FFmpeg
|
Reference: libavcodec/mpegaudiodsp_template.c libavcodec/dct32.c. More...
Go to the source code of this file.
Functions | |
static void | ff_mpadsp_apply_window_mips_float (float *synth_buf, float *window, int *dither_state, float *samples, int incr) |
static void | ff_dct32_mips_float (float *out, const float *tab) |
static void | imdct36_mips_float (float *out, float *buf, float *in, float *win) |
static void | ff_imdct36_blocks_mips_float (float *out, float *buf, float *in, int count, int switch_point, int block_type) |
void | ff_mpadsp_init_mipsfpu (MPADSPContext *s) |
Reference: libavcodec/mpegaudiodsp_template.c libavcodec/dct32.c.
Definition in file mpegaudiodsp_mips_float.c.
|
static |
instructions are scheduled to minimize pipeline stall. use of round_sample function from the original code is changed with appropriate assembly instructions.
Definition at line 60 of file mpegaudiodsp_mips_float.c.
Referenced by ff_mpadsp_init_mipsfpu().
|
static |
instructions are scheduled to minimize pipeline stall.
Definition at line 279 of file mpegaudiodsp_mips_float.c.
Referenced by ff_mpadsp_init_mipsfpu().
|
static |
all loops are unrolled totally, and instructions are scheduled to minimize pipeline stall. instructions of the first two loops are reorganized, in order to eliminate unnecessary readings and writings into array. values defined in macros and tables are eliminated - they are directly loaded in appropriate variables
Definition at line 784 of file mpegaudiodsp_mips_float.c.
Referenced by ff_imdct36_blocks_mips_float().
|
static |
Definition at line 1218 of file mpegaudiodsp_mips_float.c.
Referenced by ff_mpadsp_init_mipsfpu().
void ff_mpadsp_init_mipsfpu | ( | MPADSPContext * | s | ) |
Definition at line 1237 of file mpegaudiodsp_mips_float.c.
Referenced by ff_mpadsp_init().