3];
/*The list of fragments, indexed in image order.*/
oc_fragment *frags;
/*The the offset into the reference frame buffer to the upper-left pixel of
each fragment.*/
ptrdiff_t *frag_buf_offs;
/*The motion vector for each fragment.*/
oc_mv *frag_mvs;
/*The total number of fragments in a single frame.*/
ptrdiff_t nfrags;
/*The list of super block maps, indexed in image order.*/
oc_sb_map *sb_maps;
/*The list of super block flags, indexed in image order.*/
oc_sb_flags *sb_flags;
/*The total number of super blocks in a single frame.*/
unsigned nsbs;
/*The fragments from each color plane that belong to each macro block.
Fragments are stored in image order (left to right then top to bottom).
When chroma components are decimated, the extra fragments have an index of
-1.*/
oc_mb_map *mb_maps;
/*The list of macro block modes.
A negative number indicates the macro block lies entirely outside the
coded frame.*/
signed char *mb_modes;
/*The number of macro blocks in the X direction.*/
unsigned nhmbs;
/*The number of macro blocks in the Y direction.*/
unsigned nvmbs;
/*The total number of macro blocks.*/
size_t nmbs;
/*The list of coded fragments, in coded order.
Uncoded fragments are stored in reverse order from the end of the list.*/
ptrdiff_t *coded_fragis;
/*The number of coded fragments in each plane.*/
ptrdiff_t ncoded_fragis[3];
/*The total number of coded fragments.*/
ptrdiff_t ntotal_coded_fragis;
/*The index of the buffers being used for each OC_FRAME_* reference frame.*/
int ref_frame_idx[4];
/*The actual buffers used for the previously decoded frames.*/
th_ycbcr_buffer ref_frame_bufs[4];
/*The storage for the reference frame buffers.*/
unsigned char *ref_frame_data[4];
/*The strides for each plane in the reference frames.*/
int ref_ystride[3];
/*The number of unique border patterns.*/
int nborders;
/*The unique border patterns for all border fragments.
The borderi field of fragments which straddle the border indexes this
list.*/
oc_border_info borders[16];
/*The frame number of the last keyframe.*/
ogg_int64_t keyframe_num;
/*The frame number of the current frame.*/
ogg_int64_t curframe_num;
/*The granpos of the current frame.*/
ogg_int64_t granpos;
/*The type of the current frame.*/
unsigned char frame_type;
/*The bias to add to the frame count when computing granule positions.*/
unsigned char granpos_bias;
/*The number of quality indices used in the current frame.*/
unsigned char nqis;
/*The quality indices of the current frame.*/
unsigned char qis[3];
/*The dequantization tables, stored in zig-zag order, and indexed by
qi, pli, qti, and zzi.*/
ogg_uint16_t *dequant_tables[64][3][2];
OC_ALIGN16(oc_quant_table dequant_table_data[64][3][2]);
/*Loop filter strength parameters.*/
unsigned char loop_filter_limits[64];
};
/*The function type used to fill in the chroma plane motion vectors for a
macro block when 4 different motion vectors are specified in the luma
plane.
_cbmvs: The chroma block-level motion vectors to fill in.
_lmbmv: The luma macro-block level motion vector to fill in for use in
prediction.
_lbmvs: The luma block-level motion vectors.*/
typedef void (*oc_set_chroma_mvs_func)(oc_mv _cbmvs[4],const oc_mv _lbmvs[4]);
/*A map from the index in the zig zag scan to the coefficient number in a
@ -409,14 +100,12 @@ extern const unsigned char OC_MB_MAP_IDXS[TH_PF_NFORMATS][12];
/*The number of indices in the oc_mb_map array that can be valid for each of
the various chroma decimation types.*/
extern const unsigned char OC_MB_MAP_NIDXS[TH_PF_NFORMATS];
/*A table of functions used to fill in the Cb,Cr plane motion vectors for a
macro block when 4 different motion vectors are specified in the luma
plane.*/
extern const oc_set_chroma_mvs_func OC_SET_CHROMA_MVS_TABLE[TH_PF_NFORMATS];
int oc_ilog(unsigned _v);
void *oc_aligned_malloc(size_t _sz,size_t _align);
void oc_aligned_free(void *_ptr);
void **oc_malloc_2d(size_t _height,size_t _width,size_t _sz);
void **oc_calloc_2d(size_t _height,size_t _width,size_t _sz);
void oc_free_2d(void *_ptr);
@ -424,86 +113,4 @@ void oc_free_2d(void *_ptr);
void oc_ycbcr_buffer_flip(th_ycbcr_buffer _dst,
const th_ycbcr_buffer _src);
int oc_state_init(oc_theora_state *_state,const th_info *_info,int _nrefs);
void oc_state_clear(oc_theora_state *_state);
void oc_state_vtable_init_c(oc_theora_state *_state);
void oc_state_borders_fill_rows(oc_theora_state *_state,int _refi,int _pli,
int _y0,int _yend);
void oc_state_borders_fill_caps(oc_theora_state *_state,int _refi,int _pli);
void oc_state_borders_fill(oc_theora_state *_state,int _refi);
void oc_state_fill_buffer_ptrs(oc_theora_state *_state,int _buf_idx,
th_ycbcr_buffer _img);
int oc_state_mbi_for_pos(oc_theora_state *_state,int _mbx,int _mby);
int oc_state_get_mv_offsets(const oc_theora_state *_state,int _offsets[2],
int _pli,int _dx,int _dy);
int oc_state_loop_filter_init(oc_theora_state *_state,int *_bv);
void oc_state_loop_filter(oc_theora_state *_state,int _frame);
#if defined(OC_DUMP_IMAGES)
int oc_state_dump_frame(const oc_theora_state *_state,int _frame,
const char *_suf);
#endif
/*Shared accelerated functions.*/
void oc_frag_copy(const oc_theora_state *_state,unsigned char *_dst,
const unsigned char *_src,int _ystride);
void oc_frag_recon_intra(const oc_theora_state *_state,
unsigned char *_dst,int _dst_ystride,const ogg_int16_t _residue[64]);
void oc_frag_recon_inter(const oc_theora_state *_state,unsigned char *_dst,
const unsigned char *_src,int _ystride,const ogg_int16_t _residue[64]);
void oc_frag_recon_inter2(const oc_theora_state *_state,
unsigned char *_dst,const unsigned char *_src1,const unsigned char *_src2,
int _ystride,const ogg_int16_t _residue[64]);
void oc_idct8x8(const oc_theora_state *_state,ogg_int16_t _y[64],int _last_zzi);
void oc_state_frag_recon(const oc_theora_state *_state,ptrdiff_t _fragi,
int _pli,ogg_int16_t _dct_coeffs[64],int _last_zzi,ogg_uint16_t _dc_quant);
void oc_state_frag_copy_list(const oc_theora_state *_state,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,
int _dst_frame,int _src_frame,int _pli);
void oc_state_loop_filter_frag_rows(const oc_theora_state *_state,
int _bv[256],int _refi,int _pli,int _fragy0,int _fragy_end);
void oc_restore_fpu(const oc_theora_state *_state);
/*Default pure-C implementations.*/
void oc_frag_copy_c(unsigned char *_dst,
const unsigned char *_src,int _src_ystride);
void oc_frag_recon_intra_c(unsigned char *_dst,int _dst_ystride,
const ogg_int16_t _residue[64]);
void oc_frag_recon_inter_c(unsigned char *_dst,
const unsigned char *_src,int _ystride,const ogg_int16_t _residue[64]);
void oc_frag_recon_inter2_c(unsigned char *_dst,const unsigned char *_src1,
const unsigned char *_src2,int _ystride,const ogg_int16_t _residue[64]);
void oc_idct8x8_c(ogg_int16_t _y[64],int _last_zzi);
void oc_state_frag_recon_c(const oc_theora_state *_state,ptrdiff_t _fragi,
int _pli,ogg_int16_t _dct_coeffs[64],int _last_zzi,ogg_uint16_t _dc_quant);
void oc_state_frag_copy_list_c(const oc_theora_state *_state,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,
int _dst_frame,int _src_frame,int _pli);
void oc_state_loop_filter_frag_rows_c(const oc_theora_state *_state,
int _bv[256],int _refi,int _pli,int _fragy0,int _fragy_end);
void oc_restore_fpu_c(void);
/*We need a way to call a few encoder functions without introducing a link-time
dependency into the decoder, while still allowing the old alpha API which
does not distinguish between encoder and decoder objects to be used.
We do this by placing a function table at the start of the encoder object
which can dispatch into the encoder library.
We do a similar thing for the decoder in case we ever decide to split off a
common base library.*/
typedef void (*oc_state_clear_func)(theora_state *_th);
typedef int (*oc_state_control_func)(theora_state *th,int _req,
void *_buf,size_t _buf_sz);
typedef ogg_int64_t (*oc_state_granule_frame_func)(theora_state *_th,
ogg_int64_t _granulepos);
typedef double (*oc_state_granule_time_func)(theora_state *_th,
ogg_int64_t _granulepos);
struct oc_state_dispatch_vtable{
oc_state_clear_func clear;
oc_state_control_func control;
oc_state_granule_frame_func granule_frame;
oc_state_granule_time_func granule_time;
};
#endif

44
media/libtheora/lib/mathops.h
View File

@ -2,29 +2,27 @@
# define _mathops_H (1)
# include <ogg/ogg.h>
# ifdef __GNUC_PREREQ
# if __GNUC_PREREQ(3,4)
# include <limits.h>
# if __GNUC_PREREQ(3,4)
# include <limits.h>
/*Note the casts to (int) below: this prevents OC_CLZ{32|64}_OFFS from
"upgrading" the type of an entire expression to an (unsigned) size_t.*/
# if INT_MAX>=2147483647
# define OC_CLZ32_OFFS ((int)sizeof(unsigned)*CHAR_BIT)
# define OC_CLZ32(_x) (__builtin_clz(_x))
# elif LONG_MAX>=2147483647L
# define OC_CLZ32_OFFS ((int)sizeof(unsigned long)*CHAR_BIT)
# define OC_CLZ32(_x) (__builtin_clzl(_x))
# endif
# if INT_MAX>=9223372036854775807LL
# define OC_CLZ64_OFFS ((int)sizeof(unsigned)*CHAR_BIT)
# define OC_CLZ64(_x) (__builtin_clz(_x))
# elif LONG_MAX>=9223372036854775807LL
# define OC_CLZ64_OFFS ((int)sizeof(unsigned long)*CHAR_BIT)
# define OC_CLZ64(_x) (__builtin_clzl(_x))
# elif LLONG_MAX>=9223372036854775807LL|| \
__LONG_LONG_MAX__>=9223372036854775807LL
# define OC_CLZ64_OFFS ((int)sizeof(unsigned long long)*CHAR_BIT)
# define OC_CLZ64(_x) (__builtin_clzll(_x))
# endif
# if INT_MAX>=2147483647
# define OC_CLZ32_OFFS ((int)sizeof(unsigned)*CHAR_BIT)
# define OC_CLZ32(_x) (__builtin_clz(_x))
# elif LONG_MAX>=2147483647L
# define OC_CLZ32_OFFS ((int)sizeof(unsigned long)*CHAR_BIT)
# define OC_CLZ32(_x) (__builtin_clzl(_x))
# endif
# if INT_MAX>=9223372036854775807LL
# define OC_CLZ64_OFFS ((int)sizeof(unsigned)*CHAR_BIT)
# define OC_CLZ64(_x) (__builtin_clz(_x))
# elif LONG_MAX>=9223372036854775807LL
# define OC_CLZ64_OFFS ((int)sizeof(unsigned long)*CHAR_BIT)
# define OC_CLZ64(_x) (__builtin_clzl(_x))
# elif LLONG_MAX>=9223372036854775807LL|| \
__LONG_LONG_MAX__>=9223372036854775807LL
# define OC_CLZ64_OFFS ((int)sizeof(unsigned long long)*CHAR_BIT)
# define OC_CLZ64(_x) (__builtin_clzll(_x))
# endif
# endif
@ -134,8 +132,12 @@ int oc_ilog64(ogg_int64_t _v);
# define OC_STATIC_ILOG_64(_v) (OC_STATIC_ILOG6((ogg_int64_t)(_v)))
#define OC_Q57(_v) ((ogg_int64_t)(_v)<<57)
#define OC_Q10(_v) ((_v)<<10)
ogg_int64_t oc_bexp64(ogg_int64_t _z);
ogg_int64_t oc_blog64(ogg_int64_t _w);
ogg_uint32_t oc_bexp32_q10(int _z);
int oc_blog32_q10(ogg_uint32_t _w);
#endif

10
media/libtheora/lib/quant.c
View File

@ -11,7 +11,7 @@
********************************************************************
function:
last mod: $Id: quant.c 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: quant.c 17307 2010-06-27 06:02:15Z tterribe $
********************************************************************/
@ -21,6 +21,14 @@
#include "quant.h"
#include "decint.h"
/*The maximum output of the DCT with +/- 255 inputs is +/- 8157.
These minimum quantizers ensure the result after quantization (and after
prediction for DC) will be no more than +/- 510.
The tokenization system can handle values up to +/- 580, so there is no need
to do any coefficient clamping.
I would rather have allowed smaller quantizers and had to clamp, but these
minimums were required when constructing the original VP3 matrices and have
been formalized in the spec.*/
static const unsigned OC_DC_QUANT_MIN[2]={4<<2,8<<2};
static const unsigned OC_AC_QUANT_MIN[2]={2<<2,4<<2};

277
media/libtheora/lib/state.c
View File

@ -11,25 +11,92 @@
********************************************************************
function:
last mod: $Id: state.c 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: state.c 17576 2010-10-29 01:07:51Z tterribe $
********************************************************************/
#include <stdlib.h>
#include <string.h>
#include "internal.h"
#if defined(OC_X86_ASM)
#if defined(_MSC_VER)
# include "x86_vc/x86int.h"
#else
# include "x86/x86int.h"
#endif
#endif
#include "state.h"
#if defined(OC_DUMP_IMAGES)
# include <stdio.h>
# include "png.h"
#endif
/*The function used to fill in the chroma plane motion vectors for a macro
block when 4 different motion vectors are specified in the luma plane.
This version is for use with chroma decimated in the X and Y directions
(4:2:0).
_cbmvs: The chroma block-level motion vectors to fill in.
_lbmvs: The luma block-level motion vectors.*/
static void oc_set_chroma_mvs00(oc_mv _cbmvs[4],const oc_mv _lbmvs[4]){
int dx;
int dy;
dx=OC_MV_X(_lbmvs[0])+OC_MV_X(_lbmvs[1])
+OC_MV_X(_lbmvs[2])+OC_MV_X(_lbmvs[3]);
dy=OC_MV_Y(_lbmvs[0])+OC_MV_Y(_lbmvs[1])
+OC_MV_Y(_lbmvs[2])+OC_MV_Y(_lbmvs[3]);
_cbmvs[0]=OC_MV(OC_DIV_ROUND_POW2(dx,2,2),OC_DIV_ROUND_POW2(dy,2,2));
}
/*The function used to fill in the chroma plane motion vectors for a macro
block when 4 different motion vectors are specified in the luma plane.
This version is for use with chroma decimated in the Y direction.
_cbmvs: The chroma block-level motion vectors to fill in.
_lbmvs: The luma block-level motion vectors.*/
static void oc_set_chroma_mvs01(oc_mv _cbmvs[4],const oc_mv _lbmvs[4]){
int dx;
int dy;
dx=OC_MV_X(_lbmvs[0])+OC_MV_X(_lbmvs[2]);
dy=OC_MV_Y(_lbmvs[0])+OC_MV_Y(_lbmvs[2]);
_cbmvs[0]=OC_MV(OC_DIV_ROUND_POW2(dx,1,1),OC_DIV_ROUND_POW2(dy,1,1));
dx=OC_MV_X(_lbmvs[1])+OC_MV_X(_lbmvs[3]);
dy=OC_MV_Y(_lbmvs[1])+OC_MV_Y(_lbmvs[3]);
_cbmvs[1]=OC_MV(OC_DIV_ROUND_POW2(dx,1,1),OC_DIV_ROUND_POW2(dy,1,1));
}
/*The function used to fill in the chroma plane motion vectors for a macro
block when 4 different motion vectors are specified in the luma plane.
This version is for use with chroma decimated in the X direction (4:2:2).
_cbmvs: The chroma block-level motion vectors to fill in.
_lbmvs: The luma block-level motion vectors.*/
static void oc_set_chroma_mvs10(oc_mv _cbmvs[4],const oc_mv _lbmvs[4]){
int dx;
int dy;
dx=OC_MV_X(_lbmvs[0])+OC_MV_X(_lbmvs[1]);
dy=OC_MV_Y(_lbmvs[0])+OC_MV_Y(_lbmvs[1]);
_cbmvs[0]=OC_MV(OC_DIV_ROUND_POW2(dx,1,1),OC_DIV_ROUND_POW2(dy,1,1));
dx=OC_MV_X(_lbmvs[2])+OC_MV_X(_lbmvs[3]);
dy=OC_MV_Y(_lbmvs[2])+OC_MV_Y(_lbmvs[3]);
_cbmvs[2]=OC_MV(OC_DIV_ROUND_POW2(dx,1,1),OC_DIV_ROUND_POW2(dy,1,1));
}
/*The function used to fill in the chroma plane motion vectors for a macro
block when 4 different motion vectors are specified in the luma plane.
This version is for use with no chroma decimation (4:4:4).
_cbmvs: The chroma block-level motion vectors to fill in.
_lmbmv: The luma macro-block level motion vector to fill in for use in
prediction.
_lbmvs: The luma block-level motion vectors.*/
static void oc_set_chroma_mvs11(oc_mv _cbmvs[4],const oc_mv _lbmvs[4]){
_cbmvs[0]=_lbmvs[0];
_cbmvs[1]=_lbmvs[1];
_cbmvs[2]=_lbmvs[2];
_cbmvs[3]=_lbmvs[3];
}
/*A table of functions used to fill in the chroma plane motion vectors for a
macro block when 4 different motion vectors are specified in the luma
plane.*/
const oc_set_chroma_mvs_func OC_SET_CHROMA_MVS_TABLE[TH_PF_NFORMATS]={
(oc_set_chroma_mvs_func)oc_set_chroma_mvs00,
(oc_set_chroma_mvs_func)oc_set_chroma_mvs01,
(oc_set_chroma_mvs_func)oc_set_chroma_mvs10,
(oc_set_chroma_mvs_func)oc_set_chroma_mvs11
};
/*Returns the fragment index of the top-left block in a macro block.
This can be used to test whether or not the whole macro block is valid.
_sb_map: The super block map.
@ -469,7 +536,7 @@ static void oc_state_frarray_clear(oc_theora_state *_state){
unrestricted motion vectors without special casing the boundary.
If chroma is decimated in either direction, the padding is reduced by a
factor of 2 on the appropriate sides.
_nrefs: The number of reference buffers to init; must be 3 or 4.*/
_nrefs: The number of reference buffers to init; must be in the range 3...6.*/
static int oc_state_ref_bufs_init(oc_theora_state *_state,int _nrefs){
th_info *info;
unsigned char *ref_frame_data;
@ -481,6 +548,7 @@ static int oc_state_ref_bufs_init(oc_theora_state *_state,int _nrefs){
int yheight;
int chstride;
int cheight;
ptrdiff_t align;
ptrdiff_t yoffset;
ptrdiff_t coffset;
ptrdiff_t *frag_buf_offs;
@ -489,28 +557,33 @@ static int oc_state_ref_bufs_init(oc_theora_state *_state,int _nrefs){
int vdec;
int rfi;
int pli;
if(_nrefs<3||_nrefs>4)return TH_EINVAL;
if(_nrefs<3||_nrefs>6)return TH_EINVAL;
info=&_state->info;
/*Compute the image buffer parameters for each plane.*/
hdec=!(info->pixel_fmt&1);
vdec=!(info->pixel_fmt&2);
yhstride=info->frame_width+2*OC_UMV_PADDING;
yheight=info->frame_height+2*OC_UMV_PADDING;
chstride=yhstride>>hdec;
/*Require 16-byte aligned rows in the chroma planes.*/
chstride=(yhstride>>hdec)+15&~15;
cheight=yheight>>vdec;
yplane_sz=yhstride*(size_t)yheight;
cplane_sz=chstride*(size_t)cheight;
yoffset=OC_UMV_PADDING+OC_UMV_PADDING*(ptrdiff_t)yhstride;
coffset=(OC_UMV_PADDING>>hdec)+(OC_UMV_PADDING>>vdec)*(ptrdiff_t)chstride;
ref_frame_sz=yplane_sz+2*cplane_sz;
/*Although we guarantee the rows of the chroma planes are a multiple of 16
bytes, the initial padding on the first row may only be 8 bytes.
Compute the offset needed to the actual image data to a multiple of 16.*/
align=-coffset&15;
ref_frame_sz=yplane_sz+2*cplane_sz+16;
ref_frame_data_sz=_nrefs*ref_frame_sz;
/*Check for overflow.
The same caveats apply as for oc_state_frarray_init().*/
if(yplane_sz/yhstride!=yheight||2*cplane_sz<cplane_sz||
if(yplane_sz/yhstride!=(size_t)yheight||2*cplane_sz+16<cplane_sz||
ref_frame_sz<yplane_sz||ref_frame_data_sz/_nrefs!=ref_frame_sz){
return TH_EIMPL;
}
ref_frame_data=_ogg_malloc(ref_frame_data_sz);
ref_frame_data=oc_aligned_malloc(ref_frame_data_sz,16);
frag_buf_offs=_state->frag_buf_offs=
_ogg_malloc(_state->nfrags*sizeof(*frag_buf_offs));
if(ref_frame_data==NULL||frag_buf_offs==NULL){
@ -532,15 +605,15 @@ static int oc_state_ref_bufs_init(oc_theora_state *_state,int _nrefs){
memcpy(_state->ref_frame_bufs[rfi],_state->ref_frame_bufs[0],
sizeof(_state->ref_frame_bufs[0]));
}
_state->ref_frame_handle=ref_frame_data;
/*Set up the data pointers for the image buffers.*/
for(rfi=0;rfi<_nrefs;rfi++){
_state->ref_frame_data[rfi]=ref_frame_data;
_state->ref_frame_bufs[rfi][0].data=ref_frame_data+yoffset;
ref_frame_data+=yplane_sz;
ref_frame_data+=yplane_sz+align;
_state->ref_frame_bufs[rfi][1].data=ref_frame_data+coffset;
ref_frame_data+=cplane_sz;
_state->ref_frame_bufs[rfi][2].data=ref_frame_data+coffset;
ref_frame_data+=cplane_sz;
ref_frame_data+=cplane_sz+(16-align);
/*Flip the buffer upside down.
This allows us to decode Theora's bottom-up frames in their natural
order, yet return a top-down buffer with a positive stride to the user.*/
@ -550,7 +623,7 @@ static int oc_state_ref_bufs_init(oc_theora_state *_state,int _nrefs){
_state->ref_ystride[0]=-yhstride;
_state->ref_ystride[1]=_state->ref_ystride[2]=-chstride;
/*Initialize the fragment buffer offsets.*/
ref_frame_data=_state->ref_frame_data[0];
ref_frame_data=_state->ref_frame_bufs[0][0].data;
fragi=0;
for(pli=0;pli<3;pli++){
th_img_plane *iplane;
@ -576,41 +649,44 @@ static int oc_state_ref_bufs_init(oc_theora_state *_state,int _nrefs){
vpix+=stride<<3;
}
}
/*Initialize the reference frame indices.*/
/*Initialize the reference frame pointers and indices.*/
_state->ref_frame_idx[OC_FRAME_GOLD]=
_state->ref_frame_idx[OC_FRAME_PREV]=
_state->ref_frame_idx[OC_FRAME_SELF]=-1;
_state->ref_frame_idx[OC_FRAME_IO]=_nrefs>3?3:-1;
_state->ref_frame_idx[OC_FRAME_GOLD_ORIG]=
_state->ref_frame_idx[OC_FRAME_PREV_ORIG]=
_state->ref_frame_idx[OC_FRAME_SELF]=
_state->ref_frame_idx[OC_FRAME_IO]=-1;
_state->ref_frame_data[OC_FRAME_GOLD]=
_state->ref_frame_data[OC_FRAME_PREV]=
_state->ref_frame_data[OC_FRAME_GOLD_ORIG]=
_state->ref_frame_data[OC_FRAME_PREV_ORIG]=
_state->ref_frame_data[OC_FRAME_SELF]=
_state->ref_frame_data[OC_FRAME_IO]=NULL;
return 0;
}
static void oc_state_ref_bufs_clear(oc_theora_state *_state){
_ogg_free(_state->frag_buf_offs);
_ogg_free(_state->ref_frame_data[0]);
oc_aligned_free(_state->ref_frame_handle);
}
void oc_state_vtable_init_c(oc_theora_state *_state){
void oc_state_accel_init_c(oc_theora_state *_state){
_state->cpu_flags=0;
#if defined(OC_STATE_USE_VTABLE)
_state->opt_vtable.frag_copy=oc_frag_copy_c;
_state->opt_vtable.frag_copy_list=oc_frag_copy_list_c;
_state->opt_vtable.frag_recon_intra=oc_frag_recon_intra_c;
_state->opt_vtable.frag_recon_inter=oc_frag_recon_inter_c;
_state->opt_vtable.frag_recon_inter2=oc_frag_recon_inter2_c;
_state->opt_vtable.idct8x8=oc_idct8x8_c;
_state->opt_vtable.state_frag_recon=oc_state_frag_recon_c;
_state->opt_vtable.state_frag_copy_list=oc_state_frag_copy_list_c;
_state->opt_vtable.loop_filter_init=oc_loop_filter_init_c;
_state->opt_vtable.state_loop_filter_frag_rows=
oc_state_loop_filter_frag_rows_c;
_state->opt_vtable.restore_fpu=oc_restore_fpu_c;
_state->opt_data.dct_fzig_zag=OC_FZIG_ZAG;
}
/*Initialize the accelerated function pointers.*/
void oc_state_vtable_init(oc_theora_state *_state){
#if defined(OC_X86_ASM)
oc_state_vtable_init_x86(_state);
#else
oc_state_vtable_init_c(_state);
#endif
_state->opt_data.dct_fzig_zag=OC_FZIG_ZAG;
}
@ -648,7 +724,7 @@ int oc_state_init(oc_theora_state *_state,const th_info *_info,int _nrefs){
system.*/
_state->info.pic_y=_info->frame_height-_info->pic_height-_info->pic_y;
_state->frame_type=OC_UNKWN_FRAME;
oc_state_vtable_init(_state);
oc_state_accel_init(_state);
ret=oc_state_frarray_init(_state);
if(ret>=0)ret=oc_state_ref_bufs_init(_state,_nrefs);
if(ret<0){
@ -758,11 +834,10 @@ void oc_state_borders_fill(oc_theora_state *_state,int _refi){
_offsets[1] is set if the motion vector has non-zero fractional
components.
_pli: The color plane index.
_dx: The X component of the motion vector.
_dy: The Y component of the motion vector.
_mv: The motion vector.
Return: The number of offsets returned: 1 or 2.*/
int oc_state_get_mv_offsets(const oc_theora_state *_state,int _offsets[2],
int _pli,int _dx,int _dy){
int _pli,oc_mv _mv){
/*Here is a brief description of how Theora handles motion vectors:
Motion vector components are specified to half-pixel accuracy in
undecimated directions of each plane, and quarter-pixel accuracy in
@ -785,21 +860,25 @@ int oc_state_get_mv_offsets(const oc_theora_state *_state,int _offsets[2],
int xfrac;
int yfrac;
int offs;
int dx;
int dy;
ystride=_state->ref_ystride[_pli];
/*These two variables decide whether we are in half- or quarter-pixel
precision in each component.*/
xprec=1+(_pli!=0&&!(_state->info.pixel_fmt&1));
yprec=1+(_pli!=0&&!(_state->info.pixel_fmt&2));
dx=OC_MV_X(_mv);
dy=OC_MV_Y(_mv);
/*These two variables are either 0 if all the fractional bits are zero or -1
if any of them are non-zero.*/
xfrac=OC_SIGNMASK(-(_dx&(xprec|1)));
yfrac=OC_SIGNMASK(-(_dy&(yprec|1)));
offs=(_dx>>xprec)+(_dy>>yprec)*ystride;
xfrac=OC_SIGNMASK(-(dx&(xprec|1)));
yfrac=OC_SIGNMASK(-(dy&(yprec|1)));
offs=(dx>>xprec)+(dy>>yprec)*ystride;
if(xfrac||yfrac){
int xmask;
int ymask;
xmask=OC_SIGNMASK(_dx);
ymask=OC_SIGNMASK(_dy);
xmask=OC_SIGNMASK(dx);
ymask=OC_SIGNMASK(dy);
yfrac&=ystride;
_offsets[0]=offs-(xfrac&xmask)+(yfrac&ymask);
_offsets[1]=offs-(xfrac&~xmask)+(yfrac&~ymask);
@ -848,13 +927,17 @@ int oc_state_get_mv_offsets(const oc_theora_state *_state,int _offsets[2],
int mx2;
int my2;
int offs;
int dx;
int dy;
ystride=_state->ref_ystride[_pli];
qpy=_pli!=0&&!(_state->info.pixel_fmt&2);
my=OC_MVMAP[qpy][_dy+31];
my2=OC_MVMAP2[qpy][_dy+31];
dx=OC_MV_X(_mv);
dy=OC_MV_Y(_mv);
my=OC_MVMAP[qpy][dy+31];
my2=OC_MVMAP2[qpy][dy+31];
qpx=_pli!=0&&!(_state->info.pixel_fmt&1);
mx=OC_MVMAP[qpx][_dx+31];
mx2=OC_MVMAP2[qpx][_dx+31];
mx=OC_MVMAP[qpx][dx+31];
mx2=OC_MVMAP2[qpx][dx+31];
offs=my*ystride+mx;
if(mx2||my2){
_offsets[1]=offs+my2*ystride+mx2;
@ -866,18 +949,12 @@ int oc_state_get_mv_offsets(const oc_theora_state *_state,int _offsets[2],
#endif
}
void oc_state_frag_recon(const oc_theora_state *_state,ptrdiff_t _fragi,
int _pli,ogg_int16_t _dct_coeffs[64],int _last_zzi,ogg_uint16_t _dc_quant){
_state->opt_vtable.state_frag_recon(_state,_fragi,_pli,_dct_coeffs,
_last_zzi,_dc_quant);
}
void oc_state_frag_recon_c(const oc_theora_state *_state,ptrdiff_t _fragi,
int _pli,ogg_int16_t _dct_coeffs[64],int _last_zzi,ogg_uint16_t _dc_quant){
int _pli,ogg_int16_t _dct_coeffs[128],int _last_zzi,ogg_uint16_t _dc_quant){
unsigned char *dst;
ptrdiff_t frag_buf_off;
int ystride;
int mb_mode;
int refi;
/*Apply the inverse transform.*/
/*Special case only having a DC component.*/
if(_last_zzi<2){
@ -887,69 +964,35 @@ void oc_state_frag_recon_c(const oc_theora_state *_state,ptrdiff_t _fragi,
no iDCT rounding.*/
p=(ogg_int16_t)(_dct_coeffs[0]*(ogg_int32_t)_dc_quant+15>>5);
/*LOOP VECTORIZES.*/
for(ci=0;ci<64;ci++)_dct_coeffs[ci]=p;
for(ci=0;ci<64;ci++)_dct_coeffs[64+ci]=p;
}
else{
/*First, dequantize the DC coefficient.*/
_dct_coeffs[0]=(ogg_int16_t)(_dct_coeffs[0]*(int)_dc_quant);
oc_idct8x8(_state,_dct_coeffs,_last_zzi);
oc_idct8x8(_state,_dct_coeffs+64,_dct_coeffs,_last_zzi);
}
/*Fill in the target buffer.*/
frag_buf_off=_state->frag_buf_offs[_fragi];
mb_mode=_state->frags[_fragi].mb_mode;
refi=_state->frags[_fragi].refi;
ystride=_state->ref_ystride[_pli];
dst=_state->ref_frame_data[_state->ref_frame_idx[OC_FRAME_SELF]]+frag_buf_off;
if(mb_mode==OC_MODE_INTRA)oc_frag_recon_intra(_state,dst,ystride,_dct_coeffs);
dst=_state->ref_frame_data[OC_FRAME_SELF]+frag_buf_off;
if(refi==OC_FRAME_SELF)oc_frag_recon_intra(_state,dst,ystride,_dct_coeffs+64);
else{
const unsigned char *ref;
int mvoffsets[2];
ref=
_state->ref_frame_data[_state->ref_frame_idx[OC_FRAME_FOR_MODE(mb_mode)]]
+frag_buf_off;
ref=_state->ref_frame_data[refi]+frag_buf_off;
if(oc_state_get_mv_offsets(_state,mvoffsets,_pli,
_state->frag_mvs[_fragi][0],_state->frag_mvs[_fragi][1])>1){
_state->frag_mvs[_fragi])>1){
oc_frag_recon_inter2(_state,
dst,ref+mvoffsets[0],ref+mvoffsets[1],ystride,_dct_coeffs);
dst,ref+mvoffsets[0],ref+mvoffsets[1],ystride,_dct_coeffs+64);
}
else{
oc_frag_recon_inter(_state,dst,ref+mvoffsets[0],ystride,_dct_coeffs+64);
}
else oc_frag_recon_inter(_state,dst,ref+mvoffsets[0],ystride,_dct_coeffs);
}
}
/*Copies the fragments specified by the lists of fragment indices from one
frame to another.
_fragis: A pointer to a list of fragment indices.
_nfragis: The number of fragment indices to copy.
_dst_frame: The reference frame to copy to.
_src_frame: The reference frame to copy from.
_pli: The color plane the fragments lie in.*/
void oc_state_frag_copy_list(const oc_theora_state *_state,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,
int _dst_frame,int _src_frame,int _pli){
_state->opt_vtable.state_frag_copy_list(_state,_fragis,_nfragis,_dst_frame,
_src_frame,_pli);
}
void oc_state_frag_copy_list_c(const oc_theora_state *_state,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,
int _dst_frame,int _src_frame,int _pli){
const ptrdiff_t *frag_buf_offs;
const unsigned char *src_frame_data;
unsigned char *dst_frame_data;
ptrdiff_t fragii;
int ystride;
dst_frame_data=_state->ref_frame_data[_state->ref_frame_idx[_dst_frame]];
src_frame_data=_state->ref_frame_data[_state->ref_frame_idx[_src_frame]];
ystride=_state->ref_ystride[_pli];
frag_buf_offs=_state->frag_buf_offs;
for(fragii=0;fragii<_nfragis;fragii++){
ptrdiff_t frag_buf_off;
frag_buf_off=frag_buf_offs[_fragis[fragii]];
oc_frag_copy(_state,dst_frame_data+frag_buf_off,
src_frame_data+frag_buf_off,ystride);
}
}
static void loop_filter_h(unsigned char *_pix,int _ystride,int *_bv){
static void loop_filter_h(unsigned char *_pix,int _ystride,signed char *_bv){
int y;
_pix-=2;
for(y=0;y<8;y++){
@ -965,7 +1008,7 @@ static void loop_filter_h(unsigned char *_pix,int _ystride,int *_bv){
}
}
static void loop_filter_v(unsigned char *_pix,int _ystride,int *_bv){
static void loop_filter_v(unsigned char *_pix,int _ystride,signed char *_bv){
int x;
_pix-=_ystride*2;
for(x=0;x<8;x++){
@ -982,20 +1025,16 @@ static void loop_filter_v(unsigned char *_pix,int _ystride,int *_bv){
/*Initialize the bounding values array used by the loop filter.
_bv: Storage for the array.
Return: 0 on success, or a non-zero value if no filtering need be applied.*/
int oc_state_loop_filter_init(oc_theora_state *_state,int _bv[256]){
int flimit;
_flimit: The filter limit as defined in Section 7.10 of the spec.*/
void oc_loop_filter_init_c(signed char _bv[256],int _flimit){
int i;
flimit=_state->loop_filter_limits[_state->qis[0]];
if(flimit==0)return 1;
memset(_bv,0,sizeof(_bv[0])*256);
for(i=0;i<flimit;i++){
if(127-i-flimit>=0)_bv[127-i-flimit]=i-flimit;
_bv[127-i]=-i;
_bv[127+i]=i;
if(127+i+flimit<256)_bv[127+i+flimit]=flimit-i;
for(i=0;i<_flimit;i++){
if(127-i-_flimit>=0)_bv[127-i-_flimit]=(signed char)(i-_flimit);
_bv[127-i]=(signed char)(-i);
_bv[127+i]=(signed char)(i);
if(127+i+_flimit<256)_bv[127+i+_flimit]=(signed char)(_flimit-i);
}
return 0;
}
/*Apply the loop filter to a given set of fragment rows in the given plane.
@ -1006,14 +1045,8 @@ int oc_state_loop_filter_init(oc_theora_state *_state,int _bv[256]){
_pli: The color plane to filter.
_fragy0: The Y coordinate of the first fragment row to filter.
_fragy_end: The Y coordinate of the fragment row to stop filtering at.*/
void oc_state_loop_filter_frag_rows(const oc_theora_state *_state,int _bv[256],
int _refi,int _pli,int _fragy0,int _fragy_end){
_state->opt_vtable.state_loop_filter_frag_rows(_state,_bv,_refi,_pli,
_fragy0,_fragy_end);
}
void oc_state_loop_filter_frag_rows_c(const oc_theora_state *_state,int *_bv,
int _refi,int _pli,int _fragy0,int _fragy_end){
void oc_state_loop_filter_frag_rows_c(const oc_theora_state *_state,
signed char *_bv,int _refi,int _pli,int _fragy0,int _fragy_end){
const oc_fragment_plane *fplane;
const oc_fragment *frags;
const ptrdiff_t *frag_buf_offs;
@ -1030,7 +1063,7 @@ void oc_state_loop_filter_frag_rows_c(const oc_theora_state *_state,int *_bv,
fragi_top=fplane->froffset;
fragi_bot=fragi_top+fplane->nfrags;
fragi0=fragi_top+_fragy0*(ptrdiff_t)nhfrags;
fragi0_end=fragi0+(_fragy_end-_fragy0)*(ptrdiff_t)nhfrags;
fragi0_end=fragi_top+_fragy_end*(ptrdiff_t)nhfrags;
ystride=_state->ref_ystride[_pli];
frags=_state->frags;
frag_buf_offs=_state->frag_buf_offs;

