Signed 16-bit Multiply 32-bit Add/Subtract Instructions

__STATIC_FORCEINLINE long __RV_KMABB (long t, unsigned long a, unsigned long b)
__STATIC_FORCEINLINE long __RV_KMABT (long t, unsigned long a, unsigned long b)
__STATIC_FORCEINLINE long __RV_KMATT (long t, unsigned long a, unsigned long b)
__STATIC_FORCEINLINE long __RV_KMADA (long t, unsigned long a, unsigned long b)
__STATIC_FORCEINLINE long __RV_KMAXDA (long t, unsigned long a, unsigned long b)
__STATIC_FORCEINLINE long __RV_KMADS (long t, unsigned long a, unsigned long b)
__STATIC_FORCEINLINE long __RV_KMADRS (long t, unsigned long a, unsigned long b)
__STATIC_FORCEINLINE long __RV_KMAXDS (long t, unsigned long a, unsigned long b)
__STATIC_FORCEINLINE long __RV_KMDA (unsigned long a, unsigned long b)
__STATIC_FORCEINLINE long __RV_KMXDA (unsigned long a, unsigned long b)
__STATIC_FORCEINLINE long __RV_KMSDA (long t, unsigned long a, unsigned long b)
__STATIC_FORCEINLINE long __RV_KMSXDA (long t, unsigned long a, unsigned long b)
__STATIC_FORCEINLINE long __RV_SMBB16 (unsigned long a, unsigned long b)
__STATIC_FORCEINLINE long __RV_SMBT16 (unsigned long a, unsigned long b)
__STATIC_FORCEINLINE long __RV_SMTT16 (unsigned long a, unsigned long b)
__STATIC_FORCEINLINE long __RV_SMDS (unsigned long a, unsigned long b)
__STATIC_FORCEINLINE long __RV_SMDRS (unsigned long a, unsigned long b)
__STATIC_FORCEINLINE long __RV_SMXDS (unsigned long a, unsigned long b)
group Signed 16-bit Multiply 32-bit Add/Subtract Instructions

Signed 16-bit Multiply 32-bit Add/Subtract Instructions.

there are 18 Signed 16-bit Multiply 32-bit Add/Subtract Instructions

Functions

__STATIC_FORCEINLINE long __RV_KMABB (long t, unsigned long a, unsigned long b)

KMABB (SIMD Saturating Signed Multiply Bottom Halfs & Add)

Type: SIMD

Syntax:

KMABB Rd, Rs1, Rs2
KMABT Rd, Rs1, Rs2
KMATT Rd, Rs1, Rs2

Purpose

:

Multiply the signed 16-bit content of 32-bit elements in a register with the 16-bit content of 32-bit elements in another register and add the result to the content of 32-bit elements in the third register. The addition result may be saturated and is written to the third register.

  • KMABB: rd.W[x] + bottom*bottom (per 32-bit element)

  • KMABT rd.W[x] + bottom*top (per 32-bit element)

  • KMATT rd.W[x] + top*top (per 32-bit element)

Description

:

For the

KMABB instruction, it multiplies the bottom 16-bit content of 32-bit elements in Rs1 with the bottom 16-bit content of 32-bit elements in Rs2. For the KMABT instruction, it multiplies the bottom 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2. For the KMATT instruction, it multiplies the top 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2. The multiplication result is added to the content of 32-bit elements in Rd. If the addition result is beyond the Q31 number range (-2^31 <= Q31 <= 2^31-1), it is saturated to the range and the OV bit is set to

  1. The results after saturation are written to Rd. The 16-bit contents of Rs1 and Rs2 are treated as signed integers.

Operations:

res[x] = Rd.W[x] + (Rs1.W[x].H[0] * Rs2.W[x].H[0]); // KMABB
res[x] = Rd.W[x] + (Rs1.W[x].H[0] * Rs2.W[x].H[1]); // KMABT
res[x] = Rd.W[x] + (Rs1.W[x].H[1] * Rs2.W[x].H[1]); // KMATT
if (res[x] > (2^31)-1) {
  res[x] = (2^31)-1;
  OV = 1;
} else if (res[x] < -2^31) {
  res[x] = -2^31;
  OV = 1;
}
Rd.W[x] = res[x];
for RV32: x=0
for RV64: x=1...0

Parameters:

  • t – [in] long type of value stored in t

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type

__STATIC_FORCEINLINE long __RV_KMABT (long t, unsigned long a, unsigned long b)

KMABT (SIMD Saturating Signed Multiply Bottom & Top Halfs & Add)

Type: SIMD

Syntax:

KMABB Rd, Rs1, Rs2
KMABT Rd, Rs1, Rs2
KMATT Rd, Rs1, Rs2

Purpose

:

Multiply the signed 16-bit content of 32-bit elements in a register with the 16-bit content of 32-bit elements in another register and add the result to the content of 32-bit elements in the third register. The addition result may be saturated and is written to the third register.

