Signed-off-by: Paul B Mahol <one...@gmail.com>
---
 libavcodec/dxv.c | 1032 ++++++++++++++++++++++++++++++++++++++++++++++++++++--
 1 file changed, 1003 insertions(+), 29 deletions(-)

diff --git a/libavcodec/dxv.c b/libavcodec/dxv.c
index 529e211258..cd7b693663 100644
--- a/libavcodec/dxv.c
+++ b/libavcodec/dxv.c
@@ -1,6 +1,7 @@
 /*
  * Resolume DXV decoder
  * Copyright (C) 2015 Vittorio Giovara <vittorio.giov...@gmail.com>
+ * Copyright (C) 2018 Paul B Mahol
  *
  * This file is part of FFmpeg.
  *
@@ -23,6 +24,7 @@
 
 #include "libavutil/imgutils.h"
 
+#include "mathops.h"
 #include "avcodec.h"
 #include "bytestream.h"
 #include "internal.h"
@@ -34,53 +36,250 @@ typedef struct DXVContext {
     TextureDSPContext texdsp;
     GetByteContext gbc;
 
-    uint8_t *tex_data;  // Compressed texture
-    int tex_rat;        // Compression ratio
-    int tex_step;       // Distance between blocks
-    int64_t tex_size;   // Texture size
+    uint8_t *tex_data;   // Compressed texture
+    uint8_t *ctex_data;  // Compressed texture
+    int tex_rat;         // Compression ratio
+    int tex_step;        // Distance between blocks
+    int ctex_step;       // Distance between blocks
+    int64_t tex_size;    // Texture size
+    int64_t ctex_size;   // Texture size
 
     /* Optimal number of slices for parallel decoding */
     int slice_count;
 
+    uint8_t *op_data[4]; // Opcodes
+    int64_t op_size[4];  // Opcodes size
+
+    int texture_block_w;
+    int texture_block_h;
+
+    int ctexture_block_w;
+    int ctexture_block_h;
+
     /* Pointer to the selected decompression function */
     int (*tex_funct)(uint8_t *dst, ptrdiff_t stride, const uint8_t *block);
+    int (*ctex_funct)(uint8_t *dst, ptrdiff_t stride, const uint8_t *block);
 } DXVContext;
 
+static void decompress_indices(uint8_t *dst, const uint8_t *src)
+{
+    int block, i;
+
+    for (block = 0; block < 2; block++) {
+        int tmp = AV_RL24(src);
+
+        /* Unpack 8x3 bit from last 3 byte block */
+        for (i = 0; i < 8; i++)
+            dst[i] = (tmp >> (i * 3)) & 0x7;
+
+        src += 3;
+        dst += 8;
+    }
+}
+
+static int extract_component(int yo0, int yo1, int code)
+{
+    int yo;
+
+    if (yo0 == yo1) {
+        yo = yo0;
+    } else if (code == 0) {
+        yo = yo0;
+    } else if (code == 1) {
+        yo = yo1;
+    } else {
+        if (yo0 > yo1) {
+            yo = (uint8_t) (((8 - code) * yo0 +
+                             (code - 1) * yo1) / 7);
+        } else {
+            if (code == 6) {
+                yo = 0;
+            } else if (code == 7) {
+                yo = 255;
+            } else {
+                yo = (uint8_t) (((6 - code) * yo0 +
+                                 (code - 1) * yo1) / 5);
+            }
+        }
+    }
+
+    return yo;
+}
+
+static av_always_inline uint32_t yacocg2rgba(int yo, int co, int cg, int a)
+{
+    int r, g, b;
+
+    co = co - 127;
+    cg = cg - 127;
+
+    r = av_clip_uint8(yo + co - cg);
+    g = av_clip_uint8(yo + cg);
+    b = av_clip_uint8(yo - co - cg);
+
+    return (a << 24) | (b << 16) | (g << 8) | (r);
+}
+
+static int cocg_block(uint8_t *dst, ptrdiff_t stride,
+                      const uint8_t *block)
+{
+    uint8_t co_indices[16];
+    uint8_t cg_indices[16];
+    uint8_t co0 = *(block);
+    uint8_t co1 = *(block + 1);
+    uint8_t cg0 = *(block + 8);
+    uint8_t cg1 = *(block + 9);
+    int x, y;
+
+    decompress_indices(co_indices, block + 2);
+    decompress_indices(cg_indices, block + 10);
+
+    for (y = 0; y < 4; y++) {
+        for (x = 0; x < 4; x++) {
+            int co_code = co_indices[x + y * 4];
+            int cg_code = cg_indices[x + y * 4];
+            uint8_t co, cg;
+
+            co = extract_component(co0, co1, co_code);
+            cg = extract_component(cg0, cg1, cg_code);
+
+            dst[x * 8 + 1] = co;
+            dst[x * 8 + 2] = cg;
+            dst[x * 8 + 5] = co;
+            dst[x * 8 + 6] = cg;
+            dst[x * 8 + stride + 1] = co;
+            dst[x * 8 + stride + 2] = cg;
+            dst[x * 8 + stride + 5] = co;
+            dst[x * 8 + stride + 6] = cg;
+        }
+        dst += 2 * stride;
+    }
+
+    return 16;
+}
+
+static void yo_subblock(uint8_t *dst, uint8_t *yo_indices,
+                        ptrdiff_t stride, const uint8_t *block)
+{
+    uint8_t yo0 = *(block);
+    uint8_t yo1 = *(block + 1);
+    int x, y;
+
+    decompress_indices(yo_indices, block + 2);
+
+    for (y = 0; y < 4; y++) {
+        for (x = 0; x < 4; x++) {
+            int yo_code = yo_indices[x + y * 4];
+            uint8_t yo;
+
+            yo = extract_component(yo0, yo1, yo_code);
+
+            AV_WL32(dst + x * 4, 255u << 24 | yo);
+        }
+        dst += stride;
+    }
+}
+
+static int yo_block(uint8_t *dst, ptrdiff_t stride,
+                    const uint8_t *block)
+{
+    uint8_t yo_indices[16];
+
+    yo_subblock(dst,      yo_indices, stride, block);
+    yo_subblock(dst + 16, yo_indices, stride, block + 8);
+    yo_subblock(dst + 32, yo_indices, stride, block + 16);
+    yo_subblock(dst + 48, yo_indices, stride, block + 24);
+
+    return 32;
+}
+
+static void a_subblock(uint8_t *dst, uint8_t *a_indices,
+                       ptrdiff_t stride, const uint8_t *block)
+{
+    uint8_t a0 = *(block);
+    uint8_t a1 = *(block + 1);
+    int x, y;
+
+    decompress_indices(a_indices, block + 2);
+
+    for (y = 0; y < 4; y++) {
+        for (x = 0; x < 4; x++) {
+            int a_code = a_indices[x + y * 4];
+            uint8_t a;
+
+            a = extract_component(a0, a1, a_code);
+            dst[x * 4 + 3] = a;
+        }
+        dst += stride;
+    }
+}
+
+static int yao_block(uint8_t *dst, ptrdiff_t stride,
+                    const uint8_t *block)
+{
+    uint8_t yo_indices[16];
+    uint8_t a_indices[16];
+
+    yo_subblock(dst,      yo_indices, stride, block);
+    a_subblock(dst,       a_indices,  stride, block + 8);
+    yo_subblock(dst + 16, yo_indices, stride, block + 16);
+    a_subblock(dst  + 16, a_indices,  stride, block + 24);
+    yo_subblock(dst + 32, yo_indices, stride, block + 32);
+    a_subblock(dst  + 32, a_indices,  stride, block + 40);
+    yo_subblock(dst + 48, yo_indices, stride, block + 48);
+    a_subblock(dst  + 48, a_indices,  stride, block + 56);
+
+    return 64;
+}
+
 static int decompress_texture_thread(AVCodecContext *avctx, void *arg,
                                      int slice, int thread_nb)
 {
     DXVContext *ctx = avctx->priv_data;
     AVFrame *frame = arg;
     const uint8_t *d = ctx->tex_data;
-    int w_block = avctx->coded_width / TEXTURE_BLOCK_W;
-    int h_block = avctx->coded_height / TEXTURE_BLOCK_H;
+    const uint8_t *c = ctx->ctex_data;
+    int w_block = avctx->coded_width / ctx->texture_block_w;
+    int h_block = avctx->coded_height / ctx->texture_block_h;
     int x, y;
     int start_slice, end_slice;
-    int base_blocks_per_slice = h_block / ctx->slice_count;
-    int remainder_blocks = h_block % ctx->slice_count;
-
-    /* When the frame height (in blocks) doesn't divide evenly between the
-     * number of slices, spread the remaining blocks evenly between the first
-     * operations */
-    start_slice = slice * base_blocks_per_slice;
-    /* Add any extra blocks (one per slice) that have been added
-     * before this slice */
-    start_slice += FFMIN(slice, remainder_blocks);
-
-    end_slice = start_slice + base_blocks_per_slice;
-    /* Add an extra block if there are remainder blocks to be accounted for */
-    if (slice < remainder_blocks)
-        end_slice++;
+
+    start_slice = h_block * slice / ctx->slice_count;
+    end_slice = h_block * (slice + 1) / ctx->slice_count;
 
