From a910049aea5e2be97b232dbdd7700d078eb7ecd0 Mon Sep 17 00:00:00 2001 From: John Koleszar Date: Fri, 19 Aug 2011 14:06:00 -0400 Subject: [PATCH] New RTCD implementation This is a proof of concept RTCD implementation to replace the current system of nested includes, prototypes, INVOKE macros, etc. Currently only the decoder specific functions are implemented in the new system. Additional functions will be added in subsequent commits. Overview: RTCD "functions" are implemented as either a global function pointer or a macro (when only one eligible specialization available). Functions which have RTCD specializations are listed using a simple DSL identifying the function's base name, its prototype, and the architecture extensions that specializations are available for. Advantages over the old system: - No INVOKE macros. A call to an RTCD function looks like an ordinary function call. - No need to pass vtables around. - If there is only one eligible function to call, the function is called directly, rather than indirecting through a function pointer. - Supports the notion of "required" extensions, so in combination with the above, on x86_64 if the best function available is sse2 or lower it will be called directly, since all x86_64 platforms implement sse2. - Elides all references to functions which will never be called, which could reduce binary size. For example if sse2 is required and there are both mmx and sse2 implementations of a certain function, the code will have no link time references to the mmx code. - Significantly easier to add a new function, just one file to edit. Disadvantages: - Requires global writable data (though this is not a new requirement) - 1 new generated source file. Change-Id: Iae6edab65315f79c168485c96872641c5aa09d55 --- build/make/configure.sh | 16 +- build/make/rtcd.sh | 330 ++++++++++++++++++ libs.mk | 15 +- vp8/common/arm/arm_systemdependent.c | 12 - vp8/common/arm/armv6/idct_blk_v6.c | 1 - vp8/common/arm/dequantize_arm.c | 3 +- vp8/common/arm/dequantize_arm.h | 59 ---- vp8/common/arm/neon/idct_blk_neon.c | 1 - vp8/common/blockd.h | 2 +- vp8/common/dequantize.c | 3 +- vp8/common/dequantize.h | 85 ----- vp8/common/generic/systemdependent.c | 10 +- vp8/common/idct_blk.c | 1 - vp8/common/invtrans.h | 3 +- vp8/common/onyxc_int.h | 2 - .../x86_dsystemdependent.c => common/rtcd.c} | 13 +- vp8/common/rtcd_defs.sh | 22 ++ vp8/common/x86/dequantize_x86.h | 58 --- vp8/common/x86/idct_blk_mmx.c | 3 +- vp8/common/x86/idct_blk_sse2.c | 1 - vp8/common/x86/x86_systemdependent.c | 8 - vp8/decoder/arm/arm_dsystemdependent.c | 39 --- vp8/decoder/decodframe.c | 13 +- vp8/decoder/generic/dsystemdependent.c | 33 -- vp8/decoder/onyxd_if.c | 4 +- vp8/decoder/onyxd_int.h | 3 - vp8/decoder/threading.c | 10 +- vp8/encoder/encodeframe.c | 4 +- vp8/encoder/onyx_if.c | 2 - vp8/vp8_common.mk | 5 +- vp8/vp8dx.mk | 7 - vp8/vp8dx_arm.mk | 14 - 32 files changed, 414 insertions(+), 368 deletions(-) create mode 100755 build/make/rtcd.sh delete mode 100644 vp8/common/arm/dequantize_arm.h delete mode 100644 vp8/common/dequantize.h rename vp8/{decoder/x86/x86_dsystemdependent.c => common/rtcd.c} (67%) create mode 100644 vp8/common/rtcd_defs.sh delete mode 100644 vp8/common/x86/dequantize_x86.h delete mode 100644 vp8/decoder/arm/arm_dsystemdependent.c delete mode 100644 vp8/decoder/generic/dsystemdependent.c delete mode 100644 vp8/vp8dx_arm.mk diff --git a/build/make/configure.sh b/build/make/configure.sh index 799a4397b..15134ab6a 100755 --- a/build/make/configure.sh +++ b/build/make/configure.sh @@ -391,6 +391,7 @@ LDFLAGS = ${LDFLAGS} ASFLAGS = ${ASFLAGS} extralibs = ${extralibs} AS_SFX = ${AS_SFX:-.asm} +RTCD_OPTIONS = ${RTCD_OPTIONS} EOF if enabled rvct; then cat >> $1 << EOF @@ -454,9 +455,22 @@ process_common_cmdline() { ;; --enable-?*|--disable-?*) eval `echo "$opt" | sed 's/--/action=/;s/-/ option=/;s/-/_/g'` - echo "${CMDLINE_SELECT} ${ARCH_EXT_LIST}" | grep "^ *$option\$" >/dev/null || die_unknown $opt + if echo "${ARCH_EXT_LIST}" | grep "^ *$option\$" >/dev/null; then + [ $action = "disable" ] && RTCD_OPTIONS="${RTCD_OPTIONS}${opt} " + else + echo "${CMDLINE_SELECT}" | grep "^ *$option\$" >/dev/null || + die_unknown $opt + fi $action $option ;; + --require-?