gpsp/sound.c

820 lines
31 KiB
C

/* gameplaySP
*
* Copyright (C) 2006 Exophase <exophase@gmail.com>
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License as
* published by the Free Software Foundation; either version 2 of
* the License, or (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include "common.h"
#include <SDL.h>
u32 global_enable_audio = 1;
direct_sound_struct direct_sound_channel[2];
gbc_sound_struct gbc_sound_channel[4];
#ifdef RPI_BUILD
u32 sound_frequency = 22050;
#else
u32 sound_frequency = 44100;
#endif
SDL_mutex *sound_mutex;
static SDL_cond *sound_cv;
#ifdef PSP_BUILD
u32 audio_buffer_size_number = 1;
#elif defined(POLLUX_BUILD)
u32 audio_buffer_size_number = 7;
#else
u32 audio_buffer_size_number = 8;
#endif
u32 sound_on;
static u32 audio_buffer_size;
static s16 sound_buffer[BUFFER_SIZE];
static u32 sound_buffer_base;
static u32 sound_last_cpu_ticks;
static fixed16_16 gbc_sound_tick_step;
static u32 sound_exit_flag;
// Queue 1, 2, or 4 samples to the top of the DS FIFO, wrap around circularly
#define sound_timer_queue(size, value) \
*((s##size *)(ds->fifo + ds->fifo_top)) = value; \
ds->fifo_top = (ds->fifo_top + 1) % 32; \
void sound_timer_queue8(u32 channel, u8 value)
{
direct_sound_struct *ds = direct_sound_channel + channel;
sound_timer_queue(8, value);
}
void sound_timer_queue16(u32 channel, u16 value)
{
direct_sound_struct *ds = direct_sound_channel + channel;
sound_timer_queue(8, value & 0xFF);
sound_timer_queue(8, value >> 8);
}
void sound_timer_queue32(u32 channel, u32 value)
{
direct_sound_struct *ds = direct_sound_channel + channel;
sound_timer_queue(8, value & 0xFF);
sound_timer_queue(8, (value >> 8) & 0xFF);
sound_timer_queue(8, (value >> 16) & 0xFF);
sound_timer_queue(8, value >> 24);
}
// Unqueue 1 sample from the base of the DS FIFO and place it on the audio
// buffer for as many samples as necessary. If the DS FIFO is 16 bytes or
// smaller and if DMA is enabled for the sound channel initiate a DMA transfer
// to the DS FIFO.
#define render_sample_null() \
#define render_sample_left() \
sound_buffer[buffer_index] += current_sample + \
fp16_16_to_u32((next_sample - current_sample) * (fifo_fractional >> 8)) \
#define render_sample_right() \
sound_buffer[buffer_index + 1] += current_sample + \
fp16_16_to_u32((next_sample - current_sample) * (fifo_fractional >> 8)) \
#define render_sample_both() \
dest_sample = current_sample + \
fp16_16_to_u32((next_sample - current_sample) * (fifo_fractional >> 8)); \
sound_buffer[buffer_index] += dest_sample; \
sound_buffer[buffer_index + 1] += dest_sample \
#define render_samples(type) \
while(fifo_fractional <= 0xFFFFFF) \
{ \
render_sample_##type(); \
fifo_fractional += frequency_step; \
buffer_index = (buffer_index + 2) % BUFFER_SIZE; \
} \
void sound_timer(fixed8_24 frequency_step, u32 channel)
{
direct_sound_struct *ds = direct_sound_channel + channel;
fixed8_24 fifo_fractional = ds->fifo_fractional;
u32 buffer_index = ds->buffer_index;
s16 current_sample, next_sample, dest_sample;
current_sample = ds->fifo[ds->fifo_base] << 4;
ds->fifo_base = (ds->fifo_base + 1) % 32;
next_sample = ds->fifo[ds->fifo_base] << 4;
if(sound_on == 1)
{
if(ds->volume == DIRECT_SOUND_VOLUME_50)
{
current_sample >>= 1;
next_sample >>= 1;
}
switch(ds->status)
{
case DIRECT_SOUND_INACTIVE:
render_samples(null);
break;
case DIRECT_SOUND_RIGHT:
render_samples(right);
break;
case DIRECT_SOUND_LEFT:
render_samples(left);
break;
case DIRECT_SOUND_LEFTRIGHT:
render_samples(both);
break;
}
}
else
{
render_samples(null);
}
ds->buffer_index = buffer_index;
ds->fifo_fractional = fp8_24_fractional_part(fifo_fractional);
if(((ds->fifo_top - ds->fifo_base) % 32) <= 16)
{
if(dma[1].direct_sound_channel == channel)
dma_transfer(dma + 1);
if(dma[2].direct_sound_channel == channel)
dma_transfer(dma + 2);
}
}
void sound_reset_fifo(u32 channel)
{
direct_sound_struct *ds = direct_sound_channel;
memset(ds->fifo, 0, 32);
}
// Initial pattern data = 4bits (signed)
// Channel volume = 12bits
// Envelope volume = 14bits
// Master volume = 2bits
// Recalculate left and right volume as volume changes.
