Fix building with C++14 (default in GCC 6). Changes types (mostly float into
double). Add necessary casts.
See also: https://bugs.gentoo.org/show_bug.cgi?id=594116

--- a/src/analyzer.cpp
+++ b/src/analyzer.cpp
@@ -470,13 +470,13 @@
                                 lastoutL = fft_outL[_iter];
                                 //pumping up actual signal an erase surrounding
                                 // sounds
-                                fft_outL[_iter] = 0.25f * std::max(n * 0.6f * \
+                                fft_outL[_iter] = 0.25 * std::max(n * 0.6 * \
                                     fabs(fft_outL[_iter]) - var1L , 1e-20);
                                 if(_mode == 3 or _mode == 4) {
                                     // do the same with R channel if needed
                                     lastoutR = fft_outR[_iter];
-                                    fft_outR[_iter] = 0.25f * std::max(n * \
-                                        0.6f * fabs(fft_outR[_iter]) - var1R , 1e-20);
+                                    fft_outR[_iter] = 0.25 * std::max(n * \
+                                        0.6 * fabs(fft_outR[_iter]) - var1R , 1e-20);
                                 }
                                 break;
                         }
--- a/src/jack_client.cpp
+++ b/src/jack_client.cpp
@@ -226,7 +226,7 @@
                 map<string, int>::const_iterator p = port_to_plugin.find((*k) + cnlen + 1);
                 if (p != port_to_plugin.end())
                 {
-                    run_before.insert(make_pair<int, int>(p->second, i));
+                    run_before.insert(make_pair<int, int>((int)p->second, (int)i));
                 }
             }
             jack_free(conns);
--- a/src/modules_dist.cpp
+++ b/src/modules_dist.cpp
@@ -794,8 +794,8 @@
             lfo2.advance(1);
             
             // dot
-            rms = std::max((double)rms, (fabs(Lo) + fabs(Ro)) / 2);
-            input = std::max((double)input, (fabs(Lc) + fabs(Rc)) / 2);
+            rms = std::max((double)rms, (fabs(Lo) + fabs(Ro)) / 2.0);
+            input = std::max((double)input, (fabs(Lc) + fabs(Rc)) / 2.0);
             
             float values[] = {inL, inR, outs[0][i], outs[1][i]};
             meters.process(values);
--- a/src/modules_limit.cpp
+++ b/src/modules_limit.cpp
@@ -429,7 +429,7 @@
                 }
                 
                 // write multiband coefficient to buffer
-                buffer[pos] = std::min(*params[param_limit] / std::max(fabs(tmpL), fabs(tmpR)), 1.0);
+                buffer[pos] = std::min((double)*params[param_limit] / std::max(fabs(tmpL), fabs(tmpR)), 1.0);
                 
                 // step forward in multiband buffer
                 pos = (pos + channels) % buffer_size;
@@ -811,7 +811,7 @@
                 }
                 
                 // write multiband coefficient to buffer
-                buffer[pos] = std::min(*params[param_limit] / std::max(fabs(tmpL), fabs(tmpR)), 1.0);
+                buffer[pos] = std::min((double)*params[param_limit] / std::max(fabs(tmpL), fabs(tmpR)), 1.0);
                 
                 // step forward in multiband buffer
                 pos = (pos + channels) % buffer_size;