summaryrefslogtreecommitdiffstats
path: root/media/libstagefright/codecs/avc/dec/src/itrans.cpp
diff options
context:
space:
mode:
Diffstat (limited to 'media/libstagefright/codecs/avc/dec/src/itrans.cpp')
-rw-r--r--media/libstagefright/codecs/avc/dec/src/itrans.cpp307
1 files changed, 307 insertions, 0 deletions
diff --git a/media/libstagefright/codecs/avc/dec/src/itrans.cpp b/media/libstagefright/codecs/avc/dec/src/itrans.cpp
new file mode 100644
index 0000000..02c550d
--- /dev/null
+++ b/media/libstagefright/codecs/avc/dec/src/itrans.cpp
@@ -0,0 +1,307 @@
+/* ------------------------------------------------------------------
+ * Copyright (C) 1998-2009 PacketVideo
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either
+ * express or implied.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ * -------------------------------------------------------------------
+ */
+#include "avclib_common.h"
+
+/* input are in the first 16 elements of block,
+ output must be in the location specified in Figure 8-6. */
+/* subclause 8.5.6 */
+void Intra16DCTrans(int16 *block, int Qq, int Rq)
+{
+ int m0, m1, m2, m3;
+ int j, offset;
+ int16 *inout;
+ int scale = dequant_coefres[Rq][0];
+
+ inout = block;
+ for (j = 0; j < 4; j++)
+ {
+ m0 = inout[0] + inout[4];
+ m1 = inout[0] - inout[4];
+ m2 = inout[8] + inout[12];
+ m3 = inout[8] - inout[12];
+
+
+ inout[0] = m0 + m2;
+ inout[4] = m0 - m2;
+ inout[8] = m1 - m3;
+ inout[12] = m1 + m3;
+ inout += 64;
+ }
+
+ inout = block;
+
+ if (Qq >= 2) /* this way should be faster than JM */
+ { /* they use (((m4*scale)<<(QPy/6))+2)>>2 for both cases. */
+ Qq -= 2;
+ for (j = 0; j < 4; j++)
+ {
+ m0 = inout[0] + inout[64];
+ m1 = inout[0] - inout[64];
+ m2 = inout[128] + inout[192];
+ m3 = inout[128] - inout[192];
+
+ inout[0] = ((m0 + m2) * scale) << Qq;
+ inout[64] = ((m0 - m2) * scale) << Qq;
+ inout[128] = ((m1 - m3) * scale) << Qq;
+ inout[192] = ((m1 + m3) * scale) << Qq;
+ inout += 4;
+ }
+ }
+ else
+ {
+ Qq = 2 - Qq;
+ offset = 1 << (Qq - 1);
+
+ for (j = 0; j < 4; j++)
+ {
+ m0 = inout[0] + inout[64];
+ m1 = inout[0] - inout[64];
+ m2 = inout[128] + inout[192];
+ m3 = inout[128] - inout[192];
+
+ inout[0] = (((m0 + m2) * scale + offset) >> Qq);
+ inout[64] = (((m0 - m2) * scale + offset) >> Qq);
+ inout[128] = (((m1 - m3) * scale + offset) >> Qq);
+ inout[192] = (((m1 + m3) * scale + offset) >> Qq);
+ inout += 4;
+ }
+ }
+
+ return ;
+}
+
+/* see subclase 8.5.8 */
+void itrans(int16 *block, uint8 *pred, uint8 *cur, int width)