*) + eval `echo "$opt" | sed 's/--/action=/;s/-/ option=/;s/-/_/g'` + if echo "${ARCH_EXT_LIST}" none | grep "^ *$option\$" >/dev/null; then + RTCD_OPTIONS="${RTCD_OPTIONS}${opt} " + else + die_unknown $opt + fi + ;; --force-enable-?*|--force-disable-?*) eval `echo "$opt" | sed 's/--force-/action=/;s/-/ option=/;s/-/_/g'` $action $option diff --git a/build/make/rtcd.sh b/build/make/rtcd.sh new file mode 100755 index 000000000..a5f1e6d73 --- /dev/null +++ b/build/make/rtcd.sh @@ -0,0 +1,330 @@ +#!/bin/sh +self=$0 + +usage() { + cat <&2 +Usage: $self [options] FILE + +Reads the Run Time CPU Detections definitions from FILE and generates a +C header file on stdout. + +Options: + --arch=ARCH Architecture to generate defs for (required) + --disable-EXT Disable support for EXT extensions + --require-EXT Require support for EXT extensions + --sym=SYMBOL Unique symbol to use for RTCD initialization function + --config=FILE File with CONFIG_FOO=yes lines to parse +EOF + exit 1 +} + +die() { + echo "$@" >&2 + exit 1 +} + +die_argument_required() { + die "Option $opt requires argument" +} + +for opt; do + optval="${opt#*=}" + case "$opt" in + --arch) die_argument_required;; + --arch=*) arch=${optval};; + --disable-*) eval "disable_${opt#--disable-}=true";; + --require-*) REQUIRES="${REQUIRES}${opt#--require-} ";; + --sym) die_argument_required;; + --sym=*) symbol=${optval};; + --config=*) config_file=${optval};; + -h|--help) + usage + ;; + -*) + die "Unrecognized option: ${opt%%=*}" + ;; + *) + defs_file="$defs_file $opt" + ;; + esac + shift +done +for f in $defs_file; do [ -f "$f" ] || usage; done +[ -n "$arch" ] || usage + +# Import the configuration +[ -f "$config_file" ] && eval $(grep CONFIG_ "$config_file") + +# +# Routines for the RTCD DSL to call +# +prototype() { + local rtyp + case "$1" in + unsigned) rtyp="$1 "; shift;; + esac + rtyp="${rtyp}$1" + local fn="$2" + local args="$3" + + eval "${2}_rtyp='$rtyp'" + eval "${2}_args='$3'" + ALL_FUNCS="$ALL_FUNCS $fn" + specialize $fn c +} + +specialize() { + local fn="$1" + shift + for opt in "$@"; do + eval "${fn}_${opt}=${fn}_${opt}" + done +} + +require() { + for fn in $ALL_FUNCS; do + for opt in "$@"; do + local ofn=$(eval "echo \$${fn}_${opt}") + [ -z "$ofn" ] && continue + + # if we already have a default, then we can disable it, as we know + # we can do better. + local best=$(eval "echo \$${fn}_default") + local best_ofn=$(eval "echo \$${best}") + [ -n "$best" ] && [ "$best_ofn" != "$ofn" ] && eval "${best}_link=false" + eval "${fn}_default=${fn}_${opt}" + eval "${fn}_${opt}_link=true" + done + done +} + +forward_decls() { + ALL_FORWARD_DECLS="$ALL_FORWARD_DECLS $1" +} + +# +# Include the user's directives +# +for f in $defs_file; do + . $f +done + +# +# Process the directives according to the command line +# +process_forward_decls() { + for fn in $ALL_FORWARD_DECLS; do + eval $fn + done +} + +determine_indirection() { + [ "$CONFIG_RUNTIME_CPU_DETECT" = "yes" ] || require $ALL_ARCHS + for fn in $ALL_FUNCS; do + local n="" + local rtyp="$(eval "echo \$${fn}_rtyp")" + local args="$(eval "echo \"\$${fn}_args\"")" + local dfn="$(eval "echo \$${fn}_default")" + dfn=$(eval "echo \$${dfn}") + for opt in "$@"; do + local ofn=$(eval "echo \$${fn}_${opt}") + [ -z "$ofn" ] && continue + local link=$(eval "echo \$${fn}_${opt}_link") + [ "$link" = "false" ] && continue + n="${n}x" + done + if [ "$n" = "x" ]; then + eval "${fn}_indirect=false" + else + eval "${fn}_indirect=true" + fi + done +} + +declare_function_pointers() { + for fn in $ALL_FUNCS; do + local rtyp="$(eval "echo \$${fn}_rtyp")" + local args="$(eval "echo \"\$${fn}_args\"")" + local dfn="$(eval "echo \$${fn}_default")" + dfn=$(eval "echo \$${dfn}") + for opt in "$@"; do + local ofn=$(eval "echo \$${fn}_${opt}") + [ -z "$ofn" ] && continue + echo "$rtyp ${ofn}($args);" + done + if [ "$(eval "echo \$${fn}_indirect")" = "false" ]; then + echo "#define ${fn} ${dfn}" + else + echo "RTCD_EXTERN $rtyp (*${fn})($args);" + fi + echo + done +} + +set_function_pointers() { + for fn