Merge commit '6151e9128ce2a84a443c82b78f5b5cb364ba2ab4'
[ffmpeg.git] / libavutil / mathematics.c
index c8f1e1c..1bf044c 100644 (file)
 #include <limits.h>
 
 #include "mathematics.h"
+#include "libavutil/intmath.h"
 #include "libavutil/common.h"
 #include "avassert.h"
 #include "version.h"
 
-#if FF_API_AV_REVERSE
-const uint8_t av_reverse[256] = {
-0x00,0x80,0x40,0xC0,0x20,0xA0,0x60,0xE0,0x10,0x90,0x50,0xD0,0x30,0xB0,0x70,0xF0,
-0x08,0x88,0x48,0xC8,0x28,0xA8,0x68,0xE8,0x18,0x98,0x58,0xD8,0x38,0xB8,0x78,0xF8,
-0x04,0x84,0x44,0xC4,0x24,0xA4,0x64,0xE4,0x14,0x94,0x54,0xD4,0x34,0xB4,0x74,0xF4,
-0x0C,0x8C,0x4C,0xCC,0x2C,0xAC,0x6C,0xEC,0x1C,0x9C,0x5C,0xDC,0x3C,0xBC,0x7C,0xFC,
-0x02,0x82,0x42,0xC2,0x22,0xA2,0x62,0xE2,0x12,0x92,0x52,0xD2,0x32,0xB2,0x72,0xF2,
-0x0A,0x8A,0x4A,0xCA,0x2A,0xAA,0x6A,0xEA,0x1A,0x9A,0x5A,0xDA,0x3A,0xBA,0x7A,0xFA,
-0x06,0x86,0x46,0xC6,0x26,0xA6,0x66,0xE6,0x16,0x96,0x56,0xD6,0x36,0xB6,0x76,0xF6,
-0x0E,0x8E,0x4E,0xCE,0x2E,0xAE,0x6E,0xEE,0x1E,0x9E,0x5E,0xDE,0x3E,0xBE,0x7E,0xFE,
-0x01,0x81,0x41,0xC1,0x21,0xA1,0x61,0xE1,0x11,0x91,0x51,0xD1,0x31,0xB1,0x71,0xF1,
-0x09,0x89,0x49,0xC9,0x29,0xA9,0x69,0xE9,0x19,0x99,0x59,0xD9,0x39,0xB9,0x79,0xF9,
-0x05,0x85,0x45,0xC5,0x25,0xA5,0x65,0xE5,0x15,0x95,0x55,0xD5,0x35,0xB5,0x75,0xF5,
-0x0D,0x8D,0x4D,0xCD,0x2D,0xAD,0x6D,0xED,0x1D,0x9D,0x5D,0xDD,0x3D,0xBD,0x7D,0xFD,
-0x03,0x83,0x43,0xC3,0x23,0xA3,0x63,0xE3,0x13,0x93,0x53,0xD3,0x33,0xB3,0x73,0xF3,
-0x0B,0x8B,0x4B,0xCB,0x2B,0xAB,0x6B,0xEB,0x1B,0x9B,0x5B,0xDB,0x3B,0xBB,0x7B,0xFB,
-0x07,0x87,0x47,0xC7,0x27,0xA7,0x67,0xE7,0x17,0x97,0x57,0xD7,0x37,0xB7,0x77,0xF7,
-0x0F,0x8F,0x4F,0xCF,0x2F,0xAF,0x6F,0xEF,0x1F,0x9F,0x5F,0xDF,0x3F,0xBF,0x7F,0xFF,
-};
-#endif
-
-int64_t av_gcd(int64_t a, int64_t b)
-{
-    if (b)
-        return av_gcd(b, a % b);
-    else
+/* Stein's binary GCD algorithm:
+ * https://en.wikipedia.org/wiki/Binary_GCD_algorithm */
+int64_t av_gcd(int64_t a, int64_t b) {
+    int za, zb, k;
+    int64_t u, v;
+    if (a == 0)
+        return b;
+    if (b == 0)
         return a;
+    za = ff_ctzll(a);
+    zb = ff_ctzll(b);
+    k  = FFMIN(za, zb);
+    u = llabs(a >> za);
+    v = llabs(b >> zb);
+    while (u != v) {
+        if (u > v)
+            FFSWAP(int64_t, v, u);
+        v -= u;
+        v >>= ff_ctzll(v);
+    }
+    return (uint64_t)u << k;
 }
 
 int64_t av_rescale_rnd(int64_t a, int64_t b, int64_t c, enum AVRounding rnd)
@@ -76,8 +71,8 @@ int64_t av_rescale_rnd(int64_t a, int64_t b, int64_t c, enum AVRounding rnd)
         rnd -= AV_ROUND_PASS_MINMAX;
     }
 
-    if (a < 0 && a != INT64_MIN)
-        return -av_rescale_rnd(-a, b, c, rnd ^ ((rnd >> 1) & 1));
+    if (a < 0)
+        return -(uint64_t)av_rescale_rnd(-FFMAX(a, -INT64_MAX), b, c, rnd ^ ((rnd >> 1) & 1));
 
     if (rnd == AV_ROUND_NEAR_INF)
         r = c / 2;
@@ -87,8 +82,13 @@ int64_t av_rescale_rnd(int64_t a, int64_t b, int64_t c, enum AVRounding rnd)
     if (b <= INT_MAX && c <= INT_MAX) {
         if (a <= INT_MAX)
             return (a * b + r) / c;
-        else
-            return a / c * b + (a % c * b + r) / c;
+        else {
+            int64_t ad = a / c;
+            int64_t a2 = (a % c * b + r) / c;
+            if (ad >= INT32_MAX && b && ad > (INT64_MAX - a2) / b)
+                return INT64_MIN;
+            return ad * b + a2;
+        }
     } else {
 #if 1
         uint64_t a0  = a & 0xFFFFFFFF;
@@ -112,16 +112,18 @@ int64_t av_rescale_rnd(int64_t a, int64_t b, int64_t c, enum AVRounding rnd)
                 t1++;
             }
         }
+        if (t1 > INT64_MAX)
+            return INT64_MIN;
         return t1;
-    }
 #else
+        /* reference code doing (a*b + r) / c, requires libavutil/integer.h */
         AVInteger ai;
         ai = av_mul_i(av_int2i(a), av_int2i(b));
         ai = av_add_i(ai, av_int2i(r));
 
         return av_i2int(av_div_i(ai, av_int2i(c)));
-    }
 #endif
+    }
 }
 
 int64_t av_rescale(int64_t a, int64_t b, int64_t c)
@@ -188,13 +190,20 @@ simple_round:
 
 int64_t av_add_stable(AVRational ts_tb, int64_t ts, AVRational inc_tb, int64_t inc)
 {
+    int64_t m, d;
+
     if (inc != 1)
         inc_tb = av_mul_q(inc_tb, (AVRational) {inc, 1});
 
-    if (av_cmp_q(inc_tb, ts_tb) < 0) {
-        //increase step is too small for even 1 step to be representable
+    m = inc_tb.num * (int64_t)ts_tb.den;
+    d = inc_tb.den * (int64_t)ts_tb.num;
+
+    if (m % d == 0)
+        return ts + m / d;
+    if (m < d)
         return ts;
-    } else {
+
+    {
         int64_t old = av_rescale_q(ts, ts_tb, inc_tb);
         int64_t old_ts = av_rescale_q(old, inc_tb, ts_tb);
         return av_rescale_q(old + 1, inc_tb, ts_tb) + (ts - old_ts);