1 files changed, 335 insertions, 0 deletions

diff --git a/lib/rbcodec/codecs/libgme/blip_buffer.h b/lib/rbcodec/codecs/libgme/blip_buffer.h
new file mode 100644
index 0000000000..5fe1f4b9c6
--- /dev/null
+++ b/lib/rbcodec/codecs/libgme/blip_buffer.h
@@ -0,0 +1,335 @@
+// Band-limited sound synthesis buffer
+
+// Blip_Buffer 0.4.1
+#ifndef BLIP_BUFFER_H
+#define BLIP_BUFFER_H
+
+#include "blargg_common.h"
+
+typedef unsigned blip_resampled_time_t;
+typedef int blip_time_t;
+typedef int clocks_t;
+
+// Output samples are 16-bit signed, with a range of -32768 to 32767
+typedef short blip_sample_t;
+
+static int const blip_default_length = 1000 / 4;   // Default Blip_Buffer length (1/4 second)
+
+#ifndef BLIP_MAX_QUALITY
+        #define BLIP_MAX_QUALITY 2
+#endif
+
+#ifndef BLIP_BUFFER_ACCURACY
+        #define BLIP_BUFFER_ACCURACY 16
+#endif
+
+// linear interpolation needs 8 bits
+#ifndef BLIP_PHASE_BITS
+        #define BLIP_PHASE_BITS 8
+#endif
+
+static int const blip_res           = 1 << BLIP_PHASE_BITS;
+static int const blip_buffer_extra_ = BLIP_MAX_QUALITY + 2;
+
+// Properties of fixed-point sample position
+typedef unsigned ufixed_t; // unsigned for more range, optimized shifts
+enum { fixed_bits = BLIP_BUFFER_ACCURACY };             // bits in fraction
+enum { fixed_unit = 1 << fixed_bits };  // 1.0 samples
+
+// Deltas in buffer are fixed-point with this many fraction bits.
+// Less than 16 for extra range.
+enum { delta_bits = 14 };
+
+// Pointer to first committed delta sample
+typedef int delta_t;
+
+// Maximun buffer size (48Khz, 50 ms)
+enum { blip_buffer_max = 2466 };
+
+struct Blip_Buffer {
+        unsigned factor_;
+        ufixed_t  offset_;
+        delta_t* buffer_center_;
+        int      buffer_size_;
+        int      reader_accum_;
+        int      bass_shift_;
+        int      bass_freq_;
+        int      sample_rate_;
+        int      clock_rate_;
+        int      length_;
+        bool     modified;
+
+        delta_t  buffer_ [blip_buffer_max];
+};
+
+// Blip_Buffer_ implementation
+static inline ufixed_t to_fixed( struct Blip_Buffer *this, clocks_t t )
+{
+        return t * this->factor_ + this->offset_;
+}
+
+static inline delta_t* delta_at( struct Blip_Buffer *this, ufixed_t f )
+{
+        assert( (f >> fixed_bits) < (unsigned) this->buffer_size_ );
+        return this->buffer_center_ + (f >> fixed_bits);
+}
+
+// Number of samples available for reading with read_samples()
+static inline int Blip_samples_avail( struct Blip_Buffer* this )
+{
+        return (int) (this->offset_ >> BLIP_BUFFER_ACCURACY);
+}
+
+static inline void Blip_remove_silence( struct Blip_Buffer* this, int count )
+{
+        assert( count <= Blip_samples_avail( this ) ); // tried to remove more samples than available
+        this->offset_ -= (blip_resampled_time_t) count << BLIP_BUFFER_ACCURACY;
+}
+
+// Initializes Blip_Buffer structure
+void Blip_init( struct Blip_Buffer* this );
+
+// Sets output sample rate and resizes and clears sample buffer
+blargg_err_t Blip_set_sample_rate( struct Blip_Buffer* this, int samples_per_sec, int msec_length );
+
+// Current output sample rate
+static inline int Blip_sample_rate( struct Blip_Buffer* this )
+{
+        return this->sample_rate_;
+}
+
+// Sets number of source time units per second
+blip_resampled_time_t Blip_clock_rate_factor( struct Blip_Buffer* this, int clock_rate );
+static inline void Blip_set_clock_rate( struct Blip_Buffer* this, int clocks_per_sec )
+{
+        this->factor_ = Blip_clock_rate_factor( this, this->clock_rate_ = clocks_per_sec );
+}
+
+// Number of source time units per second
+static inline int Blip_clock_rate( struct Blip_Buffer* this )
+{
+        return this->clock_rate_;
+}
+
+static inline int Blip_length( struct Blip_Buffer* this )
+{
+        return this->length_;
+}
+
+// Clears buffer and removes all samples
+void Blip_clear( struct Blip_Buffer* this );
+
+// Use Blip_Synth to add waveform to buffer
+
+// Resamples to time t, then subtracts t from current time. Appends result of resampling
+// to buffer for reading.