  • KMABB: rd.W[x] + bottom*bottom (per 32-bit element)

  • KMABT rd.W[x] + bottom*top (per 32-bit element)

  • KMATT rd.W[x] + top*top (per 32-bit element)

Description

:

For the

KMABB instruction, it multiplies the bottom 16-bit content of 32-bit elements in Rs1 with the bottom 16-bit content of 32-bit elements in Rs2. For the KMABT instruction, it multiplies the bottom 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2. For the KMATT instruction, it multiplies the top 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2. The multiplication result is added to the content of 32-bit elements in Rd. If the addition result is beyond the Q31 number range (-2^31 <= Q31 <= 2^31-1), it is saturated to the range and the OV bit is set to

  1. The results after saturation are written to Rd. The 16-bit contents of Rs1 and Rs2 are treated as signed integers.

Operations:

res[x] = Rd.W[x] + (Rs1.W[x].H[0] * Rs2.W[x].H[0]); // KMABB
res[x] = Rd.W[x] + (Rs1.W[x].H[0] * Rs2.W[x].H[1]); // KMABT
res[x] = Rd.W[x] + (Rs1.W[x].H[1] * Rs2.W[x].H[1]); // KMATT
if (res[x] > (2^31)-1) {
  res[x] = (2^31)-1;
  OV = 1;
} else if (res[x] < -2^31) {
  res[x] = -2^31;
  OV = 1;
}
Rd.W[x] = res[x];
for RV32: x=0
for RV64: x=1...0

Parameters:

  • t – [in] long type of value stored in t

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type

__STATIC_FORCEINLINE long __RV_KMATT (long t, unsigned long a, unsigned long b)

KMATT (SIMD Saturating Signed Multiply Top Halfs & Add)

Type: SIMD

Syntax:

KMABB Rd, Rs1, Rs2
KMABT Rd, Rs1, Rs2
KMATT Rd, Rs1, Rs2

Purpose

:

Multiply the signed 16-bit content of 32-bit elements in a register with the 16-bit content of 32-bit elements in another register and add the result to the content of 32-bit elements in the third register. The addition result may be saturated and is written to the third register.

  • KMABB: rd.W[x] + bottom*bottom (per 32-bit element)

  • KMABT rd.W[x] + bottom*top (per 32-bit element)

  • KMATT rd.W[x] + top*top (per 32-bit element)

Description

:

For the

KMABB instruction, it multiplies the bottom 16-bit content of 32-bit elements in Rs1 with the bottom 16-bit content of 32-bit elements in Rs2. For the KMABT instruction, it multiplies the bottom 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2. For the KMATT instruction, it multiplies the top 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2. The multiplication result is added to the content of 32-bit elements in Rd. If the addition result is beyond the Q31 number range (-2^31 <= Q31 <= 2^31-1), it is saturated to the range and the OV bit is set to

  1. The results after saturation are written to Rd. The 16-bit contents of Rs1 and Rs2 are treated as signed integers.

Operations:

res[x] = Rd.W[x] + (Rs1.W[x].H[0] * Rs2.W[x].H[0]); // KMABB
res[x] = Rd.W[x] + (Rs1.W[x].H[0] * Rs2.W[x].H[1]); // KMABT
res[x] = Rd.W[x] + (Rs1.W[x].H[1] * Rs2.W[x].H[1]); // KMATT
if (res[x] > (2^31)-1) {
  res[x] = (2^31)-1;
  OV = 1;
} else if (res[x] < -2^31) {
  res[x] = -2^31;
  OV = 1;
}
Rd.W[x] = res[x];
for RV32: x=0
for RV64: x=1...0

Parameters:

  • t – [in] long type of value stored in t

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type

__STATIC_FORCEINLINE long __RV_KMADA (long t, unsigned long a, unsigned long b)

KMADA (SIMD Saturating Signed Multiply Two Halfs and Two Adds)

Type: SIMD

Syntax:

KMADA Rd, Rs1, Rs2
KMAXDA Rd, Rs1, Rs2

Purpose

:

Do two signed 16-bit multiplications from 32-bit elements in two registers; and then adds the two 32-bit results and 32-bit elements in a third register together. The addition result may be saturated.

  • KMADA: rd.W[x] + top*top + bottom*bottom (per 32-bit element)

  • KMAXDA: rd.W[x] + top*bottom + bottom*top (per 32-bit element)

Description

:

For the `KMADA instruction, it multiplies the bottom 16-bit content of 32-bit elements in Rs1 with the bottom 16-bit content of 32-bit elements in Rs2 and then adds the result to the result of multiplying the top 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2. For the

KMAXDA instruction, it multiplies the top 16-bit content of 32-bit elements in Rs1 with the bottom 16-bit content of 32-bit elements in Rs2 and then adds the result to the result of multiplying the bottom 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2. The result is added to the content of 32-bit elements in Rd. If the addition result is beyond the Q31 number range (-2^31 <= Q31 <= 2^31-1), it is saturated to the range and the OV bit is set to 1. The 32-bit results after saturation are written to Rd. The 16-bit contents of Rs1 and Rs2 are treated as signed integers.