in $ALL_FUNCS; do + local n="" + local rtyp="$(eval "echo \$${fn}_rtyp")" + local args="$(eval "echo \"\$${fn}_args\"")" + local dfn="$(eval "echo \$${fn}_default")" + dfn=$(eval "echo \$${dfn}") + if $(eval "echo \$${fn}_indirect"); then + echo " $fn = $dfn;" + for opt in "$@"; do + local ofn=$(eval "echo \$${fn}_${opt}") + [ -z "$ofn" ] && continue + [ "$ofn" = "$dfn" ] && continue; + local link=$(eval "echo \$${fn}_${opt}_link") + [ "$link" = "false" ] && continue + local cond="$(eval "echo \$have_${opt}")" + echo " if (${cond}) $fn = $ofn;" + done + fi + echo + done +} + +filter() { + local filtered + for opt in "$@"; do + [ -z $(eval "echo \$disable_${opt}") ] && filtered="$filtered $opt" + done + echo $filtered +} + +# +# Helper functions for generating the arch specific RTCD files +# +common_top() { + local outfile_basename=$(basename ${outfile:-rtcd.h}) + local include_guard=$(echo -n $outfile_basename | tr '[a-z]' '[A-Z]' | tr -c '[A-Z]' _) + cat < $@ +CLEAN-OBJS += $(BUILD_PFX)vpx_rtcd.h + CODEC_DOC_SRCS += vpx/vpx_codec.h \ vpx/vpx_decoder.h \ vpx/vpx_encoder.h \ diff --git a/vp8/common/arm/arm_systemdependent.c b/vp8/common/arm/arm_systemdependent.c index d16ff2bb2..43a0b7745 100644 --- a/vp8/common/arm/arm_systemdependent.c +++ b/vp8/common/arm/arm_systemdependent.c @@ -62,12 +62,6 @@ void vp8_arch_arm_common_init(VP8_COMMON *ctx) rtcd->recon.copy8x8 = vp8_copy_mem8x8_v6; rtcd->recon.copy8x4 = vp8_copy_mem8x4_v6; rtcd->recon.intra4x4_predict = vp8_intra4x4_predict_armv6; - - rtcd->dequant.block = vp8_dequantize_b_v6; - rtcd->dequant.idct_add = vp8_dequant_idct_add_v6; - rtcd->dequant.idct_add_y_block = vp8_dequant_idct_add_y_block_v6; - rtcd->dequant.idct_add_uv_block = vp8_dequant_idct_add_uv_block_v6; - } #endif @@ -102,12 +96,6 @@ void vp8_arch_arm_common_init(VP8_COMMON *ctx) vp8_build_intra_predictors_mby_neon; rtcd->recon.build_intra_predictors_mby_s = vp8_build_intra_predictors_mby_s_neon; - - rtcd->dequant.block = vp8_dequantize_b_neon; - rtcd->dequant.idct_add = vp8_dequant_idct_add_neon; - rtcd->dequant.idct_add_y_block = vp8_dequant_idct_add_y_block_neon; - rtcd->dequant.idct_add_uv_block = vp8_dequant_idct_add_uv_block_neon; - } #endif diff --git a/vp8/common/arm/armv6/idct_blk_v6.c b/vp8/common/arm/armv6/idct_blk_v6.c index 9108929f5..578f7668f 100644 --- a/vp8/common/arm/armv6/idct_blk_v6.c +++ b/vp8/common/arm/armv6/idct_blk_v6.c @@ -10,7 +10,6 @@ #include "vpx_config.h" #include "vp8/common/idct.h" -#include "vp8/common/dequantize.h" void vp8_dequant_idct_add_y_block_v6(short *q, short *dq, diff --git a/vp8/common/arm/dequantize_arm.c b/vp8/common/arm/dequantize_arm.c index 66a5dce26..70e72aa47 100644 --- a/vp8/common/arm/dequantize_arm.c +++ b/vp8/common/arm/dequantize_arm.c @@ -10,8 +10,7 @@ #include "vpx_config.h" -#include "vp8/common/dequantize.h" -#include "vp8/common/idct.h" +#include "vp8/common/blockd.h" #if HAVE_NEON extern void vp8_dequantize_b_loop_neon(short *Q, short *DQC, short *DQ); diff --git a/vp8/common/arm/dequantize_arm.h b/vp8/common/arm/dequantize_arm.h deleted file mode 100644 index e330260bf..000000000 --- a/vp8/common/arm/dequantize_arm.h +++ /dev/null @@ -1,59 +0,0 @@ -/* - * Copyright (c) 2010 The WebM project authors. All Rights Reserved. - * - * Use of this source code is governed by a BSD-style license - * that can be found in the LICENSE file in the root of the source - * tree. An additional intellectual property rights grant can be found - * in the file PATENTS. All contributing project authors may - * be found in the AUTHORS file in the root of the source tree. - */ - - -#ifndef DEQUANTIZE_ARM_H -#define DEQUANTIZE_ARM_H - -#if HAVE_MEDIA -extern prototype_dequant_block(vp8_dequantize_b_v6); -extern prototype_dequant_idct_add(vp8_dequant_idct_add_v6); -extern prototype_dequant_idct_add_y_block(vp8_dequant_idct_add_y_block_v6); -extern prototype_dequant_idct_add_uv_block(vp8_dequant_idct_add_uv_block_v6); - -#if !CONFIG_RUNTIME_CPU_DETECT -#undef vp8_dequant_block -#define vp8_dequant_block vp8_dequantize_b_v6 - -#undef vp8_dequant_idct_add -#define vp8_dequant_idct_add vp8_dequant_idct_add_v6 - -#undef