// To calculate the current sample, use (sample * volume) >> 16
// Square waves range from -8 (low) to 7 (high)
s8 square_pattern_duty[4][8] =
{
{ 0xF8, 0xF8, 0xF8, 0xF8, 0x07, 0xF8, 0xF8, 0xF8 },
{ 0xF8, 0xF8, 0xF8, 0xF8, 0x07, 0x07, 0xF8, 0xF8 },
{ 0xF8, 0xF8, 0x07, 0x07, 0x07, 0x07, 0xF8, 0xF8 },
{ 0x07, 0x07, 0x07, 0x07, 0xF8, 0xF8, 0x07, 0x07 },
};
s8 wave_samples[64];
u32 noise_table15[1024];
u32 noise_table7[4];
u32 gbc_sound_master_volume_table[4] = { 1, 2, 4, 0 };
u32 gbc_sound_channel_volume_table[8] =
{
fixed_div(0, 7, 12),
fixed_div(1, 7, 12),
fixed_div(2, 7, 12),
fixed_div(3, 7, 12),
fixed_div(4, 7, 12),
fixed_div(5, 7, 12),
fixed_div(6, 7, 12),
fixed_div(7, 7, 12)
};
u32 gbc_sound_envelope_volume_table[16] =
{
fixed_div(0, 15, 14),
fixed_div(1, 15, 14),
fixed_div(2, 15, 14),
fixed_div(3, 15, 14),
fixed_div(4, 15, 14),
fixed_div(5, 15, 14),
fixed_div(6, 15, 14),
fixed_div(7, 15, 14),
fixed_div(8, 15, 14),
fixed_div(9, 15, 14),
fixed_div(10, 15, 14),
fixed_div(11, 15, 14),
fixed_div(12, 15, 14),
fixed_div(13, 15, 14),
fixed_div(14, 15, 14),
fixed_div(15, 15, 14)
};
u32 gbc_sound_buffer_index = 0;
u32 gbc_sound_last_cpu_ticks = 0;
u32 gbc_sound_partial_ticks = 0;
u32 gbc_sound_master_volume_left;
u32 gbc_sound_master_volume_right;
u32 gbc_sound_master_volume;
#define update_volume_channel_envelope(channel) \
volume_##channel = gbc_sound_envelope_volume_table[envelope_volume] * \
gbc_sound_channel_volume_table[gbc_sound_master_volume_##channel] * \
gbc_sound_master_volume_table[gbc_sound_master_volume] \
#define update_volume_channel_noenvelope(channel) \
volume_##channel = gs->wave_volume * \
gbc_sound_channel_volume_table[gbc_sound_master_volume_##channel] * \
gbc_sound_master_volume_table[gbc_sound_master_volume] \
#define update_volume(type) \
update_volume_channel_##type(left); \
update_volume_channel_##type(right) \
#define update_tone_sweep() \
if(gs->sweep_status) \
{ \
u32 sweep_ticks = gs->sweep_ticks - 1; \
\
if(sweep_ticks == 0) \
{ \
u32 rate = gs->rate; \
\
if(gs->sweep_direction) \
rate = rate - (rate >> gs->sweep_shift); \
else \
rate = rate + (rate >> gs->sweep_shift); \
\
if(rate > 2048) \
rate = 2048; \
\
frequency_step = float_to_fp16_16(((131072.0f / (2048 - rate)) * 8.0f) \
/ sound_frequency); \
\
gs->frequency_step = frequency_step; \
gs->rate = rate; \
\
sweep_ticks = gs->sweep_initial_ticks; \
} \
gs->sweep_ticks = sweep_ticks; \
} \
#define update_tone_nosweep() \
#define update_tone_envelope() \
if(gs->envelope_status) \
{ \
u32 envelope_ticks = gs->envelope_ticks - 1; \
envelope_volume = gs->envelope_volume; \
\
if(envelope_ticks == 0) \
{ \
if(gs->envelope_direction) \
{ \
if(envelope_volume != 15) \
envelope_volume = gs->envelope_volume + 1; \
} \
else \
{ \
if(envelope_volume != 0) \
envelope_volume = gs->envelope_volume - 1; \
} \
\
update_volume(envelope); \
\