552
media/libtheora/lib/state.h Normal file
View File

@ -0,0 +1,552 @@
/********************************************************************
* *
* THIS FILE IS PART OF THE OggTheora SOFTWARE CODEC SOURCE CODE. *
* USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS *
* GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
* IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING. *
* *
* THE Theora SOURCE CODE IS COPYRIGHT (C) 2002-2009 *
* by the Xiph.Org Foundation and contributors http://www.xiph.org/ *
* *
********************************************************************
function:
last mod: $Id: internal.h 17337 2010-07-19 16:08:54Z tterribe $
********************************************************************/
#if !defined(_state_H)
# define _state_H (1)
# include "internal.h"
# include "huffman.h"
# include "quant.h"
/*A single quadrant of the map from a super block to fragment numbers.*/
typedef ptrdiff_t oc_sb_map_quad[4];
/*A map from a super block to fragment numbers.*/
typedef oc_sb_map_quad oc_sb_map[4];
/*A single plane of the map from a macro block to fragment numbers.*/
typedef ptrdiff_t oc_mb_map_plane[4];
/*A map from a macro block to fragment numbers.*/
typedef oc_mb_map_plane oc_mb_map[3];
/*A motion vector.*/
typedef ogg_int16_t oc_mv;
typedef struct oc_sb_flags oc_sb_flags;
typedef struct oc_border_info oc_border_info;
typedef struct oc_fragment oc_fragment;
typedef struct oc_fragment_plane oc_fragment_plane;
typedef struct oc_base_opt_vtable oc_base_opt_vtable;
typedef struct oc_base_opt_data oc_base_opt_data;
typedef struct oc_state_dispatch_vtable oc_state_dispatch_vtable;
typedef struct oc_theora_state oc_theora_state;
/*Shared accelerated functions.*/
# if defined(OC_X86_ASM)
# if defined(_MSC_VER)
# include "x86_vc/x86int.h"
# else
# include "x86/x86int.h"
# endif
# endif
# if defined(OC_ARM_ASM)
# include "arm/armint.h"
# endif
# if defined(OC_C64X_ASM)
# include "c64x/c64xint.h"
# endif
# if !defined(oc_state_accel_init)
# define oc_state_accel_init oc_state_accel_init_c
# endif
# if defined(OC_STATE_USE_VTABLE)
# if !defined(oc_frag_copy)
# define oc_frag_copy(_state,_dst,_src,_ystride) \
((*(_state)->opt_vtable.frag_copy)(_dst,_src,_ystride))
# endif
# if !defined(oc_frag_copy_list)
# define oc_frag_copy_list(_state,_dst_frame,_src_frame,_ystride, \
_fragis,_nfragis,_frag_buf_offs) \
((*(_state)->opt_vtable.frag_copy_list)(_dst_frame,_src_frame,_ystride, \
_fragis,_nfragis,_frag_buf_offs))
# endif
# if !defined(oc_frag_recon_intra)
# define oc_frag_recon_intra(_state,_dst,_dst_ystride,_residue) \
((*(_state)->opt_vtable.frag_recon_intra)(_dst,_dst_ystride,_residue))
# endif
# if !defined(oc_frag_recon_inter)
# define oc_frag_recon_inter(_state,_dst,_src,_ystride,_residue) \
((*(_state)->opt_vtable.frag_recon_inter)(_dst,_src,_ystride,_residue))
# endif
# if !defined(oc_frag_recon_inter2)
# define oc_frag_recon_inter2(_state,_dst,_src1,_src2,_ystride,_residue) \
((*(_state)->opt_vtable.frag_recon_inter2)(_dst, \
_src1,_src2,_ystride,_residue))
# endif
# if !defined(oc_idct8x8)
# define oc_idct8x8(_state,_y,_x,_last_zzi) \
((*(_state)->opt_vtable.idct8x8)(_y,_x,_last_zzi))
# endif
# if !defined(oc_state_frag_recon)
# define oc_state_frag_recon(_state,_fragi, \
_pli,_dct_coeffs,_last_zzi,_dc_quant) \
((*(_state)->opt_vtable.state_frag_recon)(_state,_fragi, \
_pli,_dct_coeffs,_last_zzi,_dc_quant))
# endif
# if !defined(oc_loop_filter_init)
# define oc_loop_filter_init(_state,_bv,_flimit) \
((*(_state)->opt_vtable.loop_filter_init)(_bv,_flimit))
# endif
# if !defined(oc_state_loop_filter_frag_rows)
# define oc_state_loop_filter_frag_rows(_state, \
_bv,_refi,_pli,_fragy0,_fragy_end) \
((*(_state)->opt_vtable.state_loop_filter_frag_rows)(_state, \
_bv,_refi,_pli,_fragy0,_fragy_end))
# endif
# if !defined(oc_restore_fpu)
# define oc_restore_fpu(_state) \
((*(_state)->opt_vtable.restore_fpu)())
# endif
# else
# if !defined(oc_frag_copy)
# define oc_frag_copy(_state,_dst,_src,_ystride) \
oc_frag_copy_c(_dst,_src,_ystride)
# endif
# if !defined(oc_frag_copy_list)
# define oc_frag_copy_list(_state,_dst_frame,_src_frame,_ystride, \
_fragis,_nfragis,_frag_buf_offs) \
oc_frag_copy_list_c(_dst_frame,_src_frame,_ystride, \
_fragis,_nfragis,_frag_buf_offs)
# endif
# if !defined(oc_frag_recon_intra)
# define oc_frag_recon_intra(_state,_dst,_dst_ystride,_residue) \
oc_frag_recon_intra_c(_dst,_dst_ystride,_residue)
# endif
# if !defined(oc_frag_recon_inter)
# define oc_frag_recon_inter(_state,_dst,_src,_ystride,_residue) \
oc_frag_recon_inter_c(_dst,_src,_ystride,_residue)
# endif
# if !defined(oc_frag_recon_inter2)
# define oc_frag_recon_inter2(_state,_dst,_src1,_src2,_ystride,_residue) \
oc_frag_recon_inter2_c(_dst,_src1,_src2,_ystride,_residue)
# endif
# if !defined(oc_idct8x8)
# define oc_idct8x8(_state,_y,_x,_last_zzi) oc_idct8x8_c(_y,_x,_last_zzi)
# endif
# if !defined(oc_state_frag_recon)
# define oc_state_frag_recon oc_state_frag_recon_c
# endif
# if !defined(oc_loop_filter_init)
# define oc_loop_filter_init(_state,_bv,_flimit) \
oc_loop_filter_init_c(_bv,_flimit)
# endif
# if !defined(oc_state_loop_filter_frag_rows)
# define oc_state_loop_filter_frag_rows oc_state_loop_filter_frag_rows_c
# endif
# if !defined(oc_restore_fpu)
# define oc_restore_fpu(_state) do{}while(0)
# endif
# endif
/*A keyframe.*/
# define OC_INTRA_FRAME (0)
/*A predicted frame.*/
# define OC_INTER_FRAME (1)
/*A frame of unknown type (frame type decision has not yet been made).*/
# define OC_UNKWN_FRAME (-1)
/*The amount of padding to add to the reconstructed frame buffers on all
sides.
This is used to allow unrestricted motion vectors without special casing.
This must be a multiple of 2.*/
# define OC_UMV_PADDING (16)
/*Frame classification indices.*/
/*The previous golden frame.*/
# define OC_FRAME_GOLD (0)
/*The previous frame.*/
# define OC_FRAME_PREV (1)
/*The current frame.*/
# define OC_FRAME_SELF (2)
/*Used to mark uncoded fragments (for DC prediction).*/
# define OC_FRAME_NONE (3)
/*The input or output buffer.*/
# define OC_FRAME_IO (3)
/*Uncompressed prev golden frame.*/
# define OC_FRAME_GOLD_ORIG (4)
/*Uncompressed previous frame. */
# define OC_FRAME_PREV_ORIG (5)
/*Macroblock modes.*/
/*Macro block is invalid: It is never coded.*/
# define OC_MODE_INVALID (-1)
/*Encoded difference from the same macro block in the previous frame.*/
# define OC_MODE_INTER_NOMV (0)
/*Encoded with no motion compensated prediction.*/
# define OC_MODE_INTRA (1)
/*Encoded difference from the previous frame offset by the given motion
vector.*/
# define OC_MODE_INTER_MV (2)
/*Encoded difference from the previous frame offset by the last coded motion
vector.*/
# define OC_MODE_INTER_MV_LAST (3)
/*Encoded difference from the previous frame offset by the second to last
coded motion vector.*/
# define OC_MODE_INTER_MV_LAST2 (4)
/*Encoded difference from the same macro block in the previous golden
frame.*/
# define OC_MODE_GOLDEN_NOMV (5)
/*Encoded difference from the previous golden frame offset by the given motion
vector.*/
# define OC_MODE_GOLDEN_MV (6)
/*Encoded difference from the previous frame offset by the individual motion
vectors given for each block.*/
# define OC_MODE_INTER_MV_FOUR (7)
/*The number of (coded) modes.*/
# define OC_NMODES (8)
/*Determines the reference frame used for a given MB mode.*/
# define OC_FRAME_FOR_MODE(_x) \
OC_UNIBBLE_TABLE32(OC_FRAME_PREV,OC_FRAME_SELF,OC_FRAME_PREV,OC_FRAME_PREV, \
OC_FRAME_PREV,OC_FRAME_GOLD,OC_FRAME_GOLD,OC_FRAME_PREV,(_x))
/*Constants for the packet state machine common between encoder and decoder.*/
/*Next packet to emit/read: Codec info header.*/
# define OC_PACKET_INFO_HDR (-3)
/*Next packet to emit/read: Comment header.*/
# define OC_PACKET_COMMENT_HDR (-2)
/*Next packet to emit/read: Codec setup header.*/
# define OC_PACKET_SETUP_HDR (-1)
/*No more packets to emit/read.*/
# define OC_PACKET_DONE (INT_MAX)
#define OC_MV(_x,_y) ((oc_mv)((_x)&0xFF|(_y)<<8))
#define OC_MV_X(_mv) ((signed char)(_mv))
#define OC_MV_Y(_mv) ((_mv)>>8)
#define OC_MV_ADD(_mv1,_mv2) \
OC_MV(OC_MV_X(_mv1)+OC_MV_X(_mv2), \
OC_MV_Y(_mv1)+OC_MV_Y(_mv2))
#define OC_MV_SUB(_mv1,_mv2) \
OC_MV(OC_MV_X(_mv1)-OC_MV_X(_mv2), \
OC_MV_Y(_mv1)-OC_MV_Y(_mv2))
/*Super blocks are 32x32 segments of pixels in a single color plane indexed
in image order.
Internally, super blocks are broken up into four quadrants, each of which
contains a 2x2 pattern of blocks, each of which is an 8x8 block of pixels.
Quadrants, and the blocks within them, are indexed in a special order called
a "Hilbert curve" within the super block.
In order to differentiate between the Hilbert-curve indexing strategy and
the regular image order indexing strategy, blocks indexed in image order
are called "fragments".
Fragments are indexed in image order, left to right, then bottom to top,
from Y' plane to Cb plane to Cr plane.
The co-located fragments in all image planes corresponding to the location
of a single quadrant of a luma plane super block form a macro block.
Thus there is only a single set of macro blocks for all planes, each of which
contains between 6 and 12 fragments, depending on the pixel format.
Therefore macro block information is kept in a separate set of arrays from
super blocks to avoid unused space in the other planes.
The lists are indexed in super block order.
That is, the macro block corresponding to the macro block mbi in (luma plane)
super block sbi is at index (sbi<<2|mbi).
Thus the number of macro blocks in each dimension is always twice the number
of super blocks, even when only an odd number fall inside the coded frame.
These "extra" macro blocks are just an artifact of our internal data layout,
and not part of the coded stream; they are flagged with a negative MB mode.*/
/*Super block information.*/
struct oc_sb_flags{
unsigned char coded_fully:1;
unsigned char coded_partially:1;
unsigned char quad_valid:4;
};
/*Information about a fragment which intersects the border of the displayable
region.
This marks which pixels belong to the displayable region.*/
struct oc_border_info{
/*A bit mask marking which pixels are in the displayable region.
Pixel (x,y) corresponds to bit (y<<3|x).*/
ogg_int64_t mask;
/*The number of pixels in the displayable region.
This is always positive, and always less than 64.*/
int npixels;
};
/*Fragment information.*/
struct oc_fragment{
/*A flag indicating whether or not this fragment is coded.*/
unsigned coded:1;
/*A flag indicating that this entire fragment lies outside the displayable
region of the frame.
Note the contrast with an invalid macro block, which is outside the coded
frame, not just the displayable one.
There are no fragments outside the coded frame by construction.*/
unsigned invalid:1;
/*The index of the quality index used for this fragment's AC coefficients.*/
unsigned qii:4;
/*The index of the reference frame this fragment is predicted from.*/
unsigned refi:2;
/*The mode of the macroblock this fragment belongs to.*/
unsigned mb_mode:3;
/*The index of the associated border information for fragments which lie
partially outside the displayable region.
For fragments completely inside or outside this region, this is -1.
Note that the C standard requires an explicit signed keyword for bitfield
types, since some compilers may treat them as unsigned without it.*/
signed int borderi:5;
/*The prediction-corrected DC component.
Note that the C standard requires an explicit signed keyword for bitfield
types, since some compilers may treat them as unsigned without it.*/
signed int dc:16;
};
/*A description of each fragment plane.*/
struct oc_fragment_plane{
/*The number of fragments in the horizontal direction.*/
int nhfrags;
/*The number of fragments in the vertical direction.*/
int nvfrags;
/*The offset of the first fragment in the plane.*/
ptrdiff_t froffset;
/*The total number of fragments in the plane.*/
ptrdiff_t nfrags;
/*The number of super blocks in the horizontal direction.*/
unsigned nhsbs;
/*The number of super blocks in the vertical direction.*/
unsigned nvsbs;
/*The offset of the first super block in the plane.*/
unsigned sboffset;
/*The total number of super blocks in the plane.*/
unsigned nsbs;
};
typedef void (*oc_state_loop_filter_frag_rows_func)(
const oc_theora_state *_state,signed char _bv[256],int _refi,int _pli,
int _fragy0,int _fragy_end);
/*The shared (encoder and decoder) functions that have accelerated variants.*/
struct oc_base_opt_vtable{
void (*frag_copy)(unsigned char *_dst,
const unsigned char *_src,int _ystride);
void (*frag_copy_list)(unsigned char *_dst_frame,
const unsigned char *_src_frame,int _ystride,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,const ptrdiff_t *_frag_buf_offs);
void (*frag_recon_intra)(unsigned char *_dst,int _ystride,
const ogg_int16_t _residue[64]);
void (*frag_recon_inter)(unsigned char *_dst,
const unsigned char *_src,int _ystride,const ogg_int16_t _residue[64]);
void (*frag_recon_inter2)(unsigned char *_dst,const unsigned char *_src1,
const unsigned char *_src2,int _ystride,const ogg_int16_t _residue[64]);
void (*idct8x8)(ogg_int16_t _y[64],ogg_int16_t _x[64],int _last_zzi);
void (*state_frag_recon)(const oc_theora_state *_state,ptrdiff_t _fragi,
int _pli,ogg_int16_t _dct_coeffs[128],int _last_zzi,ogg_uint16_t _dc_quant);
void (*loop_filter_init)(signed char _bv[256],int _flimit);
oc_state_loop_filter_frag_rows_func state_loop_filter_frag_rows;
void (*restore_fpu)(void);
};
/*The shared (encoder and decoder) tables that vary according to which variants
of the above functions are used.*/
struct oc_base_opt_data{
const unsigned char *dct_fzig_zag;
};
/*State information common to both the encoder and decoder.*/
struct oc_theora_state{
/*The stream information.*/
th_info info;
# if defined(OC_STATE_USE_VTABLE)
/*Table for shared accelerated functions.*/
oc_base_opt_vtable opt_vtable;
# endif
/*Table for shared data used by accelerated functions.*/
oc_base_opt_data opt_data;
/*CPU flags to detect the presence of extended instruction sets.*/
ogg_uint32_t cpu_flags;
/*The fragment plane descriptions.*/
oc_fragment_plane fplanes[3];
/*The list of fragments, indexed in image order.*/
oc_fragment *frags;
/*The the offset into the reference frame buffer to the upper-left pixel of
each fragment.*/
ptrdiff_t *frag_buf_offs;
/*The motion vector for each fragment.*/
oc_mv *frag_mvs;
/*The total number of fragments in a single frame.*/
ptrdiff_t nfrags;
/*The list of super block maps, indexed in image order.*/
oc_sb_map *sb_maps;
/*The list of super block flags, indexed in image order.*/
oc_sb_flags *sb_flags;
/*The total number of super blocks in a single frame.*/
unsigned nsbs;
/*The fragments from each color plane that belong to each macro block.
Fragments are stored in image order (left to right then top to bottom).
When chroma components are decimated, the extra fragments have an index of
-1.*/
oc_mb_map *mb_maps;
/*The list of macro block modes.
A negative number indicates the macro block lies entirely outside the
coded frame.*/
signed char *mb_modes;
/*The number of macro blocks in the X direction.*/
unsigned nhmbs;
/*The number of macro blocks in the Y direction.*/
unsigned nvmbs;
/*The total number of macro blocks.*/
size_t nmbs;
/*The list of coded fragments, in coded order.
Uncoded fragments are stored in reverse order from the end of the list.*/
ptrdiff_t *coded_fragis;
/*The number of coded fragments in each plane.*/
ptrdiff_t ncoded_fragis[3];
/*The total number of coded fragments.*/
ptrdiff_t ntotal_coded_fragis;
/*The actual buffers used for the reference frames.*/
th_ycbcr_buffer ref_frame_bufs[6];
/*The index of the buffers being used for each OC_FRAME_* reference frame.*/
int ref_frame_idx[6];
/*The storage for the reference frame buffers.
This is just ref_frame_bufs[ref_frame_idx[i]][0].data, but is cached here
for faster look-up.*/
unsigned char *ref_frame_data[6];
/*The handle used to allocate the reference frame buffers.*/
unsigned char *ref_frame_handle;
/*The strides for each plane in the reference frames.*/
int ref_ystride[3];
/*The number of unique border patterns.*/
int nborders;
/*The unique border patterns for all border fragments.
The borderi field of fragments which straddle the border indexes this
list.*/
oc_border_info borders[16];
/*The frame number of the last keyframe.*/
ogg_int64_t keyframe_num;
/*The frame number of the current frame.*/
ogg_int64_t curframe_num;
/*The granpos of the current frame.*/
ogg_int64_t granpos;
/*The type of the current frame.*/
signed char frame_type;
/*The bias to add to the frame count when computing granule positions.*/
unsigned char granpos_bias;
/*The number of quality indices used in the current frame.*/
unsigned char nqis;
/*The quality indices of the current frame.*/
unsigned char qis[3];
/*The dequantization tables, stored in zig-zag order, and indexed by
qi, pli, qti, and zzi.*/
ogg_uint16_t *dequant_tables[64][3][2];
OC_ALIGN16(oc_quant_table dequant_table_data[64][3][2]);
/*Loop filter strength parameters.*/
unsigned char loop_filter_limits[64];
};
/*The function type used to fill in the chroma plane motion vectors for a
macro block when 4 different motion vectors are specified in the luma
plane.
_cbmvs: The chroma block-level motion vectors to fill in.
_lmbmv: The luma macro-block level motion vector to fill in for use in
prediction.
_lbmvs: The luma block-level motion vectors.*/
typedef void (*oc_set_chroma_mvs_func)(oc_mv _cbmvs[4],const oc_mv _lbmvs[4]);
/*A table of functions used to fill in the Cb,Cr plane motion vectors for a
macro block when 4 different motion vectors are specified in the luma
plane.*/
extern const oc_set_chroma_mvs_func OC_SET_CHROMA_MVS_TABLE[TH_PF_NFORMATS];
int oc_state_init(oc_theora_state *_state,const th_info *_info,int _nrefs);
void oc_state_clear(oc_theora_state *_state);
void oc_state_accel_init_c(oc_theora_state *_state);
void oc_state_borders_fill_rows(oc_theora_state *_state,int _refi,int _pli,
int _y0,int _yend);
void oc_state_borders_fill_caps(oc_theora_state *_state,int _refi,int _pli);
void oc_state_borders_fill(oc_theora_state *_state,int _refi);
void oc_state_fill_buffer_ptrs(oc_theora_state *_state,int _buf_idx,
th_ycbcr_buffer _img);
int oc_state_mbi_for_pos(oc_theora_state *_state,int _mbx,int _mby);
int oc_state_get_mv_offsets(const oc_theora_state *_state,int _offsets[2],
int _pli,oc_mv _mv);
void oc_loop_filter_init_c(signed char _bv[256],int _flimit);
void oc_state_loop_filter(oc_theora_state *_state,int _frame);
# if defined(OC_DUMP_IMAGES)
int oc_state_dump_frame(const oc_theora_state *_state,int _frame,
const char *_suf);
# endif
/*Default pure-C implementations of shared accelerated functions.*/
void oc_frag_copy_c(unsigned char *_dst,
const unsigned char *_src,int _src_ystride);
void oc_frag_copy_list_c(unsigned char *_dst_frame,
const unsigned char *_src_frame,int _ystride,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,const ptrdiff_t *_frag_buf_offs);
void oc_frag_recon_intra_c(unsigned char *_dst,int _dst_ystride,
const ogg_int16_t _residue[64]);
void oc_frag_recon_inter_c(unsigned char *_dst,
const unsigned char *_src,int _ystride,const ogg_int16_t _residue[64]);
void oc_frag_recon_inter2_c(unsigned char *_dst,const unsigned char *_src1,
const unsigned char *_src2,int _ystride,const ogg_int16_t _residue[64]);
void oc_idct8x8_c(ogg_int16_t _y[64],ogg_int16_t _x[64],int _last_zzi);
void oc_state_frag_recon_c(const oc_theora_state *_state,ptrdiff_t _fragi,
int _pli,ogg_int16_t _dct_coeffs[128],int _last_zzi,ogg_uint16_t _dc_quant);
void oc_state_loop_filter_frag_rows_c(const oc_theora_state *_state,
signed char _bv[256],int _refi,int _pli,int _fragy0,int _fragy_end);
void oc_restore_fpu_c(void);
/*We need a way to call a few encoder functions without introducing a link-time
dependency into the decoder, while still allowing the old alpha API which
does not distinguish between encoder and decoder objects to be used.
We do this by placing a function table at the start of the encoder object
which can dispatch into the encoder library.
We do a similar thing for the decoder in case we ever decide to split off a
common base library.*/
typedef void (*oc_state_clear_func)(theora_state *_th);
typedef int (*oc_state_control_func)(theora_state *th,int _req,
void *_buf,size_t _buf_sz);
typedef ogg_int64_t (*oc_state_granule_frame_func)(theora_state *_th,
ogg_int64_t _granulepos);
typedef double (*oc_state_granule_time_func)(theora_state *_th,
ogg_int64_t _granulepos);
struct oc_state_dispatch_vtable{
oc_state_clear_func clear;
oc_state_control_func control;
oc_state_granule_frame_func granule_frame;
oc_state_granule_time_func granule_time;
};
#endif