+void Blip_end_frame( struct Blip_Buffer* this, blip_time_t time ) ICODE_ATTR;
+
+
+// Reads at most n samples to out [0 to n-1] and returns number actually read. If stereo
+// is true, writes to out [0], out [2], out [4] etc. instead.
+int Blip_read_samples( struct Blip_Buffer* this, blip_sample_t out [], int n, bool stereo ) ICODE_ATTR;
+
+
+// More features
+
+// Sets flag that tells some Multi_Buffer types that sound was added to buffer,
+// so they know that it needs to be mixed in. Only needs to be called once
+// per time frame that sound was added. Not needed if not using Multi_Buffer.
+static inline void Blip_set_modified( struct Blip_Buffer* this ) { this->modified = true; }
+
+// Set frequency high-pass filter frequency, where higher values reduce bass more
+void Blip_bass_freq( struct Blip_Buffer* this, int frequency );
+
+
+// Low-level features
+
+// Removes the first n samples
+void Blip_remove_samples( struct Blip_Buffer* this, int n ) ICODE_ATTR;
+
+// Returns number of clocks needed until n samples will be available.
+// If buffer cannot even hold n samples, returns number of clocks
+// until buffer becomes full.
+blip_time_t Blip_count_clocks( struct Blip_Buffer* this, int count ) ICODE_ATTR;
+
+// Number of samples that should be mixed before calling Blip_end_frame( t )
+int Blip_count_samples( struct Blip_Buffer* this, blip_time_t t ) ICODE_ATTR;
+
+// Mixes n samples into buffer
+void Blip_mix_samples( struct Blip_Buffer* this, blip_sample_t const in [], int n ) ICODE_ATTR;
+
+
+// Resampled time (sorry, poor documentation right now)
+
+// Resampled time is fixed-point, in terms of output samples.
+
+// Converts clock count to resampled time
+static inline blip_resampled_time_t Blip_resampled_duration( struct Blip_Buffer* this, int t )
+{
+        return t * this->factor_;
+}
+
+// Converts clock time since beginning of current time frame to resampled time
+static inline blip_resampled_time_t Blip_resampled_time( struct Blip_Buffer* this, blip_time_t t )
+{
+        return t * this->factor_ + this->offset_;
+}
+
+
+// Range specifies the greatest expected change in amplitude. Calculate it
+// by finding the difference between the maximum and minimum expected
+// amplitudes (max - min).
+
+typedef char coeff_t;
+
+struct Blip_Synth {
+        int delta_factor;
+        int last_amp;
+        struct Blip_Buffer* buf;
+};
+
+// Blip_Synth_
+void volume_unit( struct Blip_Synth* this, int new_unit );
+
+// Initializes Blip_Synth structure
+void Synth_init( struct Blip_Synth* this );
+
+// Sets volume of amplitude delta unit
+static inline void Synth_volume( struct Blip_Synth* this, int v )
+{
+        volume_unit( this, v ); // new_unit = 1 / range * v
+}
+
+
+// Low-level interface
+
+// (in >> sh & mask) * mul
+#define BLIP_SH_AND_MUL( in, sh, mask, mul ) \
+((int) (in) / ((1U << (sh)) / (mul)) & (unsigned) ((mask) * (mul)))
+
+// (T*) ptr + (off >> sh)
+#define BLIP_PTR_OFF_SH( T, ptr, off, sh ) \
+        ((T*) (BLIP_SH_AND_MUL( off, sh, -1, sizeof (T) ) + (char*) (ptr)))
+
+// Works directly in terms of fractional output samples. Use resampled time functions in Blip_Buffer
+// to convert clock counts to resampled time.
+static inline void Synth_offset_resampled( struct Blip_Synth* this, blip_resampled_time_t time,
+        int delta, struct Blip_Buffer* blip_buf )
+{
+        int const half_width = 1;
+
+        delta_t* BLARGG_RESTRICT buf = delta_at( blip_buf, time );
+        delta *= this->delta_factor;
+
+        int const phase_shift = BLIP_BUFFER_ACCURACY - BLIP_PHASE_BITS;
+        int const phase = (half_width & (half_width - 1)) ?
+                (int) BLIP_SH_AND_MUL( time, phase_shift, blip_res - 1, sizeof (coeff_t) ) * half_width :
+                (int) BLIP_SH_AND_MUL( time, phase_shift, blip_res - 1, sizeof (coeff_t) * half_width );
+
+        int left = buf [0] + delta;
+
+        // Kind of crappy, but doing shift after multiply results in overflow.