     for (y = start_slice; y < end_slice; y++) {
-        uint8_t *p = frame->data[0] + y * frame->linesize[0] * TEXTURE_BLOCK_H;
+        uint8_t *p = frame->data[0] + y * frame->linesize[0] * 
ctx->texture_block_h;
         int off  = y * w_block;
         for (x = 0; x < w_block; x++) {
-            ctx->tex_funct(p + x * 16, frame->linesize[0],
+            ctx->tex_funct(p + x * 4 * ctx->texture_block_w, 
frame->linesize[0],
                            d + (off + x) * ctx->tex_step);
         }
     }
 
+    if (ctx->ctex_funct) {
+        w_block = avctx->coded_width / ctx->ctexture_block_w;
+        h_block = avctx->coded_height / ctx->ctexture_block_h;
+
+        start_slice = h_block * slice / ctx->slice_count;
+        end_slice = h_block * (slice + 1) / ctx->slice_count;
+
+        for (y = start_slice; y < end_slice; y++) {
+            uint8_t *p = frame->data[0] + y * frame->linesize[0] * 
ctx->ctexture_block_h;
+            int off  = y * w_block;
+            for (x = 0; x < w_block; x++) {
+                ctx->ctex_funct(p + x * 4 * ctx->ctexture_block_w, 
frame->linesize[0],
+                                c + (off + x) * ctx->ctex_step);
+            }
+        }
+        for (y = start_slice * ctx->ctexture_block_h; y < FFMIN(end_slice * 
ctx->ctexture_block_h, avctx->height); y++) {
+            uint8_t *p8 = frame->data[0] + y * frame->linesize[0];
+            uint32_t *p = (uint32_t *)p8;
+            for (x = 0; x < avctx->width; x++) {
+                p[x] = yacocg2rgba(p8[4 * x], p8[4 * x + 1], p8[4 * x + 2], 
p8[4 * x + 3]);
+            }
+        }
+    }
+
     return 0;
 }
 
@@ -169,6 +368,728 @@ static int dxv_decompress_dxt1(AVCodecContext *avctx)
     return 0;
 }
 