Operations:

// KMADA
res[x] = Rd.W[x] + (Rs1.W[x].H[1] * Rs2.W[x].H[1]) + (Rs1.W[x].H[0] * Rs2.W[x].H[0]);
// KMAXDA
res[x] = Rd.W[x] + (Rs1.W[x].H[1] * Rs2.W[x].H[0]) + (Rs1.W[x].H[0] * Rs2.W[x].H[1]);
if (res[x] > (2^31)-1) {
  res[x] = (2^31)-1;
  OV = 1;
} else if (res[x] < -2^31) {
  res[x] = -2^31;
OV = 1;
}
Rd.W[x] = res[x];
for RV32: x=0
for RV64: x=1...0

Parameters:

  • t – [in] long type of value stored in t

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type

__STATIC_FORCEINLINE long __RV_KMAXDA (long t, unsigned long a, unsigned long b)

KMAXDA (SIMD Saturating Signed Crossed Multiply Two Halfs and Two Adds)

Type: SIMD

Syntax:

KMADA Rd, Rs1, Rs2
KMAXDA Rd, Rs1, Rs2

Purpose

:

Do two signed 16-bit multiplications from 32-bit elements in two registers; and then adds the two 32-bit results and 32-bit elements in a third register together. The addition result may be saturated.

  • KMADA: rd.W[x] + top*top + bottom*bottom (per 32-bit element)

  • KMAXDA: rd.W[x] + top*bottom + bottom*top (per 32-bit element)

Description

:

For the `KMADA instruction, it multiplies the bottom 16-bit content of 32-bit elements in Rs1 with the bottom 16-bit content of 32-bit elements in Rs2 and then adds the result to the result of multiplying the top 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2. For the

KMAXDA instruction, it multiplies the top 16-bit content of 32-bit elements in Rs1 with the bottom 16-bit content of 32-bit elements in Rs2 and then adds the result to the result of multiplying the bottom 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2. The result is added to the content of 32-bit elements in Rd. If the addition result is beyond the Q31 number range (-2^31 <= Q31 <= 2^31-1), it is saturated to the range and the OV bit is set to 1. The 32-bit results after saturation are written to Rd. The 16-bit contents of Rs1 and Rs2 are treated as signed integers.

Operations:

// KMADA
res[x] = Rd.W[x] + (Rs1.W[x].H[1] * Rs2.W[x].H[1]) + (Rs1.W[x].H[0] * Rs2.W[x].H[0]);
// KMAXDA
res[x] = Rd.W[x] + (Rs1.W[x].H[1] * Rs2.W[x].H[0]) + (Rs1.W[x].H[0] * Rs2.W[x].H[1]);
if (res[x] > (2^31)-1) {
  res[x] = (2^31)-1;
  OV = 1;
} else if (res[x] < -2^31) {
  res[x] = -2^31;
OV = 1;
}
Rd.W[x] = res[x];
for RV32: x=0
for RV64: x=1...0

Parameters:

  • t – [in] long type of value stored in t

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type

__STATIC_FORCEINLINE long __RV_KMADS (long t, unsigned long a, unsigned long b)

KMADS (SIMD Saturating Signed Multiply Two Halfs & Subtract & Add)

Type: SIMD

Syntax:

KMADS Rd, Rs1, Rs2
KMADRS Rd, Rs1, Rs2
KMAXDS Rd, Rs1, Rs2

Purpose

:

Do two signed 16-bit multiplications from 32-bit elements in two registers; and then perform a subtraction operation between the two 32-bit results. Then add the subtraction result to the corresponding 32-bit elements in a third register. The addition result may be saturated.

  • KMADS: rd.W[x] + (top*top - bottom*bottom) (per 32-bit element)

  • KMADRS: rd.W[x] + (bottom*bottom - top*top) (per 32-bit element)

  • KMAXDS: rd.W[x] + (top*bottom - bottom*top) (per 32-bit element)

Description

:

For the

KMADS instruction, it multiplies the bottom 16-bit content of 32-bit elements in Rs1 with the bottom 16-bit content of 32-bit elements in Rs2 and then subtracts the result from the result of multiplying the top 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2. For the KMADRS instruction, it multiplies the top 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2 and then subtracts the result from the result of multiplying the bottom 16-bit content of 32-bit elements in Rs1 with the bottom 16-bit content of 32- bit elements in Rs2. For the KMAXDS instruction, it multiplies the bottom 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2 and then subtracts the result from the result of multiplying the top 16-bit content of 32-bit elements in Rs1 with the bottom 16-bit content of 32-bit elements in Rs2. The subtraction result is then added to the content of the corresponding 32-bit elements in Rd. If the addition result is beyond the Q31 number range (-2^31 <= Q31 <= 2^31-1), it is saturated to the range and the OV bit is set to 1. The 32-bit results after saturation are written to Rd. The 16-bit contents of Rs1 and Rs2 are treated as signed integers.