79
media/libtheora/lib/x86/mmxfrag.c
View File

@ -11,7 +11,7 @@
********************************************************************
function:
last mod: $Id: mmxfrag.c 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: mmxfrag.c 17410 2010-09-21 21:53:48Z tterribe $
********************************************************************/
@ -22,10 +22,64 @@
The iteration each instruction belongs to is marked in the comments as #i.*/
#include <stddef.h>
#include "x86int.h"
#include "mmxfrag.h"
#if defined(OC_X86_ASM)
/*Copies an 8x8 block of pixels from _src to _dst, assuming _ystride bytes
between rows.*/
# define OC_FRAG_COPY_MMX(_dst,_src,_ystride) \
do{ \
const unsigned char *src; \
unsigned char *dst; \
ptrdiff_t ystride3; \
src=(_src); \
dst=(_dst); \
__asm__ __volatile__( \
/*src+0*ystride*/ \
"movq (%[src]),%%mm0\n\t" \
/*src+1*ystride*/ \
"movq (%[src],%[ystride]),%%mm1\n\t" \
/*ystride3=ystride*3*/ \
"lea (%[ystride],%[ystride],2),%[ystride3]\n\t" \
/*src+2*ystride*/ \
"movq (%[src],%[ystride],2),%%mm2\n\t" \
/*src+3*ystride*/ \
"movq (%[src],%[ystride3]),%%mm3\n\t" \
/*dst+0*ystride*/ \
"movq %%mm0,(%[dst])\n\t" \
/*dst+1*ystride*/ \
"movq %%mm1,(%[dst],%[ystride])\n\t" \
/*Pointer to next 4.*/ \
"lea (%[src],%[ystride],4),%[src]\n\t" \
/*dst+2*ystride*/ \
"movq %%mm2,(%[dst],%[ystride],2)\n\t" \
/*dst+3*ystride*/ \
"movq %%mm3,(%[dst],%[ystride3])\n\t" \
/*Pointer to next 4.*/ \
"lea (%[dst],%[ystride],4),%[dst]\n\t" \
/*src+0*ystride*/ \
"movq (%[src]),%%mm0\n\t" \
/*src+1*ystride*/ \
"movq (%[src],%[ystride]),%%mm1\n\t" \
/*src+2*ystride*/ \
"movq (%[src],%[ystride],2),%%mm2\n\t" \
/*src+3*ystride*/ \
"movq (%[src],%[ystride3]),%%mm3\n\t" \
/*dst+0*ystride*/ \
"movq %%mm0,(%[dst])\n\t" \
/*dst+1*ystride*/ \
"movq %%mm1,(%[dst],%[ystride])\n\t" \
/*dst+2*ystride*/ \
"movq %%mm2,(%[dst],%[ystride],2)\n\t" \
/*dst+3*ystride*/ \
"movq %%mm3,(%[dst],%[ystride3])\n\t" \
:[dst]"+r"(dst),[src]"+r"(src),[ystride3]"=&r"(ystride3) \
:[ystride]"r"((ptrdiff_t)(_ystride)) \
:"memory" \
); \
} \
while(0)
/*Copies an 8x8 block of pixels from _src to _dst, assuming _ystride bytes
between rows.*/
void oc_frag_copy_mmx(unsigned char *_dst,
@ -33,6 +87,27 @@ void oc_frag_copy_mmx(unsigned char *_dst,
OC_FRAG_COPY_MMX(_dst,_src,_ystride);
}
/*Copies the fragments specified by the lists of fragment indices from one
frame to another.
_dst_frame: The reference frame to copy to.
_src_frame: The reference frame to copy from.
_ystride: The row stride of the reference frames.
_fragis: A pointer to a list of fragment indices.
_nfragis: The number of fragment indices to copy.
_frag_buf_offs: The offsets of fragments in the reference frames.*/
void oc_frag_copy_list_mmx(unsigned char *_dst_frame,
const unsigned char *_src_frame,int _ystride,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,const ptrdiff_t *_frag_buf_offs){
ptrdiff_t fragii;
for(fragii=0;fragii<_nfragis;fragii++){
ptrdiff_t frag_buf_off;
frag_buf_off=_frag_buf_offs[_fragis[fragii]];
OC_FRAG_COPY_MMX(_dst_frame+frag_buf_off,
_src_frame+frag_buf_off,_ystride);
}
}
void oc_frag_recon_intra_mmx(unsigned char *_dst,int _ystride,
const ogg_int16_t *_residue){
__asm__ __volatile__(

64
media/libtheora/lib/x86/mmxfrag.h
View File

@ -1,64 +0,0 @@
#if !defined(_x86_mmxfrag_H)
# define _x86_mmxfrag_H (1)
# include <stddef.h>
# include "x86int.h"
#if defined(OC_X86_ASM)
/*Copies an 8x8 block of pixels from _src to _dst, assuming _ystride bytes
between rows.*/
#define OC_FRAG_COPY_MMX(_dst,_src,_ystride) \
do{ \
const unsigned char *src; \
unsigned char *dst; \
ptrdiff_t ystride3; \
src=(_src); \
dst=(_dst); \
__asm__ __volatile__( \
/*src+0*ystride*/ \
"movq (%[src]),%%mm0\n\t" \
/*src+1*ystride*/ \
"movq (%[src],%[ystride]),%%mm1\n\t" \
/*ystride3=ystride*3*/ \
"lea (%[ystride],%[ystride],2),%[ystride3]\n\t" \
/*src+2*ystride*/ \
"movq (%[src],%[ystride],2),%%mm2\n\t" \
/*src+3*ystride*/ \
"movq (%[src],%[ystride3]),%%mm3\n\t" \
/*dst+0*ystride*/ \
"movq %%mm0,(%[dst])\n\t" \
/*dst+1*ystride*/ \
"movq %%mm1,(%[dst],%[ystride])\n\t" \
/*Pointer to next 4.*/ \
"lea (%[src],%[ystride],4),%[src]\n\t" \
/*dst+2*ystride*/ \
"movq %%mm2,(%[dst],%[ystride],2)\n\t" \
/*dst+3*ystride*/ \
"movq %%mm3,(%[dst],%[ystride3])\n\t" \
/*Pointer to next 4.*/ \
"lea (%[dst],%[ystride],4),%[dst]\n\t" \
/*src+0*ystride*/ \
"movq (%[src]),%%mm0\n\t" \
/*src+1*ystride*/ \
"movq (%[src],%[ystride]),%%mm1\n\t" \
/*src+2*ystride*/ \
"movq (%[src],%[ystride],2),%%mm2\n\t" \
/*src+3*ystride*/ \
"movq (%[src],%[ystride3]),%%mm3\n\t" \
/*dst+0*ystride*/ \
"movq %%mm0,(%[dst])\n\t" \
/*dst+1*ystride*/ \
"movq %%mm1,(%[dst],%[ystride])\n\t" \
/*dst+2*ystride*/ \
"movq %%mm2,(%[dst],%[ystride],2)\n\t" \
/*dst+3*ystride*/ \
"movq %%mm3,(%[dst],%[ystride3])\n\t" \
:[dst]"+r"(dst),[src]"+r"(src),[ystride3]"=&r"(ystride3) \
:[ystride]"r"((ptrdiff_t)(_ystride)) \
:"memory" \
); \
} \
while(0)
# endif
#endif

296
media/libtheora/lib/x86/mmxidct.c
View File

@ -11,7 +11,7 @@
********************************************************************
function:
last mod: $Id: mmxidct.c 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: mmxidct.c 17446 2010-09-23 20:06:20Z tterribe $
********************************************************************/
@ -30,89 +30,66 @@
/*A table of constants used by the MMX routines.*/
static const ogg_uint16_t __attribute__((aligned(8),used))
OC_IDCT_CONSTS[(7+1)*4]={
(ogg_uint16_t)OC_C1S7,(ogg_uint16_t)OC_C1S7,
(ogg_uint16_t)OC_C1S7,(ogg_uint16_t)OC_C1S7,
(ogg_uint16_t)OC_C2S6,(ogg_uint16_t)OC_C2S6,
(ogg_uint16_t)OC_C2S6,(ogg_uint16_t)OC_C2S6,
(ogg_uint16_t)OC_C3S5,(ogg_uint16_t)OC_C3S5,
(ogg_uint16_t)OC_C3S5,(ogg_uint16_t)OC_C3S5,
(ogg_uint16_t)OC_C4S4,(ogg_uint16_t)OC_C4S4,
(ogg_uint16_t)OC_C4S4,(ogg_uint16_t)OC_C4S4,
(ogg_uint16_t)OC_C5S3,(ogg_uint16_t)OC_C5S3,
(ogg_uint16_t)OC_C5S3,(ogg_uint16_t)OC_C5S3,
(ogg_uint16_t)OC_C6S2,(ogg_uint16_t)OC_C6S2,
(ogg_uint16_t)OC_C6S2,(ogg_uint16_t)OC_C6S2,
(ogg_uint16_t)OC_C7S1,(ogg_uint16_t)OC_C7S1,
(ogg_uint16_t)OC_C7S1,(ogg_uint16_t)OC_C7S1,
8, 8, 8, 8
};
/*Converts the expression in the argument to a string.*/
#define OC_M2STR(_s) #_s
/*38 cycles*/
#define OC_IDCT_BEGIN \
#define OC_IDCT_BEGIN(_y,_x) \
"#OC_IDCT_BEGIN\n\t" \
"movq "OC_I(3)",%%mm2\n\t" \
"movq "OC_C(3)",%%mm6\n\t" \
"movq "OC_I(3,_x)",%%mm2\n\t" \
"movq "OC_MEM_OFFS(0x30,c)",%%mm6\n\t" \
"movq %%mm2,%%mm4\n\t" \
"movq "OC_J(5)",%%mm7\n\t" \
"movq "OC_J(5,_x)",%%mm7\n\t" \
"pmulhw %%mm6,%%mm4\n\t" \
"movq "OC_C(5)",%%mm1\n\t" \
"movq "OC_MEM_OFFS(0x50,c)",%%mm1\n\t" \
"pmulhw %%mm7,%%mm6\n\t" \
"movq %%mm1,%%mm5\n\t" \
"pmulhw %%mm2,%%mm1\n\t" \
"movq "OC_I(1)",%%mm3\n\t" \
"movq "OC_I(1,_x)",%%mm3\n\t" \
"pmulhw %%mm7,%%mm5\n\t" \
"movq "OC_C(1)",%%mm0\n\t" \
"movq "OC_MEM_OFFS(0x10,c)",%%mm0\n\t" \
"paddw %%mm2,%%mm4\n\t" \
"paddw %%mm7,%%mm6\n\t" \
"paddw %%mm1,%%mm2\n\t" \
"movq "OC_J(7)",%%mm1\n\t" \
"movq "OC_J(7,_x)",%%mm1\n\t" \
"paddw %%mm5,%%mm7\n\t" \
"movq %%mm0,%%mm5\n\t" \
"pmulhw %%mm3,%%mm0\n\t" \
"paddw %%mm7,%%mm4\n\t" \
"pmulhw %%mm1,%%mm5\n\t" \
"movq "OC_C(7)",%%mm7\n\t" \
"movq "OC_MEM_OFFS(0x70,c)",%%mm7\n\t" \
"psubw %%mm2,%%mm6\n\t" \
"paddw %%mm3,%%mm0\n\t" \
"pmulhw %%mm7,%%mm3\n\t" \
"movq "OC_I(2)",%%mm2\n\t" \
"movq "OC_I(2,_x)",%%mm2\n\t" \
"pmulhw %%mm1,%%mm7\n\t" \
"paddw %%mm1,%%mm5\n\t" \
"movq %%mm2,%%mm1\n\t" \
"pmulhw "OC_C(2)",%%mm2\n\t" \
"pmulhw "OC_MEM_OFFS(0x20,c)",%%mm2\n\t" \
"psubw %%mm5,%%mm3\n\t" \
"movq "OC_J(6)",%%mm5\n\t" \
"movq "OC_J(6,_x)",%%mm5\n\t" \
"paddw %%mm7,%%mm0\n\t" \
"movq %%mm5,%%mm7\n\t" \
"psubw %%mm4,%%mm0\n\t" \
"pmulhw "OC_C(2)",%%mm5\n\t" \
"pmulhw "OC_MEM_OFFS(0x20,c)",%%mm5\n\t" \
"paddw %%mm1,%%mm2\n\t" \
"pmulhw "OC_C(6)",%%mm1\n\t" \
"pmulhw "OC_MEM_OFFS(0x60,c)",%%mm1\n\t" \
"paddw %%mm4,%%mm4\n\t" \
"paddw %%mm0,%%mm4\n\t" \
"psubw %%mm6,%%mm3\n\t" \
"paddw %%mm7,%%mm5\n\t" \
"paddw %%mm6,%%mm6\n\t" \
"pmulhw "OC_C(6)",%%mm7\n\t" \
"pmulhw "OC_MEM_OFFS(0x60,c)",%%mm7\n\t" \
"paddw %%mm3,%%mm6\n\t" \
"movq %%mm4,"OC_I(1)"\n\t" \
"movq %%mm4,"OC_I(1,_y)"\n\t" \
"psubw %%mm5,%%mm1\n\t" \
"movq "OC_C(4)",%%mm4\n\t" \
"movq "OC_MEM_OFFS(0x40,c)",%%mm4\n\t" \
"movq %%mm3,%%mm5\n\t" \
"pmulhw %%mm4,%%mm3\n\t" \
"paddw %%mm2,%%mm7\n\t" \
"movq %%mm6,"OC_I(2)"\n\t" \
"movq %%mm6,"OC_I(2,_y)"\n\t" \
"movq %%mm0,%%mm2\n\t" \
"movq "OC_I(0)",%%mm6\n\t" \
"movq "OC_I(0,_x)",%%mm6\n\t" \
"pmulhw %%mm4,%%mm0\n\t" \
"paddw %%mm3,%%mm5\n\t" \
"movq "OC_J(4)",%%mm3\n\t" \
"movq "OC_J(4,_x)",%%mm3\n\t" \
"psubw %%mm1,%%mm5\n\t" \
"paddw %%mm0,%%mm2\n\t" \
"psubw %%mm3,%%mm6\n\t" \
@ -126,18 +103,18 @@ static const ogg_uint16_t __attribute__((aligned(8),used))
"paddw %%mm0,%%mm6\n\t" \
"psubw %%mm2,%%mm6\n\t" \
"paddw %%mm2,%%mm2\n\t" \
"movq "OC_I(1)",%%mm0\n\t" \
"movq "OC_I(1,_y)",%%mm0\n\t" \
"paddw %%mm6,%%mm2\n\t" \
"paddw %%mm3,%%mm4\n\t" \
"psubw %%mm1,%%mm2\n\t" \
"#end OC_IDCT_BEGIN\n\t" \
/*38+8=46 cycles.*/
#define OC_ROW_IDCT \
#define OC_ROW_IDCT(_y,_x) \
"#OC_ROW_IDCT\n" \
OC_IDCT_BEGIN \
OC_IDCT_BEGIN(_y,_x) \
/*r3=D'*/ \
"movq "OC_I(2)",%%mm3\n\t" \
"movq "OC_I(2,_y)",%%mm3\n\t" \
/*r4=E'=E-G*/ \
"psubw %%mm7,%%mm4\n\t" \
/*r1=H'+H'*/ \
@ -162,7 +139,7 @@ static const ogg_uint16_t __attribute__((aligned(8),used))
"psubw %%mm0,%%mm7\n\t" \
"paddw %%mm0,%%mm0\n\t" \
/*Save R1.*/ \
"movq %%mm1,"OC_I(1)"\n\t" \
"movq %%mm1,"OC_I(1,_y)"\n\t" \
/*r0=R0=G.+C.*/ \
"paddw %%mm7,%%mm0\n\t" \
"#end OC_ROW_IDCT\n\t" \
@ -195,11 +172,11 @@ static const ogg_uint16_t __attribute__((aligned(8),used))
Since r1 is free at entry, we calculate the Js first.*/
/*19 cycles.*/
#define OC_TRANSPOSE \
#define OC_TRANSPOSE(_y) \
"#OC_TRANSPOSE\n\t" \
"movq %%mm4,%%mm1\n\t" \
"punpcklwd %%mm5,%%mm4\n\t" \
"movq %%mm0,"OC_I(0)"\n\t" \
"movq %%mm0,"OC_I(0,_y)"\n\t" \
"punpckhwd %%mm5,%%mm1\n\t" \
"movq %%mm6,%%mm0\n\t" \
"punpcklwd %%mm7,%%mm6\n\t" \
@ -207,17 +184,17 @@ static const ogg_uint16_t __attribute__((aligned(8),used))
"punpckldq %%mm6,%%mm4\n\t" \
"punpckhdq %%mm6,%%mm5\n\t" \
"movq %%mm1,%%mm6\n\t" \
"movq %%mm4,"OC_J(4)"\n\t" \
"movq %%mm4,"OC_J(4,_y)"\n\t" \
"punpckhwd %%mm7,%%mm0\n\t" \
"movq %%mm5,"OC_J(5)"\n\t" \
"movq %%mm5,"OC_J(5,_y)"\n\t" \
"punpckhdq %%mm0,%%mm6\n\t" \
"movq "OC_I(0)",%%mm4\n\t" \
"movq "OC_I(0,_y)",%%mm4\n\t" \
"punpckldq %%mm0,%%mm1\n\t" \
"movq "OC_I(1)",%%mm5\n\t" \
"movq "OC_I(1,_y)",%%mm5\n\t" \
"movq %%mm4,%%mm0\n\t" \
"movq %%mm6,"OC_J(7)"\n\t" \
"movq %%mm6,"OC_J(7,_y)"\n\t" \
"punpcklwd %%mm5,%%mm0\n\t" \
"movq %%mm1,"OC_J(6)"\n\t" \
"movq %%mm1,"OC_J(6,_y)"\n\t" \
"punpckhwd %%mm5,%%mm4\n\t" \
"movq %%mm2,%%mm5\n\t" \
"punpcklwd %%mm3,%%mm2\n\t" \
@ -225,20 +202,20 @@ static const ogg_uint16_t __attribute__((aligned(8),used))
"punpckldq %%mm2,%%mm0\n\t" \
"punpckhdq %%mm2,%%mm1\n\t" \
"movq %%mm4,%%mm2\n\t" \
"movq %%mm0,"OC_I(0)"\n\t" \
"movq %%mm0,"OC_I(0,_y)"\n\t" \
"punpckhwd %%mm3,%%mm5\n\t" \
"movq %%mm1,"OC_I(1)"\n\t" \
"movq %%mm1,"OC_I(1,_y)"\n\t" \
"punpckhdq %%mm5,%%mm4\n\t" \
"punpckldq %%mm5,%%mm2\n\t" \
"movq %%mm4,"OC_I(3)"\n\t" \
"movq %%mm2,"OC_I(2)"\n\t" \
"movq %%mm4,"OC_I(3,_y)"\n\t" \
"movq %%mm2,"OC_I(2,_y)"\n\t" \
"#end OC_TRANSPOSE\n\t" \
/*38+19=57 cycles.*/
#define OC_COLUMN_IDCT \
#define OC_COLUMN_IDCT(_y) \
"#OC_COLUMN_IDCT\n" \
OC_IDCT_BEGIN \
"paddw "OC_8",%%mm2\n\t" \
OC_IDCT_BEGIN(_y,_y) \
"paddw "OC_MEM_OFFS(0x00,c)",%%mm2\n\t" \
/*r1=H'+H'*/ \
"paddw %%mm1,%%mm1\n\t" \
/*r1=R1=A''+H'*/ \
@ -250,18 +227,18 @@ static const ogg_uint16_t __attribute__((aligned(8),used))
/*r1=NR1*/ \
"psraw $4,%%mm1\n\t" \
/*r3=D'*/ \
"movq "OC_I(2)",%%mm3\n\t" \
"movq "OC_I(2,_y)",%%mm3\n\t" \
/*r7=G+G*/ \
"paddw %%mm7,%%mm7\n\t" \
/*Store NR2 at I(2).*/ \
"movq %%mm2,"OC_I(2)"\n\t" \
"movq %%mm2,"OC_I(2,_y)"\n\t" \
/*r7=G'=E+G*/ \
"paddw %%mm4,%%mm7\n\t" \
/*Store NR1 at I(1).*/ \
"movq %%mm1,"OC_I(1)"\n\t" \
"movq %%mm1,"OC_I(1,_y)"\n\t" \
/*r4=R4=E'-D'*/ \
"psubw %%mm3,%%mm4\n\t" \
"paddw "OC_8",%%mm4\n\t" \
"paddw "OC_MEM_OFFS(0x00,c)",%%mm4\n\t" \
/*r3=D'+D'*/ \
"paddw %%mm3,%%mm3\n\t" \
/*r3=R3=E'+D'*/ \
@ -272,7 +249,7 @@ static const ogg_uint16_t __attribute__((aligned(8),used))
"psubw %%mm5,%%mm6\n\t" \
/*r3=NR3*/ \
"psraw $4,%%mm3\n\t" \
"paddw "OC_8",%%mm6\n\t" \
"paddw "OC_MEM_OFFS(0x00,c)",%%mm6\n\t" \
/*r5=B''+B''*/ \
"paddw %%mm5,%%mm5\n\t" \
/*r5=R5=F'+B''*/ \
@ -280,14 +257,14 @@ static const ogg_uint16_t __attribute__((aligned(8),used))
/*r6=NR6*/ \
"psraw $4,%%mm6\n\t" \
/*Store NR4 at J(4).*/ \
"movq %%mm4,"OC_J(4)"\n\t" \
"movq %%mm4,"OC_J(4,_y)"\n\t" \
/*r5=NR5*/ \
"psraw $4,%%mm5\n\t" \
/*Store NR3 at I(3).*/ \
"movq %%mm3,"OC_I(3)"\n\t" \
"movq %%mm3,"OC_I(3,_y)"\n\t" \
/*r7=R7=G'-C'*/ \
"psubw %%mm0,%%mm7\n\t" \
"paddw "OC_8",%%mm7\n\t" \
"paddw "OC_MEM_OFFS(0x00,c)",%%mm7\n\t" \
/*r0=C'+C'*/ \
"paddw %%mm0,%%mm0\n\t" \
/*r0=R0=G'+C'*/ \
@ -295,113 +272,123 @@ static const ogg_uint16_t __attribute__((aligned(8),used))
/*r7=NR7*/ \
"psraw $4,%%mm7\n\t" \
/*Store NR6 at J(6).*/ \
"movq %%mm6,"OC_J(6)"\n\t" \
"movq %%mm6,"OC_J(6,_y)"\n\t" \
/*r0=NR0*/ \
"psraw $4,%%mm0\n\t" \
/*Store NR5 at J(5).*/ \
"movq %%mm5,"OC_J(5)"\n\t" \
"movq %%mm5,"OC_J(5,_y)"\n\t" \
/*Store NR7 at J(7).*/ \
"movq %%mm7,"OC_J(7)"\n\t" \
"movq %%mm7,"OC_J(7,_y)"\n\t" \
/*Store NR0 at I(0).*/ \
"movq %%mm0,"OC_I(0)"\n\t" \
"movq %%mm0,"OC_I(0,_y)"\n\t" \
"#end OC_COLUMN_IDCT\n\t" \
#define OC_MID(_m,_i) OC_M2STR(_m+(_i)*8)"(%[c])"
#define OC_C(_i) OC_MID(OC_COSINE_OFFSET,_i-1)
#define OC_8 OC_MID(OC_EIGHT_OFFSET,0)
static void oc_idct8x8_slow(ogg_int16_t _y[64]){
static void oc_idct8x8_slow_mmx(ogg_int16_t _y[64],ogg_int16_t _x[64]){
/*This routine accepts an 8x8 matrix, but in partially transposed form.
Every 4x4 block is transposed.*/
__asm__ __volatile__(
#define OC_I(_k) OC_M2STR((_k*16))"(%[y])"
#define OC_J(_k) OC_M2STR(((_k-4)*16)+8)"(%[y])"
OC_ROW_IDCT
OC_TRANSPOSE
#define OC_I(_k,_y) OC_MEM_OFFS((_k)*16,_y)
#define OC_J(_k,_y) OC_MEM_OFFS(((_k)-4)*16+8,_y)
OC_ROW_IDCT(y,x)
OC_TRANSPOSE(y)
#undef OC_I
#undef OC_J
#define OC_I(_k) OC_M2STR((_k*16)+64)"(%[y])"
#define OC_J(_k) OC_M2STR(((_k-4)*16)+72)"(%[y])"
OC_ROW_IDCT
OC_TRANSPOSE
#define OC_I(_k,_y) OC_MEM_OFFS((_k)*16+64,_y)
#define OC_J(_k,_y) OC_MEM_OFFS(((_k)-4)*16+72,_y)
OC_ROW_IDCT(y,x)
OC_TRANSPOSE(y)
#undef OC_I
#undef OC_J
#define OC_I(_k) OC_M2STR((_k*16))"(%[y])"
#define OC_J(_k) OC_I(_k)
OC_COLUMN_IDCT
#define OC_I(_k,_y) OC_MEM_OFFS((_k)*16,_y)
#define OC_J(_k,_y) OC_I(_k,_y)
OC_COLUMN_IDCT(y)
#undef OC_I
#undef OC_J
#define OC_I(_k) OC_M2STR((_k*16)+8)"(%[y])"
#define OC_J(_k) OC_I(_k)
OC_COLUMN_IDCT
#define OC_I(_k,_y) OC_MEM_OFFS((_k)*16+8,_y)
#define OC_J(_k,_y) OC_I(_k,_y)
OC_COLUMN_IDCT(y)
#undef OC_I
#undef OC_J
:
:[y]"r"(_y),[c]"r"(OC_IDCT_CONSTS)
:[y]"=m"OC_ARRAY_OPERAND(ogg_int16_t,_y,64)
:[x]"m"OC_CONST_ARRAY_OPERAND(ogg_int16_t,_x,64),
[c]"m"OC_CONST_ARRAY_OPERAND(ogg_int16_t,OC_IDCT_CONSTS,128)
);
if(_x!=_y){
int i;
__asm__ __volatile__("pxor %%mm0,%%mm0\n\t"::);
for(i=0;i<4;i++){
__asm__ __volatile__(
"movq %%mm0,"OC_MEM_OFFS(0x00,x)"\n\t"
"movq %%mm0,"OC_MEM_OFFS(0x08,x)"\n\t"
"movq %%mm0,"OC_MEM_OFFS(0x10,x)"\n\t"
"movq %%mm0,"OC_MEM_OFFS(0x18,x)"\n\t"
:[x]"=m"OC_ARRAY_OPERAND(ogg_int16_t,_x+16*i,16)
);
}
}
}
/*25 cycles.*/
#define OC_IDCT_BEGIN_10 \
#define OC_IDCT_BEGIN_10(_y,_x) \
"#OC_IDCT_BEGIN_10\n\t" \
"movq "OC_I(3)",%%mm2\n\t" \
"movq "OC_I(3,_x)",%%mm2\n\t" \
"nop\n\t" \
"movq "OC_C(3)",%%mm6\n\t" \
"movq "OC_MEM_OFFS(0x30,c)",%%mm6\n\t" \
"movq %%mm2,%%mm4\n\t" \
"movq "OC_C(5)",%%mm1\n\t" \
"movq "OC_MEM_OFFS(0x50,c)",%%mm1\n\t" \
"pmulhw %%mm6,%%mm4\n\t" \
"movq "OC_I(1)",%%mm3\n\t" \
"movq "OC_I(1,_x)",%%mm3\n\t" \
"pmulhw %%mm2,%%mm1\n\t" \
"movq "OC_C(1)",%%mm0\n\t" \
"movq "OC_MEM_OFFS(0x10,c)",%%mm0\n\t" \
"paddw %%mm2,%%mm4\n\t" \
"pxor %%mm6,%%mm6\n\t" \
"paddw %%mm1,%%mm2\n\t" \
"movq "OC_I(2)",%%mm5\n\t" \
"movq "OC_I(2,_x)",%%mm5\n\t" \
"pmulhw %%mm3,%%mm0\n\t" \
"movq %%mm5,%%mm1\n\t" \
"paddw %%mm3,%%mm0\n\t" \
"pmulhw "OC_C(7)",%%mm3\n\t" \
"pmulhw "OC_MEM_OFFS(0x70,c)",%%mm3\n\t" \
"psubw %%mm2,%%mm6\n\t" \
"pmulhw "OC_C(2)",%%mm5\n\t" \
"pmulhw "OC_MEM_OFFS(0x20,c)",%%mm5\n\t" \
"psubw %%mm4,%%mm0\n\t" \
"movq "OC_I(2)",%%mm7\n\t" \
"movq "OC_I(2,_x)",%%mm7\n\t" \
"paddw %%mm4,%%mm4\n\t" \
"paddw %%mm5,%%mm7\n\t" \
"paddw %%mm0,%%mm4\n\t" \
"pmulhw "OC_C(6)",%%mm1\n\t" \
"pmulhw "OC_MEM_OFFS(0x60,c)",%%mm1\n\t" \
"psubw %%mm6,%%mm3\n\t" \
"movq %%mm4,"OC_I(1)"\n\t" \
"movq %%mm4,"OC_I(1,_y)"\n\t" \
"paddw %%mm6,%%mm6\n\t" \
"movq "OC_C(4)",%%mm4\n\t" \
"movq "OC_MEM_OFFS(0x40,c)",%%mm4\n\t" \
"paddw %%mm3,%%mm6\n\t" \
"movq %%mm3,%%mm5\n\t" \
"pmulhw %%mm4,%%mm3\n\t" \
"movq %%mm6,"OC_I(2)"\n\t" \
"movq %%mm6,"OC_I(2,_y)"\n\t" \
"movq %%mm0,%%mm2\n\t" \
"movq "OC_I(0)",%%mm6\n\t" \
"movq "OC_I(0,_x)",%%mm6\n\t" \
"pmulhw %%mm4,%%mm0\n\t" \
"paddw %%mm3,%%mm5\n\t" \
"paddw %%mm0,%%mm2\n\t" \
"psubw %%mm1,%%mm5\n\t" \
"pmulhw %%mm4,%%mm6\n\t" \
"paddw "OC_I(0)",%%mm6\n\t" \
"paddw "OC_I(0,_x)",%%mm6\n\t" \
"paddw %%mm1,%%mm1\n\t" \
"movq %%mm6,%%mm4\n\t" \
"paddw %%mm5,%%mm1\n\t" \
"psubw %%mm2,%%mm6\n\t" \
"paddw %%mm2,%%mm2\n\t" \
"movq "OC_I(1)",%%mm0\n\t" \
"movq "OC_I(1,_y)",%%mm0\n\t" \
"paddw %%mm6,%%mm2\n\t" \
"psubw %%mm1,%%mm2\n\t" \
"nop\n\t" \
"#end OC_IDCT_BEGIN_10\n\t" \
/*25+8=33 cycles.*/
#define OC_ROW_IDCT_10 \
#define OC_ROW_IDCT_10(_y,_x) \
"#OC_ROW_IDCT_10\n\t" \
OC_IDCT_BEGIN_10 \
OC_IDCT_BEGIN_10(_y,_x) \
/*r3=D'*/ \
"movq "OC_I(2)",%%mm3\n\t" \
"movq "OC_I(2,_y)",%%mm3\n\t" \
/*r4=E'=E-G*/ \
"psubw %%mm7,%%mm4\n\t" \
/*r1=H'+H'*/ \
@ -426,16 +413,16 @@ static void oc_idct8x8_slow(ogg_int16_t _y[64]){
"psubw %%mm0,%%mm7\n\t" \
"paddw %%mm0,%%mm0\n\t" \
/*Save R1.*/ \
"movq %%mm1,"OC_I(1)"\n\t" \
"movq %%mm1,"OC_I(1,_y)"\n\t" \
/*r0=R0=G'+C'*/ \
"paddw %%mm7,%%mm0\n\t" \
"#end OC_ROW_IDCT_10\n\t" \
/*25+19=44 cycles'*/
#define OC_COLUMN_IDCT_10 \
#define OC_COLUMN_IDCT_10(_y) \
"#OC_COLUMN_IDCT_10\n\t" \
OC_IDCT_BEGIN_10 \
"paddw "OC_8",%%mm2\n\t" \
OC_IDCT_BEGIN_10(_y,_y) \
"paddw "OC_MEM_OFFS(0x00,c)",%%mm2\n\t" \
/*r1=H'+H'*/ \
"paddw %%mm1,%%mm1\n\t" \
/*r1=R1=A''+H'*/ \
@ -447,18 +434,18 @@ static void oc_idct8x8_slow(ogg_int16_t _y[64]){
/*r1=NR1*/ \
"psraw $4,%%mm1\n\t" \
/*r3=D'*/ \
"movq "OC_I(2)",%%mm3\n\t" \
"movq "OC_I(2,_y)",%%mm3\n\t" \
/*r7=G+G*/ \
"paddw %%mm7,%%mm7\n\t" \
/*Store NR2 at I(2).*/ \
"movq %%mm2,"OC_I(2)"\n\t" \
"movq %%mm2,"OC_I(2,_y)"\n\t" \
/*r7=G'=E+G*/ \
"paddw %%mm4,%%mm7\n\t" \
/*Store NR1 at I(1).*/ \
"movq %%mm1,"OC_I(1)"\n\t" \
"movq %%mm1,"OC_I(1,_y)"\n\t" \
/*r4=R4=E'-D'*/ \
"psubw %%mm3,%%mm4\n\t" \
"paddw "OC_8",%%mm4\n\t" \
"paddw "OC_MEM_OFFS(0x00,c)",%%mm4\n\t" \
/*r3=D'+D'*/ \
"paddw %%mm3,%%mm3\n\t" \
/*r3=R3=E'+D'*/ \
@ -469,7 +456,7 @@ static void oc_idct8x8_slow(ogg_int16_t _y[64]){
"psubw %%mm5,%%mm6\n\t" \
/*r3=NR3*/ \
"psraw $4,%%mm3\n\t" \
"paddw "OC_8",%%mm6\n\t" \
"paddw "OC_MEM_OFFS(0x00,c)",%%mm6\n\t" \
/*r5=B''+B''*/ \
"paddw %%mm5,%%mm5\n\t" \
/*r5=R5=F'+B''*/ \
@ -477,14 +464,14 @@ static void oc_idct8x8_slow(ogg_int16_t _y[64]){
/*r6=NR6*/ \
"psraw $4,%%mm6\n\t" \
/*Store NR4 at J(4).*/ \
"movq %%mm4,"OC_J(4)"\n\t" \
"movq %%mm4,"OC_J(4,_y)"\n\t" \
/*r5=NR5*/ \
"psraw $4,%%mm5\n\t" \
/*Store NR3 at I(3).*/ \
"movq %%mm3,"OC_I(3)"\n\t" \
"movq %%mm3,"OC_I(3,_y)"\n\t" \
/*r7=R7=G'-C'*/ \
"psubw %%mm0,%%mm7\n\t" \
"paddw "OC_8",%%mm7\n\t" \
"paddw "OC_MEM_OFFS(0x00,c)",%%mm7\n\t" \
/*r0=C'+C'*/ \
"paddw %%mm0,%%mm0\n\t" \
/*r0=R0=G'+C'*/ \
@ -492,46 +479,57 @@ static void oc_idct8x8_slow(ogg_int16_t _y[64]){
/*r7=NR7*/ \
"psraw $4,%%mm7\n\t" \
/*Store NR6 at J(6).*/ \
"movq %%mm6,"OC_J(6)"\n\t" \
"movq %%mm6,"OC_J(6,_y)"\n\t" \
/*r0=NR0*/ \
"psraw $4,%%mm0\n\t" \
/*Store NR5 at J(5).*/ \
"movq %%mm5,"OC_J(5)"\n\t" \
"movq %%mm5,"OC_J(5,_y)"\n\t" \
/*Store NR7 at J(7).*/ \
"movq %%mm7,"OC_J(7)"\n\t" \
"movq %%mm7,"OC_J(7,_y)"\n\t" \
/*Store NR0 at I(0).*/ \
"movq %%mm0,"OC_I(0)"\n\t" \
"movq %%mm0,"OC_I(0,_y)"\n\t" \
"#end OC_COLUMN_IDCT_10\n\t" \
static void oc_idct8x8_10(ogg_int16_t _y[64]){
static void oc_idct8x8_10_mmx(ogg_int16_t _y[64],ogg_int16_t _x[64]){
__asm__ __volatile__(
#define OC_I(_k) OC_M2STR((_k*16))"(%[y])"
#define OC_J(_k) OC_M2STR(((_k-4)*16)+8)"(%[y])"
#define OC_I(_k,_y) OC_MEM_OFFS((_k)*16,_y)
#define OC_J(_k,_y) OC_MEM_OFFS(((_k)-4)*16+8,_y)
/*Done with dequant, descramble, and partial transpose.
Now do the iDCT itself.*/
OC_ROW_IDCT_10
OC_TRANSPOSE
OC_ROW_IDCT_10(y,x)
OC_TRANSPOSE(y)
#undef OC_I
#undef OC_J
#define OC_I(_k) OC_M2STR((_k*16))"(%[y])"
#define OC_J(_k) OC_I(_k)
OC_COLUMN_IDCT_10
#define OC_I(_k,_y) OC_MEM_OFFS((_k)*16,_y)
#define OC_J(_k,_y) OC_I(_k,_y)
OC_COLUMN_IDCT_10(y)
#undef OC_I
#undef OC_J
#define OC_I(_k) OC_M2STR((_k*16)+8)"(%[y])"
#define OC_J(_k) OC_I(_k)
OC_COLUMN_IDCT_10
#define OC_I(_k,_y) OC_MEM_OFFS((_k)*16+8,_y)
#define OC_J(_k,_y) OC_I(_k,_y)
OC_COLUMN_IDCT_10(y)
#undef OC_I
#undef OC_J
:
:[y]"r"(_y),[c]"r"(OC_IDCT_CONSTS)
:[y]"=m"OC_ARRAY_OPERAND(ogg_int16_t,_y,64)
:[x]"m"OC_CONST_ARRAY_OPERAND(ogg_int16_t,_x,64),
[c]"m"OC_CONST_ARRAY_OPERAND(ogg_int16_t,OC_IDCT_CONSTS,128)
);
if(_x!=_y){
__asm__ __volatile__(
"pxor %%mm0,%%mm0\n\t"
"movq %%mm0,"OC_MEM_OFFS(0x00,x)"\n\t"
"movq %%mm0,"OC_MEM_OFFS(0x10,x)"\n\t"
"movq %%mm0,"OC_MEM_OFFS(0x20,x)"\n\t"
"movq %%mm0,"OC_MEM_OFFS(0x30,x)"\n\t"
:[x]"+m"OC_ARRAY_OPERAND(ogg_int16_t,_x,28)
);
}
}
/*Performs an inverse 8x8 Type-II DCT transform.
The input is assumed to be scaled by a factor of 4 relative to orthonormal
version of the transform.*/
void oc_idct8x8_mmx(ogg_int16_t _y[64],int _last_zzi){
void oc_idct8x8_mmx(ogg_int16_t _y[64],ogg_int16_t _x[64],int _last_zzi){
/*_last_zzi is subtly different from an actual count of the number of
coefficients we decoded for this block.
It contains the value of zzi BEFORE the final token in the block was
@ -557,8 +555,8 @@ void oc_idct8x8_mmx(ogg_int16_t _y[64],int _last_zzi){
gets.
Needless to say we inherited this approach from VP3.*/
/*Then perform the iDCT.*/
if(_last_zzi<10)oc_idct8x8_10(_y);
else oc_idct8x8_slow(_y);
if(_last_zzi<=10)oc_idct8x8_10_mmx(_y,_x);
else oc_idct8x8_slow_mmx(_y,_x);
}
#endif