gs->envelope_volume = envelope_volume; \
gs->envelope_ticks = gs->envelope_initial_ticks; \
} \
else \
{ \
gs->envelope_ticks = envelope_ticks; \
} \
} \
#define update_tone_noenvelope() \
#define update_tone_counters(envelope_op, sweep_op) \
tick_counter += gbc_sound_tick_step; \
if(tick_counter > 0xFFFF) \
{ \
if(gs->length_status) \
{ \
u32 length_ticks = gs->length_ticks - 1; \
gs->length_ticks = length_ticks; \
\
if(length_ticks == 0) \
{ \
gs->active_flag = 0; \
break; \
} \
} \
\
update_tone_##envelope_op(); \
update_tone_##sweep_op(); \
\
tick_counter &= 0xFFFF; \
} \
#define gbc_sound_render_sample_right() \
sound_buffer[buffer_index + 1] += (current_sample * volume_right) >> 22 \
#define gbc_sound_render_sample_left() \
sound_buffer[buffer_index] += (current_sample * volume_left) >> 22 \
#define gbc_sound_render_sample_both() \
gbc_sound_render_sample_right(); \
gbc_sound_render_sample_left() \
#define gbc_sound_render_samples(type, sample_length, envelope_op, sweep_op) \
for(i = 0; i < buffer_ticks; i++) \
{ \
current_sample = \
sample_data[fp16_16_to_u32(sample_index) % sample_length]; \
gbc_sound_render_sample_##type(); \
\
sample_index += frequency_step; \
buffer_index = (buffer_index + 2) % BUFFER_SIZE; \
\
update_tone_counters(envelope_op, sweep_op); \
} \
#define gbc_noise_wrap_full 32767
#define gbc_noise_wrap_half 126
#define get_noise_sample_full() \
current_sample = \
((s32)(noise_table15[fp16_16_to_u32(sample_index) >> 5] << \
(fp16_16_to_u32(sample_index) & 0x1F)) >> 31) & 0x0F \
#define get_noise_sample_half() \
current_sample = \
((s32)(noise_table7[fp16_16_to_u32(sample_index) >> 5] << \
(fp16_16_to_u32(sample_index) & 0x1F)) >> 31) & 0x0F \
#define gbc_sound_render_noise(type, noise_type, envelope_op, sweep_op) \
for(i = 0; i < buffer_ticks; i++) \
{ \
get_noise_sample_##noise_type(); \
gbc_sound_render_sample_##type(); \
\
sample_index += frequency_step; \
\
if(sample_index >= u32_to_fp16_16(gbc_noise_wrap_##noise_type)) \
sample_index -= u32_to_fp16_16(gbc_noise_wrap_##noise_type); \
\
buffer_index = (buffer_index + 2) % BUFFER_SIZE; \
update_tone_counters(envelope_op, sweep_op); \
} \
#define gbc_sound_render_channel(type, sample_length, envelope_op, sweep_op) \
buffer_index = gbc_sound_buffer_index; \
sample_index = gs->sample_index; \
frequency_step = gs->frequency_step; \
tick_counter = gs->tick_counter; \
\
update_volume(envelope_op); \
\
switch(gs->status) \
{ \
case GBC_SOUND_INACTIVE: \
break; \
\
case GBC_SOUND_LEFT: \
gbc_sound_render_##type(left, sample_length, envelope_op, sweep_op); \
break; \
\
case GBC_SOUND_RIGHT: \
gbc_sound_render_##type(right, sample_length, envelope_op, sweep_op); \
break; \
\
case GBC_SOUND_LEFTRIGHT: \
gbc_sound_render_##type(both, sample_length, envelope_op, sweep_op); \
break; \
} \
\
gs->sample_index = sample_index; \
gs->tick_counter = tick_counter; \