vp8_dequant_idct_add_y_block -#define vp8_dequant_idct_add_y_block vp8_dequant_idct_add_y_block_v6 - -#undef vp8_dequant_idct_add_uv_block -#define vp8_dequant_idct_add_uv_block vp8_dequant_idct_add_uv_block_v6 -#endif -#endif - -#if HAVE_NEON -extern prototype_dequant_block(vp8_dequantize_b_neon); -extern prototype_dequant_idct_add(vp8_dequant_idct_add_neon); -extern prototype_dequant_idct_add_y_block(vp8_dequant_idct_add_y_block_neon); -extern prototype_dequant_idct_add_uv_block(vp8_dequant_idct_add_uv_block_neon); - - -#if !CONFIG_RUNTIME_CPU_DETECT -#undef vp8_dequant_block -#define vp8_dequant_block vp8_dequantize_b_neon - -#undef vp8_dequant_idct_add -#define vp8_dequant_idct_add vp8_dequant_idct_add_neon - -#undef vp8_dequant_idct_add_y_block -#define vp8_dequant_idct_add_y_block vp8_dequant_idct_add_y_block_neon - -#undef vp8_dequant_idct_add_uv_block -#define vp8_dequant_idct_add_uv_block vp8_dequant_idct_add_uv_block_neon -#endif - -#endif - -#endif diff --git a/vp8/common/arm/neon/idct_blk_neon.c b/vp8/common/arm/neon/idct_blk_neon.c index cc55843d5..7424b029c 100644 --- a/vp8/common/arm/neon/idct_blk_neon.c +++ b/vp8/common/arm/neon/idct_blk_neon.c @@ -10,7 +10,6 @@ #include "vpx_config.h" #include "vp8/common/idct.h" -#include "vp8/common/dequantize.h" /* place these declarations here because we don't want to maintain them * outside of this scope diff --git a/vp8/common/blockd.h b/vp8/common/blockd.h index b237206e6..6b58397cd 100644 --- a/vp8/common/blockd.h +++ b/vp8/common/blockd.h @@ -179,7 +179,7 @@ typedef struct } LOWER_RES_INFO; #endif -typedef struct +typedef struct blockd { short *qcoeff; short *dqcoeff; diff --git a/vp8/common/dequantize.c b/vp8/common/dequantize.c index 96245162f..66425da1f 100644 --- a/vp8/common/dequantize.c +++ b/vp8/common/dequantize.c @@ -10,7 +10,8 @@ #include "vpx_config.h" -#include "dequantize.h" +#include "vpx_rtcd.h" +#include "vp8/common/blockd.h" #include "vp8/common/idct.h" #include "vpx_mem/vpx_mem.h" diff --git a/vp8/common/dequantize.h b/vp8/common/dequantize.h deleted file mode 100644 index 429359190..000000000 --- a/vp8/common/dequantize.h +++ /dev/null @@ -1,85 +0,0 @@ -/* - * Copyright (c) 2010 The WebM project authors. All Rights Reserved. - * - * Use of this source code is governed by a BSD-style license - * that can be found in the LICENSE file in the root of the source - * tree. An additional intellectual property rights grant can be found - * in the file PATENTS. All contributing project authors may - * be found in the AUTHORS file in the root of the source tree. - */ - - -#ifndef DEQUANTIZE_H -#define DEQUANTIZE_H -#include "vp8/common/blockd.h" - -#define prototype_dequant_block(sym) \ - void sym(BLOCKD *x, short *DQC) - -#define prototype_dequant_idct_add(sym) \ - void sym(short *input, short *dq, \ - unsigned char *output, \ - int stride) - -#define prototype_dequant_idct_add_y_block(sym) \ - void sym(short *q, short *dq, \ - unsigned char *dst, \ - int stride, char *eobs) - -#define prototype_dequant_idct_add_uv_block(sym) \ - void sym(short *q, short *dq, \ - unsigned char *dst_u, \ - unsigned char *dst_v, int stride, char *eobs) - -#if ARCH_X86 || ARCH_X86_64 -#include "x86/dequantize_x86.h" -#endif - -#if ARCH_ARM -#include "arm/dequantize_arm.h" -#endif - -#ifndef vp8_dequant_block -#define vp8_dequant_block vp8_dequantize_b_c -#endif -extern prototype_dequant_block(vp8_dequant_block); - -#ifndef vp8_dequant_idct_add -#define vp8_dequant_idct_add vp8_dequant_idct_add_c -#endif -extern prototype_dequant_idct_add(vp8_dequant_idct_add); - -#ifndef vp8_dequant_idct_add_y_block -#define vp8_dequant_idct_add_y_block vp8_dequant_idct_add_y_block_c -#endif -extern prototype_dequant_idct_add_y_block(vp8_dequant_idct_add_y_block); - -#ifndef vp8_dequant_idct_add_uv_block -#define vp8_dequant_idct_add_uv_block vp8_dequant_idct_add_uv_block_c -#endif -extern prototype_dequant_idct_add_uv_block(vp8_dequant_idct_add_uv_block); - - -typedef prototype_dequant_block((*vp8_dequant_block_fn_t)); - -typedef prototype_dequant_idct_add((*vp8_dequant_idct_add_fn_t)); - -typedef