271
media/libtheora/lib/x86/mmxloop.h
View File

@ -9,88 +9,191 @@
On exit, mm1={b0+lflim(R_0,L),...,b7+lflim(R_7,L)} and
mm2={c0-lflim(R_0,L),...,c7-lflim(R_7,L)}; mm0 and mm3 are clobbered.*/
#define OC_LOOP_FILTER8_MMX \
"#OC_LOOP_FILTER8_MMX\n\t" \
/*mm7=0*/ \
"pxor %%mm7,%%mm7\n\t" \
/*mm6:mm0={a0,...,a7}*/ \
"movq %%mm0,%%mm6\n\t" \
"punpcklbw %%mm7,%%mm0\n\t" \
"punpckhbw %%mm7,%%mm6\n\t" \
/*mm3:mm5={d0,...,d7}*/ \
"movq %%mm3,%%mm5\n\t" \
"punpcklbw %%mm7,%%mm3\n\t" \
"punpckhbw %%mm7,%%mm5\n\t" \
/*mm6:mm0={a0-d0,...,a7-d7}*/ \
"psubw %%mm3,%%mm0\n\t" \
"psubw %%mm5,%%mm6\n\t" \
/*mm3:mm1={b0,...,b7}*/ \
"movq %%mm1,%%mm3\n\t" \
"punpcklbw %%mm7,%%mm1\n\t" \
"movq %%mm2,%%mm4\n\t" \
"punpckhbw %%mm7,%%mm3\n\t" \
/*mm5:mm4={c0,...,c7}*/ \
"movq %%mm2,%%mm5\n\t" \
"punpcklbw %%mm7,%%mm4\n\t" \
"punpckhbw %%mm7,%%mm5\n\t" \
/*mm7={3}x4 \
mm5:mm4={c0-b0,...,c7-b7}*/ \
"pcmpeqw %%mm7,%%mm7\n\t" \
"psubw %%mm1,%%mm4\n\t" \
"psrlw $14,%%mm7\n\t" \
"psubw %%mm3,%%mm5\n\t" \
/*Scale by 3.*/ \
"pmullw %%mm7,%%mm4\n\t" \
"pmullw %%mm7,%%mm5\n\t" \
/*mm7={4}x4 \
mm5:mm4=f={a0-d0+3*(c0-b0),...,a7-d7+3*(c7-b7)}*/ \
"psrlw $1,%%mm7\n\t" \
"paddw %%mm0,%%mm4\n\t" \
"psllw $2,%%mm7\n\t" \
"movq (%[ll]),%%mm0\n\t" \
"paddw %%mm6,%%mm5\n\t" \
/*R_i has the range [-127,128], so we compute -R_i instead. \
mm4=-R_i=-(f+4>>3)=0xFF^(f-4>>3)*/ \
"psubw %%mm7,%%mm4\n\t" \
"psubw %%mm7,%%mm5\n\t" \
"psraw $3,%%mm4\n\t" \
"psraw $3,%%mm5\n\t" \
"pcmpeqb %%mm7,%%mm7\n\t" \
"packsswb %%mm5,%%mm4\n\t" \
"pxor %%mm6,%%mm6\n\t" \
"pxor %%mm7,%%mm4\n\t" \
"packuswb %%mm3,%%mm1\n\t" \
/*Now compute lflim of -mm4 cf. Section 7.10 of the sepc.*/ \
/*There's no unsigned byte+signed byte with unsigned saturation op code, so \
we have to split things by sign (the other option is to work in 16 bits, \
but working in 8 bits gives much better parallelism). \
We compute abs(R_i), but save a mask of which terms were negative in mm6. \
Then we compute mm4=abs(lflim(R_i,L))=min(abs(R_i),max(2*L-abs(R_i),0)). \
Finally, we split mm4 into positive and negative pieces using the mask in \
mm6, and add and subtract them as appropriate.*/ \
/*mm4=abs(-R_i)*/ \
/*mm7=255-2*L*/ \
"pcmpgtb %%mm4,%%mm6\n\t" \
"psubb %%mm0,%%mm7\n\t" \
"pxor %%mm6,%%mm4\n\t" \
"psubb %%mm0,%%mm7\n\t" \
"psubb %%mm6,%%mm4\n\t" \
/*mm7=255-max(2*L-abs(R_i),0)*/ \
"paddusb %%mm4,%%mm7\n\t" \
/*mm4=min(abs(R_i),max(2*L-abs(R_i),0))*/ \
"paddusb %%mm7,%%mm4\n\t" \
"psubusb %%mm7,%%mm4\n\t" \
/*Now split mm4 by the original sign of -R_i.*/ \
"movq %%mm4,%%mm5\n\t" \
"pand %%mm6,%%mm4\n\t" \
"pandn %%mm5,%%mm6\n\t" \
/*mm1={b0+lflim(R_0,L),...,b7+lflim(R_7,L)}*/ \
/*mm2={c0-lflim(R_0,L),...,c7-lflim(R_7,L)}*/ \
"paddusb %%mm4,%%mm1\n\t" \
"psubusb %%mm4,%%mm2\n\t" \
"psubusb %%mm6,%%mm1\n\t" \
"paddusb %%mm6,%%mm2\n\t" \
"#OC_LOOP_FILTER8_MMX\n\t" \
/*mm7=0*/ \
"pxor %%mm7,%%mm7\n\t" \
/*mm6:mm0={a0,...,a7}*/ \
"movq %%mm0,%%mm6\n\t" \
"punpcklbw %%mm7,%%mm0\n\t" \
"punpckhbw %%mm7,%%mm6\n\t" \
/*mm3:mm5={d0,...,d7}*/ \
"movq %%mm3,%%mm5\n\t" \
"punpcklbw %%mm7,%%mm3\n\t" \
"punpckhbw %%mm7,%%mm5\n\t" \
/*mm6:mm0={a0-d0,...,a7-d7}*/ \
"psubw %%mm3,%%mm0\n\t" \
"psubw %%mm5,%%mm6\n\t" \
/*mm3:mm1={b0,...,b7}*/ \
"movq %%mm1,%%mm3\n\t" \
"punpcklbw %%mm7,%%mm1\n\t" \
"movq %%mm2,%%mm4\n\t" \
"punpckhbw %%mm7,%%mm3\n\t" \
/*mm5:mm4={c0,...,c7}*/ \
"movq %%mm2,%%mm5\n\t" \
"punpcklbw %%mm7,%%mm4\n\t" \
"punpckhbw %%mm7,%%mm5\n\t" \
/*mm7={3}x4 \
mm5:mm4={c0-b0,...,c7-b7}*/ \
"pcmpeqw %%mm7,%%mm7\n\t" \
"psubw %%mm1,%%mm4\n\t" \
"psrlw $14,%%mm7\n\t" \
"psubw %%mm3,%%mm5\n\t" \
/*Scale by 3.*/ \
"pmullw %%mm7,%%mm4\n\t" \
"pmullw %%mm7,%%mm5\n\t" \
/*mm7={4}x4 \
mm5:mm4=f={a0-d0+3*(c0-b0),...,a7-d7+3*(c7-b7)}*/ \
"psrlw $1,%%mm7\n\t" \
"paddw %%mm0,%%mm4\n\t" \
"psllw $2,%%mm7\n\t" \
"movq (%[ll]),%%mm0\n\t" \
"paddw %%mm6,%%mm5\n\t" \
/*R_i has the range [-127,128], so we compute -R_i instead. \
mm4=-R_i=-(f+4>>3)=0xFF^(f-4>>3)*/ \
"psubw %%mm7,%%mm4\n\t" \
"psubw %%mm7,%%mm5\n\t" \
"psraw $3,%%mm4\n\t" \
"psraw $3,%%mm5\n\t" \
"pcmpeqb %%mm7,%%mm7\n\t" \
"packsswb %%mm5,%%mm4\n\t" \
"pxor %%mm6,%%mm6\n\t" \
"pxor %%mm7,%%mm4\n\t" \
"packuswb %%mm3,%%mm1\n\t" \
/*Now compute lflim of -mm4 cf. Section 7.10 of the sepc.*/ \
/*There's no unsigned byte+signed byte with unsigned saturation op code, so \
we have to split things by sign (the other option is to work in 16 bits, \
but working in 8 bits gives much better parallelism). \
We compute abs(R_i), but save a mask of which terms were negative in mm6. \
Then we compute mm4=abs(lflim(R_i,L))=min(abs(R_i),max(2*L-abs(R_i),0)). \
Finally, we split mm4 into positive and negative pieces using the mask in \
mm6, and add and subtract them as appropriate.*/ \
/*mm4=abs(-R_i)*/ \
/*mm7=255-2*L*/ \
"pcmpgtb %%mm4,%%mm6\n\t" \
"psubb %%mm0,%%mm7\n\t" \
"pxor %%mm6,%%mm4\n\t" \
"psubb %%mm0,%%mm7\n\t" \
"psubb %%mm6,%%mm4\n\t" \
/*mm7=255-max(2*L-abs(R_i),0)*/ \
"paddusb %%mm4,%%mm7\n\t" \
/*mm4=min(abs(R_i),max(2*L-abs(R_i),0))*/ \
"paddusb %%mm7,%%mm4\n\t" \
"psubusb %%mm7,%%mm4\n\t" \
/*Now split mm4 by the original sign of -R_i.*/ \
"movq %%mm4,%%mm5\n\t" \
"pand %%mm6,%%mm4\n\t" \
"pandn %%mm5,%%mm6\n\t" \
/*mm1={b0+lflim(R_0,L),...,b7+lflim(R_7,L)}*/ \
/*mm2={c0-lflim(R_0,L),...,c7-lflim(R_7,L)}*/ \
"paddusb %%mm4,%%mm1\n\t" \
"psubusb %%mm4,%%mm2\n\t" \
"psubusb %%mm6,%%mm1\n\t" \
"paddusb %%mm6,%%mm2\n\t" \
#define OC_LOOP_FILTER_V_MMX(_pix,_ystride,_ll) \
/*On entry, mm0={a0,...,a7}, mm1={b0,...,b7}, mm2={c0,...,c7}, mm3={d0,...d7}.
On exit, mm1={b0+lflim(R_0,L),...,b7+lflim(R_7,L)} and
mm2={c0-lflim(R_0,L),...,c7-lflim(R_7,L)}.
All other MMX registers are clobbered.*/
#define OC_LOOP_FILTER8_MMXEXT \
"#OC_LOOP_FILTER8_MMXEXT\n\t" \
/*R_i=(a_i-3*b_i+3*c_i-d_i+4>>3) has the range [-127,128], so we compute \
-R_i=(-a_i+3*b_i-3*c_i+d_i+3>>3) instead.*/ \
/*This first part is based on the transformation \
f = -(3*(c-b)+a-d+4>>3) \
= -(3*(c+255-b)+(a+255-d)+4-1020>>3) \
= -(3*(c+~b)+(a+~d)-1016>>3) \
= 127-(3*(c+~b)+(a+~d)>>3) \
= 128+~(3*(c+~b)+(a+~d)>>3) (mod 256). \
Although pavgb(a,b) = (a+b+1>>1) (biased up), we rely heavily on the \
fact that ~pavgb(~a,~b) = (a+b>>1) (biased down). \
Using this, the last expression above can be computed in 8 bits of working \
precision via: \
u = ~pavgb(~b,c); \
v = pavgb(b,~c); \
This mask is 0 or 0xFF, and controls whether t is biased up or down: \
m = u-v; \
t = m^pavgb(m^~a,m^d); \
f = 128+pavgb(pavgb(t,u),v); \
This required some careful analysis to ensure that carries are propagated \
correctly in all cases, but has been checked exhaustively.*/ \
/*input (a, b, c, d, ., ., ., .)*/ \
/*ff=0xFF; \
u=b; \
v=c; \
ll=255-2*L;*/ \
"pcmpeqb %%mm7,%%mm7\n\t" \
"movq %%mm1,%%mm4\n\t" \
"movq %%mm2,%%mm5\n\t" \
"movq (%[ll]),%%mm6\n\t" \
/*allocated u, v, ll, ff: (a, b, c, d, u, v, ll, ff)*/ \
/*u^=ff; \
v^=ff;*/ \
"pxor %%mm7,%%mm4\n\t" \
"pxor %%mm7,%%mm5\n\t" \
/*allocated ll: (a, b, c, d, u, v, ll, ff)*/ \
/*u=pavgb(u,c); \
v=pavgb(v,b);*/ \
"pavgb %%mm2,%%mm4\n\t" \
"pavgb %%mm1,%%mm5\n\t" \
/*u^=ff; \
a^=ff;*/ \
"pxor %%mm7,%%mm4\n\t" \
"pxor %%mm7,%%mm0\n\t" \
/*m=u-v;*/ \
"psubb %%mm5,%%mm4\n\t" \
/*freed u, allocated m: (a, b, c, d, m, v, ll, ff)*/ \
/*a^=m; \
d^=m;*/ \
"pxor %%mm4,%%mm0\n\t" \
"pxor %%mm4,%%mm3\n\t" \
/*t=pavgb(a,d);*/ \
"pavgb %%mm3,%%mm0\n\t" \
"psllw $7,%%mm7\n\t" \
/*freed a, d, ff, allocated t, of: (t, b, c, ., m, v, ll, of)*/ \
/*t^=m; \
u=m+v;*/ \
"pxor %%mm4,%%mm0\n\t" \
"paddb %%mm5,%%mm4\n\t" \
/*freed t, m, allocated f, u: (f, b, c, ., u, v, ll, of)*/ \
/*f=pavgb(f,u); \
of=128;*/ \
"pavgb %%mm4,%%mm0\n\t" \
"packsswb %%mm7,%%mm7\n\t" \
/*freed u, ff, allocated ll: (f, b, c, ., ll, v, ll, of)*/ \
/*f=pavgb(f,v);*/ \
"pavgb %%mm5,%%mm0\n\t" \
"movq %%mm7,%%mm3\n\t" \
"movq %%mm6,%%mm4\n\t" \
/*freed v, allocated of: (f, b, c, of, ll, ., ll, of)*/ \
/*Now compute lflim of R_i=-(128+mm0) cf. Section 7.10 of the sepc.*/ \
/*There's no unsigned byte+signed byte with unsigned saturation op code, so \
we have to split things by sign (the other option is to work in 16 bits, \
but staying in 8 bits gives much better parallelism).*/ \
/*Instead of adding the offset of 128 in mm3, we use it to split mm0. \
This is the same number of instructions as computing a mask and splitting \
after the lflim computation, but has shorter dependency chains.*/ \
/*mm0=R_i<0?-R_i:0 (denoted abs(R_i<0))\
mm3=R_i>0?R_i:0* (denoted abs(R_i>0))*/ \
"psubusb %%mm0,%%mm3\n\t" \
"psubusb %%mm7,%%mm0\n\t" \
/*mm6=255-max(2*L-abs(R_i<0),0) \
mm4=255-max(2*L-abs(R_i>0),0)*/ \
"paddusb %%mm3,%%mm4\n\t" \
"paddusb %%mm0,%%mm6\n\t" \
/*mm0=min(abs(R_i<0),max(2*L-abs(R_i<0),0)) \
mm3=min(abs(R_i>0),max(2*L-abs(R_i>0),0))*/ \
"paddusb %%mm4,%%mm3\n\t" \
"paddusb %%mm6,%%mm0\n\t" \
"psubusb %%mm4,%%mm3\n\t" \
"psubusb %%mm6,%%mm0\n\t" \
/*mm1={b0+lflim(R_0,L),...,b7+lflim(R_7,L)}*/ \
/*mm2={c0-lflim(R_0,L),...,c7-lflim(R_7,L)}*/ \
"paddusb %%mm3,%%mm1\n\t" \
"psubusb %%mm3,%%mm2\n\t" \
"psubusb %%mm0,%%mm1\n\t" \
"paddusb %%mm0,%%mm2\n\t" \
#define OC_LOOP_FILTER_V(_filter,_pix,_ystride,_ll) \
do{ \
ptrdiff_t ystride3__; \
__asm__ __volatile__( \
@ -104,7 +207,7 @@
"movq (%[pix],%[ystride]),%%mm1\n\t" \
/*mm2={c0,...,c7}*/ \
"movq (%[pix],%[ystride],2),%%mm2\n\t" \
OC_LOOP_FILTER8_MMX \
_filter \
/*Write it back out.*/ \
"movq %%mm1,(%[pix],%[ystride])\n\t" \
"movq %%mm2,(%[pix],%[ystride],2)\n\t" \
@ -116,7 +219,7 @@
} \
while(0)
#define OC_LOOP_FILTER_H_MMX(_pix,_ystride,_ll) \
#define OC_LOOP_FILTER_H(_filter,_pix,_ystride,_ll) \
do{ \
unsigned char *pix__; \
ptrdiff_t ystride3__; \
@ -174,7 +277,7 @@
"punpckldq %%mm5,%%mm2\n\t" \
/*mm3=d7 d6 d5 d4 d3 d2 d1 d0*/ \
"punpckhdq %%mm5,%%mm3\n\t" \
OC_LOOP_FILTER8_MMX \
_filter \
/*mm2={b0+R_0'',...,b7+R_7''}*/ \
"movq %%mm1,%%mm0\n\t" \
/*mm1={b0+R_0'',c0-R_0'',...,b3+R_3'',c3-R_3''}*/ \