#define gbc_sound_load_wave_ram(bank) \
wave_bank = wave_samples + (bank * 32); \
for(i = 0, i2 = 0; i < 16; i++, i2 += 2) \
{ \
current_sample = wave_ram[i]; \
wave_bank[i2] = (((current_sample >> 4) & 0x0F) - 8); \
wave_bank[i2 + 1] = ((current_sample & 0x0F) - 8); \
} \
void update_gbc_sound(u32 cpu_ticks)
{
fixed16_16 buffer_ticks = float_to_fp16_16((float)(cpu_ticks -
gbc_sound_last_cpu_ticks) * sound_frequency / GBC_BASE_RATE);
u32 i, i2;
gbc_sound_struct *gs = gbc_sound_channel;
fixed16_16 sample_index, frequency_step;
fixed16_16 tick_counter;
u32 buffer_index;
s32 volume_left, volume_right;
u32 envelope_volume;
s32 current_sample;
u32 sound_status = address16(io_registers, 0x84) & 0xFFF0;
s8 *sample_data;
s8 *wave_bank;
u8 *wave_ram = ((u8 *)io_registers) + 0x90;
gbc_sound_partial_ticks += fp16_16_fractional_part(buffer_ticks);
buffer_ticks = fp16_16_to_u32(buffer_ticks);
if(gbc_sound_partial_ticks > 0xFFFF)
{
buffer_ticks += 1;
gbc_sound_partial_ticks &= 0xFFFF;
}
SDL_LockMutex(sound_mutex);
if(synchronize_flag)
{
if(((gbc_sound_buffer_index - sound_buffer_base) % BUFFER_SIZE) >=
(audio_buffer_size * 2))
{
while(((gbc_sound_buffer_index - sound_buffer_base) % BUFFER_SIZE) >
(audio_buffer_size * 3 / 2))
{
SDL_CondWait(sound_cv, sound_mutex);
}
#ifdef PSP_BUILD
if(current_frameskip_type == auto_frameskip)
{
sceDisplayWaitVblankStart();
real_frame_count = 0;
virtual_frame_count = 0;
}
#else
if(current_frameskip_type == auto_frameskip)
{
/*
u64 current_ticks;
u64 next_ticks;
get_ticks_us(&current_ticks);
next_ticks = ((current_ticks + 16666) / 16667) * 16667;
delay_us(next_ticks - current_ticks);
get_ticks_us(&frame_count_initial_timestamp);
*/
/* prevent frameskip, or it will cause more audio,
* then more waiting here, then frame skip again, ... */
num_skipped_frames = 100;
}
#endif
}
}
if(sound_on == 1)
{
gs = gbc_sound_channel + 0;
if(gs->active_flag)
{
sound_status |= 0x01;
sample_data = gs->sample_data;
envelope_volume = gs->envelope_volume;
gbc_sound_render_channel(samples, 8, envelope, sweep);
}
gs = gbc_sound_channel + 1;
if(gs->active_flag)
{
sound_status |= 0x02;
sample_data = gs->sample_data;
envelope_volume = gs->envelope_volume;
gbc_sound_render_channel(samples, 8, envelope, nosweep);
}
gs = gbc_sound_channel + 2;
if(gbc_sound_wave_update)
{
if(gs->wave_bank == 1)
{
gbc_sound_load_wave_ram(1);
}
else
{
gbc_sound_load_wave_ram(0);
}
gbc_sound_wave_update = 0;
}
if((gs->active_flag) && (gs->master_enable))
{
sound_status |= 0x04;
sample_data = wave_samples;
if(gs->wave_type == 0)
{
if(gs->wave_bank == 1)
sample_data += 32;
gbc_sound_render_channel(samples, 32, noenvelope, nosweep);
}
else
{
gbc_sound_render_channel(samples, 64, noenvelope, nosweep);
}
}
gs = gbc_sound_channel + 3;
if(gs->active_flag)
{
sound_status |= 0x08;
envelope_volume = gs->envelope_volume;
if(gs->noise_type == 1)
{