Operations:

// KMADS
res[x] = Rd.W[x] + (Rs1.W[x].H[1] * Rs2.W[x].H[1]) - (Rs1.W[x].H[0] * Rs2.W[x].H[0]);
// KMADRS
res[x] = Rd.W[x] + (Rs1.W[x].H[0] * Rs2.W[x].H[0]) - (Rs1.W[x].H[1] * Rs2.W[x].H[1]);
// KMAXDS
res[x] = Rd.W[x] + (Rs1.W[x].H[1] * Rs2.W[x].H[0]) - (Rs1.W[x].H[0] * Rs2.W[x].H[1]);
if (res[x] > (2^31)-1) {
  res[x] = (2^31)-1;
  OV = 1;
} else if (res[x] < -2^31) {
  res[x] = -2^31;
  OV = 1;
}
Rd.W[x] = res[x];
for RV32: x=0
for RV64: x=1...0

Parameters:

  • t – [in] long type of value stored in t

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type

__STATIC_FORCEINLINE long __RV_KMADRS (long t, unsigned long a, unsigned long b)

KMADRS (SIMD Saturating Signed Multiply Two Halfs & Reverse Subtract & Add)

Type: SIMD

Syntax:

KMADS Rd, Rs1, Rs2
KMADRS Rd, Rs1, Rs2
KMAXDS Rd, Rs1, Rs2

Purpose

:

Do two signed 16-bit multiplications from 32-bit elements in two registers; and then perform a subtraction operation between the two 32-bit results. Then add the subtraction result to the corresponding 32-bit elements in a third register. The addition result may be saturated.

  • KMADS: rd.W[x] + (top*top - bottom*bottom) (per 32-bit element)

  • KMADRS: rd.W[x] + (bottom*bottom - top*top) (per 32-bit element)

  • KMAXDS: rd.W[x] + (top*bottom - bottom*top) (per 32-bit element)

Description

:

For the

KMADS instruction, it multiplies the bottom 16-bit content of 32-bit elements in Rs1 with the bottom 16-bit content of 32-bit elements in Rs2 and then subtracts the result from the result of multiplying the top 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2. For the KMADRS instruction, it multiplies the top 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2 and then subtracts the result from the result of multiplying the bottom 16-bit content of 32-bit elements in Rs1 with the bottom 16-bit content of 32- bit elements in Rs2. For the KMAXDS instruction, it multiplies the bottom 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2 and then subtracts the result from the result of multiplying the top 16-bit content of 32-bit elements in Rs1 with the bottom 16-bit content of 32-bit elements in Rs2. The subtraction result is then added to the content of the corresponding 32-bit elements in Rd. If the addition result is beyond the Q31 number range (-2^31 <= Q31 <= 2^31-1), it is saturated to the range and the OV bit is set to 1. The 32-bit results after saturation are written to Rd. The 16-bit contents of Rs1 and Rs2 are treated as signed integers.

Operations:

// KMADS
res[x] = Rd.W[x] + (Rs1.W[x].H[1] * Rs2.W[x].H[1]) - (Rs1.W[x].H[0] * Rs2.W[x].H[0]);
// KMADRS
res[x] = Rd.W[x] + (Rs1.W[x].H[0] * Rs2.W[x].H[0]) - (Rs1.W[x].H[1] * Rs2.W[x].H[1]);
// KMAXDS
res[x] = Rd.W[x] + (Rs1.W[x].H[1] * Rs2.W[x].H[0]) - (Rs1.W[x].H[0] * Rs2.W[x].H[1]);
if (res[x] > (2^31)-1) {
  res[x] = (2^31)-1;
  OV = 1;
} else if (res[x] < -2^31) {
  res[x] = -2^31;
  OV = 1;
}
Rd.W[x] = res[x];
for RV32: x=0
for RV64: x=1...0

Parameters:

  • t – [in] long type of value stored in t

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type

__STATIC_FORCEINLINE long __RV_KMAXDS (long t, unsigned long a, unsigned long b)

KMAXDS (SIMD Saturating Signed Crossed Multiply Two Halfs & Subtract & Add)

Type: SIMD

Syntax:

KMADS Rd, Rs1, Rs2
KMADRS Rd, Rs1, Rs2
KMAXDS Rd, Rs1, Rs2

Purpose

:

Do two signed 16-bit multiplications from 32-bit elements in two registers; and then perform a subtraction operation between the two 32-bit results. Then add the subtraction result to the corresponding 32-bit elements in a third register. The addition result may be saturated.