162
media/libtheora/lib/x86/mmxstate.c
View File

@ -11,7 +11,7 @@
********************************************************************
function:
last mod: $Id: mmxstate.c 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: mmxstate.c 17563 2010-10-25 17:40:54Z tterribe $
********************************************************************/
@ -19,23 +19,23 @@
Originally written by Rudolf Marek.*/
#include <string.h>
#include "x86int.h"
#include "mmxfrag.h"
#include "mmxloop.h"
#if defined(OC_X86_ASM)
void oc_state_frag_recon_mmx(const oc_theora_state *_state,ptrdiff_t _fragi,
int _pli,ogg_int16_t _dct_coeffs[64],int _last_zzi,ogg_uint16_t _dc_quant){
int _pli,ogg_int16_t _dct_coeffs[128],int _last_zzi,ogg_uint16_t _dc_quant){
unsigned char *dst;
ptrdiff_t frag_buf_off;
int ystride;
int mb_mode;
int refi;
/*Apply the inverse transform.*/
/*Special case only having a DC component.*/
if(_last_zzi<2){
/*Note that this value must be unsigned, to keep the __asm__ block from
sign-extending it when it puts it in a register.*/
ogg_uint16_t p;
int i;
/*We round this dequant product (and not any of the others) because there's
no iDCT rounding.*/
p=(ogg_int16_t)(_dct_coeffs[0]*(ogg_int32_t)_dc_quant+15>>5);
@ -47,81 +47,48 @@ void oc_state_frag_recon_mmx(const oc_theora_state *_state,ptrdiff_t _fragi,
"punpcklwd %%mm0,%%mm0\n\t"
/*mm0=AAAA AAAA AAAA AAAA*/
"punpckldq %%mm0,%%mm0\n\t"
"movq %%mm0,(%[y])\n\t"
"movq %%mm0,8(%[y])\n\t"
"movq %%mm0,16(%[y])\n\t"
"movq %%mm0,24(%[y])\n\t"
"movq %%mm0,32(%[y])\n\t"
"movq %%mm0,40(%[y])\n\t"
"movq %%mm0,48(%[y])\n\t"
"movq %%mm0,56(%[y])\n\t"
"movq %%mm0,64(%[y])\n\t"
"movq %%mm0,72(%[y])\n\t"
"movq %%mm0,80(%[y])\n\t"
"movq %%mm0,88(%[y])\n\t"
"movq %%mm0,96(%[y])\n\t"
"movq %%mm0,104(%[y])\n\t"
"movq %%mm0,112(%[y])\n\t"
"movq %%mm0,120(%[y])\n\t"
:
:[y]"r"(_dct_coeffs),[p]"r"((unsigned)p)
:"memory"
:[p]"r"((unsigned)p)
);
for(i=0;i<4;i++){
__asm__ __volatile__(
"movq %%mm0,"OC_MEM_OFFS(0x00,y)"\n\t"
"movq %%mm0,"OC_MEM_OFFS(0x08,y)"\n\t"
"movq %%mm0,"OC_MEM_OFFS(0x10,y)"\n\t"
"movq %%mm0,"OC_MEM_OFFS(0x18,y)"\n\t"
:[y]"=m"OC_ARRAY_OPERAND(ogg_int16_t,_dct_coeffs+64+16*i,16)
);
}
}
else{
/*Dequantize the DC coefficient.*/
_dct_coeffs[0]=(ogg_int16_t)(_dct_coeffs[0]*(int)_dc_quant);
oc_idct8x8_mmx(_dct_coeffs,_last_zzi);
oc_idct8x8(_state,_dct_coeffs+64,_dct_coeffs,_last_zzi);
}
/*Fill in the target buffer.*/
frag_buf_off=_state->frag_buf_offs[_fragi];
mb_mode=_state->frags[_fragi].mb_mode;
refi=_state->frags[_fragi].refi;
ystride=_state->ref_ystride[_pli];
dst=_state->ref_frame_data[_state->ref_frame_idx[OC_FRAME_SELF]]+frag_buf_off;
if(mb_mode==OC_MODE_INTRA)oc_frag_recon_intra_mmx(dst,ystride,_dct_coeffs);
dst=_state->ref_frame_data[OC_FRAME_SELF]+frag_buf_off;
if(refi==OC_FRAME_SELF)oc_frag_recon_intra_mmx(dst,ystride,_dct_coeffs+64);
else{
const unsigned char *ref;
int mvoffsets[2];
ref=
_state->ref_frame_data[_state->ref_frame_idx[OC_FRAME_FOR_MODE(mb_mode)]]
+frag_buf_off;
ref=_state->ref_frame_data[refi]+frag_buf_off;
if(oc_state_get_mv_offsets(_state,mvoffsets,_pli,
_state->frag_mvs[_fragi][0],_state->frag_mvs[_fragi][1])>1){
_state->frag_mvs[_fragi])>1){
oc_frag_recon_inter2_mmx(dst,ref+mvoffsets[0],ref+mvoffsets[1],ystride,
_dct_coeffs);
_dct_coeffs+64);
}
else oc_frag_recon_inter_mmx(dst,ref+mvoffsets[0],ystride,_dct_coeffs);
else oc_frag_recon_inter_mmx(dst,ref+mvoffsets[0],ystride,_dct_coeffs+64);
}
}
/*We copy these entire function to inline the actual MMX routines so that we
use only a single indirect call.*/
/*Copies the fragments specified by the lists of fragment indices from one
frame to another.
_fragis: A pointer to a list of fragment indices.
_nfragis: The number of fragment indices to copy.
_dst_frame: The reference frame to copy to.
_src_frame: The reference frame to copy from.
_pli: The color plane the fragments lie in.*/
void oc_state_frag_copy_list_mmx(const oc_theora_state *_state,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,
int _dst_frame,int _src_frame,int _pli){
const ptrdiff_t *frag_buf_offs;
const unsigned char *src_frame_data;
unsigned char *dst_frame_data;
ptrdiff_t fragii;
int ystride;
dst_frame_data=_state->ref_frame_data[_state->ref_frame_idx[_dst_frame]];
src_frame_data=_state->ref_frame_data[_state->ref_frame_idx[_src_frame]];
ystride=_state->ref_ystride[_pli];
frag_buf_offs=_state->frag_buf_offs;
for(fragii=0;fragii<_nfragis;fragii++){
ptrdiff_t frag_buf_off;
frag_buf_off=frag_buf_offs[_fragis[fragii]];
OC_FRAG_COPY_MMX(dst_frame_data+frag_buf_off,
src_frame_data+frag_buf_off,ystride);
}
void oc_loop_filter_init_mmx(signed char _bv[256],int _flimit){
memset(_bv,_flimit,8);
}
/*Apply the loop filter to a given set of fragment rows in the given plane.
@ -133,7 +100,7 @@ void oc_state_frag_copy_list_mmx(const oc_theora_state *_state,
_fragy0: The Y coordinate of the first fragment row to filter.
_fragy_end: The Y coordinate of the fragment row to stop filtering at.*/
void oc_state_loop_filter_frag_rows_mmx(const oc_theora_state *_state,
int _bv[256],int _refi,int _pli,int _fragy0,int _fragy_end){
signed char _bv[256],int _refi,int _pli,int _fragy0,int _fragy_end){
OC_ALIGN8(unsigned char ll[8]);
const oc_fragment_plane *fplane;
const oc_fragment *frags;
@ -170,13 +137,84 @@ void oc_state_loop_filter_frag_rows_mmx(const oc_theora_state *_state,
if(frags[fragi].coded){
unsigned char *ref;
ref=ref_frame_data+frag_buf_offs[fragi];
if(fragi>fragi0)OC_LOOP_FILTER_H_MMX(ref,ystride,ll);
if(fragi0>fragi_top)OC_LOOP_FILTER_V_MMX(ref,ystride,ll);
if(fragi>fragi0){
OC_LOOP_FILTER_H(OC_LOOP_FILTER8_MMX,ref,ystride,ll);
}
if(fragi0>fragi_top){
OC_LOOP_FILTER_V(OC_LOOP_FILTER8_MMX,ref,ystride,ll);
}
if(fragi+1<fragi_end&&!frags[fragi+1].coded){
OC_LOOP_FILTER_H_MMX(ref+8,ystride,ll);
OC_LOOP_FILTER_H(OC_LOOP_FILTER8_MMX,ref+8,ystride,ll);
}
if(fragi+nhfrags<fragi_bot&&!frags[fragi+nhfrags].coded){
OC_LOOP_FILTER_V_MMX(ref+(ystride<<3),ystride,ll);
OC_LOOP_FILTER_V(OC_LOOP_FILTER8_MMX,ref+(ystride<<3),ystride,ll);
}
}
fragi++;
}
fragi0+=nhfrags;
}
}
void oc_loop_filter_init_mmxext(signed char _bv[256],int _flimit){
memset(_bv,~(_flimit<<1),8);
}
/*Apply the loop filter to a given set of fragment rows in the given plane.
The filter may be run on the bottom edge, affecting pixels in the next row of
fragments, so this row also needs to be available.
_bv: The bounding values array.
_refi: The index of the frame buffer to filter.
_pli: The color plane to filter.
_fragy0: The Y coordinate of the first fragment row to filter.
_fragy_end: The Y coordinate of the fragment row to stop filtering at.*/
void oc_state_loop_filter_frag_rows_mmxext(const oc_theora_state *_state,
signed char _bv[256],int _refi,int _pli,int _fragy0,int _fragy_end){
const oc_fragment_plane *fplane;
const oc_fragment *frags;
const ptrdiff_t *frag_buf_offs;
unsigned char *ref_frame_data;
ptrdiff_t fragi_top;
ptrdiff_t fragi_bot;
ptrdiff_t fragi0;
ptrdiff_t fragi0_end;
int ystride;
int nhfrags;
fplane=_state->fplanes+_pli;
nhfrags=fplane->nhfrags;
fragi_top=fplane->froffset;
fragi_bot=fragi_top+fplane->nfrags;
fragi0=fragi_top+_fragy0*(ptrdiff_t)nhfrags;
fragi0_end=fragi_top+_fragy_end*(ptrdiff_t)nhfrags;
ystride=_state->ref_ystride[_pli];
frags=_state->frags;
frag_buf_offs=_state->frag_buf_offs;
ref_frame_data=_state->ref_frame_data[_refi];
/*The following loops are constructed somewhat non-intuitively on purpose.
The main idea is: if a block boundary has at least one coded fragment on
it, the filter is applied to it.
However, the order that the filters are applied in matters, and VP3 chose
the somewhat strange ordering used below.*/
while(fragi0<fragi0_end){
ptrdiff_t fragi;
ptrdiff_t fragi_end;
fragi=fragi0;
fragi_end=fragi+nhfrags;
while(fragi<fragi_end){
if(frags[fragi].coded){
unsigned char *ref;
ref=ref_frame_data+frag_buf_offs[fragi];
if(fragi>fragi0){
OC_LOOP_FILTER_H(OC_LOOP_FILTER8_MMXEXT,ref,ystride,_bv);
}
if(fragi0>fragi_top){
OC_LOOP_FILTER_V(OC_LOOP_FILTER8_MMXEXT,ref,ystride,_bv);
}
if(fragi+1<fragi_end&&!frags[fragi+1].coded){
OC_LOOP_FILTER_H(OC_LOOP_FILTER8_MMXEXT,ref+8,ystride,_bv);
}
if(fragi+nhfrags<fragi_bot&&!frags[fragi+nhfrags].coded){
OC_LOOP_FILTER_V(OC_LOOP_FILTER8_MMXEXT,ref+(ystride<<3),ystride,_bv);
}
}
fragi++;

460
media/libtheora/lib/x86/sse2idct.c Normal file
View File

@ -0,0 +1,460 @@
/********************************************************************
* *
* THIS FILE IS PART OF THE OggTheora SOFTWARE CODEC SOURCE CODE. *
* USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS *
* GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
* IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING. *
* *
* THE Theora SOURCE CODE IS COPYRIGHT (C) 2002-2009 *
* by the Xiph.Org Foundation and contributors http://www.xiph.org/ *
* *
********************************************************************
function:
last mod: $Id: mmxidct.c 16503 2009-08-22 18:14:02Z giles $
********************************************************************/
/*SSE2 acceleration of Theora's iDCT.*/
#include "x86int.h"
#include "sse2trans.h"
#include "../dct.h"
#if defined(OC_X86_ASM)
/*A table of constants used by the MMX routines.*/
const unsigned short __attribute__((aligned(16),used)) OC_IDCT_CONSTS[64]={
8, 8, 8, 8, 8, 8, 8, 8,
OC_C1S7,OC_C1S7,OC_C1S7,OC_C1S7,OC_C1S7,OC_C1S7,OC_C1S7,OC_C1S7,
OC_C2S6,OC_C2S6,OC_C2S6,OC_C2S6,OC_C2S6,OC_C2S6,OC_C2S6,OC_C2S6,
OC_C3S5,OC_C3S5,OC_C3S5,OC_C3S5,OC_C3S5,OC_C3S5,OC_C3S5,OC_C3S5,
OC_C4S4,OC_C4S4,OC_C4S4,OC_C4S4,OC_C4S4,OC_C4S4,OC_C4S4,OC_C4S4,
OC_C5S3,OC_C5S3,OC_C5S3,OC_C5S3,OC_C5S3,OC_C5S3,OC_C5S3,OC_C5S3,
OC_C6S2,OC_C6S2,OC_C6S2,OC_C6S2,OC_C6S2,OC_C6S2,OC_C6S2,OC_C6S2,
OC_C7S1,OC_C7S1,OC_C7S1,OC_C7S1,OC_C7S1,OC_C7S1,OC_C7S1,OC_C7S1
};
/*Performs the first three stages of the iDCT.
xmm2, xmm6, xmm3, and xmm5 must contain the corresponding rows of the input
(accessed in that order).
The remaining rows must be in _x at their corresponding locations.
On output, xmm7 down to xmm4 contain rows 0 through 3, and xmm0 up to xmm3
contain rows 4 through 7.*/
#define OC_IDCT_8x8_ABC(_x) \
"#OC_IDCT_8x8_ABC\n\t" \
/*Stage 1:*/ \
/*2-3 rotation by 6pi/16. \
xmm4=xmm7=C6, xmm0=xmm1=C2, xmm2=X2, xmm6=X6.*/ \
"movdqa "OC_MEM_OFFS(0x20,c)",%%xmm1\n\t" \
"movdqa "OC_MEM_OFFS(0x60,c)",%%xmm4\n\t" \
"movdqa %%xmm1,%%xmm0\n\t" \
"pmulhw %%xmm2,%%xmm1\n\t" \
"movdqa %%xmm4,%%xmm7\n\t" \
"pmulhw %%xmm6,%%xmm0\n\t" \
"pmulhw %%xmm2,%%xmm7\n\t" \
"pmulhw %%xmm6,%%xmm4\n\t" \
"paddw %%xmm6,%%xmm0\n\t" \
"movdqa "OC_MEM_OFFS(0x30,c)",%%xmm6\n\t" \
"paddw %%xmm1,%%xmm2\n\t" \
"psubw %%xmm0,%%xmm7\n\t" \
"movdqa %%xmm7,"OC_MEM_OFFS(0x00,buf)"\n\t" \
"paddw %%xmm4,%%xmm2\n\t" \
"movdqa "OC_MEM_OFFS(0x50,c)",%%xmm4\n\t" \
"movdqa %%xmm2,"OC_MEM_OFFS(0x10,buf)"\n\t" \
/*5-6 rotation by 3pi/16. \
xmm4=xmm2=C5, xmm1=xmm6=C3, xmm3=X3, xmm5=X5.*/ \
"movdqa %%xmm4,%%xmm2\n\t" \
"movdqa %%xmm6,%%xmm1\n\t" \
"pmulhw %%xmm3,%%xmm4\n\t" \
"pmulhw %%xmm5,%%xmm1\n\t" \
"pmulhw %%xmm3,%%xmm6\n\t" \
"pmulhw %%xmm5,%%xmm2\n\t" \
"paddw %%xmm3,%%xmm4\n\t" \
"paddw %%xmm5,%%xmm3\n\t" \
"paddw %%xmm6,%%xmm3\n\t" \
"movdqa "OC_MEM_OFFS(0x70,_x)",%%xmm6\n\t" \
"paddw %%xmm5,%%xmm1\n\t" \
"movdqa "OC_MEM_OFFS(0x10,_x)",%%xmm5\n\t" \
"paddw %%xmm3,%%xmm2\n\t" \
"movdqa "OC_MEM_OFFS(0x70,c)",%%xmm3\n\t" \
"psubw %%xmm4,%%xmm1\n\t" \
"movdqa "OC_MEM_OFFS(0x10,c)",%%xmm4\n\t" \
/*4-7 rotation by 7pi/16. \
xmm4=xmm7=C1, xmm3=xmm0=C7, xmm5=X1, xmm6=X7.*/ \
"movdqa %%xmm3,%%xmm0\n\t" \
"movdqa %%xmm4,%%xmm7\n\t" \
"pmulhw %%xmm5,%%xmm3\n\t" \
"pmulhw %%xmm5,%%xmm7\n\t" \
"pmulhw %%xmm6,%%xmm4\n\t" \
"pmulhw %%xmm6,%%xmm0\n\t" \
"paddw %%xmm6,%%xmm4\n\t" \
"movdqa "OC_MEM_OFFS(0x40,_x)",%%xmm6\n\t" \
"paddw %%xmm5,%%xmm7\n\t" \
"psubw %%xmm4,%%xmm3\n\t" \
"movdqa "OC_MEM_OFFS(0x40,c)",%%xmm4\n\t" \
"paddw %%xmm7,%%xmm0\n\t" \
"movdqa "OC_MEM_OFFS(0x00,_x)",%%xmm7\n\t" \
/*0-1 butterfly. \
xmm4=xmm5=C4, xmm7=X0, xmm6=X4.*/ \
"paddw %%xmm7,%%xmm6\n\t" \
"movdqa %%xmm4,%%xmm5\n\t" \
"pmulhw %%xmm6,%%xmm4\n\t" \
"paddw %%xmm7,%%xmm7\n\t" \
"psubw %%xmm6,%%xmm7\n\t" \
"paddw %%xmm6,%%xmm4\n\t" \
/*Stage 2:*/ \
/*4-5 butterfly: xmm3=t[4], xmm1=t[5] \
7-6 butterfly: xmm2=t[6], xmm0=t[7]*/ \
"movdqa %%xmm3,%%xmm6\n\t" \
"paddw %%xmm1,%%xmm3\n\t" \
"psubw %%xmm1,%%xmm6\n\t" \
"movdqa %%xmm5,%%xmm1\n\t" \
"pmulhw %%xmm7,%%xmm5\n\t" \
"paddw %%xmm7,%%xmm5\n\t" \
"movdqa %%xmm0,%%xmm7\n\t" \
"paddw %%xmm2,%%xmm0\n\t" \
"psubw %%xmm2,%%xmm7\n\t" \
"movdqa %%xmm1,%%xmm2\n\t" \
"pmulhw %%xmm6,%%xmm1\n\t" \
"pmulhw %%xmm7,%%xmm2\n\t" \
"paddw %%xmm6,%%xmm1\n\t" \
"movdqa "OC_MEM_OFFS(0x00,buf)",%%xmm6\n\t" \
"paddw %%xmm7,%%xmm2\n\t" \
"movdqa "OC_MEM_OFFS(0x10,buf)",%%xmm7\n\t" \
/*Stage 3: \
6-5 butterfly: xmm1=t[5], xmm2=t[6] -> xmm1=t[6]+t[5], xmm2=t[6]-t[5] \
0-3 butterfly: xmm4=t[0], xmm7=t[3] -> xmm7=t[0]+t[3], xmm4=t[0]-t[3] \
1-2 butterfly: xmm5=t[1], xmm6=t[2] -> xmm6=t[1]+t[2], xmm5=t[1]-t[2]*/ \
"paddw %%xmm2,%%xmm1\n\t" \
"paddw %%xmm5,%%xmm6\n\t" \
"paddw %%xmm4,%%xmm7\n\t" \
"paddw %%xmm2,%%xmm2\n\t" \
"paddw %%xmm4,%%xmm4\n\t" \
"paddw %%xmm5,%%xmm5\n\t" \
"psubw %%xmm1,%%xmm2\n\t" \
"psubw %%xmm7,%%xmm4\n\t" \
"psubw %%xmm6,%%xmm5\n\t" \
/*Performs the last stage of the iDCT.
On input, xmm7 down to xmm4 contain rows 0 through 3, and xmm0 up to xmm3
contain rows 4 through 7.
On output, xmm0 through xmm7 contain the corresponding rows.*/
#define OC_IDCT_8x8_D \
"#OC_IDCT_8x8_D\n\t" \
/*Stage 4: \
0-7 butterfly: xmm7=t[0], xmm0=t[7] -> xmm0=t[0]+t[7], xmm7=t[0]-t[7] \
1-6 butterfly: xmm6=t[1], xmm1=t[6] -> xmm1=t[1]+t[6], xmm6=t[1]-t[6] \
2-5 butterfly: xmm5=t[2], xmm2=t[5] -> xmm2=t[2]+t[5], xmm5=t[2]-t[5] \
3-4 butterfly: xmm4=t[3], xmm3=t[4] -> xmm3=t[3]+t[4], xmm4=t[3]-t[4]*/ \
"psubw %%xmm0,%%xmm7\n\t" \
"psubw %%xmm1,%%xmm6\n\t" \
"psubw %%xmm2,%%xmm5\n\t" \
"psubw %%xmm3,%%xmm4\n\t" \
"paddw %%xmm0,%%xmm0\n\t" \
"paddw %%xmm1,%%xmm1\n\t" \
"paddw %%xmm2,%%xmm2\n\t" \
"paddw %%xmm3,%%xmm3\n\t" \
"paddw %%xmm7,%%xmm0\n\t" \
"paddw %%xmm6,%%xmm1\n\t" \
"paddw %%xmm5,%%xmm2\n\t" \
"paddw %%xmm4,%%xmm3\n\t" \
/*Performs the last stage of the iDCT.
On input, xmm7 down to xmm4 contain rows 0 through 3, and xmm0 up to xmm3
contain rows 4 through 7.
On output, xmm0 through xmm7 contain the corresponding rows.*/
#define OC_IDCT_8x8_D_STORE \
"#OC_IDCT_8x8_D_STORE\n\t" \
/*Stage 4: \
0-7 butterfly: xmm7=t[0], xmm0=t[7] -> xmm0=t[0]+t[7], xmm7=t[0]-t[7] \
1-6 butterfly: xmm6=t[1], xmm1=t[6] -> xmm1=t[1]+t[6], xmm6=t[1]-t[6] \
2-5 butterfly: xmm5=t[2], xmm2=t[5] -> xmm2=t[2]+t[5], xmm5=t[2]-t[5] \
3-4 butterfly: xmm4=t[3], xmm3=t[4] -> xmm3=t[3]+t[4], xmm4=t[3]-t[4]*/ \
"psubw %%xmm3,%%xmm4\n\t" \
"movdqa %%xmm4,"OC_MEM_OFFS(0x40,y)"\n\t" \
"movdqa "OC_MEM_OFFS(0x00,c)",%%xmm4\n\t" \
"psubw %%xmm0,%%xmm7\n\t" \
"psubw %%xmm1,%%xmm6\n\t" \
"psubw %%xmm2,%%xmm5\n\t" \
"paddw %%xmm4,%%xmm7\n\t" \
"paddw %%xmm4,%%xmm6\n\t" \
"paddw %%xmm4,%%xmm5\n\t" \
"paddw "OC_MEM_OFFS(0x40,y)",%%xmm4\n\t" \
"paddw %%xmm0,%%xmm0\n\t" \
"paddw %%xmm1,%%xmm1\n\t" \
"paddw %%xmm2,%%xmm2\n\t" \
"paddw %%xmm3,%%xmm3\n\t" \
"paddw %%xmm7,%%xmm0\n\t" \
"paddw %%xmm6,%%xmm1\n\t" \
"psraw $4,%%xmm0\n\t" \
"paddw %%xmm5,%%xmm2\n\t" \
"movdqa %%xmm0,"OC_MEM_OFFS(0x00,y)"\n\t" \
"psraw $4,%%xmm1\n\t" \
"paddw %%xmm4,%%xmm3\n\t" \
"movdqa %%xmm1,"OC_MEM_OFFS(0x10,y)"\n\t" \
"psraw $4,%%xmm2\n\t" \
"movdqa %%xmm2,"OC_MEM_OFFS(0x20,y)"\n\t" \
"psraw $4,%%xmm3\n\t" \
"movdqa %%xmm3,"OC_MEM_OFFS(0x30,y)"\n\t" \
"psraw $4,%%xmm4\n\t" \
"movdqa %%xmm4,"OC_MEM_OFFS(0x40,y)"\n\t" \
"psraw $4,%%xmm5\n\t" \
"movdqa %%xmm5,"OC_MEM_OFFS(0x50,y)"\n\t" \
"psraw $4,%%xmm6\n\t" \
"movdqa %%xmm6,"OC_MEM_OFFS(0x60,y)"\n\t" \
"psraw $4,%%xmm7\n\t" \
"movdqa %%xmm7,"OC_MEM_OFFS(0x70,y)"\n\t" \
static void oc_idct8x8_slow_sse2(ogg_int16_t _y[64],ogg_int16_t _x[64]){
OC_ALIGN16(ogg_int16_t buf[16]);
/*This routine accepts an 8x8 matrix pre-transposed.*/
__asm__ __volatile__(
/*Load rows 2, 3, 5, and 6 for the first stage of the iDCT.*/
"movdqa "OC_MEM_OFFS(0x20,x)",%%xmm2\n\t"
"movdqa "OC_MEM_OFFS(0x60,x)",%%xmm6\n\t"
"movdqa "OC_MEM_OFFS(0x30,x)",%%xmm3\n\t"
"movdqa "OC_MEM_OFFS(0x50,x)",%%xmm5\n\t"
OC_IDCT_8x8_ABC(x)
OC_IDCT_8x8_D
OC_TRANSPOSE_8x8
/*Clear out rows 0, 1, 4, and 7 for the first stage of the iDCT.*/
"movdqa %%xmm7,"OC_MEM_OFFS(0x70,y)"\n\t"
"movdqa %%xmm4,"OC_MEM_OFFS(0x40,y)"\n\t"
"movdqa %%xmm1,"OC_MEM_OFFS(0x10,y)"\n\t"
"movdqa %%xmm0,"OC_MEM_OFFS(0x00,y)"\n\t"
OC_IDCT_8x8_ABC(y)
OC_IDCT_8x8_D_STORE
:[buf]"=m"(OC_ARRAY_OPERAND(ogg_int16_t,buf,16)),
[y]"=m"(OC_ARRAY_OPERAND(ogg_int16_t,_y,64))
:[x]"m"(OC_CONST_ARRAY_OPERAND(ogg_int16_t,_x,64)),
[c]"m"(OC_CONST_ARRAY_OPERAND(ogg_int16_t,OC_IDCT_CONSTS,128))
);
if(_x!=_y){
int i;
__asm__ __volatile__("pxor %%xmm0,%%xmm0\n\t"::);
/*Clear input data for next block (decoder only).*/
for(i=0;i<2;i++){
__asm__ __volatile__(
"movdqa %%xmm0,"OC_MEM_OFFS(0x00,x)"\n\t"
"movdqa %%xmm0,"OC_MEM_OFFS(0x10,x)"\n\t"
"movdqa %%xmm0,"OC_MEM_OFFS(0x20,x)"\n\t"
"movdqa %%xmm0,"OC_MEM_OFFS(0x30,x)"\n\t"
:[x]"=m"(OC_ARRAY_OPERAND(ogg_int16_t,_x+i*32,32))
);
}
}
}
/*For the first step of the 10-coefficient version of the 8x8 iDCT, we only
need to work with four columns at a time.
Doing this in MMX is faster on processors with a 64-bit data path.*/
#define OC_IDCT_8x8_10_MMX \
"#OC_IDCT_8x8_10_MMX\n\t" \
/*Stage 1:*/ \
/*2-3 rotation by 6pi/16. \
mm7=C6, mm6=C2, mm2=X2, X6=0.*/ \
"movq "OC_MEM_OFFS(0x60,c)",%%mm7\n\t" \
"movq "OC_MEM_OFFS(0x20,c)",%%mm6\n\t" \
"pmulhw %%mm2,%%mm6\n\t" \
"pmulhw %%mm2,%%mm7\n\t" \
"movq "OC_MEM_OFFS(0x50,c)",%%mm5\n\t" \
"paddw %%mm6,%%mm2\n\t" \
"movq %%mm2,"OC_MEM_OFFS(0x10,buf)"\n\t" \
"movq "OC_MEM_OFFS(0x30,c)",%%mm2\n\t" \
"movq %%mm7,"OC_MEM_OFFS(0x00,buf)"\n\t" \
/*5-6 rotation by 3pi/16. \
mm5=C5, mm2=C3, mm3=X3, X5=0.*/ \
"pmulhw %%mm3,%%mm5\n\t" \
"pmulhw %%mm3,%%mm2\n\t" \
"movq "OC_MEM_OFFS(0x10,c)",%%mm7\n\t" \
"paddw %%mm3,%%mm5\n\t" \
"paddw %%mm3,%%mm2\n\t" \
"movq "OC_MEM_OFFS(0x70,c)",%%mm3\n\t" \
/*4-7 rotation by 7pi/16. \
mm7=C1, mm3=C7, mm1=X1, X7=0.*/ \
"pmulhw %%mm1,%%mm3\n\t" \
"pmulhw %%mm1,%%mm7\n\t" \
"movq "OC_MEM_OFFS(0x40,c)",%%mm4\n\t" \
"movq %%mm3,%%mm6\n\t" \
"paddw %%mm1,%%mm7\n\t" \
/*0-1 butterfly. \
mm4=C4, mm0=X0, X4=0.*/ \
/*Stage 2:*/ \
/*4-5 butterfly: mm3=t[4], mm5=t[5] \
7-6 butterfly: mm2=t[6], mm7=t[7]*/ \
"psubw %%mm5,%%mm3\n\t" \
"paddw %%mm5,%%mm6\n\t" \
"movq %%mm4,%%mm1\n\t" \
"pmulhw %%mm0,%%mm4\n\t" \
"paddw %%mm0,%%mm4\n\t" \
"movq %%mm7,%%mm0\n\t" \
"movq %%mm4,%%mm5\n\t" \
"paddw %%mm2,%%mm0\n\t" \
"psubw %%mm2,%%mm7\n\t" \
"movq %%mm1,%%mm2\n\t" \
"pmulhw %%mm6,%%mm1\n\t" \
"pmulhw %%mm7,%%mm2\n\t" \
"paddw %%mm6,%%mm1\n\t" \
"movq "OC_MEM_OFFS(0x00,buf)",%%mm6\n\t" \
"paddw %%mm7,%%mm2\n\t" \
"movq "OC_MEM_OFFS(0x10,buf)",%%mm7\n\t" \
/*Stage 3: \
6-5 butterfly: mm1=t[5], mm2=t[6] -> mm1=t[6]+t[5], mm2=t[6]-t[5] \
0-3 butterfly: mm4=t[0], mm7=t[3] -> mm7=t[0]+t[3], mm4=t[0]-t[3] \
1-2 butterfly: mm5=t[1], mm6=t[2] -> mm6=t[1]+t[2], mm5=t[1]-t[2]*/ \
"paddw %%mm2,%%mm1\n\t" \
"paddw %%mm5,%%mm6\n\t" \
"paddw %%mm4,%%mm7\n\t" \
"paddw %%mm2,%%mm2\n\t" \
"paddw %%mm4,%%mm4\n\t" \
"paddw %%mm5,%%mm5\n\t" \
"psubw %%mm1,%%mm2\n\t" \
"psubw %%mm7,%%mm4\n\t" \
"psubw %%mm6,%%mm5\n\t" \
/*Stage 4: \
0-7 butterfly: mm7=t[0], mm0=t[7] -> mm0=t[0]+t[7], mm7=t[0]-t[7] \
1-6 butterfly: mm6=t[1], mm1=t[6] -> mm1=t[1]+t[6], mm6=t[1]-t[6] \
2-5 butterfly: mm5=t[2], mm2=t[5] -> mm2=t[2]+t[5], mm5=t[2]-t[5] \
3-4 butterfly: mm4=t[3], mm3=t[4] -> mm3=t[3]+t[4], mm4=t[3]-t[4]*/ \
"psubw %%mm0,%%mm7\n\t" \
"psubw %%mm1,%%mm6\n\t" \
"psubw %%mm2,%%mm5\n\t" \
"psubw %%mm3,%%mm4\n\t" \
"paddw %%mm0,%%mm0\n\t" \
"paddw %%mm1,%%mm1\n\t" \
"paddw %%mm2,%%mm2\n\t" \
"paddw %%mm3,%%mm3\n\t" \
"paddw %%mm7,%%mm0\n\t" \
"paddw %%mm6,%%mm1\n\t" \
"paddw %%mm5,%%mm2\n\t" \
"paddw %%mm4,%%mm3\n\t" \
#define OC_IDCT_8x8_10_ABC \
"#OC_IDCT_8x8_10_ABC\n\t" \
/*Stage 1:*/ \
/*2-3 rotation by 6pi/16. \
xmm7=C6, xmm6=C2, xmm2=X2, X6=0.*/ \
"movdqa "OC_MEM_OFFS(0x60,c)",%%xmm7\n\t" \
"movdqa "OC_MEM_OFFS(0x20,c)",%%xmm6\n\t" \
"pmulhw %%xmm2,%%xmm6\n\t" \
"pmulhw %%xmm2,%%xmm7\n\t" \
"movdqa "OC_MEM_OFFS(0x50,c)",%%xmm5\n\t" \
"paddw %%xmm6,%%xmm2\n\t" \
"movdqa %%xmm2,"OC_MEM_OFFS(0x10,buf)"\n\t" \
"movdqa "OC_MEM_OFFS(0x30,c)",%%xmm2\n\t" \
"movdqa %%xmm7,"OC_MEM_OFFS(0x00,buf)"\n\t" \
/*5-6 rotation by 3pi/16. \
xmm5=C5, xmm2=C3, xmm3=X3, X5=0.*/ \
"pmulhw %%xmm3,%%xmm5\n\t" \
"pmulhw %%xmm3,%%xmm2\n\t" \
"movdqa "OC_MEM_OFFS(0x10,c)",%%xmm7\n\t" \
"paddw %%xmm3,%%xmm5\n\t" \
"paddw %%xmm3,%%xmm2\n\t" \
"movdqa "OC_MEM_OFFS(0x70,c)",%%xmm3\n\t" \
/*4-7 rotation by 7pi/16. \
xmm7=C1, xmm3=C7, xmm1=X1, X7=0.*/ \
"pmulhw %%xmm1,%%xmm3\n\t" \
"pmulhw %%xmm1,%%xmm7\n\t" \
"movdqa "OC_MEM_OFFS(0x40,c)",%%xmm4\n\t" \
"movdqa %%xmm3,%%xmm6\n\t" \
"paddw %%xmm1,%%xmm7\n\t" \
/*0-1 butterfly. \
xmm4=C4, xmm0=X0, X4=0.*/ \
/*Stage 2:*/ \
/*4-5 butterfly: xmm3=t[4], xmm5=t[5] \
7-6 butterfly: xmm2=t[6], xmm7=t[7]*/ \
"psubw %%xmm5,%%xmm3\n\t" \
"paddw %%xmm5,%%xmm6\n\t" \
"movdqa %%xmm4,%%xmm1\n\t" \
"pmulhw %%xmm0,%%xmm4\n\t" \
"paddw %%xmm0,%%xmm4\n\t" \
"movdqa %%xmm7,%%xmm0\n\t" \
"movdqa %%xmm4,%%xmm5\n\t" \
"paddw %%xmm2,%%xmm0\n\t" \
"psubw %%xmm2,%%xmm7\n\t" \
"movdqa %%xmm1,%%xmm2\n\t" \
"pmulhw %%xmm6,%%xmm1\n\t" \
"pmulhw %%xmm7,%%xmm2\n\t" \
"paddw %%xmm6,%%xmm1\n\t" \
"movdqa "OC_MEM_OFFS(0x00,buf)",%%xmm6\n\t" \
"paddw %%xmm7,%%xmm2\n\t" \
"movdqa "OC_MEM_OFFS(0x10,buf)",%%xmm7\n\t" \
/*Stage 3: \
6-5 butterfly: xmm1=t[5], xmm2=t[6] -> xmm1=t[6]+t[5], xmm2=t[6]-t[5] \
0-3 butterfly: xmm4=t[0], xmm7=t[3] -> xmm7=t[0]+t[3], xmm4=t[0]-t[3] \
1-2 butterfly: xmm5=t[1], xmm6=t[2] -> xmm6=t[1]+t[2], xmm5=t[1]-t[2]*/ \
"paddw %%xmm2,%%xmm1\n\t" \
"paddw %%xmm5,%%xmm6\n\t" \
"paddw %%xmm4,%%xmm7\n\t" \
"paddw %%xmm2,%%xmm2\n\t" \
"paddw %%xmm4,%%xmm4\n\t" \
"paddw %%xmm5,%%xmm5\n\t" \
"psubw %%xmm1,%%xmm2\n\t" \
"psubw %%xmm7,%%xmm4\n\t" \
"psubw %%xmm6,%%xmm5\n\t" \
static void oc_idct8x8_10_sse2(ogg_int16_t _y[64],ogg_int16_t _x[64]){
OC_ALIGN16(ogg_int16_t buf[16]);
/*This routine accepts an 8x8 matrix pre-transposed.*/
__asm__ __volatile__(
"movq "OC_MEM_OFFS(0x20,x)",%%mm2\n\t"
"movq "OC_MEM_OFFS(0x30,x)",%%mm3\n\t"
"movq "OC_MEM_OFFS(0x10,x)",%%mm1\n\t"
"movq "OC_MEM_OFFS(0x00,x)",%%mm0\n\t"
OC_IDCT_8x8_10_MMX
OC_TRANSPOSE_8x4_MMX2SSE
OC_IDCT_8x8_10_ABC
OC_IDCT_8x8_D_STORE
:[buf]"=m"(OC_ARRAY_OPERAND(short,buf,16)),
[y]"=m"(OC_ARRAY_OPERAND(ogg_int16_t,_y,64))
:[x]"m"OC_CONST_ARRAY_OPERAND(ogg_int16_t,_x,64),
[c]"m"(OC_CONST_ARRAY_OPERAND(ogg_int16_t,OC_IDCT_CONSTS,128))
);
if(_x!=_y){
/*Clear input data for next block (decoder only).*/
__asm__ __volatile__(
"pxor %%mm0,%%mm0\n\t"
"movq %%mm0,"OC_MEM_OFFS(0x00,x)"\n\t"
"movq %%mm0,"OC_MEM_OFFS(0x10,x)"\n\t"
"movq %%mm0,"OC_MEM_OFFS(0x20,x)"\n\t"
"movq %%mm0,"OC_MEM_OFFS(0x30,x)"\n\t"
:[x]"+m"(OC_ARRAY_OPERAND(ogg_int16_t,_x,28))
);
}
}
/*Performs an inverse 8x8 Type-II DCT transform.
The input is assumed to be scaled by a factor of 4 relative to orthonormal
version of the transform.*/
void oc_idct8x8_sse2(ogg_int16_t _y[64],ogg_int16_t _x[64],int _last_zzi){
/*_last_zzi is subtly different from an actual count of the number of
coefficients we decoded for this block.
It contains the value of zzi BEFORE the final token in the block was
decoded.
In most cases this is an EOB token (the continuation of an EOB run from a
previous block counts), and so this is the same as the coefficient count.
However, in the case that the last token was NOT an EOB token, but filled
the block up with exactly 64 coefficients, _last_zzi will be less than 64.
Provided the last token was not a pure zero run, the minimum value it can
be is 46, and so that doesn't affect any of the cases in this routine.
However, if the last token WAS a pure zero run of length 63, then _last_zzi
will be 1 while the number of coefficients decoded is 64.
Thus, we will trigger the following special case, where the real
coefficient count would not.
Note also that a zero run of length 64 will give _last_zzi a value of 0,
but we still process the DC coefficient, which might have a non-zero value
due to DC prediction.
Although convoluted, this is arguably the correct behavior: it allows us to
use a smaller transform when the block ends with a long zero run instead
of a normal EOB token.
It could be smarter... multiple separate zero runs at the end of a block
will fool it, but an encoder that generates these really deserves what it
gets.
Needless to say we inherited this approach from VP3.*/
/*Then perform the iDCT.*/
if(_last_zzi<=10)oc_idct8x8_10_sse2(_y,_x);
else oc_idct8x8_slow_sse2(_y,_x);
}
#endif