gbc_sound_render_channel(noise, half, envelope, nosweep);
}
else
{
gbc_sound_render_channel(noise, full, envelope, nosweep);
}
}
}
address16(io_registers, 0x84) = sound_status;
gbc_sound_last_cpu_ticks = cpu_ticks;
gbc_sound_buffer_index =
(gbc_sound_buffer_index + (buffer_ticks * 2)) % BUFFER_SIZE;
SDL_UnlockMutex(sound_mutex);
SDL_CondSignal(sound_cv);
}
#define sound_copy_normal() \
current_sample = source[i] \
#define sound_copy(source_offset, length, render_type) \
_length = (length) / 2; \
source = (s16 *)(sound_buffer + source_offset); \
for(i = 0; i < _length; i++) \
{ \
sound_copy_##render_type(); \
if(current_sample > 2047) \
current_sample = 2047; \
if(current_sample < -2048) \
current_sample = -2048; \
\
stream_base[i] = current_sample << 4; \
source[i] = 0; \
} \
#define sound_copy_null(source_offset, length) \
_length = (length) / 2; \
source = (s16 *)(sound_buffer + source_offset); \
for(i = 0; i < _length; i++) \
{ \
stream_base[i] = 0; \
source[i] = 0; \
} \
void sound_callback(void *userdata, Uint8 *stream, int length)
{
u32 sample_length = length / 2;
u32 _length;
u32 i;
s16 *stream_base = (s16 *)stream;
s16 *source;
s32 current_sample;
SDL_LockMutex(sound_mutex);
while(((gbc_sound_buffer_index - sound_buffer_base) % BUFFER_SIZE) <
length && !sound_exit_flag)
{
SDL_CondWait(sound_cv, sound_mutex);
}
if(global_enable_audio)
{
if((sound_buffer_base + sample_length) >= BUFFER_SIZE)
{
u32 partial_length = (BUFFER_SIZE - sound_buffer_base) * 2;
sound_copy(sound_buffer_base, partial_length, normal);
source = (s16 *)sound_buffer;
sound_copy(0, length - partial_length, normal);
sound_buffer_base = (length - partial_length) / 2;
}
else
{
sound_copy(sound_buffer_base, length, normal);
sound_buffer_base += sample_length;
}
}
else
{
if((sound_buffer_base + sample_length) >= BUFFER_SIZE)
{
u32 partial_length = (BUFFER_SIZE - sound_buffer_base) * 2;
sound_copy_null(sound_buffer_base, partial_length);
source = (s16 *)sound_buffer;
sound_copy(0, length - partial_length, normal);
sound_buffer_base = (length - partial_length) / 2;
}
else
{
sound_copy_null(sound_buffer_base, length);
sound_buffer_base += sample_length;
}
}
SDL_CondSignal(sound_cv);
SDL_UnlockMutex(sound_mutex);
}
// Special thanks to blarrg for the LSFR frequency used in Meridian, as posted
// on the forum at http://meridian.overclocked.org:
// http://meridian.overclocked.org/cgi-bin/wwwthreads/showpost.pl?Board=merid
// angeneraldiscussion&Number=2069&page=0&view=expanded&mode=threaded&sb=4
// Hope you don't mind me borrowing it ^_-
void init_noise_table(u32 *table, u32 period, u32 bit_length)
{
u32 shift_register = 0xFF;
u32 mask = ~(1 << bit_length);
s32 table_pos, bit_pos;
u32 current_entry;
u32 table_period = (period + 31) / 32;
// Bits are stored in reverse order so they can be more easily moved to
// bit 31, for sign extended shift down.
