FreeCalypso > hg > gsm-codec-lib
view libgsmhr1/mathdp31.c @ 581:e2d5cad04cbf
libgsmhr1 RxFE: store CN R0+LPC separately from speech
In the original GSM 06.06 code the ECU for speech mode is entirely
separate from the CN generator, maintaining separate state. (The
main intertie between them is the speech vs CN state variable,
distinguishing between speech and CN BFIs, in addition to the
CN-specific function of distinguishing between initial and update
SIDs.)
In the present RxFE implementation I initially thought that we could
use the same saved_frame buffer for both ECU and CN, overwriting
just the first 4 params (R0 and LPC) when a valid SID comes in.
However, I now realize it was a bad idea: the original code has a
corner case (long sequence of speech-mode BFIs to put the ECU in
state 6, then SID and CN-mode BFIs, then a good speech frame) that
would be broken by that buffer reuse approach. We could eliminate
this corner case by resetting the ECU state when passing through
a CN insertion period, but doing so would needlessly increase
the behavioral diffs between GSM 06.06 and our version.
Solution: use a separate CN-specific buffer for CN R0+LPC parameters,
and match the behavior of GSM 06.06 code in this regard.
author | Mychaela Falconia <falcon@freecalypso.org> |
---|---|
date | Thu, 13 Feb 2025 10:02:45 +0000 |
parents | 024615de06fe |
children |
line wrap: on
line source
/*************************************************************************** * * File Name: mathdp31.c * * Purpose: Contains functions increased-precision arithmetic operations. * * Below is a listing of all the functions in this file. There * is no interdependence among the functions. * * L_mpy_ls() * L_mpy_ll() * isLwLimit() * isSwLimit() * ***************************************************************************/ /*_________________________________________________________________________ | | | Include Files | |_________________________________________________________________________| */ #include "typedefs.h" #include "namespace.h" #include "mathhalf.h" #include "mathdp31.h" /**************************************************************************** * * FUNCTION NAME: isLwLimit * * PURPOSE: * * Check to see if the input Longword is at the * upper or lower limit of its range. i.e. * 0x7fff ffff or -0x8000 0000 * * Ostensibly this is a check for an overflow. * This does not truly mean an overflow occurred, * it means the value reached is the * maximum/minimum value representable. It may * have come about due to an overflow. * * INPUTS: * * L_In A Longword input variable * * * OUTPUTS: none * * RETURN VALUE: 1 if input == 0x7fff ffff or -0x8000 0000 * 0 otherwise * * KEYWORDS: saturation, limit * ***************************************************************************/ short isLwLimit(Longword L_In) { Longword L_ls; short siOut; if (L_In != 0) { L_ls = L_shl(L_In, 1); if (L_sub(L_In, L_ls) == 0) siOut = 1; else siOut = 0; } else { siOut = 0; } return (siOut); } /**************************************************************************** * * FUNCTION NAME: isSwLimit * * PURPOSE: * * Check to see if the input Shortword is at the * upper or lower limit of its range. i.e. * 0x7fff or -0x8000 * * Ostensibly this is a check for an overflow. * This does not truly mean an overflow occurred, * it means the value reached is the * maximum/minimum value representable. It may * have come about due to an overflow. * * INPUTS: * * swIn A Shortword input variable * * * OUTPUTS: none * * RETURN VALUE: 1 if input == 0x7fff or -0x8000 * 0 otherwise * * KEYWORDS: saturation, limit * ***************************************************************************/ short isSwLimit(Shortword swIn) { Shortword swls; short siOut; if (swIn != 0) { swls = shl(swIn, 1); if (sub(swIn, swls) == 0) /* logical compare outputs 1/0 */ siOut = 1; else siOut = 0; } else { siOut = 0; } return (siOut); } /**************************************************************************** * * FUNCTION NAME: L_mpy_ll * * PURPOSE: Multiply a 32 bit number (L_var1) and a 32 bit number * (L_var2), and return a 32 bit result. * * INPUTS: * * L_var1 A Longword input variable * * L_var2 A Longword input variable * * OUTPUTS: none * * IMPLEMENTATION: * * Performs a 31x31 bit multiply, Complexity=24 Ops. * * Let x1x0, or y1y0, be the two constituent halves * of a 32 bit number. This function performs the * following: * * low = ((x0 >> 1)*(y0 >> 1)) >> 16 (low * low) * mid1 = [(x1 * (y0 >> 1)) >> 1 ] (high * low) * mid2 = [(y1 * (x0 >> 1)) >> 1] (high * low) * mid = (mid1 + low + mid2) >> 14 (sum so far) * output = (y1*x1) + mid (high * high) * * * RETURN VALUE: A Longword value * * KEYWORDS: mult,mpy,multiplication * ***************************************************************************/ Longword L_mpy_ll(Longword L_var1, Longword L_var2) { Shortword swLow1, swLow2, swHigh1, swHigh2; Longword L_varOut, L_low, L_mid1, L_mid2, L_mid; swLow1 = shr(extract_l(L_var1), 1); swLow1 = SW_MAX & swLow1; swLow2 = shr(extract_l(L_var2), 1); swLow2 = SW_MAX & swLow2; swHigh1 = extract_h(L_var1); swHigh2 = extract_h(L_var2); L_low = L_mult(swLow1, swLow2); L_low = L_shr(L_low, 16); L_mid1 = L_mult(swHigh1, swLow2); L_mid1 = L_shr(L_mid1, 1); L_mid = L_add(L_mid1, L_low); L_mid2 = L_mult(swHigh2, swLow1); L_mid2 = L_shr(L_mid2, 1); L_mid = L_add(L_mid, L_mid2); L_mid = L_shr(L_mid, 14); L_varOut = L_mac(L_mid, swHigh1, swHigh2); return (L_varOut); } /**************************************************************************** * * FUNCTION NAME: L_mpy_ls * * PURPOSE: Multiply a 32 bit number (L_var2) and a 16 bit * number (var1) returning a 32 bit result. L_var2 * is truncated to 31 bits prior to executing the * multiply. * * INPUTS: * * L_var2 A Longword input variable * * var1 A Shortword input variable * * OUTPUTS: none * * RETURN VALUE: A Longword value * * KEYWORDS: mult,mpy,multiplication * ***************************************************************************/ Longword L_mpy_ls(Longword L_var2, Shortword var1) { Longword L_varOut; Shortword swtemp; swtemp = shr(extract_l(L_var2), 1); swtemp = (short) 32767 & (short) swtemp; L_varOut = L_mult(var1, swtemp); L_varOut = L_shr(L_varOut, 15); L_varOut = L_mac(L_varOut, var1, extract_h(L_var2)); return (L_varOut); }