242
media/libtheora/lib/x86/sse2trans.h Normal file
View File

@ -0,0 +1,242 @@
/********************************************************************
* *
* THIS FILE IS PART OF THE OggTheora SOFTWARE CODEC SOURCE CODE. *
* USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS *
* GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
* IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING. *
* *
* THE Theora SOURCE CODE IS COPYRIGHT (C) 2002-2009 *
* by the Xiph.Org Foundation and contributors http://www.xiph.org/ *
* *
********************************************************************
function:
last mod: $Id: sse2trans.h 15675 2009-02-06 09:43:27Z tterribe $
********************************************************************/
#if !defined(_x86_sse2trans_H)
# define _x86_sse2trans_H (1)
# include "x86int.h"
# if defined(OC_X86_64_ASM)
/*On x86-64 we can transpose in-place without spilling registers.
By clever choices of the order to apply the butterflies and the order of
their outputs, we can take the rows in order and output the columns in order
without any extra operations and using just one temporary register.*/
# define OC_TRANSPOSE_8x8 \
"#OC_TRANSPOSE_8x8\n\t" \
"movdqa %%xmm4,%%xmm8\n\t" \
/*xmm4 = f3 e3 f2 e2 f1 e1 f0 e0*/ \
"punpcklwd %%xmm5,%%xmm4\n\t" \
/*xmm8 = f7 e7 f6 e6 f5 e5 f4 e4*/ \
"punpckhwd %%xmm5,%%xmm8\n\t" \
/*xmm5 is free.*/ \
"movdqa %%xmm0,%%xmm5\n\t" \
/*xmm0 = b3 a3 b2 a2 b1 a1 b0 a0*/ \
"punpcklwd %%xmm1,%%xmm0\n\t" \
/*xmm5 = b7 a7 b6 a6 b5 a5 b4 a4*/ \
"punpckhwd %%xmm1,%%xmm5\n\t" \
/*xmm1 is free.*/ \
"movdqa %%xmm6,%%xmm1\n\t" \
/*xmm6 = h3 g3 h2 g2 h1 g1 h0 g0*/ \
"punpcklwd %%xmm7,%%xmm6\n\t" \
/*xmm1 = h7 g7 h6 g6 h5 g5 h4 g4*/ \
"punpckhwd %%xmm7,%%xmm1\n\t" \
/*xmm7 is free.*/ \
"movdqa %%xmm2,%%xmm7\n\t" \
/*xmm2 = d7 c7 d6 c6 d5 c5 d4 c4*/ \
"punpckhwd %%xmm3,%%xmm2\n\t" \
/*xmm7 = d3 c3 d2 c2 d1 c1 d0 c0*/ \
"punpcklwd %%xmm3,%%xmm7\n\t" \
/*xmm3 is free.*/ \
"movdqa %%xmm0,%%xmm3\n\t" \
/*xmm0 = d1 c1 b1 a1 d0 c0 b0 a0*/ \
"punpckldq %%xmm7,%%xmm0\n\t" \
/*xmm3 = d3 c3 b3 a3 d2 c2 b2 a2*/ \
"punpckhdq %%xmm7,%%xmm3\n\t" \
/*xmm7 is free.*/ \
"movdqa %%xmm5,%%xmm7\n\t" \
/*xmm5 = d5 c5 b5 a5 d4 c4 b4 a4*/ \
"punpckldq %%xmm2,%%xmm5\n\t" \
/*xmm7 = d7 c7 b7 a7 d6 c6 b6 a6*/ \
"punpckhdq %%xmm2,%%xmm7\n\t" \
/*xmm2 is free.*/ \
"movdqa %%xmm4,%%xmm2\n\t" \
/*xmm4 = h3 g3 f3 e3 h2 g2 f2 e2*/ \
"punpckhdq %%xmm6,%%xmm4\n\t" \
/*xmm2 = h1 g1 f1 e1 h0 g0 f0 e0*/ \
"punpckldq %%xmm6,%%xmm2\n\t" \
/*xmm6 is free.*/ \
"movdqa %%xmm8,%%xmm6\n\t" \
/*xmm6 = h5 g5 f5 e5 h4 g4 f4 e4*/ \
"punpckldq %%xmm1,%%xmm6\n\t" \
/*xmm8 = h7 g7 f7 e7 h6 g6 f6 e6*/ \
"punpckhdq %%xmm1,%%xmm8\n\t" \
/*xmm1 is free.*/ \
"movdqa %%xmm0,%%xmm1\n\t" \
/*xmm0 = h0 g0 f0 e0 d0 c0 b0 a0*/ \
"punpcklqdq %%xmm2,%%xmm0\n\t" \
/*xmm1 = h1 g1 f1 e1 d1 c1 b1 a1*/ \
"punpckhqdq %%xmm2,%%xmm1\n\t" \
/*xmm2 is free.*/ \
"movdqa %%xmm3,%%xmm2\n\t" \
/*xmm3 = h3 g3 f3 e3 d3 c3 b3 a3*/ \
"punpckhqdq %%xmm4,%%xmm3\n\t" \
/*xmm2 = h2 g2 f2 e2 d2 c2 b2 a2*/ \
"punpcklqdq %%xmm4,%%xmm2\n\t" \
/*xmm4 is free.*/ \
"movdqa %%xmm5,%%xmm4\n\t" \
/*xmm5 = h5 g5 f5 e5 d5 c5 b5 a5*/ \
"punpckhqdq %%xmm6,%%xmm5\n\t" \
/*xmm4 = h4 g4 f4 e4 d4 c4 b4 a4*/ \
"punpcklqdq %%xmm6,%%xmm4\n\t" \
/*xmm6 is free.*/ \
"movdqa %%xmm7,%%xmm6\n\t" \
/*xmm7 = h7 g7 f7 e7 d7 c7 b7 a7*/ \
"punpckhqdq %%xmm8,%%xmm7\n\t" \
/*xmm6 = h6 g6 f6 e6 d6 c6 b6 a6*/ \
"punpcklqdq %%xmm8,%%xmm6\n\t" \
/*xmm8 is free.*/ \
# else
/*Otherwise, we need to spill some values to %[buf] temporarily.
Again, the butterflies are carefully arranged to get the columns to come out
in order, minimizing register spills and maximizing the delay between a load
and when the value loaded is actually used.*/
# define OC_TRANSPOSE_8x8 \
"#OC_TRANSPOSE_8x8\n\t" \
/*buf[0] = a7 a6 a5 a4 a3 a2 a1 a0*/ \
"movdqa %%xmm0,"OC_MEM_OFFS(0x00,buf)"\n\t" \
/*xmm0 is free.*/ \
"movdqa %%xmm2,%%xmm0\n\t" \
/*xmm2 = d7 c7 d6 c6 d5 c5 d4 c4*/ \
"punpckhwd %%xmm3,%%xmm2\n\t" \
/*xmm0 = d3 c3 d2 c2 d1 c1 d0 c0*/ \
"punpcklwd %%xmm3,%%xmm0\n\t" \
/*xmm3 = a7 a6 a5 a4 a3 a2 a1 a0*/ \
"movdqa "OC_MEM_OFFS(0x00,buf)",%%xmm3\n\t" \
/*buf[1] = d7 c7 d6 c6 d5 c5 d4 c4*/ \
"movdqa %%xmm2,"OC_MEM_OFFS(0x10,buf)"\n\t" \
/*xmm2 is free.*/ \
"movdqa %%xmm6,%%xmm2\n\t" \
/*xmm6 = h3 g3 h2 g2 h1 g1 h0 g0*/ \
"punpcklwd %%xmm7,%%xmm6\n\t" \
/*xmm2 = h7 g7 h6 g6 h5 g5 h4 g4*/ \
"punpckhwd %%xmm7,%%xmm2\n\t" \
/*xmm7 is free.*/ \
"movdqa %%xmm4,%%xmm7\n\t" \
/*xmm4 = f3 e3 f2 e2 f1 e1 f0 e0*/ \
"punpcklwd %%xmm5,%%xmm4\n\t" \
/*xmm7 = f7 e7 f6 e6 f5 e5 f4 e4*/ \
"punpckhwd %%xmm5,%%xmm7\n\t" \
/*xmm5 is free.*/ \
"movdqa %%xmm3,%%xmm5\n\t" \
/*xmm3 = b3 a3 b2 a2 b1 a1 b0 a0*/ \
"punpcklwd %%xmm1,%%xmm3\n\t" \
/*xmm5 = b7 a7 b6 a6 b5 a5 b4 a4*/ \
"punpckhwd %%xmm1,%%xmm5\n\t" \
/*xmm1 is free.*/ \
"movdqa %%xmm7,%%xmm1\n\t" \
/*xmm7 = h5 g5 f5 e5 h4 g4 f4 e4*/ \
"punpckldq %%xmm2,%%xmm7\n\t" \
/*xmm1 = h7 g7 f7 e7 h6 g6 f6 e6*/ \
"punpckhdq %%xmm2,%%xmm1\n\t" \
/*xmm2 = d7 c7 d6 c6 d5 c5 d4 c4*/ \
"movdqa "OC_MEM_OFFS(0x10,buf)",%%xmm2\n\t" \
/*buf[0] = h7 g7 f7 e7 h6 g6 f6 e6*/ \
"movdqa %%xmm1,"OC_MEM_OFFS(0x00,buf)"\n\t" \
/*xmm1 is free.*/ \
"movdqa %%xmm3,%%xmm1\n\t" \
/*xmm3 = d3 c3 b3 a3 d2 c2 b2 a2*/ \
"punpckhdq %%xmm0,%%xmm3\n\t" \
/*xmm1 = d1 c1 b1 a1 d0 c0 b0 a0*/ \
"punpckldq %%xmm0,%%xmm1\n\t" \
/*xmm0 is free.*/ \
"movdqa %%xmm4,%%xmm0\n\t" \
/*xmm4 = h3 g3 f3 e3 h2 g2 f2 e2*/ \
"punpckhdq %%xmm6,%%xmm4\n\t" \
/*xmm0 = h1 g1 f1 e1 h0 g0 f0 e0*/ \
"punpckldq %%xmm6,%%xmm0\n\t" \
/*xmm6 is free.*/ \
"movdqa %%xmm5,%%xmm6\n\t" \
/*xmm5 = d5 c5 b5 a5 d4 c4 b4 a4*/ \
"punpckldq %%xmm2,%%xmm5\n\t" \
/*xmm6 = d7 c7 b7 a7 d6 c6 b6 a6*/ \
"punpckhdq %%xmm2,%%xmm6\n\t" \
/*xmm2 is free.*/ \
"movdqa %%xmm1,%%xmm2\n\t" \
/*xmm1 = h1 g1 f1 e1 d1 c1 b1 a1*/ \
"punpckhqdq %%xmm0,%%xmm1\n\t" \
/*xmm2 = h0 g0 f0 e0 d0 c0 b0 a0*/ \
"punpcklqdq %%xmm0,%%xmm2\n\t" \
/*xmm0 = h7 g7 f7 e7 h6 g6 f6 e6*/ \
"movdqa "OC_MEM_OFFS(0x00,buf)",%%xmm0\n\t" \
/*buf[1] = h0 g0 f0 e0 d0 c0 b0 a0*/ \
"movdqa %%xmm2,"OC_MEM_OFFS(0x10,buf)"\n\t" \
/*xmm2 is free.*/ \
"movdqa %%xmm3,%%xmm2\n\t" \
/*xmm3 = h3 g3 f3 e3 d3 c3 b3 a3*/ \
"punpckhqdq %%xmm4,%%xmm3\n\t" \
/*xmm2 = h2 g2 f2 e2 d2 c2 b2 a2*/ \
"punpcklqdq %%xmm4,%%xmm2\n\t" \
/*xmm4 is free.*/ \
"movdqa %%xmm5,%%xmm4\n\t" \
/*xmm5 = h5 g5 f5 e5 d5 c5 b5 a5*/ \
"punpckhqdq %%xmm7,%%xmm5\n\t" \
/*xmm4 = h4 g4 f4 e4 d4 c4 b4 a4*/ \
"punpcklqdq %%xmm7,%%xmm4\n\t" \
/*xmm7 is free.*/ \
"movdqa %%xmm6,%%xmm7\n\t" \
/*xmm6 = h6 g6 f6 e6 d6 c6 b6 a6*/ \
"punpcklqdq %%xmm0,%%xmm6\n\t" \
/*xmm7 = h7 g7 f7 e7 d7 c7 b7 a7*/ \
"punpckhqdq %%xmm0,%%xmm7\n\t" \
/*xmm0 = h0 g0 f0 e0 d0 c0 b0 a0*/ \
"movdqa "OC_MEM_OFFS(0x10,buf)",%%xmm0\n\t" \
# endif
/*Transpose 4 values in each of 8 MMX registers into 8 values in the first
four SSE registers.
No need to be clever here; we have plenty of room.*/
# define OC_TRANSPOSE_8x4_MMX2SSE \
"#OC_TRANSPOSE_8x4_MMX2SSE\n\t" \
"movq2dq %%mm0,%%xmm0\n\t" \
"movq2dq %%mm1,%%xmm1\n\t" \
/*xmmA = b3 a3 b2 a2 b1 a1 b0 a0*/ \
"punpcklwd %%xmm1,%%xmm0\n\t" \
"movq2dq %%mm2,%%xmm3\n\t" \
"movq2dq %%mm3,%%xmm2\n\t" \
/*xmmC = d3 c3 d2 c2 d1 c1 d0 c0*/ \
"punpcklwd %%xmm2,%%xmm3\n\t" \
"movq2dq %%mm4,%%xmm4\n\t" \
"movq2dq %%mm5,%%xmm5\n\t" \
/*xmmE = f3 e3 f2 e2 f1 e1 f0 e0*/ \
"punpcklwd %%xmm5,%%xmm4\n\t" \
"movq2dq %%mm6,%%xmm7\n\t" \
"movq2dq %%mm7,%%xmm6\n\t" \
/*xmmG = h3 g3 h2 g2 h1 g1 h0 g0*/ \
"punpcklwd %%xmm6,%%xmm7\n\t" \
"movdqa %%xmm0,%%xmm2\n\t" \
/*xmm0 = d1 c1 b1 a1 d0 c0 b0 a0*/ \
"punpckldq %%xmm3,%%xmm0\n\t" \
/*xmm2 = d3 c3 b3 a3 d2 c2 b2 a2*/ \
"punpckhdq %%xmm3,%%xmm2\n\t" \
"movdqa %%xmm4,%%xmm5\n\t" \
/*xmm4 = h1 g1 f1 e1 h0 g0 f0 e0*/ \
"punpckldq %%xmm7,%%xmm4\n\t" \
/*xmm3 = h3 g3 f3 e3 h2 g2 f2 e2*/ \
"punpckhdq %%xmm7,%%xmm5\n\t" \
"movdqa %%xmm0,%%xmm1\n\t" \
/*xmm0 = h0 g0 f0 e0 d0 c0 b0 a0*/ \
"punpcklqdq %%xmm4,%%xmm0\n\t" \
/*xmm1 = h1 g1 f1 e1 d1 c1 b1 a1*/ \
"punpckhqdq %%xmm4,%%xmm1\n\t" \
"movdqa %%xmm2,%%xmm3\n\t" \
/*xmm2 = h2 g2 f2 e2 d2 c2 b2 a2*/ \
"punpcklqdq %%xmm5,%%xmm2\n\t" \
/*xmm3 = h3 g3 f3 e3 d3 c3 b3 a3*/ \
"punpckhqdq %%xmm5,%%xmm3\n\t" \
#endif

182
media/libtheora/lib/x86/x86cpu.c Normal file
View File

@ -0,0 +1,182 @@
/********************************************************************
* *
* THIS FILE IS PART OF THE OggTheora SOFTWARE CODEC SOURCE CODE. *
* USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS *
* GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
* IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING. *
* *
* THE Theora SOURCE CODE IS COPYRIGHT (C) 2002-2009 *
* by the Xiph.Org Foundation and contributors http://www.xiph.org/ *
* *
********************************************************************
CPU capability detection for x86 processors.
Originally written by Rudolf Marek.
function:
last mod: $Id: x86cpu.c 17410 2010-09-21 21:53:48Z tterribe $
********************************************************************/
#include "x86cpu.h"
#if !defined(OC_X86_ASM)
ogg_uint32_t oc_cpu_flags_get(void){
return 0;
}
#else
# if defined(__amd64__)||defined(__x86_64__)
/*On x86-64, gcc seems to be able to figure out how to save %rbx for us when
compiling with -fPIC.*/
# define cpuid(_op,_eax,_ebx,_ecx,_edx) \
__asm__ __volatile__( \
"cpuid\n\t" \
:[eax]"=a"(_eax),[ebx]"=b"(_ebx),[ecx]"=c"(_ecx),[edx]"=d"(_edx) \
:"a"(_op) \
:"cc" \
)
# else
/*On x86-32, not so much.*/
# define cpuid(_op,_eax,_ebx,_ecx,_edx) \
__asm__ __volatile__( \
"xchgl %%ebx,%[ebx]\n\t" \
"cpuid\n\t" \
"xchgl %%ebx,%[ebx]\n\t" \
:[eax]"=a"(_eax),[ebx]"=r"(_ebx),[ecx]"=c"(_ecx),[edx]"=d"(_edx) \
:"a"(_op) \
:"cc" \
)
# endif
static ogg_uint32_t oc_parse_intel_flags(ogg_uint32_t _edx,ogg_uint32_t _ecx){
ogg_uint32_t flags;
/*If there isn't even MMX, give up.*/
if(!(_edx&0x00800000))return 0;
flags=OC_CPU_X86_MMX;
if(_edx&0x02000000)flags|=OC_CPU_X86_MMXEXT|OC_CPU_X86_SSE;
if(_edx&0x04000000)flags|=OC_CPU_X86_SSE2;
if(_ecx&0x00000001)flags|=OC_CPU_X86_PNI;
if(_ecx&0x00000100)flags|=OC_CPU_X86_SSSE3;
if(_ecx&0x00080000)flags|=OC_CPU_X86_SSE4_1;
if(_ecx&0x00100000)flags|=OC_CPU_X86_SSE4_2;
return flags;
}
static ogg_uint32_t oc_parse_amd_flags(ogg_uint32_t _edx,ogg_uint32_t _ecx){
ogg_uint32_t flags;
/*If there isn't even MMX, give up.*/
if(!(_edx&0x00800000))return 0;
flags=OC_CPU_X86_MMX;
if(_edx&0x00400000)flags|=OC_CPU_X86_MMXEXT;
if(_edx&0x80000000)flags|=OC_CPU_X86_3DNOW;
if(_edx&0x40000000)flags|=OC_CPU_X86_3DNOWEXT;
if(_ecx&0x00000040)flags|=OC_CPU_X86_SSE4A;
if(_ecx&0x00000800)flags|=OC_CPU_X86_SSE5;
return flags;
}
ogg_uint32_t oc_cpu_flags_get(void){
ogg_uint32_t flags;
ogg_uint32_t eax;
ogg_uint32_t ebx;
ogg_uint32_t ecx;
ogg_uint32_t edx;
# if !defined(__amd64__)&&!defined(__x86_64__)
/*Not all x86-32 chips support cpuid, so we have to check.*/
__asm__ __volatile__(
"pushfl\n\t"
"pushfl\n\t"
"popl %[a]\n\t"
"movl %[a],%[b]\n\t"
"xorl $0x200000,%[a]\n\t"
"pushl %[a]\n\t"
"popfl\n\t"
"pushfl\n\t"
"popl %[a]\n\t"
"popfl\n\t"
:[a]"=r"(eax),[b]"=r"(ebx)
:
:"cc"
);
/*No cpuid.*/
if(eax==ebx)return 0;
# endif
cpuid(0,eax,ebx,ecx,edx);
/* l e t n I e n i u n e G*/
if(ecx==0x6C65746E&&edx==0x49656E69&&ebx==0x756E6547||
/* 6 8 x M T e n i u n e G*/
ecx==0x3638784D&&edx==0x54656E69&&ebx==0x756E6547){
int family;
int model;
/*Intel, Transmeta (tested with Crusoe TM5800):*/
cpuid(1,eax,ebx,ecx,edx);
flags=oc_parse_intel_flags(edx,ecx);
family=(eax>>8)&0xF;
model=(eax>>4)&0xF;
/*The SSE unit on the Pentium M and Core Duo is much slower than the MMX
unit, so don't use it.*/
if(family==6&&(model==9||model==13||model==14)){
flags&=~(OC_CPU_X86_SSE2|OC_CPU_X86_PNI);
}
}
/* D M A c i t n e h t u A*/
else if(ecx==0x444D4163&&edx==0x69746E65&&ebx==0x68747541||
/* C S N y b e d o e G*/
ecx==0x43534e20&&edx==0x79622065&&ebx==0x646f6547){
/*AMD, Geode:*/
cpuid(0x80000000,eax,ebx,ecx,edx);
if(eax<0x80000001)flags=0;
else{
cpuid(0x80000001,eax,ebx,ecx,edx);
flags=oc_parse_amd_flags(edx,ecx);
}
/*Also check for SSE.*/
cpuid(1,eax,ebx,ecx,edx);
flags|=oc_parse_intel_flags(edx,ecx);
}
/*Technically some VIA chips can be configured in the BIOS to return any
string here the user wants.
There is a special detection method that can be used to identify such
processors, but in my opinion, if the user really wants to change it, they
deserve what they get.*/
/* s l u a H r u a t n e C*/
else if(ecx==0x736C7561&&edx==0x48727561&&ebx==0x746E6543){
/*VIA:*/
/*I only have documentation for the C7 (Esther) and Isaiah (forthcoming)
chips (thanks to the engineers from Centaur Technology who provided it).
These chips support Intel-like cpuid info.
The C3-2 (Nehemiah) cores appear to, as well.*/
cpuid(1,eax,ebx,ecx,edx);
flags=oc_parse_intel_flags(edx,ecx);
if(eax>=0x80000001){
/*The (non-Nehemiah) C3 processors support AMD-like cpuid info.
We need to check this even if the Intel test succeeds to pick up 3DNow!
support on these processors.
Unlike actual AMD processors, we cannot _rely_ on this info, since
some cores (e.g., the 693 stepping of the Nehemiah) claim to support
this function, yet return edx=0, despite the Intel test indicating
MMX support.
Therefore the features detected here are strictly added to those
detected by the Intel test.*/
/*TODO: How about earlier chips?*/
cpuid(0x80000001,eax,ebx,ecx,edx);
/*Note: As of the C7, this function returns Intel-style extended feature
flags, not AMD-style.
Currently, this only defines bits 11, 20, and 29 (0x20100800), which
do not conflict with any of the AMD flags we inspect.
For the remaining bits, Intel tells us, "Do not count on their value",
but VIA assures us that they will all be zero (at least on the C7 and
Isaiah chips).
In the (unlikely) event a future processor uses bits 18, 19, 30, or 31
(0xC0C00000) for something else, we will have to add code to detect
the model to decide when it is appropriate to inspect them.*/
flags|=oc_parse_amd_flags(edx,ecx);
}
}
else{
/*Implement me.*/
flags=0;
}
return flags;
}
#endif

10
media/libtheora/lib/cpu.h → media/libtheora/lib/x86/x86cpu.h
View File

@ -10,13 +10,13 @@
* *
********************************************************************
function:
last mod: $Id: cpu.h 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: x86cpu.h 17410 2010-09-21 21:53:48Z tterribe $
********************************************************************/
#if !defined(_x86_cpu_H)
# define _x86_cpu_H (1)
#include "internal.h"
#if !defined(_x86_x86cpu_H)
# define _x86_x86cpu_H (1)
#include "../internal.h"
#define OC_CPU_X86_MMX (1<<0)
#define OC_CPU_X86_3DNOW (1<<1)
@ -31,4 +31,6 @@
#define OC_CPU_X86_SSE4A (1<<10)
#define OC_CPU_X86_SSE5 (1<<11)
ogg_uint32_t oc_cpu_flags_get(void);
#endif