  • KMADS: rd.W[x] + (top*top - bottom*bottom) (per 32-bit element)

  • KMADRS: rd.W[x] + (bottom*bottom - top*top) (per 32-bit element)

  • KMAXDS: rd.W[x] + (top*bottom - bottom*top) (per 32-bit element)

Description

:

For the

KMADS instruction, it multiplies the bottom 16-bit content of 32-bit elements in Rs1 with the bottom 16-bit content of 32-bit elements in Rs2 and then subtracts the result from the result of multiplying the top 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2. For the KMADRS instruction, it multiplies the top 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2 and then subtracts the result from the result of multiplying the bottom 16-bit content of 32-bit elements in Rs1 with the bottom 16-bit content of 32- bit elements in Rs2. For the KMAXDS instruction, it multiplies the bottom 16-bit content of 32-bit elements in Rs1 with the top 16-bit content of 32-bit elements in Rs2 and then subtracts the result from the result of multiplying the top 16-bit content of 32-bit elements in Rs1 with the bottom 16-bit content of 32-bit elements in Rs2. The subtraction result is then added to the content of the corresponding 32-bit elements in Rd. If the addition result is beyond the Q31 number range (-2^31 <= Q31 <= 2^31-1), it is saturated to the range and the OV bit is set to 1. The 32-bit results after saturation are written to Rd. The 16-bit contents of Rs1 and Rs2 are treated as signed integers.

Operations:

// KMADS
res[x] = Rd.W[x] + (Rs1.W[x].H[1] * Rs2.W[x].H[1]) - (Rs1.W[x].H[0] * Rs2.W[x].H[0]);
// KMADRS
res[x] = Rd.W[x] + (Rs1.W[x].H[0] * Rs2.W[x].H[0]) - (Rs1.W[x].H[1] * Rs2.W[x].H[1]);
// KMAXDS
res[x] = Rd.W[x] + (Rs1.W[x].H[1] * Rs2.W[x].H[0]) - (Rs1.W[x].H[0] * Rs2.W[x].H[1]);
if (res[x] > (2^31)-1) {
  res[x] = (2^31)-1;
  OV = 1;
} else if (res[x] < -2^31) {
  res[x] = -2^31;
  OV = 1;
}
Rd.W[x] = res[x];
for RV32: x=0
for RV64: x=1...0

Parameters:

  • t – [in] long type of value stored in t

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type

__STATIC_FORCEINLINE long __RV_KMDA (unsigned long a, unsigned long b)

KMDA (SIMD Signed Multiply Two Halfs and Add)

Type: SIMD

Syntax:

KMDA Rd, Rs1, Rs2
KMXDA Rd, Rs1, Rs2

Purpose

:

Do two signed 16-bit multiplications from the 32-bit elements of two registers; and then adds the two 32-bit results together. The addition result may be saturated.

  • KMDA: top*top + bottom*bottom (per 32-bit element)

  • KMXDA: top*bottom + bottom*top (per 32-bit element)

Description

:

For the

KMDA instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2 and then adds the result to the result of multiplying the top 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32- bit elements of Rs2. For the KMXDA instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2 and then adds the result to the result of multiplying the top 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2. The addition result is checked for saturation. If saturation happens, the result is saturated to 2^31-1. The final results are written to Rd. The 16-bit contents are treated as signed integers.

Operations:

if  Rs1.W[x]  !=  0x80008000)  or  (Rs2.W[x]  !=  0x80008000  {  //  KMDA  Rd.W[x]  =  Rs1.W[x].H[1]  *
Rs2.W[x].H[1]) + (Rs1.W[x].H[0] * Rs2.W[x].H[0]; // KMXDA Rd.W[x] = Rs1.W[x].H[1] * Rs2.W[x].H[0])
+  (Rs1.W[x].H[0]  *  Rs2.W[x].H[1];  }  else  {  Rd.W[x]  =  0x7fffffff;  OV  =  1;  }  for  RV32:  x=0  for  RV64:
x=1...0

Parameters:

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type

__STATIC_FORCEINLINE long __RV_KMXDA (unsigned long a, unsigned long b)

KMXDA (SIMD Signed Crossed Multiply Two Halfs and Add)

Type: SIMD

Syntax:

KMDA Rd, Rs1, Rs2
KMXDA Rd, Rs1, Rs2

Purpose

:

Do two signed 16-bit multiplications from the 32-bit elements of two registers; and then adds the two 32-bit results together. The addition result may be saturated.

  • KMDA: top*top + bottom*bottom (per 32-bit element)

  • KMXDA: top*bottom + bottom*top (per 32-bit element)

Description

:

For the

KMDA instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2 and then adds the result to the result of multiplying the top 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32- bit elements of Rs2. For the KMXDA instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2 and then adds the result to the result of multiplying the top 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2. The addition result is checked for saturation. If saturation happens, the result is saturated to 2^31-1. The final results are written to Rd. The 16-bit contents are treated as signed integers.