+typedef struct OpcodeTable {
+    int16_t next;
+    uint8_t val1;
+    uint8_t val2;
+} OpcodeTable;
+
+static int fill_ltable(GetByteContext *gb, uint32_t *table, int *nb_elements)
+{
+    unsigned half = 512, bits = 1023, left = 1024, input, mask;
+    int value, counter = 0, rshift = 10, lshift = 30;
+
+    mask = bytestream2_get_le32(gb) >> 2;
+    while (left) {
+        if (bytestream2_get_bytes_left(gb) < 0 || counter >= 256)
+            return AVERROR_INVALIDDATA;
+
+        value = bits & mask;
+        left -= bits & mask;
+        mask >>= rshift;
+        lshift -= rshift;
+        table[counter++] = value;
+        if (lshift < 16) {
+            input = bytestream2_get_le16(gb);
+            mask += input << lshift;
+            lshift += 16;
+        }
+        if (left < half) {
+            half >>= 1;
+            bits >>= 1;
+            rshift--;
+        }
+    }
+
+    for (; !table[counter - 1]; counter--)
+        if (counter <= 0)
+            return AVERROR_INVALIDDATA;
+
+    *nb_elements = counter;
+
+    if (counter < 256)
+        memset(&table[counter], 0, 4 * (256 - counter));
+
+    if (lshift >= 16)
+        bytestream2_seek(gb, -2, SEEK_CUR);
+
+    return 0;
+}
+
+static int fill_optable(unsigned *table0, OpcodeTable *table1, int nb_elements)
+{
+    unsigned table2[256] = { 0 };
+    unsigned x = 0;
+    int val0, val1, i, j = 2, k = 0;
+
+    table2[0] = table0[0];
+    for (i = 0; i < nb_elements - 1; i++, table2[i] = val0) {
+        val0 = table0[i + 1] + table2[i];
+    }
+
+    if (!table2[0]) {
+        do {
+            k++;
+        } while (!table2[k]);
+    }
+
+    j = 2;
+    for (i = 1024; i > 0; i--) {
+        for (table1[x].val1 = k; k < 256 && j > table2[k]; k++)
+            ;
+        x = (x - 383) & 0x3FF;
+        j++;
+    }
+
+    if (nb_elements > 0)
+        memcpy(&table2[0], table0, 4 * nb_elements);
+
+    for (i = 0; i < 1024; i++) {
+        val0 = table1[i].val1;
+        val1 = table2[val0];
+        table2[val0]++;
+        x = 31 - ff_clz(val1);
+        if (x > 10)
+            return AVERROR_INVALIDDATA;
+        table1[i].val2 = 10 - x;
+        table1[i].next = (val1 << table1[i].val2) - 1024;
+    }
+
+    return 0;
+}
+
+static int get_opcodes(GetByteContext *gb, uint32_t *table, uint8_t *dst, int 
op_size, int nb_elements)
+{
+    OpcodeTable optable[1024];
+    int sum, x, val, lshift, rshift, ret, size_in_bits, i, idx;
+    unsigned endoffset, newoffset, offset;
+    unsigned next;
+    uint8_t *src = (uint8_t *)gb->buffer;
+
+    ret = fill_optable(table, optable, nb_elements);
+    if (ret < 0)
+        return ret;
+
+    size_in_bits = bytestream2_get_le32(gb);
+    endoffset = ((size_in_bits + 7) >> 3) - 4;
+    if (endoffset <= 0 || bytestream2_get_bytes_left(gb) < endoffset)
+        return AVERROR_INVALIDDATA;
+
+    offset = endoffset;
+    next = AV_RL32(src + endoffset);
+    rshift = (((size_in_bits & 0xFF) - 1) & 7) + 15;
+    lshift = 32 - rshift;
+    idx = (next >> rshift) & 0x3FF;
+    for (i = 0; i < op_size; i++) {
+        dst[i] = optable[idx].val1;
+        val = optable[idx].val2;
+        sum = val + lshift;
+        x = (next << lshift) >> 1 >> (31 - val);
+        newoffset = offset - (sum >> 3);
+        lshift = sum & 7;
+        idx = x + optable[idx].next;
+        offset = newoffset;
+        if (offset > endoffset)
+            return AVERROR_INVALIDDATA;
+        next = AV_RL32(src + offset);
+    }
+
+    bytestream2_skip(gb, (size_in_bits + 7 >> 3) - 4);
+
+    return 0;
+}
+
+static int dxv_decompress_opcodes(GetByteContext *gb, void *dstp, size_t 
op_size)
+{
+    int pos = bytestream2_tell(gb);
+    int flag = bytestream2_peek_byte(gb);
+
+    if ((flag & 3) == 0) {
+        bytestream2_skip(gb, 1);
+        bytestream2_get_buffer(gb, dstp, op_size);
+    } else if ((flag & 3) == 1) {
+        bytestream2_skip(gb, 1);
+        memset(dstp, bytestream2_get_byte(gb), op_size);
+    } else {
+        uint32_t table[256];
+        int ret, elements = 0;
+
+        ret = fill_ltable(gb, table, &elements);
+        if (ret < 0)
+            return ret;
+        ret = get_opcodes(gb, table, dstp, op_size, elements);
+        if (ret < 0)
+            return ret;
+    }
+    return bytestream2_tell(gb) - pos;
+}
+
+static int dxv_decompress_cgo(DXVContext *ctx, GetByteContext *gb,
+                              uint8_t *tex_data, int tex_size,
+                              uint8_t *op_data, int *oindex,
+                              uint8_t **dstp, int *statep,
+                              uint8_t **tab0, uint8_t **tab1)
+{
+    uint8_t *dst = *dstp;
+    uint8_t *tptr0, *tptr1, *tptr3;
+    int oi = *oindex;
+    int state = *statep;
+    int opcode, v, vv;
+
+    if (state <= 0) {
+        opcode = op_data[oi++];
+        if (opcode) {
+            switch (opcode) {
+            case 1:
+                AV_WL32(dst, AV_RL32(dst - 16));
+                AV_WL32(dst + 4, AV_RL32(dst - 12));
+                break;
+            case 2:
+                vv = 16 * ((uint16_t)bytestream2_get_le16(gb) + 1);
+                if (vv < 0 || vv > dst - tex_data)
+                    return AVERROR_INVALIDDATA;
+                tptr0 = dst - vv;
+                v = AV_RL32(tptr0);
+                AV_WL32(dst, AV_RL32(tptr0));
+                AV_WL32(dst + 4, AV_RL32(tptr0 + 4));
+                tab0[0x9E3779B1 * (uint16_t)v >> 24] = dst;
+                tab1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = dst 
+ 2;
+                break;
+            case 3:
+                AV_WL32(dst, bytestream2_get_le32(gb));
+                AV_WL32(dst + 4, bytestream2_get_le32(gb));
+                tab0[0x9E3779B1 * AV_RL16(dst) >> 24] = dst;
+                tab1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = dst 
+ 2;
+                break;
+            case 4:
+                tptr3 = tab1[bytestream2_get_byte(gb)];
+                if (!tptr3)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, bytestream2_get_le16(gb));
+                AV_WL16(dst + 2, AV_RL16(tptr3));
+                dst[4] = tptr3[2];
+                AV_WL16(dst + 5, bytestream2_get_le16(gb));
+                dst[7] = bytestream2_get_byte(gb);
+                tab0[0x9E3779B1 * AV_RL16(dst) >> 24] = dst;
+                break;
+            case 5:
+                tptr3 = tab1[bytestream2_get_byte(gb)];
+                if (!tptr3)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, bytestream2_get_le16(gb));
+                AV_WL16(dst + 2, bytestream2_get_le16(gb));
+                dst[4] = bytestream2_get_byte(gb);
+                AV_WL16(dst + 5, AV_RL16(tptr3));
+                dst[7] = tptr3[2];
+                tab0[0x9E3779B1 * AV_RL16(dst) >> 24] = dst;
+                tab1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = dst 
+ 2;
+                break;
+            case 6:
+                tptr0 = tab1[bytestream2_get_byte(gb)];