for(table_pos = 0; table_pos < table_period; table_pos++)
{
current_entry = 0;
for(bit_pos = 31; bit_pos >= 0; bit_pos--)
{
current_entry |= (shift_register & 0x01) << bit_pos;
shift_register =
((1 & (shift_register ^ (shift_register >> 1))) << bit_length) |
((shift_register >> 1) & mask);
}
table[table_pos] = current_entry;
}
}
void reset_sound()
{
direct_sound_struct *ds = direct_sound_channel;
gbc_sound_struct *gs = gbc_sound_channel;
u32 i;
SDL_LockMutex(sound_mutex);
sound_on = 0;
sound_buffer_base = 0;
sound_last_cpu_ticks = 0;
memset(sound_buffer, 0, sizeof(sound_buffer));
for(i = 0; i < 2; i++, ds++)
{
ds->buffer_index = 0;
ds->status = DIRECT_SOUND_INACTIVE;
ds->fifo_top = 0;
ds->fifo_base = 0;
ds->fifo_fractional = 0;
ds->last_cpu_ticks = 0;
memset(ds->fifo, 0, 32);
}
gbc_sound_buffer_index = 0;
gbc_sound_last_cpu_ticks = 0;
gbc_sound_partial_ticks = 0;
gbc_sound_master_volume_left = 0;
gbc_sound_master_volume_right = 0;
gbc_sound_master_volume = 0;
memset(wave_samples, 0, 64);
for(i = 0; i < 4; i++, gs++)
{
gs->status = GBC_SOUND_INACTIVE;
gs->sample_data = square_pattern_duty[2];
gs->active_flag = 0;
}
SDL_UnlockMutex(sound_mutex);
}
void sound_exit()
{
gbc_sound_buffer_index =
(sound_buffer_base + audio_buffer_size) % BUFFER_SIZE;
SDL_PauseAudio(1);
sound_exit_flag = 1;
SDL_CondSignal(sound_cv);
SDL_CloseAudio();
SDL_Delay(200);
SDL_DestroyMutex(sound_mutex);
sound_mutex = NULL;
SDL_DestroyCond(sound_cv);
sound_cv = NULL;
}
void init_sound(int need_reset)
{
SDL_AudioSpec sound_settings;
sound_exit_flag = 0;
#ifdef PSP_BUILD
audio_buffer_size = (audio_buffer_size_number * 1024) + 3072;
#else
audio_buffer_size = 16 << audio_buffer_size_number;
// audio_buffer_size = 16384;
#endif
SDL_AudioSpec desired_spec =
{
sound_frequency,
AUDIO_S16,
2,
0,
audio_buffer_size / 4,
0,
0,
sound_callback,
NULL
};
sound_mutex = SDL_CreateMutex();
sound_cv = SDL_CreateCond();
SDL_OpenAudio(&desired_spec, &sound_settings);
sound_frequency = sound_settings.freq;
audio_buffer_size = sound_settings.size;
u32 i = audio_buffer_size / 16;
for (audio_buffer_size_number = 0; i && (i & 1) == 0; i >>= 1)
audio_buffer_size_number++;
#ifndef PSP_BUILD
printf("audio: freq %d, size %d\n", sound_frequency, audio_buffer_size);
#endif
gbc_sound_tick_step =
float_to_fp16_16(256.0f / sound_frequency);
init_noise_table(noise_table15, 32767, 14);
init_noise_table(noise_table7, 127, 6);
if (need_reset)
reset_sound();
SDL_PauseAudio(0);
}
#define sound_savestate_builder(type) \
void sound_##type##_savestate(file_tag_type savestate_file) \
{ \
file_##type##_variable(savestate_file, sound_on); \
file_##type##_variable(savestate_file, sound_buffer_base); \
file_##type##_variable(savestate_file, sound_last_cpu_ticks); \
file_##type##_variable(savestate_file, gbc_sound_buffer_index); \
file_##type##_variable(savestate_file, gbc_sound_last_cpu_ticks); \
file_##type##_variable(savestate_file, gbc_sound_partial_ticks); \
file_##type##_variable(savestate_file, gbc_sound_master_volume_left); \
file_##type##_variable(savestate_file, gbc_sound_master_volume_right); \
file_##type##_variable(savestate_file, gbc_sound_master_volume); \
file_##type##_array(savestate_file, wave_samples); \
file_##type##_array(savestate_file, direct_sound_channel); \
file_##type##_array(savestate_file, gbc_sound_channel); \
} \
sound_savestate_builder(read);
sound_savestate_builder(write_mem);