Operations:

if  Rs1.W[x]  !=  0x80008000)  or  (Rs2.W[x]  !=  0x80008000  {  //  KMDA  Rd.W[x]  =  Rs1.W[x].H[1]  *
Rs2.W[x].H[1]) + (Rs1.W[x].H[0] * Rs2.W[x].H[0]; // KMXDA Rd.W[x] = Rs1.W[x].H[1] * Rs2.W[x].H[0])
+  (Rs1.W[x].H[0]  *  Rs2.W[x].H[1];  }  else  {  Rd.W[x]  =  0x7fffffff;  OV  =  1;  }  for  RV32:  x=0  for  RV64:
x=1...0

Parameters:

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type

__STATIC_FORCEINLINE long __RV_KMSDA (long t, unsigned long a, unsigned long b)

KMSDA (SIMD Saturating Signed Multiply Two Halfs & Add & Subtract)

Type: SIMD

Syntax:

KMSDA Rd, Rs1, Rs2
KMSXDA Rd, Rs1, Rs2

Purpose

:

Do two signed 16-bit multiplications from the 32-bit elements of two registers; and then subtracts the two 32-bit results from the corresponding 32-bit elements of a third register. The subtraction result may be saturated.

  • KMSDA: rd.W[x] - top*top - bottom*bottom (per 32-bit element)

  • KMSXDA: rd.W[x] - top*bottom - bottom*top (per 32-bit element)

Description

:

For the

KMSDA instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2 and multiplies the top 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2. For the KMSXDA instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2 and multiplies the top 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2. The two 32-bit multiplication results are then subtracted from the content of the corresponding 32- bit elements of Rd. If the subtraction result is beyond the Q31 number range (-2^31 <= Q31 <= 2^31-1), it is saturated to the range and the OV bit is set to 1. The results after saturation are written to Rd. The 16-bit contents are treated as signed integers.

Operations:

// KMSDA
res[x] = Rd.W[x] - (Rs1.W[x].H[1] * Rs2.W[x].H[1]) - (Rs1.W[x].H[0] * Rs2.W[x].H[0]);
// KMSXDA
res[x] = Rd.W[x] - (Rs1.W[x].H[1] * Rs2.W[x].H[0]) - (Rs1.W[x].H[0] * Rs2.W[x].H[1]);
if (res[x] > (2^31)-1) {
  res[x] = (2^31)-1;
  OV = 1;
} else if (res[x] < -2^31) {
  res[x] = -2^31;
  OV = 1;
}
Rd.W[x] = res[x];
for RV32: x=0
for RV64: x=1...0

Parameters:

  • t – [in] long type of value stored in t

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type

__STATIC_FORCEINLINE long __RV_KMSXDA (long t, unsigned long a, unsigned long b)

KMSXDA (SIMD Saturating Signed Crossed Multiply Two Halfs & Add & Subtract)

Type: SIMD

Syntax:

KMSDA Rd, Rs1, Rs2
KMSXDA Rd, Rs1, Rs2

Purpose

:

Do two signed 16-bit multiplications from the 32-bit elements of two registers; and then subtracts the two 32-bit results from the corresponding 32-bit elements of a third register. The subtraction result may be saturated.

  • KMSDA: rd.W[x] - top*top - bottom*bottom (per 32-bit element)

  • KMSXDA: rd.W[x] - top*bottom - bottom*top (per 32-bit element)

Description

:

For the

KMSDA instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2 and multiplies the top 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2. For the KMSXDA instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2 and multiplies the top 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2. The two 32-bit multiplication results are then subtracted from the content of the corresponding 32- bit elements of Rd. If the subtraction result is beyond the Q31 number range (-2^31 <= Q31 <= 2^31-1), it is saturated to the range and the OV bit is set to 1. The results after saturation are written to Rd. The 16-bit contents are treated as signed integers.

Operations:

// KMSDA
res[x] = Rd.W[x] - (Rs1.W[x].H[1] * Rs2.W[x].H[1]) - (Rs1.W[x].H[0] * Rs2.W[x].H[0]);
// KMSXDA
res[x] = Rd.W[x] - (Rs1.W[x].H[1] * Rs2.W[x].H[0]) - (Rs1.W[x].H[0] * Rs2.W[x].H[1]);
if (res[x] > (2^31)-1) {
  res[x] = (2^31)-1;
  OV = 1;
} else if (res[x] < -2^31) {
  res[x] = -2^31;
  OV = 1;
}
Rd.W[x] = res[x];
for RV32: x=0
for RV64: x=1...0

Parameters:

  • t – [in] long type of value stored in t

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type

__STATIC_FORCEINLINE long __RV_SMBB16 (unsigned long a, unsigned long b)

SMBB16 (SIMD Signed Multiply Bottom Half & Bottom Half)

Type: SIMD

Syntax:

SMBB16 Rd, Rs1, Rs2
SMBT16 Rd, Rs1, Rs2
SMTT16 Rd, Rs1, Rs2

Purpose

:

Multiply the signed 16-bit content of the 32-bit elements of a register with the signed 16- bit content of the 32-bit elements of another register and write the result to a third register.