+                if (!tptr0)
+                    return AVERROR_INVALIDDATA;
+                tptr1 = tab1[bytestream2_get_byte(gb)];
+                if (!tptr1)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, bytestream2_get_le16(gb));
+                AV_WL16(dst + 2, AV_RL16(tptr0));
+                dst[4] = tptr0[2];
+                AV_WL16(dst + 5, AV_RL16(tptr1));
+                dst[7] = tptr1[2];
+                tab0[0x9E3779B1 * AV_RL16(dst) >> 24] = dst;
+                break;
+            case 7:
+                v = 16 * ((uint16_t)bytestream2_get_le16(gb) + 1);
+                if (v < 0 || v > dst - tex_data)
+                    return AVERROR_INVALIDDATA;
+                tptr0 = dst - v;
+                AV_WL16(dst, bytestream2_get_le16(gb));
+                AV_WL16(dst + 2, AV_RL16(tptr0 + 2));
+                AV_WL32(dst + 4, AV_RL32(tptr0 + 4));
+                tab0[0x9E3779B1 * AV_RL16(dst) >> 24] = dst;
+                tab1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = dst 
+ 2;
+                break;
+            case 8:
+                tptr1 = tab0[bytestream2_get_byte(gb)];
+                if (!tptr1)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, AV_RL16(tptr1));
+                AV_WL16(dst + 2, bytestream2_get_le16(gb));
+                AV_WL32(dst + 4, bytestream2_get_le32(gb));
+                tab1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = dst 
+ 2;
+                break;
+            case 9:
+                tptr1 = tab0[bytestream2_get_byte(gb)];
+                if (!tptr1)
+                    return AVERROR_INVALIDDATA;
+                tptr3 = tab1[bytestream2_get_byte(gb)];
+                if (!tptr3)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, AV_RL16(tptr1));
+                AV_WL16(dst + 2, AV_RL16(tptr3));
+                dst[4] = tptr3[2];
+                AV_WL16(dst + 5, bytestream2_get_le16(gb));
+                dst[7] = bytestream2_get_byte(gb);
+                tab1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = dst 
+ 2;
+                break;
+            case 10:
+                tptr1 = tab0[bytestream2_get_byte(gb)];
+                if (!tptr1)
+                    return AVERROR_INVALIDDATA;
+                tptr3 = tab1[bytestream2_get_byte(gb)];
+                if (!tptr3)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, AV_RL16(tptr1));
+                AV_WL16(dst + 2, bytestream2_get_le16(gb));
+                dst[4] = bytestream2_get_byte(gb);
+                AV_WL16(dst + 5, AV_RL16(tptr3));
+                dst[7] = tptr3[2];
+                tab1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = dst 
+ 2;
+                break;
+            case 11:
+                tptr0 = tab0[bytestream2_get_byte(gb)];
+                if (!tptr0)
+                    return AVERROR_INVALIDDATA;
+                tptr3 = tab1[bytestream2_get_byte(gb)];
+                if (!tptr3)
+                    return AVERROR_INVALIDDATA;
+                tptr1 = tab1[bytestream2_get_byte(gb)];
+                if (!tptr1)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, AV_RL16(tptr0));
+                AV_WL16(dst + 2, AV_RL16(tptr3));
+                dst[4] = tptr3[2];
+                AV_WL16(dst + 5, AV_RL16(tptr1));
+                dst[7] = tptr1[2];
+                break;
+            case 12:
+                tptr1 = tab0[bytestream2_get_byte(gb)];
+                if (!tptr1)
+                    return AVERROR_INVALIDDATA;
+                v = 16 * ((uint16_t)bytestream2_get_le16(gb) + 1);
+                if (v < 0 || v > dst - tex_data)
+                    return AVERROR_INVALIDDATA;
+                tptr0 = dst - v;
+                AV_WL16(dst, AV_RL16(tptr1));
+                AV_WL16(dst + 2, AV_RL16(tptr0 + 2));
+                AV_WL32(dst + 4, AV_RL32(tptr0 + 4));
+                tab1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = dst 
+ 2;
+                break;
+            case 13:
+                AV_WL16(dst, AV_RL16(dst - 16));
+                AV_WL16(dst + 2, bytestream2_get_le16(gb));
+                AV_WL32(dst + 4, bytestream2_get_le32(gb));
+                tab1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = dst 
+ 2;
+                break;
+            case 14:
+                tptr3 = tab1[bytestream2_get_byte(gb)];
+                if (!tptr3)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, AV_RL16(dst - 16));
+                AV_WL16(dst + 2, AV_RL16(tptr3));
+                dst[4] = tptr3[2];
+                AV_WL16(dst + 5, bytestream2_get_le16(gb));
+                dst[7] = bytestream2_get_byte(gb);
+                tab1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = dst 
+ 2;
+                break;
+            case 15:
+                tptr3 = tab1[bytestream2_get_byte(gb)];
+                if (!tptr3)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, AV_RL16(dst - 16));
+                AV_WL16(dst + 2, bytestream2_get_le16(gb));
+                dst[4] = bytestream2_get_byte(gb);
+                AV_WL16(dst + 5, AV_RL16(tptr3));
+                dst[7] = tptr3[2];
+                tab1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = dst 
+ 2;
+                break;
+            case 16:
+                tptr3 = tab1[bytestream2_get_byte(gb)];
+                if (!tptr3)
+                    return AVERROR_INVALIDDATA;
+                tptr1 = tab1[bytestream2_get_byte(gb)];
+                if (!tptr1)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, AV_RL16(dst - 16));
+                AV_WL16(dst + 2, AV_RL16(tptr3));
+                dst[4] = tptr3[2];
+                AV_WL16(dst + 5, AV_RL16(tptr1));
+                dst[7] = tptr1[2];
+                break;
+            case 17:
+                v = 16 * ((uint16_t)bytestream2_get_le16(gb) + 1);
+                if (v < 0 || v > dst - tex_data)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, AV_RL16(dst - 16));
+                AV_WL16(dst + 2, AV_RL16(&dst[-v + 2]));
+                AV_WL32(dst + 4, AV_RL32(&dst[-v + 4]));
+                tab1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = dst 
+ 2;
+                break;
+            default:
+                break;
+            }
+        } else {
+            v = bytestream2_get_byte(gb);
+            if (v == 255) {
+                do {
+                    if (bytestream2_get_bytes_left(gb) <= 0)
+                        return AVERROR_INVALIDDATA;
+                    opcode = bytestream2_get_le16(gb);
+                    v += opcode;
+                } while (opcode == 0xFFFF);
+            }
+            AV_WL32(dst, AV_RL32(dst - 16));
+            AV_WL32(dst + 4, AV_RL32(dst - 12));
+            state = v + 3;
+        }
+    } else {
+        AV_WL32(dst, AV_RL32(dst - 16));
+        AV_WL32(dst + 4, AV_RL32(dst - 12));
+        state--;
+    }
+    if (dst - tex_data + 8 > tex_size)