+{
+ int e0, e1, e2, e3; /* note, at every step of the calculation, these values */
+ /* shall never exceed 16bit sign value, but we don't check */
+ int i; /* to save the cycles. */
+ int16 *inout;
+
+ inout = block;
+
+ for (i = 4; i > 0; i--)
+ {
+ e0 = inout[0] + inout[2];
+ e1 = inout[0] - inout[2];
+ e2 = (inout[1] >> 1) - inout[3];
+ e3 = inout[1] + (inout[3] >> 1);
+
+ inout[0] = e0 + e3;
+ inout[1] = e1 + e2;
+ inout[2] = e1 - e2;
+ inout[3] = e0 - e3;
+
+ inout += 16;
+ }
+
+ for (i = 4; i > 0; i--)
+ {
+ e0 = block[0] + block[32];
+ e1 = block[0] - block[32];
+ e2 = (block[16] >> 1) - block[48];
+ e3 = block[16] + (block[48] >> 1);
+
+ e0 += e3;
+ e3 = (e0 - (e3 << 1)); /* e0-e3 */
+ e1 += e2;
+ e2 = (e1 - (e2 << 1)); /* e1-e2 */
+ e0 += 32;
+ e1 += 32;
+ e2 += 32;
+ e3 += 32;
+#ifdef USE_PRED_BLOCK
+ e0 = pred[0] + (e0 >> 6);
+ if ((uint)e0 > 0xFF) e0 = 0xFF & (~(e0 >> 31)); /* clip */
+ e1 = pred[20] + (e1 >> 6);
+ if ((uint)e1 > 0xFF) e1 = 0xFF & (~(e1 >> 31)); /* clip */
+ e2 = pred[40] + (e2 >> 6);
+ if ((uint)e2 > 0xFF) e2 = 0xFF & (~(e2 >> 31)); /* clip */
+ e3 = pred[60] + (e3 >> 6);
+ if ((uint)e3 > 0xFF) e3 = 0xFF & (~(e3 >> 31)); /* clip */
+ *cur = e0;
+ *(cur += width) = e1;
+ *(cur += width) = e2;
+ cur[width] = e3;
+ cur -= (width << 1);
+ cur++;
+ pred++;
+#else
+ OSCL_UNUSED_ARG(pred);
+
+ e0 = *cur + (e0 >> 6);
+ if ((uint)e0 > 0xFF) e0 = 0xFF & (~(e0 >> 31)); /* clip */
+ *cur = e0;
+ e1 = *(cur += width) + (e1 >> 6);
+ if ((uint)e1 > 0xFF) e1 = 0xFF & (~(e1 >> 31)); /* clip */
+ *cur = e1;
+ e2 = *(cur += width) + (e2 >> 6);
+ if ((uint)e2 > 0xFF) e2 = 0xFF & (~(e2 >> 31)); /* clip */
+ *cur = e2;
+ e3 = cur[width] + (e3 >> 6);
+ if ((uint)e3 > 0xFF) e3 = 0xFF & (~(e3 >> 31)); /* clip */
+ cur[width] = e3;
+ cur -= (width << 1);
+ cur++;
+#endif
+ block++;
+ }
+
+ return ;
+}
+
+/* see subclase 8.5.8 */
+void ictrans(int16 *block, uint8 *pred, uint8 *cur, int width)
+{
+ int e0, e1, e2, e3; /* note, at every step of the calculation, these values */
+ /* shall never exceed 16bit sign value, but we don't check */
+ int i; /* to save the cycles. */
+ int16 *inout;
+
+ inout = block;
+
+ for (i = 4; i > 0; i--)
+ {
+ e0 = inout[0] + inout[2];
+ e1 = inout[0] - inout[2];
+ e2 = (inout[1] >> 1) - inout[3];
+ e3 = inout[1] + (inout[3] >> 1);
+
+ inout[0] = e0 + e3;
+ inout[1] = e1 + e2;
+ inout[2] = e1 - e2;
+ inout[3] = e0 - e3;
+
+ inout += 16;
+ }
+
+ for (i = 4; i > 0; i--)
+ {
+ e0 = block[0] + block[32];
