Skip to content
Henk-Jan Lebbink edited this page Jun 5, 2018 · 12 revisions

DIVPD — Divide Packed Double-Precision Floating-Point Values

Opcode/ Instruction Op / En 64/32 bit Mode Support CPUID Feature Flag Description
66 0F 5E /r DIVPD xmm1, xmm2/m128 A V/V SSE2 Divide packed double-precision floating-point values in xmm1 by packed double-precision floating-point values in xmm2/mem.
VEX.NDS.128.66.0F.WIG 5E /r VDIVPD xmm1, xmm2, xmm3/m128 B V/V AVX Divide packed double-precision floating-point values in xmm2 by packed double-precision floating-point values in xmm3/mem.
VEX.NDS.256.66.0F.WIG 5E /r VDIVPD ymm1, ymm2, ymm3/m256 B V/V AVX Divide packed double-precision floating-point values in ymm2 by packed double-precision floating-point values in ymm3/mem.
EVEX.NDS.128.66.0F.W1 5E /r VDIVPD xmm1 {k1}{z}, xmm2, xmm3/m128/m64bcst C V/V AVX512VL AVX512F Divide packed double-precision floating-point values in xmm2 by packed double-precision floating-point values in xmm3/m128/m64bcst and write results to xmm1 subject to writemask k1.
EVEX.NDS.256.66.0F.W1 5E /r VDIVPD ymm1 {k1}{z}, ymm2, ymm3/m256/m64bcst C V/V AVX512VL AVX512F Divide packed double-precision floating-point values in ymm2 by packed double-precision floating-point values in ymm3/m256/m64bcst and write results to ymm1 subject to writemask k1.
EVEX.NDS.512.66.0F.W1 5E /r VDIVPD zmm1 {k1}{z}, zmm2, zmm3/m512/m64bcst{er} C V/V AVX512F Divide packed double-precision floating-point values in zmm2 by packed double-precision FP values in zmm3/m512/m64bcst and write results to zmm1 subject to writemask k1.

Instruction Operand Encoding

Op/En Tuple Type Operand 1 Operand 2 Operand 3 Operand 4
A NA ModRM:reg (r, w) ModRM:r/m (r) NA NA
B NA ModRM:reg (w) VEX.vvvv ModRM:r/m (r) NA
C Full ModRM:reg (w) EVEX.vvvv ModRM:r/m (r) NA

Description

Performs a SIMD divide of the double-precision floating-point values in the first source operand by the floating- point values in the second source operand (the third operand). Results are written to the destination operand (the first operand).

EVEX encoded versions: The first source operand (the second operand) is a ZMM/YMM/XMM register. The second source operand can be a ZMM/YMM/XMM register, a 512/256/128-bit memory location or a 512/256/128-bit vector broadcasted from a 64-bit memory location. The destination operand is a ZMM/YMM/XMM register conditionally updated with writemask k1.

VEX.256 encoded version: The first source operand (the second operand) is a YMM register. The second source operand can be a YMM register or a 256-bit memory location. The destination operand is a YMM register. The upper bits (MAXVL-1:256) of the corresponding destination are zeroed.

VEX.128 encoded version: The first source operand (the second operand) is a XMM register. The second source operand can be a XMM register or a 128-bit memory location. The destination operand is a XMM register. The upper bits (MAXVL-1:128) of the corresponding destination are zeroed.

128-bit Legacy SSE version: The second source operand (the second operand) can be an XMM register or an 128- bit memory location. The destination is the same as the first source operand. The upper bits (MAXVL-1:128) of the corresponding destination are unmodified.

Operation

VDIVPD (EVEX encoded versions)

(KL, VL) = (2, 128), (4, 256), (8, 512)
IF (VL = 512) AND (EVEX.b = 1) AND SRC2 *is a register*
    THEN
        SET_RM(EVEX.RC);  ; refer to Table 15-4 in the Intel® 64 and IA-32 Architectures Software Developers Manual, Volume 1
    ELSE 
        SET_RM(MXCSR.RM);
FI;
FOR j0 TO KL-1
    ij * 64
    IF k1[j] OR *no writemask*
        THEN 
            IF (EVEX.b = 1) AND (SRC2 *is memory*)
                THEN
                    DEST[i+63:i] ← SRC1[i+63:i] / SRC2[63:0]
                ELSE 
                    DEST[i+63:i] ← SRC1[i+63:i] / SRC2[i+63:i]
            FI;
        ELSE 
            IF *merging-masking*
                            ; merging-masking
                THEN *DEST[i+63:i] remains unchanged*
                ELSE 
                            ; zeroing-masking
                    DEST[i+63:i] ← 0
            FI
    FI;
ENDFOR
DEST[MAXVL-1:VL] ← 0

VDIVPD (VEX.256 encoded version)

DEST[63:0] ←SRC1[63:0] / SRC2[63:0]
DEST[127:64] ←SRC1[127:64] / SRC2[127:64]
DEST[191:128] ←SRC1[191:128] / SRC2[191:128]
DEST[255:192] ←SRC1[255:192] / SRC2[255:192]
DEST[MAXVL-1:256] ←0;

VDIVPD (VEX.128 encoded version)

DEST[63:0] ←SRC1[63:0] / SRC2[63:0]
DEST[127:64] ←SRC1[127:64] / SRC2[127:64]
DEST[MAXVL-1:128] ←0;

DIVPD (128-bit Legacy SSE version)

DEST[63:0] ←SRC1[63:0] / SRC2[63:0]
DEST[127:64] ←SRC1[127:64] / SRC2[127:64]
DEST[MAXVL-1:128] (Unmodified)

Intel C/C++ Compiler Intrinsic Equivalent

VDIVPD __m512d _mm512_div_pd( __m512d a, __m512d b);
VDIVPD __m512d _mm512_mask_div_pd(__m512d s, __mmask8 k, __m512d a, __m512d b);
VDIVPD __m512d _mm512_maskz_div_pd( __mmask8 k, __m512d a, __m512d b);
VDIVPD __m256d _mm256_mask_div_pd(__m256d s, __mmask8 k, __m256d a, __m256d b);
VDIVPD __m256d _mm256_maskz_div_pd( __mmask8 k, __m256d a, __m256d b);
VDIVPD __m128d _mm_mask_div_pd(__m128d s, __mmask8 k, __m128d a, __m128d b);
VDIVPD __m128d _mm_maskz_div_pd( __mmask8 k, __m128d a, __m128d b);
VDIVPD __m512d _mm512_div_round_pd( __m512d a, __m512d b, int);
VDIVPD __m512d _mm512_mask_div_round_pd(__m512d s, __mmask8 k, __m512d a, __m512d b, int);
VDIVPD __m512d _mm512_maskz_div_round_pd( __mmask8 k, __m512d a, __m512d b, int);
VDIVPD __m256d _mm256_div_pd (__m256d a, __m256d b);
DIVPD __m128d _mm_div_pd (__m128d a, __m128d b);

SIMD Floating-Point Exceptions

Overflow, Underflow, Invalid, Divide-by-Zero, Precision, Denormal

Other Exceptions

VEX-encoded instructions, see Exceptions Type 2.

EVEX-encoded instructions, see Exceptions Type E2.


Source: Intel® Architecture Software Developer's Manual (May 2018)
Generated: 5-6-2018

Clone this wiki locally