+        return AVERROR_INVALIDDATA;
+    dst += 8;
+
+    *oindex = oi;
+    *dstp = dst;
+    *statep = state;
+
+    return 0;
+}
+
+static int dxv_decompress_cocg(DXVContext *ctx, GetByteContext *gb,
+                               uint8_t *tex_data, int tex_size,
+                               uint8_t *op_data0, uint8_t *op_data1,
+                               int max_op_size0, int max_op_size1)
+{
+    uint8_t *dst, *tab2[256] = { 0 }, *tab0[256] = { 0 }, *tab3[256] = { 0 }, 
*tab1[256] = { 0 };
+    int op_offset = bytestream2_get_le32(gb);
+    unsigned op_size0 = bytestream2_get_le32(gb);
+    unsigned op_size1 = bytestream2_get_le32(gb);
+    int data_start = bytestream2_tell(gb);
+    int skip0, skip1, oi0 = 0, oi1 = 0;
+    int ret, state0 = 0, state1 = 0;
+
+    dst = tex_data;
+    bytestream2_skip(gb, op_offset - 12);
+    if (op_size0 > max_op_size0)
+        return AVERROR_INVALIDDATA;
+    skip0 = dxv_decompress_opcodes(gb, op_data0, op_size0);
+    if (skip0 < 0)
+        return skip0;
+    bytestream2_seek(gb, data_start + op_offset + skip0 - 12, SEEK_SET);
+    if (op_size1 > max_op_size1)
+        return AVERROR_INVALIDDATA;
+    skip1 = dxv_decompress_opcodes(gb, op_data1, op_size1);
+    if (skip1 < 0)
+        return skip1;
+    bytestream2_seek(gb, data_start, SEEK_SET);
+
+    AV_WL32(dst, bytestream2_get_le32(gb));
+    AV_WL32(dst + 4, bytestream2_get_le32(gb));
+    AV_WL32(dst + 8, bytestream2_get_le32(gb));
+    AV_WL32(dst + 12, bytestream2_get_le32(gb));
+
+    tab0[0x9E3779B1 * AV_RL16(dst) >> 24] = dst;
+    tab1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFF) >> 24] = dst + 2;
+    tab2[0x9E3779B1 * AV_RL16(dst + 8) >> 24] = dst + 8;
+    tab3[0x9E3779B1 * (AV_RL32(dst + 10) & 0xFFFFFF) >> 24] = dst + 10;
+    dst += 16;
+    while (dst + 10 < tex_data + tex_size) {
+        ret = dxv_decompress_cgo(ctx, gb, tex_data, tex_size, op_data0, &oi0, 
&dst, &state0, tab0, tab1);
+        if (ret < 0)
+            return ret;
+        ret = dxv_decompress_cgo(ctx, gb, tex_data, tex_size, op_data1, &oi1, 
&dst, &state1, tab2, tab3);
+        if (ret < 0)
+            return ret;
+    }
+
+    bytestream2_seek(gb, data_start + op_offset + skip0 + skip1 - 12, 
SEEK_SET);
+
+    return 0;
+}
+
+static int dxv_decompress_yo(DXVContext *ctx, GetByteContext *gb,
+                             uint8_t *tex_data, int tex_size,
+                             uint8_t *op_data, int max_op_size)
+{
+    int op_offset = bytestream2_get_le32(gb);
+    unsigned op_size = bytestream2_get_le32(gb);
+    int data_start = bytestream2_tell(gb);
+    uint8_t *dst, *tptr0, *tptr1, *table0[256] = { 0 }, *table1[256] = { 0 };
+    int opcode, skip, oi = 0, v, vv;
+
+    dst = tex_data;
+    bytestream2_skip(gb, op_offset - 8);
+    if (op_size > max_op_size)
+        return AVERROR_INVALIDDATA;
+    skip = dxv_decompress_opcodes(gb, op_data, op_size);
+    if (skip < 0)
+        return skip;
+    bytestream2_seek(gb, data_start, SEEK_SET);
+
+    v = bytestream2_get_le32(gb);
+    AV_WL32(dst, v);
+    vv = bytestream2_get_le32(gb);
+    table0[0x9E3779B1 * (uint16_t)v >> 24] = dst;
+    AV_WL32(dst + 4, vv);
+    table1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFF) >> 24] = dst + 2;
+    dst += 8;
+
+    while (dst < tex_data + tex_size) {
+        opcode = op_data[oi++];
+        if (opcode) {
+            switch (opcode) {
+            case 1:
+                AV_WL32(dst, AV_RL32(dst - 8));
+                AV_WL32(dst + 4, AV_RL32(dst - 4));
+                break;
+            case 2:
+                vv = 8 * (uint16_t)bytestream2_get_le16(gb) + 8;
+                if (vv > dst - tex_data)
+                    return AVERROR_INVALIDDATA;
+                tptr0 = dst - vv;
+                v = AV_RL32(tptr0);
+                AV_WL32(dst, AV_RL32(tptr0));
+                AV_WL32(dst + 4, AV_RL32(tptr0 + 4));
+                table0[0x9E3779B1 * (uint16_t)v >> 24] = dst;
+                table1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFU) >> 24] = 
dst + 2;
+                break;
+            case 3:
+                v = bytestream2_get_le32(gb);
+                AV_WL32(dst, v);
+                AV_WL32(dst + 4, bytestream2_get_le32(gb));
+                vv = AV_RL32(dst + 2);
+                table0[0x9E3779B1 * (uint16_t)v >> 24] = dst;
+                table1[0x9E3779B1 * (vv & 0xFFFFFF) >> 24] = dst + 2;
+                break;
+            case 4:
+                tptr1 = table1[bytestream2_get_byte(gb)];
+                if (!tptr1)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, bytestream2_get_le16(gb));
+                AV_WL16(dst + 2, AV_RL16(tptr1));
+                dst[4] = tptr1[2];
+                AV_WL16(dst + 5, bytestream2_get_le16(gb));
+                dst[7] = bytestream2_get_byte(gb);
+                v = 0x9E3779B1 * AV_RL16(dst) >> 24;
+                table1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = 
dst + 2;
+                table0[v] = dst;
+                break;
+            case 5:
+                tptr1 = table1[bytestream2_get_byte(gb)];
+                if (!tptr1)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, bytestream2_get_le16(gb));
+                AV_WL16(dst + 2, bytestream2_get_le16(gb));
+                dst[4] = bytestream2_get_byte(gb);
+                AV_WL16(dst + 5, AV_RL16(tptr1));
+                dst[7] = tptr1[2];
+                table0[0x9E3779B1 * AV_RL16(dst) >> 24] = dst;
+                table1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = 
dst + 2;
+                break;
+            case 6:
+                tptr0 = table1[bytestream2_get_byte(gb)];
+                if (!tptr0)
+                    return AVERROR_INVALIDDATA;
+                tptr1 = table1[bytestream2_get_byte(gb)];
+                if (!tptr1)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, bytestream2_get_le16(gb));
+                AV_WL16(dst + 2, AV_RL16(tptr0));
+                dst[4] = tptr0[2];
+                AV_WL16(dst + 5, AV_RL16(tptr1));
+                dst[7] = tptr1[2];
+                table0[0x9E3779B1 * AV_RL16(dst) >> 24] = dst;
+                break;
+            case 7:
+                v = 8 * (uint16_t)bytestream2_get_le16(gb) + 8;
+                if (v > dst - tex_data)
+                    return AVERROR_INVALIDDATA;
+                tptr0 = dst - v;
+                AV_WL16(dst, bytestream2_get_le16(gb));
+                AV_WL16(dst + 2, AV_RL16(tptr0 + 2));
+                AV_WL32(dst + 4, AV_RL32(tptr0 + 4));
+                table0[0x9E3779B1 * AV_RL16(dst) >> 24] = dst;
+                table1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = 
dst + 2;
+                break;
+            case 8:
+                tptr0 = table0[bytestream2_get_byte(gb)];
+                if (!tptr0)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, AV_RL16(tptr0));
+                AV_WL16(dst + 2, bytestream2_get_le16(gb));