prototype_dequant_idct_add_y_block((*vp8_dequant_idct_add_y_block_fn_t)); - -typedef prototype_dequant_idct_add_uv_block((*vp8_dequant_idct_add_uv_block_fn_t)); - -typedef struct -{ - vp8_dequant_block_fn_t block; - vp8_dequant_idct_add_fn_t idct_add; - vp8_dequant_idct_add_y_block_fn_t idct_add_y_block; - vp8_dequant_idct_add_uv_block_fn_t idct_add_uv_block; -} vp8_dequant_rtcd_vtable_t; - -#if CONFIG_RUNTIME_CPU_DETECT -#define DEQUANT_INVOKE(ctx,fn) (ctx)->fn -#else -#define DEQUANT_INVOKE(ctx,fn) vp8_dequant_##fn -#endif - -#endif diff --git a/vp8/common/generic/systemdependent.c b/vp8/common/generic/systemdependent.c index 01d76206d..05c54fb60 100644 --- a/vp8/common/generic/systemdependent.c +++ b/vp8/common/generic/systemdependent.c @@ -10,6 +10,7 @@ #include "vpx_config.h" +#include "vpx_rtcd.h" #include "vp8/common/subpixel.h" #include "vp8/common/loopfilter.h" #include "vp8/common/recon.h" @@ -70,13 +71,6 @@ void vp8_machine_specific_config(VP8_COMMON *ctx) VP8_COMMON_RTCD *rtcd = &ctx->rtcd; - rtcd->dequant.block = vp8_dequantize_b_c; - rtcd->dequant.idct_add = vp8_dequant_idct_add_c; - rtcd->dequant.idct_add_y_block = vp8_dequant_idct_add_y_block_c; - rtcd->dequant.idct_add_uv_block = - vp8_dequant_idct_add_uv_block_c; - - rtcd->idct.idct16 = vp8_short_idct4x4llm_c; rtcd->idct.idct1_scalar_add = vp8_dc_only_idct_add_c; rtcd->idct.iwalsh1 = vp8_short_inv_walsh4x4_1_c; @@ -138,4 +132,6 @@ void vp8_machine_specific_config(VP8_COMMON *ctx) #if CONFIG_MULTITHREAD ctx->processor_core_count = get_cpu_count(); #endif /* CONFIG_MULTITHREAD */ + + vpx_rtcd(); } diff --git a/vp8/common/idct_blk.c b/vp8/common/idct_blk.c index 249fad4ea..b9c5d3212 100644 --- a/vp8/common/idct_blk.c +++ b/vp8/common/idct_blk.c @@ -10,7 +10,6 @@ #include "vpx_config.h" #include "vp8/common/idct.h" -#include "dequantize.h" void vp8_dequant_idct_add_c(short *input, short *dq, unsigned char *dest, int stride); diff --git a/vp8/common/invtrans.h b/vp8/common/invtrans.h index f49e2e577..c67132ca5 100644 --- a/vp8/common/invtrans.h +++ b/vp8/common/invtrans.h @@ -13,6 +13,7 @@ #define __INC_INVTRANS_H #include "vpx_config.h" +#include "vpx_rtcd.h" #include "idct.h" #include "blockd.h" #include "onyxc_int.h" @@ -55,7 +56,7 @@ static void vp8_inverse_transform_mby(MACROBLOCKD *xd, DQC = xd->dequant_y1_dc; } - DEQUANT_INVOKE (&rtcd->dequant, idct_add_y_block) + vp8_dequant_idct_add_y_block (xd->qcoeff, DQC, xd->dst.y_buffer, xd->dst.y_stride, xd->eobs); diff --git a/vp8/common/onyxc_int.h b/vp8/common/onyxc_int.h index f91383de8..63022f5e4 100644 --- a/vp8/common/onyxc_int.h +++ b/vp8/common/onyxc_int.h @@ -22,7 +22,6 @@ #if CONFIG_POSTPROC #include "postproc.h" #endif -#include "dequantize.h" /*#ifdef PACKET_TESTING*/ #include "header.h" @@ -74,7 +73,6 @@ typedef enum typedef struct VP8_COMMON_RTCD { #if CONFIG_RUNTIME_CPU_DETECT - vp8_dequant_rtcd_vtable_t dequant; vp8_idct_rtcd_vtable_t idct; vp8_recon_rtcd_vtable_t recon; vp8_subpix_rtcd_vtable_t subpix; diff --git a/vp8/decoder/x86/x86_dsystemdependent.c b/vp8/common/rtcd.c similarity index 67% rename from vp8/decoder/x86/x86_dsystemdependent.c rename to vp8/common/rtcd.c index 27bf5ddbd..232640dc8 100644 --- a/vp8/decoder/x86/x86_dsystemdependent.c +++ b/vp8/common/rtcd.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2010 The WebM project authors. All Rights Reserved. + * Copyright (c) 2011 The WebM project authors. All Rights Reserved. * * Use of this source code is governed by a BSD-style license * that can be found in the LICENSE file in the root of the source @@ -7,13 +7,6 @@ * in the file PATENTS. All contributing project authors may * be found in the AUTHORS file in the root of the source tree. */ - - #include "vpx_config.h" -#include "vpx_ports/x86.h" -#include "vp8/decoder/onyxd_int.h" - -void vp8_arch_x86_decode_init(VP8D_COMP *pbi) -{ - -} +#define RTCD_C +#include "vpx_rtcd.h" diff --git a/vp8/common/rtcd_defs.sh b/vp8/common/rtcd_defs.sh new file mode 100644 index 000000000..0fb40f731 --- /dev/null +++ b/vp8/common/rtcd_defs.sh @@ -0,0 +1,22 @@ +common_forward_decls() { +cat <dequant.block = vp8_dequantize_b_mmx; - rtcd->dequant.idct_add = vp8_dequant_idct_add_mmx; - rtcd->dequant.idct_add_y_block = vp8_dequant_idct_add_y_block_mmx; - rtcd->dequant.idct_add_uv_block = vp8_dequant_idct_add_uv_block_mmx; - rtcd->idct.idct16 = vp8_short_idct4x4llm_mmx; rtcd->idct.idct1_scalar_add = vp8_dc_only_idct_add_mmx; rtcd->idct.iwalsh16 = vp8_short_inv_walsh4x4_mmx; @@ -90,9 +85,6 @@ void vp8_arch_x86_common_init(VP8_COMMON *ctx) rtcd->recon.build_intra_predictors_mby_s = vp8_build_intra_predictors_mby_s_sse2; - rtcd->dequant.idct_add_y_block = vp8_dequant_idct_add_y_block_sse2; - rtcd->dequant.idct_add_uv_block = vp8_dequant_idct_add_uv_block_sse2; - rtcd->idct.iwalsh16 = vp8_short_inv_walsh4x4_sse2; rtcd->subpix.sixtap16x16 = vp8_sixtap_predict16x16_sse2; diff --git a/vp8/decoder/arm/arm_dsystemdependent.c b/vp8/decoder/arm/arm_dsystemdependent.c deleted file mode 100644 index aeecacb40..000000000 --- a/vp8/decoder/arm/arm_dsystemdependent.c +++ /dev/null @@ -1,39 +0,0 @@ -/* - * Copyright (c) 2010 The WebM project authors. All Rights Reserved. - * - * Use of this source code is governed by a BSD-style license - * that can be found in the LICENSE file in the root of the source - * tree. An additional intellectual property rights grant can be found - * in the file PATENTS. All contributing project authors may - * be found in the AUTHORS file in the root of the source tree. - */ - - -#include "vpx_config.h" -#include "vpx_ports/arm.h" -#include "vp8/decoder/onyxd_int.h" - -void vp8_arch_arm_decode_init(VP8D_COMP *pbi) -{ -#if CONFIG_RUNTIME_CPU_DETECT - int flags = pbi->common.rtcd.flags; - -#if HAVE_EDSP - if (flags & HAS_EDSP) - { - } -#endif - -#if HAVE_MEDIA - if (flags & HAS_MEDIA) - { - } -#endif - -#if HAVE_NEON - if (flags & HAS_NEON) - { - } -#endif -#endif -} diff --git a/vp8/decoder/decodframe.c b/vp8/decoder/decodframe.c index 917aeceb6..0de9c4314 100644 --- a/vp8/decoder/decodframe.c +++ b/vp8/decoder/decodframe.c @@ -9,13 +9,14 @@ */ +#include "vpx_config.h" +#include "vpx_rtcd.h" #include "onyxd_int.h" #include "vp8/common/header.h" #include "vp8/common/reconintra.h" #include "vp8/common/reconintra4x4.h" #include "vp8/common/recon.h" #include "vp8/common/reconinter.h" -#include "vp8/common/dequantize.h" #include "detokenize.h" #include "vp8/common/invtrans.h" #include "vp8/common/alloccommon.h" @@ -32,7 +33,6 @@ #endif #include "vpx_mem/vpx_mem.h" #include "vp8/common/idct.h" - #include "vp8/common/threading.h" #include "decoderthreading.h" #include "dboolhuff.h" @@ -194,7 +194,7 @@ static void decode_macroblock(VP8D_COMP *pbi, MACROBLOCKD *xd, { if (xd->eobs[i] > 1) { - DEQUANT_INVOKE(&pbi->common.rtcd.dequant, idct_add) + vp8_dequant_idct_add (b->qcoeff, DQC, *(b->base_dst) + b->dst, b->dst_stride); } @@ -237,8 +237,7 @@ static void decode_macroblock(VP8D_COMP *pbi, MACROBLOCKD *xd, /* do 2nd order transform on the dc block */ if (xd->eobs[24] > 1) { - DEQUANT_INVOKE(&pbi->common.rtcd.dequant, block)(b, - xd->dequant_y2); + vp8_dequantize_b(b, xd->dequant_y2); IDCT_INVOKE(RTCD_VTABLE(idct), iwalsh16)(&b->dqcoeff[0], xd->qcoeff); @@ -265,13 +264,13 @@ static void decode_macroblock(VP8D_COMP *pbi, MACROBLOCKD *xd, DQC = xd->dequant_y1_dc; } - DEQUANT_INVOKE (&pbi->common.rtcd.dequant, idct_add_y_block) + vp8_dequant_idct_add_y_block (xd->qcoeff, DQC, xd->dst.y_buffer, xd->dst.y_stride, xd->eobs); } - DEQUANT_INVOKE (&pbi->common.rtcd.dequant, idct_add_uv_block) + vp8_dequant_idct_add_uv_block (xd->qcoeff+16*16, xd->dequant_uv, xd->dst.u_buffer, xd->dst.v_buffer, xd->dst.uv_stride, xd->eobs+16); diff --git a/vp8/decoder/generic/dsystemdependent.c b/vp8/decoder/generic/dsystemdependent.c deleted file mode 100644 index 8a84e566a..000000000 --- a/vp8/decoder/generic/dsystemdependent.c +++ /dev/null @@ -1,33 +0,0 @@ -/* - * Copyright (c) 2010 The WebM project authors. All Rights Reserved. - * - * Use of this source code is governed by a BSD-style license - * that can be found in the LICENSE file in the root of the source - * tree. An additional intellectual property rights grant can be found - * in the file PATENTS. All contributing project authors may - * be found in the AUTHORS file in the root of the source tree. - */ - - -#include "vpx_config.h" -#include "vp8/common/dequantize.h" -#include "vp8/decoder/onyxd_int.h" - -extern void vp8_arch_x86_decode_init(VP8D_COMP *pbi); -extern void vp8_arch_arm_decode_init(VP8D_COMP *pbi); - -void vp8_dmachine_specific_config(VP8D_COMP *pbi) -{ - /* Pure C: */ -#if CONFIG_RUNTIME_CPU_DETECT - pbi->mb.rtcd = &pbi->common.rtcd; -#endif - -#if ARCH_X86 || ARCH_X86_64 - vp8_arch_x86_decode_init(pbi); -#endif - -#if ARCH_ARM - vp8_arch_arm_decode_init(pbi); -#endif -} diff --git a/vp8/decoder/onyxd_if.c b/vp8/decoder/onyxd_if.c index 13be34f94..6a20e5146 100644 --- a/vp8/decoder/onyxd_if.c +++ b/vp8/decoder/onyxd_if.c @@ -76,7 +76,9 @@ struct VP8D_COMP * vp8dx_create_decompressor(VP8D_CONFIG *oxcf) vp8dx_initialize(); vp8_create_common(&pbi->common); - vp8_dmachine_specific_config(pbi); +#if CONFIG_RUNTIME_CPU_DETECT + pbi->mb.rtcd = &pbi->common.rtcd; +#endif pbi->common.current_video_frame = 0; pbi->ready_for_new_data = 1; diff --git a/vp8/decoder/onyxd_int.h b/vp8/decoder/onyxd_int.h index cb2593b2c..989f68bf8 100644 --- a/vp8/decoder/onyxd_int.h +++ b/vp8/decoder/onyxd_int.h @@ -17,7 +17,6 @@ #include "vp8/common/onyxc_int.h" #include "vp8/common/threading.h" - #if CONFIG_ERROR_CONCEALMENT #include "ec_types.h" #endif @@ -114,8 +113,6 @@ typedef struct VP8D_COMP } VP8D_COMP; int vp8_decode_frame(VP8D_COMP *cpi); -void vp8_dmachine_specific_config(VP8D_COMP *pbi); - #if CONFIG_DEBUG #define CHECK_MEM_ERROR(lval,expr) do {\ diff --git a/vp8/decoder/threading.c b/vp8/decoder/threading.c index 2ce00f705..23c5da4e0 100644 --- a/vp8/decoder/threading.c +++ b/vp8/decoder/threading.c @@ -9,6 +9,8 @@ */ +#include "vpx_config.h" +#include "vpx_rtcd.h" #if !defined(WIN32) && CONFIG_OS_SUPPORT == 1 # include #endif @@ -191,7 +193,7 @@ static void decode_macroblock(VP8D_COMP *pbi, MACROBLOCKD *xd, int mb_row, int m { if (xd->eobs[i] > 1) { - DEQUANT_INVOKE(&pbi->common.rtcd.dequant, idct_add) + vp8_dequant_idct_add (b->qcoeff, DQC, *(b->base_dst) + b->dst, b->dst_stride); } @@ -217,7 +219,7 @@ static void decode_macroblock(VP8D_COMP *pbi, MACROBLOCKD *xd, int mb_row, int m /* do 2nd order transform on the dc block */ if (xd->eobs[24] > 1) { - DEQUANT_INVOKE(&pbi->common.rtcd.dequant, block)(b, xd->dequant_y2); + vp8_dequantize_b(b, xd->dequant_y2); IDCT_INVOKE(RTCD_VTABLE(idct), iwalsh16)(&b->dqcoeff[0], xd->qcoeff); @@ -241,13 +243,13 @@ static void decode_macroblock(VP8D_COMP *pbi, MACROBLOCKD *xd, int mb_row, int m DQC = xd->dequant_y1_dc; } - DEQUANT_INVOKE (&pbi->common.rtcd.dequant, idct_add_y_block) + vp8_dequant_idct_add_y_block (xd->qcoeff, DQC, xd->dst.y_buffer, xd->dst.y_stride, xd->eobs); } - DEQUANT_INVOKE (&pbi->common.rtcd.dequant, idct_add_uv_block) + vp8_dequant_idct_add_uv_block (xd->qcoeff+16*16, xd->dequant_uv, xd->dst.u_buffer, xd->dst.v_buffer, xd->dst.uv_stride, xd->eobs+16); diff --git a/vp8/encoder/encodeframe.c b/vp8/encoder/encodeframe.c index b5c5c7445..da2472325 100644 --- a/vp8/encoder/encodeframe.c +++ b/vp8/encoder/encodeframe.c @@ -1119,7 +1119,7 @@ int vp8cx_encode_intra_macro_block(VP8_COMP *cpi, MACROBLOCK *x, TOKENEXTRA **t, if (xd->mode_info_context->mbmi.mode != B_PRED) vp8_inverse_transform_mby(xd, IF_RTCD(&cpi->common.rtcd)); - DEQUANT_INVOKE (&cpi->common.rtcd.dequant, idct_add_uv_block) + vp8_dequant_idct_add_uv_block (xd->qcoeff+16*16, xd->dequant_uv, xd->dst.u_buffer, xd->dst.v_buffer, xd->dst.uv_stride, xd->eobs+16); @@ -1304,7 +1304,7 @@ int vp8cx_encode_inter_macroblock if (xd->mode_info_context->mbmi.mode != B_PRED) vp8_inverse_transform_mby(xd, IF_RTCD(&cpi->common.rtcd)); - DEQUANT_INVOKE (&cpi->common.rtcd.dequant, idct_add_uv_block) + vp8_dequant_idct_add_uv_block (xd->qcoeff+16*16, xd->dequant_uv, xd->dst.u_buffer, xd->dst.v_buffer, xd->dst.uv_stride, xd->eobs+16); diff --git a/vp8/encoder/onyx_if.c b/vp8/encoder/onyx_if.c index 9dcf71dfc..95f7cfe29 100644 --- a/vp8/encoder/onyx_if.c +++ b/vp8/encoder/onyx_if.c @@ -55,7 +55,6 @@ extern void vp8cx_pick_filter_level_fast(YV12_BUFFER_CONFIG *sd, VP8_COMP *cpi); extern void vp8cx_set_alt_lf_level(VP8_COMP *cpi, int filt_val); extern void