+        // Alternate way of delaying multiply by delta_factor results in worse
+        // sub-sample resolution.
+        int right = (delta >> BLIP_PHASE_BITS) * phase;
+        #ifdef BLIP_BUFFER_NOINTERP
+                // TODO: remove? (just a hack to see how it sounds)
+                right = 0;
+        #endif
+        left  -= right;
+        right += buf [1];
+
+        buf [0] = left;
+        buf [1] = right;
+}
+
+// Update amplitude of waveform at given time. Using this requires a separate
+// Blip_Synth for each waveform.
+static inline void Synth_update( struct Blip_Synth* this, blip_time_t t, int amp )
+{
+        int delta = amp - this->last_amp;
+        this->last_amp = amp;
+        Synth_offset_resampled( this, to_fixed(this->buf, t), delta, this->buf );
+}
+
+// Adds amplitude transition at time t. Delta can be positive or negative.
+// The actual change in amplitude is delta * volume.
+static inline void Synth_offset_inline( struct Blip_Synth* this, blip_time_t t, int delta, struct Blip_Buffer* buf )
+{
+        Synth_offset_resampled( this, to_fixed(buf, t), delta, buf );
+}
+
+#define Synth_offset( synth, time, delta, buf ) Synth_offset_inline( synth, time, delta, buf )
+
+// Number of bits in raw sample that covers normal output range. Less than 32 bits to give
+// extra amplitude range. That is,
+// +1 << (blip_sample_bits-1) = +1.0
+// -1 << (blip_sample_bits-1) = -1.0
+static int const blip_sample_bits = 30;
+
+// Optimized reading from Blip_Buffer, for use in custom sample output
+
+// Begin reading from buffer. Name should be unique to the current block.
+#define BLIP_READER_BEGIN( name, blip_buffer ) \
+        const delta_t* BLARGG_RESTRICT name##_reader_buf = (blip_buffer).buffer_;\
+        int name##_reader_accum = (blip_buffer).reader_accum_
+
+// Get value to pass to BLIP_READER_NEXT()
+#define BLIP_READER_BASS( blip_buffer ) ((blip_buffer).bass_shift_)
+
+// Constant value to use instead of BLIP_READER_BASS(), for slightly more optimal
+// code at the cost of having no bass_freq() functionality
+static int const blip_reader_default_bass = 9;
+
+// Current sample
+#define BLIP_READER_READ( name )        (name##_reader_accum >> (blip_sample_bits - 16))
+
+// Current raw sample in full internal resolution
+#define BLIP_READER_READ_RAW( name )    (name##_reader_accum)
+
+// Advance to next sample
+#define BLIP_READER_NEXT( name, bass ) \
+        (void) (name##_reader_accum += *name##_reader_buf++ - (name##_reader_accum >> (bass)))
+
+// End reading samples from buffer. The number of samples read must now be removed
+// using Blip_remove_samples().
+#define BLIP_READER_END( name, blip_buffer ) \
+        (void) ((blip_buffer).reader_accum_ = name##_reader_accum)
+
+#define BLIP_READER_ADJ_( name, offset ) (name##_reader_buf += offset)
+
+#define BLIP_READER_NEXT_IDX_( name, bass, idx ) {\
+        name##_reader_accum -= name##_reader_accum >> (bass);\
+        name##_reader_accum += name##_reader_buf [(idx)];\
+}
+
+#define BLIP_READER_NEXT_RAW_IDX_( name, bass, idx ) {\
+        name##_reader_accum -= name##_reader_accum >> (bass);\
+        name##_reader_accum +=\
+        *(delta_t const*) ((char const*) name##_reader_buf + (idx));\
+}
+
+//// BLIP_CLAMP
+
+#if defined(CPU_ARM) && (ARM_ARCH >= 6)
+        #define BLIP_CLAMP( sample, out ) \
+                ({ \
+                asm ("ssat %0, #16, %1" \
+                                : "=r" ( out ) : "r"( sample ) ); \
+                out; \
+                })
+#else
+        #if defined (_M_IX86) || defined (_M_IA64) || defined (__i486__) || \
+                        defined (__x86_64__) || defined (__ia64__) || defined (__i386__)
+                #define BLIP_X86 1
+                #define BLIP_CLAMP_( in ) in < -0x8000 || 0x7FFF < in
+        #else
+                #define BLIP_CLAMP_( in ) (blip_sample_t) in != in
+        #endif
+
+        // Clamp sample to blip_sample_t range
+        #define BLIP_CLAMP( sample, out )\
+                { if ( BLIP_CLAMP_( (sample) ) ) (out) = ((sample) >> 31) ^ 0x7FFF; }
+#endif
+
+#endif