+                AV_WL32(dst + 4, bytestream2_get_le32(gb));
+                table1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = 
dst + 2;
+                break;
+            case 9:
+                tptr0 = table0[bytestream2_get_byte(gb)];
+                if (!tptr0)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, AV_RL16(tptr0));
+                tptr1 = table1[bytestream2_get_byte(gb)];
+                if (!tptr1)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst + 2, AV_RL16(tptr1));
+                dst[4] = tptr1[2];
+                AV_WL16(dst + 5, bytestream2_get_le16(gb));
+                dst[7] = bytestream2_get_byte(gb);
+                table1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = 
dst + 2;
+                break;
+            case 10:
+                tptr0 = table0[bytestream2_get_byte(gb)];
+                if (!tptr0)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, AV_RL16(tptr0));
+                tptr1 = table1[bytestream2_get_byte(gb)];
+                AV_WL16(dst + 2, bytestream2_get_le16(gb));
+                dst[4] = bytestream2_get_byte(gb);
+                AV_WL16(dst + 5, AV_RL16(tptr1));
+                dst[7] = tptr1[2];
+                table1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = 
dst + 2;
+                break;
+            case 11:
+                tptr0 = table0[bytestream2_get_byte(gb)];
+                if (!tptr0)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, AV_RL16(tptr0));
+                tptr0 = table1[bytestream2_get_byte(gb)];
+                if (!tptr0)
+                    return AVERROR_INVALIDDATA;
+                tptr1 = table1[bytestream2_get_byte(gb)];
+                if (!tptr1)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst + 2, AV_RL16(tptr0));
+                dst[4] = tptr0[2];
+                AV_WL16(dst + 5, AV_RL16(tptr1));
+                dst[7] = tptr1[2];
+                break;
+            case 12:
+                tptr0 = table0[bytestream2_get_byte(gb)];
+                if (!tptr0)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, AV_RL16(tptr0));
+                tptr0 = dst - (8 * bytestream2_get_le16(gb) + 8);
+                AV_WL16(dst + 2, AV_RL16(tptr0 + 2));
+                AV_WL32(dst + 4, AV_RL32(tptr0 + 4));
+                table1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFF) >> 24] = dst 
+ 2;
+                break;
+            case 13:
+                AV_WL16(dst, AV_RL16(dst - 8));
+                AV_WL16(dst + 2, bytestream2_get_le16(gb));
+                AV_WL32(dst + 4, bytestream2_get_le32(gb));
+                table1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFF) >> 24] = dst 
+ 2;
+                break;
+            case 14:
+                tptr1 = table1[bytestream2_get_byte(gb)];
+                if (!tptr1)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, AV_RL16(dst - 8));
+                AV_WL16(dst + 2, AV_RL16(tptr1));
+                dst[4] = tptr1[2];
+                AV_WL16(dst + 5, bytestream2_get_le16(gb));
+                dst[7] = bytestream2_get_byte(gb);
+                table1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFF) >> 24] = dst 
+ 2;
+                break;
+            case 15:
+                tptr1 = table1[bytestream2_get_byte(gb)];
+                if (!tptr1)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, AV_RL16(dst - 8));
+                AV_WL16(dst + 2, bytestream2_get_le16(gb));
+                dst[4] = bytestream2_get_byte(gb);
+                AV_WL16(dst + 5, AV_RL16(tptr1));
+                dst[7] = tptr1[2];
+                table1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = 
dst + 2;
+                break;
+            case 16:
+                tptr0 = table1[bytestream2_get_byte(gb)];
+                if (!tptr0)
+                    return AVERROR_INVALIDDATA;
+                tptr1 = table1[bytestream2_get_byte(gb)];
+                if (!tptr1)
+                    return AVERROR_INVALIDDATA;
+                AV_WL16(dst, AV_RL16(dst - 8));
+                AV_WL16(dst + 2, AV_RL16(tptr0));
+                dst[4] = tptr0[2];
+                AV_WL16(dst + 5, AV_RL16(tptr1));
+                dst[7] = tptr1[2];
+                break;
+            case 17:
+                v = 8 * bytestream2_get_le16(gb) + 8;
+                if (v > dst - tex_data)
+                    return AVERROR_INVALIDDATA;
+                tptr0 = dst - v;
+                AV_WL16(dst, AV_RL16(dst - 8));
+                AV_WL16(dst + 2, AV_RL16(tptr0 + 2));
+                AV_WL32(dst + 4, AV_RL32(tptr0 + 4));
+                table1[0x9E3779B1 * (AV_RL32(dst + 2) & 0xFFFFFFu) >> 24] = 
dst + 2;
+                break;
+            default:
+                break;
+            }
+            dst += 8;
+        } else {
+            v = bytestream2_get_byte(gb);
+            if (v == 255) {
+                do {
+                    if (bytestream2_get_bytes_left(gb) <= 0)
+                        return AVERROR_INVALIDDATA;
+                    opcode = bytestream2_get_le16(gb);
+                    v += opcode;
+                } while (opcode == 0xFFFF);
+            }
+            vv = v + 4;
+            do {
+                if (dst - tex_data + 8 > tex_size)
+                    return AVERROR_INVALIDDATA;
+                AV_WL32(dst, AV_RL32(dst - 8));
+                AV_WL32(dst + 4, AV_RL32(dst - 4));
+                dst += 8;
+                --vv;
+            } while (vv);
+        }
+    }
+
+    bytestream2_seek(gb, data_start + op_offset + skip - 8, SEEK_SET);
+
+    return 0;
+}
+
+static int dxv_decompress_ycg6(AVCodecContext *avctx)
+{
+    DXVContext *ctx = avctx->priv_data;
+    GetByteContext *gb = &ctx->gbc;
+    int ret;
+
+    ret = dxv_decompress_yo(ctx, gb, ctx->tex_data, ctx->tex_size,
+                            ctx->op_data[0], ctx->op_size[0]);
+    if (ret < 0)
+        return ret;
+
+    return dxv_decompress_cocg(ctx, gb, ctx->ctex_data, ctx->ctex_size,
+                               ctx->op_data[1], ctx->op_data[2],
+                               ctx->op_size[1], ctx->op_size[2]);
+}
+
+static int dxv_decompress_yg10(AVCodecContext *avctx)
+{
+    DXVContext *ctx = avctx->priv_data;
+    GetByteContext *gb = &ctx->gbc;
+    int ret;
+
+    ret = dxv_decompress_cocg(ctx, gb, ctx->tex_data, ctx->tex_size,
+                              ctx->op_data[0], ctx->op_data[3],
+                              ctx->op_size[0], ctx->op_size[3]);
+    if (ret < 0)
+        return ret;
+
+    return dxv_decompress_cocg(ctx, gb, ctx->ctex_data, ctx->ctex_size,
+                               ctx->op_data[1], ctx->op_data[2],
+                               ctx->op_size[1], ctx->op_size[2]);
+}
+
 static int dxv_decompress_dxt5(AVCodecContext *avctx)
 {
     DXVContext *ctx = avctx->priv_data;
@@ -359,6 +1280,9 @@ static int dxv_decode(AVCodecContext *avctx, void *data,
 