vp8cx_pick_filter_level(YV12_BUFFER_CONFIG *sd, VP8_COMP *cpi); -extern void vp8_dmachine_specific_config(VP8_COMP *cpi); extern void vp8_cmachine_specific_config(VP8_COMP *cpi); extern void vp8_deblock_frame(YV12_BUFFER_CONFIG *source, YV12_BUFFER_CONFIG *post, int filt_lvl, int low_var_thresh, int flag); extern void print_parms(VP8_CONFIG *ocf, char *filenam); @@ -230,7 +229,6 @@ void vp8_initialize() { vp8_scale_machine_specific_config(); vp8_initialize_common(); - //vp8_dmachine_specific_config(); vp8_tokenize_initialize(); init_done = 1; diff --git a/vp8/vp8_common.mk b/vp8/vp8_common.mk index e32fc6917..b0f45f2b0 100644 --- a/vp8/vp8_common.mk +++ b/vp8/vp8_common.mk @@ -20,7 +20,6 @@ VP8_COMMON_SRCS-yes += common/coefupdateprobs.h VP8_COMMON_SRCS-yes += common/debugmodes.c VP8_COMMON_SRCS-yes += common/default_coef_probs.h VP8_COMMON_SRCS-yes += common/dequantize.c -VP8_COMMON_SRCS-yes += common/dequantize.h VP8_COMMON_SRCS-yes += common/entropy.c VP8_COMMON_SRCS-yes += common/entropymode.c VP8_COMMON_SRCS-yes += common/entropymv.c @@ -51,6 +50,8 @@ VP8_COMMON_SRCS-yes += common/recon.h VP8_COMMON_SRCS-yes += common/reconinter.h VP8_COMMON_SRCS-yes += common/reconintra.h VP8_COMMON_SRCS-yes += common/reconintra4x4.h +VP8_COMMON_SRCS-yes += common/rtcd.c +VP8_COMMON_SRCS-yes += common/rtcd_defs.sh VP8_COMMON_SRCS-yes += common/setupintrarecon.h VP8_COMMON_SRCS-yes += common/subpixel.h VP8_COMMON_SRCS-yes += common/swapyv12buffer.h @@ -74,7 +75,6 @@ VP8_COMMON_SRCS-yes += common/swapyv12buffer.c VP8_COMMON_SRCS-$(CONFIG_POSTPROC_VISUALIZER) += common/textblit.c VP8_COMMON_SRCS-yes += common/treecoder.c -VP8_COMMON_SRCS-$(ARCH_X86)$(ARCH_X86_64) += common/x86/dequantize_x86.h VP8_COMMON_SRCS-$(ARCH_X86)$(ARCH_X86_64) += common/x86/filter_x86.c VP8_COMMON_SRCS-$(ARCH_X86)$(ARCH_X86_64) += common/x86/filter_x86.h VP8_COMMON_SRCS-$(ARCH_X86)$(ARCH_X86_64) += common/x86/idct_x86.h @@ -120,7 +120,6 @@ VP8_COMMON_SRCS-$(ARCH_ARM) += common/arm/recon_arm.h VP8_COMMON_SRCS-$(ARCH_ARM) += common/arm/reconintra_arm.c VP8_COMMON_SRCS-$(ARCH_ARM) += common/arm/subpixel_arm.h VP8_COMMON_SRCS-$(ARCH_ARM) += common/arm/dequantize_arm.c -VP8_COMMON_SRCS-$(ARCH_ARM) += common/arm/dequantize_arm.h # common (media) VP8_COMMON_SRCS-$(HAVE_MEDIA) += common/arm/bilinearfilter_arm.c diff --git a/vp8/vp8dx.mk b/vp8/vp8dx.mk index d6dc15305..ce307b6d0 100644 --- a/vp8/vp8dx.mk +++ b/vp8/vp8dx.mk @@ -18,10 +18,6 @@ VP8_DX_SRCS-no += $(VP8_COMMON_SRCS-no) VP8_DX_SRCS_REMOVE-yes += $(VP8_COMMON_SRCS_REMOVE-yes) VP8_DX_SRCS_REMOVE-no += $(VP8_COMMON_SRCS_REMOVE-no) -ifeq ($(ARCH_ARM),yes) - include $(SRC_PATH_BARE)/$(VP8_PREFIX)vp8dx_arm.mk -endif - VP8_DX_SRCS-yes += vp8_dx_iface.c # common @@ -56,7 +52,6 @@ VP8_DX_SRCS-yes += decoder/detokenize.c VP8_DX_SRCS-$(CONFIG_ERROR_CONCEALMENT) += decoder/ec_types.h VP8_DX_SRCS-$(CONFIG_ERROR_CONCEALMENT) += decoder/error_concealment.h VP8_DX_SRCS-$(CONFIG_ERROR_CONCEALMENT) += decoder/error_concealment.c -VP8_DX_SRCS-yes += decoder/generic/dsystemdependent.c VP8_DX_SRCS-yes += decoder/dboolhuff.h VP8_DX_SRCS-yes += decoder/decodemv.h VP8_DX_SRCS-yes += decoder/decoderthreading.h @@ -69,5 +64,3 @@ VP8_DX_SRCS-$(CONFIG_MULTITHREAD) += decoder/reconintra_mt.h VP8_DX_SRCS-$(CONFIG_MULTITHREAD) += decoder/reconintra_mt.c VP8_DX_SRCS-yes := $(filter-out $(VP8_DX_SRCS_REMOVE-yes),$(VP8_DX_SRCS-yes)) - -VP8_DX_SRCS-$(ARCH_X86)$(ARCH_X86_64) += decoder/x86/x86_dsystemdependent.c diff --git a/vp8/vp8dx_arm.mk b/vp8/vp8dx_arm.mk deleted file mode 100644 index fa1aaea0b..000000000 --- a/vp8/vp8dx_arm.mk +++ /dev/null @@ -1,14 +0,0 @@ -## -## Copyright (c) 2010 The WebM project authors. All Rights Reserved. -## -## Use of this source code is governed by a BSD-style license -## that can be found in the LICENSE file in the root of the source -## tree. An additional intellectual property rights grant can be found -## in the file PATENTS. All contributing project authors may -## be found in the AUTHORS file in the root of the source tree. -## - - -#VP8_DX_SRCS list is modified according to different platforms. - -VP8_DX_SRCS-$(ARCH_ARM) += decoder/arm/arm_dsystemdependent.c -- 2.40.0