  • SMBB16: W[x].bottom*W[x].bottom

  • SMBT16: W[x].bottom *W[x].top

  • SMTT16: W[x].top * W[x].top

Description

:

For the

SMBB16 instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2. For the SMBT16 instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2. For the SMTT16 instruction, it multiplies the top 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2. The multiplication results are written to Rd. The 16-bit contents of Rs1 and Rs2 are treated as signed integers.

Operations:

Rd.W[x] = Rs1.W[x].H[0] * Rs2.W[x].H[0]; // SMBB16
Rd.W[x] = Rs1.W[x].H[0] * Rs2.W[x].H[1]; // SMBT16
Rd.W[x] = Rs1.W[x].H[1] * Rs2.W[x].H[1]; // SMTT16
for RV32: x=0,
for RV64: x=1...0

Parameters:

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type

__STATIC_FORCEINLINE long __RV_SMBT16 (unsigned long a, unsigned long b)

SMBT16 (SIMD Signed Multiply Bottom Half & Top Half)

Type: SIMD

Syntax:

SMBB16 Rd, Rs1, Rs2
SMBT16 Rd, Rs1, Rs2
SMTT16 Rd, Rs1, Rs2

Purpose

:

Multiply the signed 16-bit content of the 32-bit elements of a register with the signed 16- bit content of the 32-bit elements of another register and write the result to a third register.

  • SMBB16: W[x].bottom*W[x].bottom

  • SMBT16: W[x].bottom *W[x].top

  • SMTT16: W[x].top * W[x].top

Description

:

For the

SMBB16 instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2. For the SMBT16 instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2. For the SMTT16 instruction, it multiplies the top 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2. The multiplication results are written to Rd. The 16-bit contents of Rs1 and Rs2 are treated as signed integers.

Operations:

Rd.W[x] = Rs1.W[x].H[0] * Rs2.W[x].H[0]; // SMBB16
Rd.W[x] = Rs1.W[x].H[0] * Rs2.W[x].H[1]; // SMBT16
Rd.W[x] = Rs1.W[x].H[1] * Rs2.W[x].H[1]; // SMTT16
for RV32: x=0,
for RV64: x=1...0

Parameters:

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type

__STATIC_FORCEINLINE long __RV_SMTT16 (unsigned long a, unsigned long b)

SMTT16 (SIMD Signed Multiply Top Half & Top Half)

Type: SIMD

Syntax:

SMBB16 Rd, Rs1, Rs2
SMBT16 Rd, Rs1, Rs2
SMTT16 Rd, Rs1, Rs2

Purpose

:

Multiply the signed 16-bit content of the 32-bit elements of a register with the signed 16- bit content of the 32-bit elements of another register and write the result to a third register.

  • SMBB16: W[x].bottom*W[x].bottom

  • SMBT16: W[x].bottom *W[x].top

  • SMTT16: W[x].top * W[x].top

Description

:

For the

SMBB16 instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2. For the SMBT16 instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2. For the SMTT16 instruction, it multiplies the top 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2. The multiplication results are written to Rd. The 16-bit contents of Rs1 and Rs2 are treated as signed integers.

Operations:

Rd.W[x] = Rs1.W[x].H[0] * Rs2.W[x].H[0]; // SMBB16
Rd.W[x] = Rs1.W[x].H[0] * Rs2.W[x].H[1]; // SMBT16
Rd.W[x] = Rs1.W[x].H[1] * Rs2.W[x].H[1]; // SMTT16
for RV32: x=0,
for RV64: x=1...0

Parameters:

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type

__STATIC_FORCEINLINE long __RV_SMDS (unsigned long a, unsigned long b)

SMDS (SIMD Signed Multiply Two Halfs and Subtract)

Type: SIMD

Syntax:

SMDS Rd, Rs1, Rs2
SMDRS Rd, Rs1, Rs2
SMXDS Rd, Rs1, Rs2

Purpose

:

Do two signed 16-bit multiplications from the 32-bit elements of two registers; and then perform a subtraction operation between the two 32-bit results.