+ e1 = block[0] - block[32];
+ e2 = (block[16] >> 1) - block[48];
+ e3 = block[16] + (block[48] >> 1);
+
+ e0 += e3;
+ e3 = (e0 - (e3 << 1)); /* e0-e3 */
+ e1 += e2;
+ e2 = (e1 - (e2 << 1)); /* e1-e2 */
+ e0 += 32;
+ e1 += 32;
+ e2 += 32;
+ e3 += 32;
+#ifdef USE_PRED_BLOCK
+ e0 = pred[0] + (e0 >> 6);
+ if ((uint)e0 > 0xFF) e0 = 0xFF & (~(e0 >> 31)); /* clip */
+ e1 = pred[12] + (e1 >> 6);
+ if ((uint)e1 > 0xFF) e1 = 0xFF & (~(e1 >> 31)); /* clip */
+ e2 = pred[24] + (e2 >> 6);
+ if ((uint)e2 > 0xFF) e2 = 0xFF & (~(e2 >> 31)); /* clip */
+ e3 = pred[36] + (e3 >> 6);
+ if ((uint)e3 > 0xFF) e3 = 0xFF & (~(e3 >> 31)); /* clip */
+ *cur = e0;
+ *(cur += width) = e1;
+ *(cur += width) = e2;
+ cur[width] = e3;
+ cur -= (width << 1);
+ cur++;
+ pred++;
+#else
+ OSCL_UNUSED_ARG(pred);
+
+ e0 = *cur + (e0 >> 6);
+ if ((uint)e0 > 0xFF) e0 = 0xFF & (~(e0 >> 31)); /* clip */
+ *cur = e0;
+ e1 = *(cur += width) + (e1 >> 6);
+ if ((uint)e1 > 0xFF) e1 = 0xFF & (~(e1 >> 31)); /* clip */
+ *cur = e1;
+ e2 = *(cur += width) + (e2 >> 6);
+ if ((uint)e2 > 0xFF) e2 = 0xFF & (~(e2 >> 31)); /* clip */
+ *cur = e2;
+ e3 = cur[width] + (e3 >> 6);
+ if ((uint)e3 > 0xFF) e3 = 0xFF & (~(e3 >> 31)); /* clip */
+ cur[width] = e3;
+ cur -= (width << 1);
+ cur++;
+#endif
+ block++;
+ }
+
+ return ;
+}
+
+/* see subclause 8.5.7 */
+void ChromaDCTrans(int16 *block, int Qq, int Rq)
+{
+ int c00, c01, c10, c11;
+ int f0, f1, f2, f3;
+ int scale = dequant_coefres[Rq][0];
+
+ c00 = block[0] + block[4];
+ c01 = block[0] - block[4];
+ c10 = block[64] + block[68];
+ c11 = block[64] - block[68];
+
+ f0 = c00 + c10;
+ f1 = c01 + c11;
+ f2 = c00 - c10;
+ f3 = c01 - c11;
+
+ if (Qq >= 1)
+ {
+ Qq -= 1;
+ block[0] = (f0 * scale) << Qq;
+ block[4] = (f1 * scale) << Qq;
+ block[64] = (f2 * scale) << Qq;
+ block[68] = (f3 * scale) << Qq;
+ }
+ else
+ {
+ block[0] = (f0 * scale) >> 1;
+ block[4] = (f1 * scale) >> 1;
+ block[64] = (f2 * scale) >> 1;
+ block[68] = (f3 * scale) >> 1;
+ }
+
+ return ;
+}
+
+
+void copy_block(uint8 *pred, uint8 *cur, int width, int pred_pitch)
+{
+ uint32 temp;
+
+ temp = *((uint32*)pred);
+ pred += pred_pitch;
+ *((uint32*)cur) = temp;
+ cur += width;
+ temp = *((uint32*)pred);
+ pred += pred_pitch;
+ *((uint32*)cur) = temp;
+ cur += width;
+ temp = *((uint32*)pred);
+ pred += pred_pitch;
+ *((uint32*)cur) = temp;
+ cur += width;
+ temp = *((uint32*)pred);
+ *((uint32*)cur) = temp;
+
+ return ;
+}
+
+
generated by cgit v1.1 at 2024-06-08 22:27:51 +0000