     bytestream2_init(gbc, avpkt->data, avpkt->size);
 
+    ctx->texture_block_h = 4;
+    ctx->texture_block_w = 4;
+
     tag = bytestream2_get_le32(gbc);
     switch (tag) {
     case MKBETAG('D', 'X', 'T', '1'):
@@ -378,9 +1302,35 @@ static int dxv_decode(AVCodecContext *avctx, void *data,
         msgtext = "DXT5";
         break;
     case MKBETAG('Y', 'C', 'G', '6'):
+        decompress_tex = dxv_decompress_ycg6;
+        ctx->tex_funct = yo_block;
+        ctx->ctex_funct = cocg_block;
+        ctx->tex_rat   = 8;
+        ctx->tex_step  = 32;
+        ctx->ctex_step = 16;
+        msgcomp = "YOCOCG6";
+        msgtext = "YCG6";
+        ctx->ctex_size = avctx->coded_width * avctx->coded_height / 4;
+        ctx->texture_block_h = 4;
+        ctx->texture_block_w = 16;
+        ctx->ctexture_block_h = 8;
+        ctx->ctexture_block_w = 8;
+        break;
     case MKBETAG('Y', 'G', '1', '0'):
-        avpriv_report_missing_feature(avctx, "Tag 0x%08"PRIX32, tag);
-        return AVERROR_PATCHWELCOME;
+        decompress_tex = dxv_decompress_yg10;
+        ctx->tex_funct = yao_block;
+        ctx->ctex_funct = cocg_block;
+        ctx->tex_rat   = 4;
+        ctx->tex_step  = 64;
+        ctx->ctex_step = 16;
+        msgcomp = "YAOCOCG10";
+        msgtext = "YG10";
+        ctx->ctex_size = avctx->coded_width * avctx->coded_height / 4;
+        ctx->texture_block_h = 4;
+        ctx->texture_block_w = 16;
+        ctx->ctexture_block_h = 8;
+        ctx->ctexture_block_w = 8;
+        break;
     default:
         /* Old version does not have a real header, just size and type. */
         size = tag & 0x00FFFFFF;
@@ -413,6 +1363,10 @@ static int dxv_decode(AVCodecContext *avctx, void *data,
         break;
     }
 