96
media/libtheora/lib/x86/x86int.h
View File

@ -11,7 +11,7 @@
********************************************************************
function:
last mod: $Id: x86int.h 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: x86int.h 17578 2010-10-29 04:21:26Z tterribe $
********************************************************************/
@ -19,24 +19,104 @@
# define _x86_x86int_H (1)
# include "../internal.h"
void oc_state_vtable_init_x86(oc_theora_state *_state);
# if defined(OC_X86_ASM)
# define oc_state_accel_init oc_state_accel_init_x86
# if defined(OC_X86_64_ASM)
/*x86-64 guarantees SIMD support up through at least SSE2.
If the best routine we have available only needs SSE2 (which at the moment
covers all of them), then we can avoid runtime detection and the indirect
call.*/
# define oc_frag_copy(_state,_dst,_src,_ystride) \
oc_frag_copy_mmx(_dst,_src,_ystride)
# define oc_frag_copy_list(_state,_dst_frame,_src_frame,_ystride, \
_fragis,_nfragis,_frag_buf_offs) \
oc_frag_copy_list_mmx(_dst_frame,_src_frame,_ystride, \
_fragis,_nfragis,_frag_buf_offs)
# define oc_frag_recon_intra(_state,_dst,_ystride,_residue) \
oc_frag_recon_intra_mmx(_dst,_ystride,_residue)
# define oc_frag_recon_inter(_state,_dst,_src,_ystride,_residue) \
oc_frag_recon_inter_mmx(_dst,_src,_ystride,_residue)
# define oc_frag_recon_inter2(_state,_dst,_src1,_src2,_ystride,_residue) \
oc_frag_recon_inter2_mmx(_dst,_src1,_src2,_ystride,_residue)
# define oc_idct8x8(_state,_y,_x,_last_zzi) \
oc_idct8x8_sse2(_y,_x,_last_zzi)
# define oc_state_frag_recon oc_state_frag_recon_mmx
# define oc_loop_filter_init(_state,_bv,_flimit) \
oc_loop_filter_init_mmxext(_bv,_flimit)
# define oc_state_loop_filter_frag_rows oc_state_loop_filter_frag_rows_mmxext
# define oc_restore_fpu(_state) \
oc_restore_fpu_mmx()
# else
# define OC_STATE_USE_VTABLE (1)
# endif
# endif
# include "../state.h"
# include "x86cpu.h"
/*Converts the expression in the argument to a string.*/
#define OC_M2STR(_s) #_s
/*Memory operands do not always include an offset.
To avoid warnings, we force an offset with %H (which adds 8).*/
# if __GNUC_PREREQ(4,0)
# define OC_MEM_OFFS(_offs,_name) \
OC_M2STR(_offs-8+%H[_name])
# endif
/*If your gcc version does't support %H, then you get to suffer the warnings.
Note that Apple's gas breaks on things like _offs+(%esp): it throws away the
whole offset, instead of substituting in 0 for the missing operand to +.*/
# if !defined(OC_MEM_OFFS)
# define OC_MEM_OFFS(_offs,_name) \
OC_M2STR(_offs+%[_name])
# endif
/*Declare an array operand with an exact size.
This tells gcc we're going to clobber this memory region, without having to
clobber all of "memory" and lets us access local buffers directly using the
stack pointer, without allocating a separate register to point to them.*/
#define OC_ARRAY_OPERAND(_type,_ptr,_size) \
(*({ \
struct{_type array_value__[(_size)];} *array_addr__=(void *)(_ptr); \
array_addr__; \
}))
/*Declare an array operand with an exact size.
This tells gcc we're going to clobber this memory region, without having to
clobber all of "memory" and lets us access local buffers directly using the
stack pointer, without allocating a separate register to point to them.*/
#define OC_CONST_ARRAY_OPERAND(_type,_ptr,_size) \
(*({ \
const struct{_type array_value__[(_size)];} *array_addr__= \
(const void *)(_ptr); \
array_addr__; \
}))
extern const unsigned short __attribute__((aligned(16))) OC_IDCT_CONSTS[64];
void oc_state_accel_init_x86(oc_theora_state *_state);
void oc_frag_copy_mmx(unsigned char *_dst,
const unsigned char *_src,int _ystride);
void oc_frag_copy_list_mmx(unsigned char *_dst_frame,
const unsigned char *_src_frame,int _ystride,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,const ptrdiff_t *_frag_buf_offs);
void oc_frag_recon_intra_mmx(unsigned char *_dst,int _ystride,
const ogg_int16_t *_residue);
void oc_frag_recon_inter_mmx(unsigned char *_dst,
const unsigned char *_src,int _ystride,const ogg_int16_t *_residue);
void oc_frag_recon_inter2_mmx(unsigned char *_dst,const unsigned char *_src1,
const unsigned char *_src2,int _ystride,const ogg_int16_t *_residue);
void oc_idct8x8_mmx(ogg_int16_t _y[64],int _last_zzi);
void oc_idct8x8_mmx(ogg_int16_t _y[64],ogg_int16_t _x[64],int _last_zzi);
void oc_idct8x8_sse2(ogg_int16_t _y[64],ogg_int16_t _x[64],int _last_zzi);
void oc_state_frag_recon_mmx(const oc_theora_state *_state,ptrdiff_t _fragi,
int _pli,ogg_int16_t _dct_coeffs[64],int _last_zzi,ogg_uint16_t _dc_quant);
void oc_state_frag_copy_list_mmx(const oc_theora_state *_state,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,
int _dst_frame,int _src_frame,int _pli);
int _pli,ogg_int16_t _dct_coeffs[128],int _last_zzi,ogg_uint16_t _dc_quant);
void oc_loop_filter_init_mmx(signed char _bv[256],int _flimit);
void oc_loop_filter_init_mmxext(signed char _bv[256],int _flimit);
void oc_state_loop_filter_frag_rows_mmx(const oc_theora_state *_state,
int _bv[256],int _refi,int _pli,int _fragy0,int _fragy_end);
signed char _bv[256],int _refi,int _pli,int _fragy0,int _fragy_end);
void oc_state_loop_filter_frag_rows_mmxext(const oc_theora_state *_state,
signed char _bv[256],int _refi,int _pli,int _fragy0,int _fragy_end);
void oc_restore_fpu_mmx(void);
#endif

47
media/libtheora/lib/x86/x86state.c
View File

@ -11,7 +11,7 @@
********************************************************************
function:
last mod: $Id: x86state.c 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: x86state.c 17421 2010-09-22 16:46:18Z giles $
********************************************************************/
@ -19,8 +19,6 @@
#if defined(OC_X86_ASM)
#include "../cpu.c"
/*This table has been modified from OC_FZIG_ZAG by baking a 4x4 transpose into
each quadrant of the destination.*/
static const unsigned char OC_FZIG_ZAG_MMX[128]={
@ -39,24 +37,59 @@ static const unsigned char OC_FZIG_ZAG_MMX[128]={
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64
};
void oc_state_vtable_init_x86(oc_theora_state *_state){
/*This table has been modified from OC_FZIG_ZAG by baking an 8x8 transpose into
the destination.*/
static const unsigned char OC_FZIG_ZAG_SSE2[128]={
0, 8, 1, 2, 9,16,24,17,
10, 3, 4,11,18,25,32,40,
33,26,19,12, 5, 6,13,20,
27,34,41,48,56,49,42,35,
28,21,14, 7,15,22,29,36,
43,50,57,58,51,44,37,30,
23,31,38,45,52,59,60,53,
46,39,47,54,61,62,55,63,
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64,
64,64,64,64,64,64,64,64
};
void oc_state_accel_init_x86(oc_theora_state *_state){
oc_state_accel_init_c(_state);
_state->cpu_flags=oc_cpu_flags_get();
# if defined(OC_STATE_USE_VTABLE)
if(_state->cpu_flags&OC_CPU_X86_MMX){
_state->opt_vtable.frag_copy=oc_frag_copy_mmx;
_state->opt_vtable.frag_copy_list=oc_frag_copy_list_mmx;
_state->opt_vtable.frag_recon_intra=oc_frag_recon_intra_mmx;
_state->opt_vtable.frag_recon_inter=oc_frag_recon_inter_mmx;
_state->opt_vtable.frag_recon_inter2=oc_frag_recon_inter2_mmx;
_state->opt_vtable.idct8x8=oc_idct8x8_mmx;
_state->opt_vtable.state_frag_recon=oc_state_frag_recon_mmx;
_state->opt_vtable.state_frag_copy_list=oc_state_frag_copy_list_mmx;
_state->opt_vtable.loop_filter_init=oc_loop_filter_init_mmx;
_state->opt_vtable.state_loop_filter_frag_rows=
oc_state_loop_filter_frag_rows_mmx;
_state->opt_vtable.restore_fpu=oc_restore_fpu_mmx;
_state->opt_data.dct_fzig_zag=OC_FZIG_ZAG_MMX;
}
else oc_state_vtable_init_c(_state);
if(_state->cpu_flags&OC_CPU_X86_MMXEXT){
_state->opt_vtable.loop_filter_init=oc_loop_filter_init_mmxext;
_state->opt_vtable.state_loop_filter_frag_rows=
oc_state_loop_filter_frag_rows_mmxext;
}
if(_state->cpu_flags&OC_CPU_X86_SSE2){
_state->opt_vtable.idct8x8=oc_idct8x8_sse2;
# endif
_state->opt_data.dct_fzig_zag=OC_FZIG_ZAG_SSE2;
# if defined(OC_STATE_USE_VTABLE)
}
# endif
}
#endif

83
media/libtheora/lib/x86_vc/mmxfrag.c
View File

@ -11,7 +11,7 @@
********************************************************************
function:
last mod: $Id: mmxfrag.c 16578 2009-09-25 19:50:48Z cristianadam $
last mod: $Id: mmxfrag.c 17446 2010-09-23 20:06:20Z tterribe $
********************************************************************/
@ -22,10 +22,61 @@
The iteration each instruction belongs to is marked in the comments as #i.*/
#include <stddef.h>
#include "x86int.h"
#include "mmxfrag.h"
#if defined(OC_X86_ASM)
/*Copies an 8x8 block of pixels from _src to _dst, assuming _ystride bytes
between rows.*/
# define OC_FRAG_COPY_MMX(_dst,_src,_ystride) \
do{ \
const unsigned char *src; \
unsigned char *dst; \
src=(_src); \
dst=(_dst); \
__asm mov SRC,src \
__asm mov DST,dst \
__asm mov YSTRIDE,_ystride \
/*src+0*ystride*/ \
__asm movq mm0,[SRC] \
/*src+1*ystride*/ \
__asm movq mm1,[SRC+YSTRIDE] \
/*ystride3=ystride*3*/ \
__asm lea YSTRIDE3,[YSTRIDE+YSTRIDE*2] \
/*src+2*ystride*/ \
__asm movq mm2,[SRC+YSTRIDE*2] \
/*src+3*ystride*/ \
__asm movq mm3,[SRC+YSTRIDE3] \
/*dst+0*ystride*/ \
__asm movq [DST],mm0 \
/*dst+1*ystride*/ \
__asm movq [DST+YSTRIDE],mm1 \
/*Pointer to next 4.*/ \
__asm lea SRC,[SRC+YSTRIDE*4] \
/*dst+2*ystride*/ \
__asm movq [DST+YSTRIDE*2],mm2 \
/*dst+3*ystride*/ \
__asm movq [DST+YSTRIDE3],mm3 \
/*Pointer to next 4.*/ \
__asm lea DST,[DST+YSTRIDE*4] \
/*src+0*ystride*/ \
__asm movq mm0,[SRC] \
/*src+1*ystride*/ \
__asm movq mm1,[SRC+YSTRIDE] \
/*src+2*ystride*/ \
__asm movq mm2,[SRC+YSTRIDE*2] \
/*src+3*ystride*/ \
__asm movq mm3,[SRC+YSTRIDE3] \
/*dst+0*ystride*/ \
__asm movq [DST],mm0 \
/*dst+1*ystride*/ \
__asm movq [DST+YSTRIDE],mm1 \
/*dst+2*ystride*/ \
__asm movq [DST+YSTRIDE*2],mm2 \
/*dst+3*ystride*/ \
__asm movq [DST+YSTRIDE3],mm3 \
} \
while(0)
/*Copies an 8x8 block of pixels from _src to _dst, assuming _ystride bytes
between rows.*/
void oc_frag_copy_mmx(unsigned char *_dst,
@ -41,6 +92,34 @@ void oc_frag_copy_mmx(unsigned char *_dst,
#undef YSTRIDE3
}
/*Copies the fragments specified by the lists of fragment indices from one
frame to another.
_dst_frame: The reference frame to copy to.
_src_frame: The reference frame to copy from.
_ystride: The row stride of the reference frames.
_fragis: A pointer to a list of fragment indices.
_nfragis: The number of fragment indices to copy.
_frag_buf_offs: The offsets of fragments in the reference frames.*/
void oc_frag_copy_list_mmx(unsigned char *_dst_frame,
const unsigned char *_src_frame,int _ystride,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,const ptrdiff_t *_frag_buf_offs){
ptrdiff_t fragii;
for(fragii=0;fragii<_nfragis;fragii++){
ptrdiff_t frag_buf_off;
frag_buf_off=_frag_buf_offs[_fragis[fragii]];
#define SRC edx
#define DST eax
#define YSTRIDE ecx
#define YSTRIDE3 edi
OC_FRAG_COPY_MMX(_dst_frame+frag_buf_off,
_src_frame+frag_buf_off,_ystride);
#undef SRC
#undef DST
#undef YSTRIDE
#undef YSTRIDE3
}
}
void oc_frag_recon_intra_mmx(unsigned char *_dst,int _ystride,
const ogg_int16_t *_residue){
__asm{

61
media/libtheora/lib/x86_vc/mmxfrag.h
View File

@ -1,61 +0,0 @@
#if !defined(_x86_vc_mmxfrag_H)
# define _x86_vc_mmxfrag_H (1)
# include <stddef.h>
# include "x86int.h"
#if defined(OC_X86_ASM)
/*Copies an 8x8 block of pixels from _src to _dst, assuming _ystride bytes
between rows.*/
#define OC_FRAG_COPY_MMX(_dst,_src,_ystride) \
do{ \
const unsigned char *src; \
unsigned char *dst; \
src=(_src); \
dst=(_dst); \
__asm mov SRC,src \
__asm mov DST,dst \
__asm mov YSTRIDE,_ystride \
/*src+0*ystride*/ \
__asm movq mm0,[SRC] \
/*src+1*ystride*/ \
__asm movq mm1,[SRC+YSTRIDE] \
/*ystride3=ystride*3*/ \
__asm lea YSTRIDE3,[YSTRIDE+YSTRIDE*2] \
/*src+2*ystride*/ \
__asm movq mm2,[SRC+YSTRIDE*2] \
/*src+3*ystride*/ \
__asm movq mm3,[SRC+YSTRIDE3] \
/*dst+0*ystride*/ \
__asm movq [DST],mm0 \
/*dst+1*ystride*/ \
__asm movq [DST+YSTRIDE],mm1 \
/*Pointer to next 4.*/ \
__asm lea SRC,[SRC+YSTRIDE*4] \
/*dst+2*ystride*/ \
__asm movq [DST+YSTRIDE*2],mm2 \
/*dst+3*ystride*/ \
__asm movq [DST+YSTRIDE3],mm3 \
/*Pointer to next 4.*/ \
__asm lea DST,[DST+YSTRIDE*4] \
/*src+0*ystride*/ \
__asm movq mm0,[SRC] \
/*src+1*ystride*/ \
__asm movq mm1,[SRC+YSTRIDE] \
/*src+2*ystride*/ \
__asm movq mm2,[SRC+YSTRIDE*2] \
/*src+3*ystride*/ \
__asm movq mm3,[SRC+YSTRIDE3] \
/*dst+0*ystride*/ \
__asm movq [DST],mm0 \
/*dst+1*ystride*/ \
__asm movq [DST+YSTRIDE],mm1 \
/*dst+2*ystride*/ \
__asm movq [DST+YSTRIDE*2],mm2 \
/*dst+3*ystride*/ \
__asm movq [DST+YSTRIDE3],mm3 \
} \
while(0)
# endif
#endif

235
media/libtheora/lib/x86_vc/mmxidct.c
View File

@ -11,7 +11,7 @@
********************************************************************
function:
last mod: $Id: mmxidct.c 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: mmxidct.c 17446 2010-09-23 20:06:20Z tterribe $
********************************************************************/
@ -24,15 +24,15 @@
/*These are offsets into the table of constants below.*/
/*7 rows of cosines, in order: pi/16 * (1 ... 7).*/
#define OC_COSINE_OFFSET (0)
#define OC_COSINE_OFFSET (8)
/*A row of 8's.*/
#define OC_EIGHT_OFFSET (56)
#define OC_EIGHT_OFFSET (0)
/*A table of constants used by the MMX routines.*/
static const __declspec(align(16))ogg_uint16_t
OC_IDCT_CONSTS[(7+1)*4]={
static const OC_ALIGN16(ogg_uint16_t) OC_IDCT_CONSTS[(1+7)*4]={
8, 8, 8, 8,
(ogg_uint16_t)OC_C1S7,(ogg_uint16_t)OC_C1S7,
(ogg_uint16_t)OC_C1S7,(ogg_uint16_t)OC_C1S7,
(ogg_uint16_t)OC_C2S6,(ogg_uint16_t)OC_C2S6,
@ -46,28 +46,27 @@ static const __declspec(align(16))ogg_uint16_t
(ogg_uint16_t)OC_C6S2,(ogg_uint16_t)OC_C6S2,
(ogg_uint16_t)OC_C6S2,(ogg_uint16_t)OC_C6S2,
(ogg_uint16_t)OC_C7S1,(ogg_uint16_t)OC_C7S1,
(ogg_uint16_t)OC_C7S1,(ogg_uint16_t)OC_C7S1,
8, 8, 8, 8
(ogg_uint16_t)OC_C7S1,(ogg_uint16_t)OC_C7S1
};
/*38 cycles*/
#define OC_IDCT_BEGIN __asm{ \
__asm movq mm2,OC_I(3) \
#define OC_IDCT_BEGIN(_y,_x) __asm{ \
__asm movq mm2,OC_I(3,_x) \
__asm movq mm6,OC_C(3) \
__asm movq mm4,mm2 \
__asm movq mm7,OC_J(5) \
__asm movq mm7,OC_J(5,_x) \
__asm pmulhw mm4,mm6 \
__asm movq mm1,OC_C(5) \
__asm pmulhw mm6,mm7 \
__asm movq mm5,mm1 \
__asm pmulhw mm1,mm2 \
__asm movq mm3,OC_I(1) \
__asm movq mm3,OC_I(1,_x) \
__asm pmulhw mm5,mm7 \
__asm movq mm0,OC_C(1) \
__asm paddw mm4,mm2 \
__asm paddw mm6,mm7 \
__asm paddw mm2,mm1 \
__asm movq mm1,OC_J(7) \
__asm movq mm1,OC_J(7,_x) \
__asm paddw mm7,mm5 \
__asm movq mm5,mm0 \
__asm pmulhw mm0,mm3 \
@ -77,13 +76,13 @@ static const __declspec(align(16))ogg_uint16_t
__asm psubw mm6,mm2 \
__asm paddw mm0,mm3 \
__asm pmulhw mm3,mm7 \
__asm movq mm2,OC_I(2) \
__asm movq mm2,OC_I(2,_x) \
__asm pmulhw mm7,mm1 \
__asm paddw mm5,mm1 \
__asm movq mm1,mm2 \
__asm pmulhw mm2,OC_C(2) \
__asm psubw mm3,mm5 \
__asm movq mm5,OC_J(6) \
__asm movq mm5,OC_J(6,_x) \
__asm paddw mm0,mm7 \
__asm movq mm7,mm5 \
__asm psubw mm0,mm4 \
@ -97,18 +96,18 @@ static const __declspec(align(16))ogg_uint16_t
__asm paddw mm6,mm6 \
__asm pmulhw mm7,OC_C(6) \
__asm paddw mm6,mm3 \
__asm movq OC_I(1),mm4 \
__asm movq OC_I(1,_y),mm4 \
__asm psubw mm1,mm5 \
__asm movq mm4,OC_C(4) \
__asm movq mm5,mm3 \
__asm pmulhw mm3,mm4 \
__asm paddw mm7,mm2 \
__asm movq OC_I(2),mm6 \
__asm movq OC_I(2,_y),mm6 \
__asm movq mm2,mm0 \
__asm movq mm6,OC_I(0) \
__asm movq mm6,OC_I(0,_x) \
__asm pmulhw mm0,mm4 \
__asm paddw mm5,mm3 \
__asm movq mm3,OC_J(4) \
__asm movq mm3,OC_J(4,_x) \
__asm psubw mm5,mm1 \
__asm paddw mm2,mm0 \
__asm psubw mm6,mm3 \
@ -122,17 +121,17 @@ static const __declspec(align(16))ogg_uint16_t
__asm paddw mm6,mm0 \
__asm psubw mm6,mm2 \
__asm paddw mm2,mm2 \
__asm movq mm0,OC_I(1) \
__asm movq mm0,OC_I(1,_y) \
__asm paddw mm2,mm6 \
__asm paddw mm4,mm3 \
__asm psubw mm2,mm1 \
}
/*38+8=46 cycles.*/
#define OC_ROW_IDCT __asm{ \
OC_IDCT_BEGIN \
#define OC_ROW_IDCT(_y,_x) __asm{ \
OC_IDCT_BEGIN(_y,_x) \
/*r3=D'*/ \
__asm movq mm3,OC_I(2) \
__asm movq mm3,OC_I(2,_y) \
/*r4=E'=E-G*/ \
__asm psubw mm4,mm7 \
/*r1=H'+H'*/ \
@ -157,7 +156,7 @@ static const __declspec(align(16))ogg_uint16_t
__asm psubw mm7,mm0 \
__asm paddw mm0,mm0 \
/*Save R1.*/ \
__asm movq OC_I(1),mm1 \
__asm movq OC_I(1,_y),mm1 \
/*r0=R0=G.+C.*/ \
__asm paddw mm0,mm7 \
}
@ -190,10 +189,10 @@ static const __declspec(align(16))ogg_uint16_t
Since r1 is free at entry, we calculate the Js first.*/
/*19 cycles.*/
#define OC_TRANSPOSE __asm{ \
#define OC_TRANSPOSE(_y) __asm{ \
__asm movq mm1,mm4 \
__asm punpcklwd mm4,mm5 \
__asm movq OC_I(0),mm0 \
__asm movq OC_I(0,_y),mm0 \
__asm punpckhwd mm1,mm5 \
__asm movq mm0,mm6 \
__asm punpcklwd mm6,mm7 \
@ -201,17 +200,17 @@ static const __declspec(align(16))ogg_uint16_t
__asm punpckldq mm4,mm6 \
__asm punpckhdq mm5,mm6 \
__asm movq mm6,mm1 \
__asm movq OC_J(4),mm4 \
__asm movq OC_J(4,_y),mm4 \
__asm punpckhwd mm0,mm7 \
__asm movq OC_J(5),mm5 \
__asm movq OC_J(5,_y),mm5 \
__asm punpckhdq mm6,mm0 \
__asm movq mm4,OC_I(0) \
__asm movq mm4,OC_I(0,_y) \
__asm punpckldq mm1,mm0 \
__asm movq mm5,OC_I(1) \
__asm movq mm5,OC_I(1,_y) \
__asm movq mm0,mm4 \
__asm movq OC_J(7),mm6 \
__asm movq OC_J(7,_y),mm6 \
__asm punpcklwd mm0,mm5 \
__asm movq OC_J(6),mm1 \
__asm movq OC_J(6,_y),mm1 \
__asm punpckhwd mm4,mm5 \
__asm movq mm5,mm2 \
__asm punpcklwd mm2,mm3 \
@ -219,18 +218,18 @@ static const __declspec(align(16))ogg_uint16_t
__asm punpckldq mm0,mm2 \
__asm punpckhdq mm1,mm2 \
__asm movq mm2,mm4 \
__asm movq OC_I(0),mm0 \
__asm movq OC_I(0,_y),mm0 \
__asm punpckhwd mm5,mm3 \
__asm movq OC_I(1),mm1 \
__asm movq OC_I(1,_y),mm1 \
__asm punpckhdq mm4,mm5 \
__asm punpckldq mm2,mm5 \
__asm movq OC_I(3),mm4 \
__asm movq OC_I(2),mm2 \
__asm movq OC_I(3,_y),mm4 \
__asm movq OC_I(2,_y),mm2 \
}
/*38+19=57 cycles.*/
#define OC_COLUMN_IDCT __asm{ \
OC_IDCT_BEGIN \
#define OC_COLUMN_IDCT(_y) __asm{ \
OC_IDCT_BEGIN(_y,_y) \
__asm paddw mm2,OC_8 \
/*r1=H'+H'*/ \
__asm paddw mm1,mm1 \
@ -243,15 +242,15 @@ static const __declspec(align(16))ogg_uint16_t
/*r1=NR1*/ \
__asm psraw mm1,4 \
/*r3=D'*/ \
__asm movq mm3,OC_I(2) \
__asm movq mm3,OC_I(2,_y) \
/*r7=G+G*/ \
__asm paddw mm7,mm7 \
/*Store NR2 at I(2).*/ \
__asm movq OC_I(2),mm2 \
__asm movq OC_I(2,_y),mm2 \
/*r7=G'=E+G*/ \
__asm paddw mm7,mm4 \
/*Store NR1 at I(1).*/ \
__asm movq OC_I(1),mm1 \
__asm movq OC_I(1,_y),mm1 \
/*r4=R4=E'-D'*/ \
__asm psubw mm4,mm3 \
__asm paddw mm4,OC_8 \
@ -273,11 +272,11 @@ static const __declspec(align(16))ogg_uint16_t
/*r6=NR6*/ \
__asm psraw mm6,4 \
/*Store NR4 at J(4).*/ \
__asm movq OC_J(4),mm4 \
__asm movq OC_J(4,_y),mm4 \
/*r5=NR5*/ \
__asm psraw mm5,4 \
/*Store NR3 at I(3).*/ \
__asm movq OC_I(3),mm3 \
__asm movq OC_I(3,_y),mm3 \
/*r7=R7=G'-C'*/ \
__asm psubw mm7,mm0 \
__asm paddw mm7,OC_8 \
@ -288,71 +287,92 @@ static const __declspec(align(16))ogg_uint16_t
/*r7=NR7*/ \
__asm psraw mm7,4 \
/*Store NR6 at J(6).*/ \
__asm movq OC_J(6),mm6 \
__asm movq OC_J(6,_y),mm6 \
/*r0=NR0*/ \
__asm psraw mm0,4 \
/*Store NR5 at J(5).*/ \
__asm movq OC_J(5),mm5 \
__asm movq OC_J(5,_y),mm5 \
/*Store NR7 at J(7).*/ \
__asm movq OC_J(7),mm7 \
__asm movq OC_J(7,_y),mm7 \
/*Store NR0 at I(0).*/ \
__asm movq OC_I(0),mm0 \
__asm movq OC_I(0,_y),mm0 \
}
#define OC_MID(_m,_i) [CONSTS+_m+(_i)*8]
#define OC_C(_i) OC_MID(OC_COSINE_OFFSET,_i-1)
#define OC_8 OC_MID(OC_EIGHT_OFFSET,0)
static void oc_idct8x8_slow(ogg_int16_t _y[64]){
static void oc_idct8x8_slow(ogg_int16_t _y[64],ogg_int16_t _x[64]){
int i;
/*This routine accepts an 8x8 matrix, but in partially transposed form.
Every 4x4 block is transposed.*/
__asm{
#define CONSTS eax
#define Y edx
#define X ecx
mov CONSTS,offset OC_IDCT_CONSTS
mov Y,_y
#define OC_I(_k) [Y+_k*16]
#define OC_J(_k) [Y+(_k-4)*16+8]
OC_ROW_IDCT
OC_TRANSPOSE
mov X,_x
#define OC_I(_k,_y) [(_y)+(_k)*16]
#define OC_J(_k,_y) [(_y)+((_k)-4)*16+8]
OC_ROW_IDCT(Y,X)
OC_TRANSPOSE(Y)
#undef OC_I
#undef OC_J
#define OC_I(_k) [Y+(_k*16)+64]
#define OC_J(_k) [Y+(_k-4)*16+72]
OC_ROW_IDCT
OC_TRANSPOSE
#define OC_I(_k,_y) [(_y)+(_k)*16+64]
#define OC_J(_k,_y) [(_y)+((_k)-4)*16+72]
OC_ROW_IDCT(Y,X)
OC_TRANSPOSE(Y)
#undef OC_I
#undef OC_J
#define OC_I(_k) [Y+_k*16]
#define OC_J(_k) OC_I(_k)
OC_COLUMN_IDCT
#define OC_I(_k,_y) [(_y)+(_k)*16]
#define OC_J(_k,_y) OC_I(_k,_y)
OC_COLUMN_IDCT(Y)
#undef OC_I
#undef OC_J
#define OC_I(_k) [Y+_k*16+8]
#define OC_J(_k) OC_I(_k)
OC_COLUMN_IDCT
#define OC_I(_k,_y) [(_y)+(_k)*16+8]
#define OC_J(_k,_y) OC_I(_k,_y)
OC_COLUMN_IDCT(Y)
#undef OC_I
#undef OC_J
#undef CONSTS
#undef Y
#undef X
}
if(_x!=_y){
int i;
__asm pxor mm0,mm0;
for(i=0;i<4;i++){
ogg_int16_t *x;
x=_x+16*i;
#define X ecx
__asm{
mov X,x
movq [X+0x00],mm0
movq [X+0x08],mm0
movq [X+0x10],mm0
movq [X+0x18],mm0
}
#undef X
}
}
}
/*25 cycles.*/
#define OC_IDCT_BEGIN_10 __asm{ \
__asm movq mm2,OC_I(3) \
#define OC_IDCT_BEGIN_10(_y,_x) __asm{ \
__asm movq mm2,OC_I(3,_x) \
__asm nop \
__asm movq mm6,OC_C(3) \
__asm movq mm4,mm2 \
__asm movq mm1,OC_C(5) \
__asm pmulhw mm4,mm6 \
__asm movq mm3,OC_I(1) \
__asm movq mm3,OC_I(1,_x) \
__asm pmulhw mm1,mm2 \
__asm movq mm0,OC_C(1) \
__asm paddw mm4,mm2 \
__asm pxor mm6,mm6 \
__asm paddw mm2,mm1 \
__asm movq mm5,OC_I(2) \
__asm movq mm5,OC_I(2,_x) \
__asm pmulhw mm0,mm3 \
__asm movq mm1,mm5 \
__asm paddw mm0,mm3 \
@ -360,43 +380,43 @@ static void oc_idct8x8_slow(ogg_int16_t _y[64]){
__asm psubw mm6,mm2 \
__asm pmulhw mm5,OC_C(2) \
__asm psubw mm0,mm4 \
__asm movq mm7,OC_I(2) \
__asm movq mm7,OC_I(2,_x) \
__asm paddw mm4,mm4 \
__asm paddw mm7,mm5 \
__asm paddw mm4,mm0 \
__asm pmulhw mm1,OC_C(6) \
__asm psubw mm3,mm6 \
__asm movq OC_I(1),mm4 \
__asm movq OC_I(1,_y),mm4 \
__asm paddw mm6,mm6 \
__asm movq mm4,OC_C(4) \
__asm paddw mm6,mm3 \
__asm movq mm5,mm3 \
__asm pmulhw mm3,mm4 \
__asm movq OC_I(2),mm6 \
__asm movq OC_I(2,_y),mm6 \
__asm movq mm2,mm0 \
__asm movq mm6,OC_I(0) \
__asm movq mm6,OC_I(0,_x) \
__asm pmulhw mm0,mm4 \
__asm paddw mm5,mm3 \
__asm paddw mm2,mm0 \
__asm psubw mm5,mm1 \
__asm pmulhw mm6,mm4 \
__asm paddw mm6,OC_I(0) \
__asm paddw mm6,OC_I(0,_x) \
__asm paddw mm1,mm1 \
__asm movq mm4,mm6 \
__asm paddw mm1,mm5 \
__asm psubw mm6,mm2 \
__asm paddw mm2,mm2 \
__asm movq mm0,OC_I(1) \
__asm movq mm0,OC_I(1,_y) \
__asm paddw mm2,mm6 \
__asm psubw mm2,mm1 \
__asm nop \
}
/*25+8=33 cycles.*/
#define OC_ROW_IDCT_10 __asm{ \
OC_IDCT_BEGIN_10 \
#define OC_ROW_IDCT_10(_y,_x) __asm{ \
OC_IDCT_BEGIN_10(_y,_x) \
/*r3=D'*/ \
__asm movq mm3,OC_I(2) \
__asm movq mm3,OC_I(2,_y) \
/*r4=E'=E-G*/ \
__asm psubw mm4,mm7 \
/*r1=H'+H'*/ \
@ -421,14 +441,14 @@ static void oc_idct8x8_slow(ogg_int16_t _y[64]){
__asm psubw mm7,mm0 \
__asm paddw mm0,mm0 \
/*Save R1.*/ \
__asm movq OC_I(1),mm1 \
__asm movq OC_I(1,_y),mm1 \
/*r0=R0=G'+C'*/ \
__asm paddw mm0,mm7 \
}
/*25+19=44 cycles'*/
#define OC_COLUMN_IDCT_10 __asm{ \
OC_IDCT_BEGIN_10 \
#define OC_COLUMN_IDCT_10(_y) __asm{ \
OC_IDCT_BEGIN_10(_y,_y) \
__asm paddw mm2,OC_8 \
/*r1=H'+H'*/ \
__asm paddw mm1,mm1 \
@ -441,15 +461,15 @@ static void oc_idct8x8_slow(ogg_int16_t _y[64]){
/*r1=NR1*/ \
__asm psraw mm1,4 \
/*r3=D'*/ \
__asm movq mm3,OC_I(2) \
__asm movq mm3,OC_I(2,_y) \
/*r7=G+G*/ \
__asm paddw mm7,mm7 \
/*Store NR2 at I(2).*/ \
__asm movq OC_I(2),mm2 \
__asm movq OC_I(2,_y),mm2 \
/*r7=G'=E+G*/ \
__asm paddw mm7,mm4 \
/*Store NR1 at I(1).*/ \
__asm movq OC_I(1),mm1 \
__asm movq OC_I(1,_y),mm1 \
/*r4=R4=E'-D'*/ \
__asm psubw mm4,mm3 \
__asm paddw mm4,OC_8 \
@ -471,11 +491,11 @@ static void oc_idct8x8_slow(ogg_int16_t _y[64]){
/*r6=NR6*/ \
__asm psraw mm6,4 \
/*Store NR4 at J(4).*/ \
__asm movq OC_J(4),mm4 \
__asm movq OC_J(4,_y),mm4 \
/*r5=NR5*/ \
__asm psraw mm5,4 \
/*Store NR3 at I(3).*/ \
__asm movq OC_I(3),mm3 \
__asm movq OC_I(3,_y),mm3 \
/*r7=R7=G'-C'*/ \
__asm psubw mm7,mm0 \
__asm paddw mm7,OC_8 \
@ -486,50 +506,65 @@ static void oc_idct8x8_slow(ogg_int16_t _y[64]){
/*r7=NR7*/ \
__asm psraw mm7,4 \
/*Store NR6 at J(6).*/ \
__asm movq OC_J(6),mm6 \
__asm movq OC_J(6,_y),mm6 \
/*r0=NR0*/ \
__asm psraw mm0,4 \
/*Store NR5 at J(5).*/ \
__asm movq OC_J(5),mm5 \
__asm movq OC_J(5,_y),mm5 \
/*Store NR7 at J(7).*/ \
__asm movq OC_J(7),mm7 \
__asm movq OC_J(7,_y),mm7 \
/*Store NR0 at I(0).*/ \
__asm movq OC_I(0),mm0 \
__asm movq OC_I(0,_y),mm0 \
}
static void oc_idct8x8_10(ogg_int16_t _y[64]){
static void oc_idct8x8_10(ogg_int16_t _y[64],ogg_int16_t _x[64]){
__asm{
#define CONSTS eax
#define Y edx
#define X ecx
mov CONSTS,offset OC_IDCT_CONSTS
mov Y,_y
#define OC_I(_k) [Y+_k*16]
#define OC_J(_k) [Y+(_k-4)*16+8]
mov X,_x
#define OC_I(_k,_y) [(_y)+(_k)*16]
#define OC_J(_k,_y) [(_y)+((_k)-4)*16+8]
/*Done with dequant, descramble, and partial transpose.
Now do the iDCT itself.*/
OC_ROW_IDCT_10
OC_TRANSPOSE
OC_ROW_IDCT_10(Y,X)
OC_TRANSPOSE(Y)
#undef OC_I
#undef OC_J
#define OC_I(_k) [Y+_k*16]
#define OC_J(_k) OC_I(_k)
OC_COLUMN_IDCT_10
#define OC_I(_k,_y) [(_y)+(_k)*16]
#define OC_J(_k,_y) OC_I(_k,_y)
OC_COLUMN_IDCT_10(Y)
#undef OC_I
#undef OC_J
#define OC_I(_k) [Y+_k*16+8]
#define OC_J(_k) OC_I(_k)
OC_COLUMN_IDCT_10
#define OC_I(_k,_y) [(_y)+(_k)*16+8]
#define OC_J(_k,_y) OC_I(_k,_y)
OC_COLUMN_IDCT_10(Y)
#undef OC_I
#undef OC_J
#undef CONSTS
#undef Y
#undef X
}
if(_x!=_y){
#define X ecx
__asm{
pxor mm0,mm0;
mov X,_x
movq [X+0x00],mm0
movq [X+0x10],mm0
movq [X+0x20],mm0
movq [X+0x30],mm0
}
#undef X
}
}
/*Performs an inverse 8x8 Type-II DCT transform.
The input is assumed to be scaled by a factor of 4 relative to orthonormal
version of the transform.*/
void oc_idct8x8_mmx(ogg_int16_t _y[64],int _last_zzi){
void oc_idct8x8_mmx(ogg_int16_t _y[64],ogg_int16_t _x[64],int _last_zzi){
/*_last_zzi is subtly different from an actual count of the number of
coefficients we decoded for this block.
It contains the value of zzi BEFORE the final token in the block was
@ -555,8 +590,8 @@ void oc_idct8x8_mmx(ogg_int16_t _y[64],int _last_zzi){
gets.
Needless to say we inherited this approach from VP3.*/
/*Perform the iDCT.*/
if(_last_zzi<10)oc_idct8x8_10(_y);
else oc_idct8x8_slow(_y);
if(_last_zzi<=10)oc_idct8x8_10(_y,_x);
else oc_idct8x8_slow(_y,_x);
}
#endif