  • SMDS: top*top - bottom*bottom (per 32-bit element)

  • SMDRS: bottom*bottom - top*top (per 32-bit element)

  • SMXDS: top*bottom - bottom*top (per 32-bit element)

Description

:

For the

SMDS instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2 and then subtracts the result from the result of multiplying the top 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2. For the SMDRS instruction, it multiplies the top 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2 and then subtracts the result from the result of multiplying the bottom 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2. For the SMXDS instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2 and then subtracts the result from the result of multiplying the top 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2. The subtraction result is written to the corresponding 32-bit element of Rd. The 16-bit contents of multiplication are treated as signed integers.

Operations:

* SMDS:
Rd.W[x] = (Rs1.W[x].H[1] * Rs2.W[x].H[1]) - (Rs1.W[x].H[0] * Rs2.W[x].H[0]);
* SMDRS:
Rd.W[x] = (Rs1.W[x].H[0] * Rs2.W[x].H[0]) - (Rs1.W[x].H[1] * Rs2.W[x].H[1]);
* SMXDS:
Rd.W[x] = (Rs1.W[x].H[1] * Rs2.W[x].H[0]) - (Rs1.W[x].H[0] * Rs2.W[x].H[1]);

Parameters:

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type

__STATIC_FORCEINLINE long __RV_SMDRS (unsigned long a, unsigned long b)

SMDRS (SIMD Signed Multiply Two Halfs and Reverse Subtract)

Type: SIMD

Syntax:

SMDS Rd, Rs1, Rs2
SMDRS Rd, Rs1, Rs2
SMXDS Rd, Rs1, Rs2

Purpose

:

Do two signed 16-bit multiplications from the 32-bit elements of two registers; and then perform a subtraction operation between the two 32-bit results.

  • SMDS: top*top - bottom*bottom (per 32-bit element)

  • SMDRS: bottom*bottom - top*top (per 32-bit element)

  • SMXDS: top*bottom - bottom*top (per 32-bit element)

Description

:

For the

SMDS instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2 and then subtracts the result from the result of multiplying the top 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2. For the SMDRS instruction, it multiplies the top 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2 and then subtracts the result from the result of multiplying the bottom 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2. For the SMXDS instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2 and then subtracts the result from the result of multiplying the top 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2. The subtraction result is written to the corresponding 32-bit element of Rd. The 16-bit contents of multiplication are treated as signed integers.

Operations:

* SMDS:
Rd.W[x] = (Rs1.W[x].H[1] * Rs2.W[x].H[1]) - (Rs1.W[x].H[0] * Rs2.W[x].H[0]);
* SMDRS:
Rd.W[x] = (Rs1.W[x].H[0] * Rs2.W[x].H[0]) - (Rs1.W[x].H[1] * Rs2.W[x].H[1]);
* SMXDS:
Rd.W[x] = (Rs1.W[x].H[1] * Rs2.W[x].H[0]) - (Rs1.W[x].H[0] * Rs2.W[x].H[1]);

Parameters:

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type

__STATIC_FORCEINLINE long __RV_SMXDS (unsigned long a, unsigned long b)

SMXDS (SIMD Signed Crossed Multiply Two Halfs and Subtract)

Type: SIMD

Syntax:

SMDS Rd, Rs1, Rs2
SMDRS Rd, Rs1, Rs2
SMXDS Rd, Rs1, Rs2

Purpose

:

Do two signed 16-bit multiplications from the 32-bit elements of two registers; and then perform a subtraction operation between the two 32-bit results.

  • SMDS: top*top - bottom*bottom (per 32-bit element)

  • SMDRS: bottom*bottom - top*top (per 32-bit element)

  • SMXDS: top*bottom - bottom*top (per 32-bit element)

Description

:

For the

SMDS instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2 and then subtracts the result from the result of multiplying the top 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2. For the SMDRS instruction, it multiplies the top 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2 and then subtracts the result from the result of multiplying the bottom 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2. For the SMXDS instruction, it multiplies the bottom 16-bit content of the 32-bit elements of Rs1 with the top 16-bit content of the 32-bit elements of Rs2 and then subtracts the result from the result of multiplying the top 16-bit content of the 32-bit elements of Rs1 with the bottom 16-bit content of the 32-bit elements of Rs2. The subtraction result is written to the corresponding 32-bit element of Rd. The 16-bit contents of multiplication are treated as signed integers.

Operations:

* SMDS:
Rd.W[x] = (Rs1.W[x].H[1] * Rs2.W[x].H[1]) - (Rs1.W[x].H[0] * Rs2.W[x].H[0]);
* SMDRS:
Rd.W[x] = (Rs1.W[x].H[0] * Rs2.W[x].H[0]) - (Rs1.W[x].H[1] * Rs2.W[x].H[1]);
* SMXDS:
Rd.W[x] = (Rs1.W[x].H[1] * Rs2.W[x].H[0]) - (Rs1.W[x].H[0] * Rs2.W[x].H[1]);

Parameters:

  • a – [in] unsigned long type of value stored in a

  • b – [in] unsigned long type of value stored in b

Returns:

value stored in long type