avcodec/opus_silk: Change silk_lsf2lpc() slightly toward silk/NLSF2A.c

Fixes: runtime error: signed integer overflow: -1440457022 - 785819492 cannot be represented in type 'int'
Fixes: 7700/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_OPUS_fuzzer-6595838684954624

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
(cherry picked from commit e7dda51150)
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

libavcodec/opus_silk.c

@@ -955,8 +955,10 @@ static void silk_lsf2lpc(const int16_t nlsf[16], float lpcf[16], int order)
 
     /* reconstruct A(z) */
     for (k = 0; k < order>>1; k++) {
-        lpc32[k]         = -p[k + 1] - p[k] - q[k + 1] + q[k];
-        lpc32[order-k-1] = -p[k + 1] - p[k] + q[k + 1] - q[k];
+        int32_t p_tmp = p[k + 1] + p[k];
+        int32_t q_tmp = q[k + 1] - q[k];
+        lpc32[k]         = -q_tmp - p_tmp;
+        lpc32[order-k-1] =  q_tmp - p_tmp;
     }
 
     /* limit the range of the LPC coefficients to each fit within an int16_t */