+    ctx->slice_count = av_clip(avctx->thread_count, 1,
+                               avctx->coded_height / 
FFMAX(ctx->texture_block_h,
+                                                           
ctx->ctexture_block_h));
+
     /* New header is 12 bytes long. */
     if (!old_type) {
         version_major = bytestream2_get_byte(gbc) - 1;
@@ -440,10 +1394,28 @@ static int dxv_decode(AVCodecContext *avctx, void *data,
     }
 
     ctx->tex_size = avctx->coded_width * avctx->coded_height * 4 / 
ctx->tex_rat;
-    ret = av_reallocp(&ctx->tex_data, ctx->tex_size);
+    ret = av_reallocp(&ctx->tex_data, ctx->tex_size + 
AV_INPUT_BUFFER_PADDING_SIZE);
     if (ret < 0)
         return ret;
 
+    if (ctx->ctex_size) {
+        int i;
+
+        ctx->op_size[0] = avctx->coded_width * avctx->coded_height / 16;
+        ctx->op_size[1] = avctx->coded_width * avctx->coded_height / 32;
+        ctx->op_size[2] = avctx->coded_width * avctx->coded_height / 32;
+        ctx->op_size[3] = avctx->coded_width * avctx->coded_height / 16;
+
+        ret = av_reallocp(&ctx->ctex_data, ctx->ctex_size + 
AV_INPUT_BUFFER_PADDING_SIZE);
+        if (ret < 0)
+            return ret;
+        for (i = 0; i < 4; i++) {
+            ret = av_reallocp(&ctx->op_data[i], ctx->op_size[i]);
+            if (ret < 0)
+                return ret;
+        }
+    }
+
     /* Decompress texture out of the intermediate compression. */
     ret = decompress_tex(avctx);
     if (ret < 0)
@@ -484,9 +1456,6 @@ static int dxv_init(AVCodecContext *avctx)
     ff_texturedsp_init(&ctx->texdsp);
     avctx->pix_fmt = AV_PIX_FMT_RGBA;
 
-    ctx->slice_count = av_clip(avctx->thread_count, 1,
-                               avctx->coded_height / TEXTURE_BLOCK_H);
-
     return 0;
 }
 
@@ -495,6 +1464,11 @@ static int dxv_close(AVCodecContext *avctx)
     DXVContext *ctx = avctx->priv_data;
 
     av_freep(&ctx->tex_data);
+    av_freep(&ctx->ctex_data);
+    av_freep(&ctx->op_data[0]);
+    av_freep(&ctx->op_data[1]);
+    av_freep(&ctx->op_data[2]);
+    av_freep(&ctx->op_data[3]);
 
     return 0;
 }
-- 
2.11.0

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Reply via email to