75
media/libtheora/lib/x86_vc/mmxstate.c
View File

@ -11,7 +11,7 @@
********************************************************************
function:
last mod: $Id: mmxstate.c 16584 2009-09-26 19:35:55Z tterribe $
last mod: $Id: mmxstate.c 17563 2010-10-25 17:40:54Z tterribe $
********************************************************************/
@ -19,17 +19,16 @@
Originally written by Rudolf Marek.*/
#include <string.h>
#include "x86int.h"
#include "mmxfrag.h"
#include "mmxloop.h"
#if defined(OC_X86_ASM)
void oc_state_frag_recon_mmx(const oc_theora_state *_state,ptrdiff_t _fragi,
int _pli,ogg_int16_t _dct_coeffs[64],int _last_zzi,ogg_uint16_t _dc_quant){
int _pli,ogg_int16_t _dct_coeffs[128],int _last_zzi,ogg_uint16_t _dc_quant){
unsigned char *dst;
ptrdiff_t frag_buf_off;
int ystride;
int mb_mode;
int refi;
/*Apply the inverse transform.*/
/*Special case only having a DC component.*/
if(_last_zzi<2){
@ -45,6 +44,7 @@ void oc_state_frag_recon_mmx(const oc_theora_state *_state,ptrdiff_t _fragi,
#define P ecx
mov Y,_dct_coeffs
movzx P,p
lea Y,[Y+128]
/*mm0=0000 0000 0000 AAAA*/
movd mm0,P
/*mm0=0000 0000 AAAA AAAA*/
@ -74,65 +74,32 @@ void oc_state_frag_recon_mmx(const oc_theora_state *_state,ptrdiff_t _fragi,
else{
/*Dequantize the DC coefficient.*/
_dct_coeffs[0]=(ogg_int16_t)(_dct_coeffs[0]*(int)_dc_quant);
oc_idct8x8_mmx(_dct_coeffs,_last_zzi);
oc_idct8x8_mmx(_dct_coeffs+64,_dct_coeffs,_last_zzi);
}
/*Fill in the target buffer.*/
frag_buf_off=_state->frag_buf_offs[_fragi];
mb_mode=_state->frags[_fragi].mb_mode;
refi=_state->frags[_fragi].refi;
ystride=_state->ref_ystride[_pli];
dst=_state->ref_frame_data[_state->ref_frame_idx[OC_FRAME_SELF]]+frag_buf_off;
if(mb_mode==OC_MODE_INTRA)oc_frag_recon_intra_mmx(dst,ystride,_dct_coeffs);
dst=_state->ref_frame_data[OC_FRAME_SELF]+frag_buf_off;
if(refi==OC_FRAME_SELF)oc_frag_recon_intra_mmx(dst,ystride,_dct_coeffs+64);
else{
const unsigned char *ref;
int mvoffsets[2];
ref=
_state->ref_frame_data[_state->ref_frame_idx[OC_FRAME_FOR_MODE(mb_mode)]]
+frag_buf_off;
ref=_state->ref_frame_data[refi]+frag_buf_off;
if(oc_state_get_mv_offsets(_state,mvoffsets,_pli,
_state->frag_mvs[_fragi][0],_state->frag_mvs[_fragi][1])>1){
_state->frag_mvs[_fragi])>1){
oc_frag_recon_inter2_mmx(dst,ref+mvoffsets[0],ref+mvoffsets[1],ystride,
_dct_coeffs);
_dct_coeffs+64);
}
else oc_frag_recon_inter_mmx(dst,ref+mvoffsets[0],ystride,_dct_coeffs);
else oc_frag_recon_inter_mmx(dst,ref+mvoffsets[0],ystride,_dct_coeffs+64);
}
}
/*We copy these entire function to inline the actual MMX routines so that we
use only a single indirect call.*/
/*Copies the fragments specified by the lists of fragment indices from one
frame to another.
_fragis: A pointer to a list of fragment indices.
_nfragis: The number of fragment indices to copy.
_dst_frame: The reference frame to copy to.
_src_frame: The reference frame to copy from.
_pli: The color plane the fragments lie in.*/
void oc_state_frag_copy_list_mmx(const oc_theora_state *_state,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,
int _dst_frame,int _src_frame,int _pli){
const ptrdiff_t *frag_buf_offs;
const unsigned char *src_frame_data;
unsigned char *dst_frame_data;
ptrdiff_t fragii;
int ystride;
dst_frame_data=_state->ref_frame_data[_state->ref_frame_idx[_dst_frame]];
src_frame_data=_state->ref_frame_data[_state->ref_frame_idx[_src_frame]];
ystride=_state->ref_ystride[_pli];
frag_buf_offs=_state->frag_buf_offs;
for(fragii=0;fragii<_nfragis;fragii++){
ptrdiff_t frag_buf_off;
frag_buf_off=frag_buf_offs[_fragis[fragii]];
#define SRC edx
#define DST eax
#define YSTRIDE ecx
#define YSTRIDE3 edi
OC_FRAG_COPY_MMX(dst_frame_data+frag_buf_off,
src_frame_data+frag_buf_off,ystride);
#undef SRC
#undef DST
#undef YSTRIDE
#undef YSTRIDE3
}
void oc_loop_filter_init_mmx(signed char _bv[256],int _flimit){
memset(_bv,~(_flimit<<1),8);
}
/*Apply the loop filter to a given set of fragment rows in the given plane.
@ -144,8 +111,7 @@ void oc_state_frag_copy_list_mmx(const oc_theora_state *_state,
_fragy0: The Y coordinate of the first fragment row to filter.
_fragy_end: The Y coordinate of the fragment row to stop filtering at.*/
void oc_state_loop_filter_frag_rows_mmx(const oc_theora_state *_state,
int _bv[256],int _refi,int _pli,int _fragy0,int _fragy_end){
OC_ALIGN8(unsigned char ll[8]);
signed char _bv[256],int _refi,int _pli,int _fragy0,int _fragy_end){
const oc_fragment_plane *fplane;
const oc_fragment *frags;
const ptrdiff_t *frag_buf_offs;
@ -156,13 +122,12 @@ void oc_state_loop_filter_frag_rows_mmx(const oc_theora_state *_state,
ptrdiff_t fragi0_end;
int ystride;
int nhfrags;
memset(ll,_state->loop_filter_limits[_state->qis[0]],sizeof(ll));
fplane=_state->fplanes+_pli;
nhfrags=fplane->nhfrags;
fragi_top=fplane->froffset;
fragi_bot=fragi_top+fplane->nfrags;
fragi0=fragi_top+_fragy0*(ptrdiff_t)nhfrags;
fragi0_end=fragi0+(_fragy_end-_fragy0)*(ptrdiff_t)nhfrags;
fragi0_end=fragi_top+_fragy_end*(ptrdiff_t)nhfrags;
ystride=_state->ref_ystride[_pli];
frags=_state->frags;
frag_buf_offs=_state->frag_buf_offs;
@ -187,13 +152,13 @@ void oc_state_loop_filter_frag_rows_mmx(const oc_theora_state *_state,
#define LL edx
#define D esi
#define D_WORD si
if(fragi>fragi0)OC_LOOP_FILTER_H_MMX(ref,ystride,ll);
if(fragi0>fragi_top)OC_LOOP_FILTER_V_MMX(ref,ystride,ll);
if(fragi>fragi0)OC_LOOP_FILTER_H_MMX(ref,ystride,_bv);
if(fragi0>fragi_top)OC_LOOP_FILTER_V_MMX(ref,ystride,_bv);
if(fragi+1<fragi_end&&!frags[fragi+1].coded){
OC_LOOP_FILTER_H_MMX(ref+8,ystride,ll);
OC_LOOP_FILTER_H_MMX(ref+8,ystride,_bv);
}
if(fragi+nhfrags<fragi_bot&&!frags[fragi+nhfrags].coded){
OC_LOOP_FILTER_V_MMX(ref+(ystride<<3),ystride,ll);
OC_LOOP_FILTER_V_MMX(ref+(ystride<<3),ystride,_bv);
}
#undef PIX
#undef YSTRIDE3

60
media/libtheora/lib/cpu.c → media/libtheora/lib/x86_vc/x86cpu.c
View File

@ -14,41 +14,17 @@
Originally written by Rudolf Marek.
function:
last mod: $Id: cpu.c 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: x86cpu.c 17410 2010-09-21 21:53:48Z tterribe $
********************************************************************/
#include "cpu.h"
#include "x86cpu.h"
#if !defined(OC_X86_ASM)
static ogg_uint32_t oc_cpu_flags_get(void){
ogg_uint32_t oc_cpu_flags_get(void){
return 0;
}
#else
# if !defined(_MSC_VER)
# if defined(__amd64__)||defined(__x86_64__)
/*On x86-64, gcc seems to be able to figure out how to save %rbx for us when
compiling with -fPIC.*/
# define cpuid(_op,_eax,_ebx,_ecx,_edx) \
__asm__ __volatile__( \
"cpuid\n\t" \
:[eax]"=a"(_eax),[ebx]"=b"(_ebx),[ecx]"=c"(_ecx),[edx]"=d"(_edx) \
:"a"(_op) \
:"cc" \
)
# else
/*On x86-32, not so much.*/
# define cpuid(_op,_eax,_ebx,_ecx,_edx) \
__asm__ __volatile__( \
"xchgl %%ebx,%[ebx]\n\t" \
"cpuid\n\t" \
"xchgl %%ebx,%[ebx]\n\t" \
:[eax]"=a"(_eax),[ebx]"=r"(_ebx),[ecx]"=c"(_ecx),[edx]"=d"(_edx) \
:"a"(_op) \
:"cc" \
)
# endif
# else
/*Why does MSVC need this complicated rigamarole?
At this point I honestly do not care.*/
@ -95,7 +71,6 @@ static void oc_detect_cpuid_helper(ogg_uint32_t *_eax,ogg_uint32_t *_ebx){
mov [ecx],ebx
}
}
# endif
static ogg_uint32_t oc_parse_intel_flags(ogg_uint32_t _edx,ogg_uint32_t _ecx){
ogg_uint32_t flags;
@ -124,7 +99,7 @@ static ogg_uint32_t oc_parse_amd_flags(ogg_uint32_t _edx,ogg_uint32_t _ecx){
return flags;
}
static ogg_uint32_t oc_cpu_flags_get(void){
ogg_uint32_t oc_cpu_flags_get(void){
ogg_uint32_t flags;
ogg_uint32_t eax;
ogg_uint32_t ebx;
@ -132,25 +107,7 @@ static ogg_uint32_t oc_cpu_flags_get(void){
ogg_uint32_t edx;
# if !defined(__amd64__)&&!defined(__x86_64__)
/*Not all x86-32 chips support cpuid, so we have to check.*/
# if !defined(_MSC_VER)
__asm__ __volatile__(
"pushfl\n\t"
"pushfl\n\t"
"popl %[a]\n\t"
"movl %[a],%[b]\n\t"
"xorl $0x200000,%[a]\n\t"
"pushl %[a]\n\t"
"popfl\n\t"
"pushfl\n\t"
"popl %[a]\n\t"
"popfl\n\t"
:[a]"=r"(eax),[b]"=r"(ebx)
:
:"cc"
);
# else
oc_detect_cpuid_helper(&eax,&ebx);
# endif
/*No cpuid.*/
if(eax==ebx)return 0;
# endif
@ -159,9 +116,18 @@ static ogg_uint32_t oc_cpu_flags_get(void){
if(ecx==0x6C65746E&&edx==0x49656E69&&ebx==0x756E6547||
/* 6 8 x M T e n i u n e G*/
ecx==0x3638784D&&edx==0x54656E69&&ebx==0x756E6547){
int family;
int model;
/*Intel, Transmeta (tested with Crusoe TM5800):*/
cpuid(1,eax,ebx,ecx,edx);
flags=oc_parse_intel_flags(edx,ecx);
family=(eax>>8)&0xF;
model=(eax>>4)&0xF;
/*The SSE unit on the Pentium M and Core Duo is much slower than the MMX
unit, so don't use it.*/
if(family==6&&(model==9||model==13||model==14)){
flags&=~(OC_CPU_X86_SSE2|OC_CPU_X86_PNI);
}
}
/* D M A c i t n e h t u A*/
else if(ecx==0x444D4163&&edx==0x69746E65&&ebx==0x68747541||

36
media/libtheora/lib/x86_vc/x86cpu.h Normal file
View File

@ -0,0 +1,36 @@
/********************************************************************
* *
* THIS FILE IS PART OF THE OggTheora SOFTWARE CODEC SOURCE CODE. *
* USE, DISTRIBUTION AND REPRODUCTION OF THIS LIBRARY SOURCE IS *
* GOVERNED BY A BSD-STYLE SOURCE LICENSE INCLUDED WITH THIS SOURCE *
* IN 'COPYING'. PLEASE READ THESE TERMS BEFORE DISTRIBUTING. *
* *
* THE Theora SOURCE CODE IS COPYRIGHT (C) 2002-2009 *
* by the Xiph.Org Foundation and contributors http://www.xiph.org/ *
* *
********************************************************************
function:
last mod: $Id: x86cpu.h 17410 2010-09-21 21:53:48Z tterribe $
********************************************************************/
#if !defined(_x86_vc_x86cpu_H)
# define _x86_vc_x86cpu_H (1)
#include "../internal.h"
#define OC_CPU_X86_MMX (1<<0)
#define OC_CPU_X86_3DNOW (1<<1)
#define OC_CPU_X86_3DNOWEXT (1<<2)
#define OC_CPU_X86_MMXEXT (1<<3)
#define OC_CPU_X86_SSE (1<<4)
#define OC_CPU_X86_SSE2 (1<<5)
#define OC_CPU_X86_PNI (1<<6)
#define OC_CPU_X86_SSSE3 (1<<7)
#define OC_CPU_X86_SSE4_1 (1<<8)
#define OC_CPU_X86_SSE4_2 (1<<9)
#define OC_CPU_X86_SSE4A (1<<10)
#define OC_CPU_X86_SSE5 (1<<11)
ogg_uint32_t oc_cpu_flags_get(void);
#endif

23
media/libtheora/lib/x86_vc/x86int.h
View File

@ -11,32 +11,39 @@
********************************************************************
function:
last mod: $Id: x86int.h 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: x86int.h 17410 2010-09-21 21:53:48Z tterribe $
********************************************************************/
#if !defined(_x86_vc_x86int_H)
# define _x86_vc_x86int_H (1)
# include "../internal.h"
# if defined(OC_X86_ASM)
# define oc_state_accel_init oc_state_accel_init_x86
# define OC_STATE_USE_VTABLE (1)
# endif
# include "../state.h"
# include "x86cpu.h"
void oc_state_vtable_init_x86(oc_theora_state *_state);
void oc_state_accel_init_x86(oc_theora_state *_state);
void oc_frag_copy_mmx(unsigned char *_dst,
const unsigned char *_src,int _ystride);
void oc_frag_copy_list_mmx(unsigned char *_dst_frame,
const unsigned char *_src_frame,int _ystride,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,const ptrdiff_t *_frag_buf_offs);
void oc_frag_recon_intra_mmx(unsigned char *_dst,int _ystride,
const ogg_int16_t *_residue);
void oc_frag_recon_inter_mmx(unsigned char *_dst,
const unsigned char *_src,int _ystride,const ogg_int16_t *_residue);
void oc_frag_recon_inter2_mmx(unsigned char *_dst,const unsigned char *_src1,
const unsigned char *_src2,int _ystride,const ogg_int16_t *_residue);
void oc_idct8x8_mmx(ogg_int16_t _y[64],int _last_zzi);
void oc_idct8x8_mmx(ogg_int16_t _y[64],ogg_int16_t _x[64],int _last_zzi);
void oc_state_frag_recon_mmx(const oc_theora_state *_state,ptrdiff_t _fragi,
int _pli,ogg_int16_t _dct_coeffs[64],int _last_zzi,ogg_uint16_t _dc_quant);
void oc_state_frag_copy_list_mmx(const oc_theora_state *_state,
const ptrdiff_t *_fragis,ptrdiff_t _nfragis,
int _dst_frame,int _src_frame,int _pli);
int _pli,ogg_int16_t _dct_coeffs[128],int _last_zzi,ogg_uint16_t _dc_quant);
void oc_loop_filter_init_mmx(signed char _bv[256],int _flimit);
void oc_state_loop_filter_frag_rows_mmx(const oc_theora_state *_state,
int _bv[256],int _refi,int _pli,int _fragy0,int _fragy_end);
signed char _bv[256],int _refi,int _pli,int _fragy0,int _fragy_end);
void oc_restore_fpu_mmx(void);
#endif

11
media/libtheora/lib/x86_vc/x86state.c
View File

@ -11,7 +11,7 @@
********************************************************************
function:
last mod: $Id: x86state.c 16503 2009-08-22 18:14:02Z giles $
last mod: $Id: x86state.c 17410 2010-09-21 21:53:48Z tterribe $
********************************************************************/
@ -19,8 +19,6 @@
#if defined(OC_X86_ASM)
#include "../cpu.c"
/*This table has been modified from OC_FZIG_ZAG by baking a 4x4 transpose into
each quadrant of the destination.*/
static const unsigned char OC_FZIG_ZAG_MMX[128]={
@ -42,21 +40,22 @@ static const unsigned char OC_FZIG_ZAG_MMX[128]={
64,64,64,64,64,64,64,64,
};
void oc_state_vtable_init_x86(oc_theora_state *_state){
void oc_state_accel_init_x86(oc_theora_state *_state){
_state->cpu_flags=oc_cpu_flags_get();
if(_state->cpu_flags&OC_CPU_X86_MMX){
_state->opt_vtable.frag_copy=oc_frag_copy_mmx;
_state->opt_vtable.frag_copy_list=oc_frag_copy_list_mmx;
_state->opt_vtable.frag_recon_intra=oc_frag_recon_intra_mmx;
_state->opt_vtable.frag_recon_inter=oc_frag_recon_inter_mmx;
_state->opt_vtable.frag_recon_inter2=oc_frag_recon_inter2_mmx;
_state->opt_vtable.idct8x8=oc_idct8x8_mmx;
_state->opt_vtable.state_frag_recon=oc_state_frag_recon_mmx;
_state->opt_vtable.state_frag_copy_list=oc_state_frag_copy_list_mmx;
_state->opt_vtable.loop_filter_init=oc_loop_filter_init_mmx;
_state->opt_vtable.state_loop_filter_frag_rows=
oc_state_loop_filter_frag_rows_mmx;
_state->opt_vtable.restore_fpu=oc_restore_fpu_mmx;
_state->opt_data.dct_fzig_zag=OC_FZIG_ZAG_MMX;
}
else oc_state_vtable_init_c(_state);
else oc_state_accel_init_c(_state);
}
#endif

45
media/libtheora/update.sh Normal file → Executable file
View File

@ -2,11 +2,20 @@
#
# Copies the needed files from a directory containing the original
# libtheora source that we need for the Mozilla HTML5 media support.
sed s/\#define\ OC_X86_ASM//g $1/config.h >./lib/config.h
sed s/\#define\ USE_ASM//g ./lib/config.h >./lib/config.h2
sed s/\#define\ THEORA_DISABLE_ENCODE//g ./lib/config.h2 >./lib/config.h
rm ./lib/config.h2
cp ./lib/config.h ./include/theora/config.h
sed \
-e s/\#define\ OC_X86_ASM//g \
-e s/\#define\ OC_X86_64_ASM//g \
-e s/\#define\ OC_ARM_ASM_EDSP\ 1//g \
-e s/\#define\ OC_ARM_ASM_MEDIA\ 1//g \
-e s/\#define\ OC_ARM_ASM_NEON\ 1//g \
-e s/\#define\ OC_ARM_ASM//g \
-e s/\#define\ THEORA_DISABLE_ENCODE//g \
$1/config.h > lib/config.h
sed \
-e s/@HAVE_ARM_ASM_EDSP@/1/g \
-e s/@HAVE_ARM_ASM_MEDIA@/1/g \
-e s/@HAVE_ARM_ASM_NEON@/1/g \
$1/lib/arm/armopts.s.in > lib/arm/armopts.s
cp $1/LICENSE ./LICENSE
cp $1/CHANGES ./CHANGES
cp $1/COPYING ./COPYING
@ -16,8 +25,6 @@ cp $1/lib/apiwrapper.c ./lib/
cp $1/lib/apiwrapper.h ./lib/
cp $1/lib/bitpack.c ./lib/
cp $1/lib/bitpack.h ./lib/
cp $1/lib/cpu.c ./lib/
cp $1/lib/cpu.h ./lib/
cp $1/lib/dct.h ./lib/
cp $1/lib/decapiwrapper.c ./lib/
cp $1/lib/decinfo.c ./lib/
@ -25,13 +32,9 @@ cp $1/lib/decint.h ./lib/
cp $1/lib/decode.c ./lib/
cp $1/lib/dequant.c ./lib/
cp $1/lib/dequant.h ./lib/
cp $1/lib/encint.h ./lib/
cp $1/lib/encoder_disabled.c ./lib/
cp $1/lib/enquant.h ./lib/
cp $1/lib/fragment.c ./lib/
cp $1/lib/huffdec.c ./lib/
cp $1/lib/huffdec.h ./lib/
cp $1/lib/huffenc.h ./lib/
cp $1/lib/huffman.h ./lib/
cp $1/lib/idct.c ./lib/
cp $1/lib/info.c ./lib/
@ -42,22 +45,36 @@ cp $1/lib/ocintrin.h ./lib/
cp $1/lib/quant.c ./lib/
cp $1/lib/quant.h ./lib/
cp $1/lib/state.c ./lib/
cp $1/lib/state.h ./lib/
cp $1/lib/arm/arm2gnu.pl ./lib/arm/
cp $1/lib/arm/armbits.h ./lib/arm/
cp $1/lib/arm/armbits.s ./lib/arm/
cp $1/lib/arm/armcpu.c ./lib/arm/
cp $1/lib/arm/armcpu.h ./lib/arm/
cp $1/lib/arm/armfrag.s ./lib/arm/
cp $1/lib/arm/armidct.s ./lib/arm/
cp $1/lib/arm/armint.h ./lib/arm/
cp $1/lib/arm/armloop.s ./lib/arm/
cp $1/lib/arm/armstate.c ./lib/arm/
cp $1/lib/x86/mmxfrag.c ./lib/x86/
cp $1/lib/x86/mmxfrag.h ./lib/x86/
cp $1/lib/x86/mmxidct.c ./lib/x86/
cp $1/lib/x86/mmxloop.h ./lib/x86/
cp $1/lib/x86/mmxstate.c ./lib/x86/
cp $1/lib/x86/sse2idct.c ./lib/x86/
cp $1/lib/x86/sse2trans.h ./lib/x86/
cp $1/lib/x86/x86cpu.c ./lib/x86/
cp $1/lib/x86/x86cpu.h ./lib/x86/
cp $1/lib/x86/x86int.h ./lib/x86/
cp $1/lib/x86/x86state.c ./lib/x86/
cp $1/lib/x86_vc/mmxfrag.c ./lib/x86_vc/
cp $1/lib/x86_vc/mmxfrag.h ./lib/x86_vc/
cp $1/lib/x86_vc/mmxidct.c ./lib/x86_vc/
cp $1/lib/x86_vc/mmxloop.h ./lib/x86_vc/
cp $1/lib/x86_vc/mmxstate.c ./lib/x86_vc/
cp $1/lib/x86_vc/x86cpu.c ./lib/x86_vc/
cp $1/lib/x86_vc/x86cpu.h ./lib/x86_vc/
cp $1/lib/x86_vc/x86int.h ./lib/x86_vc/
cp $1/lib/x86_vc/x86state.c ./lib/x86_vc/
cp $1/include/theora/theora.h ./include/theora/theora.h
cp $1/include/theora/theoradec.h ./include/theora/theoradec.h
cp $1/include/theora/theoraenc.h ./include/theora/theoraenc.h
cp $1/include/theora/codec.h ./include/theora/codec.h
patch -p3 <bug559343.patch