forked from NRZCode/ia32-64
212 lines
8.5 KiB
HTML
212 lines
8.5 KiB
HTML
<!DOCTYPE html>
|
||
<html xmlns="http://www.w3.org/1999/xhtml" xmlns:svg="http://www.w3.org/2000/svg" xmlns:x86="http://www.felixcloutier.com/x86"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8"><link rel="stylesheet" type="text/css" href="style.css"></link><title>VPSHRDV
|
||
— Concatenate and Variable Shift Packed Data Right Logical</title></head><body><header><nav><ul><li><a href='index.html'>Index</a></li><li>December 2023</li></ul></nav></header><h1>VPSHRDV
|
||
— Concatenate and Variable Shift Packed Data Right Logical</h1>
|
||
|
||
<table>
|
||
<tr>
|
||
<th>Opcode/Instruction</th>
|
||
<th>Op/En</th>
|
||
<th>64/32 bit Mode Support</th>
|
||
<th>CPUID Feature Flag</th>
|
||
<th>Description</th></tr>
|
||
<tr>
|
||
<td>EVEX.128.66.0F38.W1 72 /r VPSHRDVW xmm1{k1}{z}, xmm2, xmm3/m128</td>
|
||
<td>A</td>
|
||
<td>V/V</td>
|
||
<td>AVX512_VBMI2 AVX512VL</td>
|
||
<td>Concatenate xmm1 and xmm2, extract result shifted to the right by value in xmm3/m128 into xmm1.</td></tr>
|
||
<tr>
|
||
<td>EVEX.256.66.0F38.W1 72 /r VPSHRDVW ymm1{k1}{z}, ymm2, ymm3/m256</td>
|
||
<td>A</td>
|
||
<td>V/V</td>
|
||
<td>AVX512_VBMI2 AVX512VL</td>
|
||
<td>Concatenate ymm1 and ymm2, extract result shifted to the right by value in xmm3/m256 into ymm1.</td></tr>
|
||
<tr>
|
||
<td>EVEX.512.66.0F38.W1 72 /r VPSHRDVW zmm1{k1}{z}, zmm2, zmm3/m512</td>
|
||
<td>A</td>
|
||
<td>V/V</td>
|
||
<td>AVX512_VBMI2</td>
|
||
<td>Concatenate zmm1 and zmm2, extract result shifted to the right by value in zmm3/m512 into zmm1.</td></tr>
|
||
<tr>
|
||
<td>EVEX.128.66.0F38.W0 73 /r VPSHRDVD xmm1{k1}{z}, xmm2, xmm3/m128/m32bcst</td>
|
||
<td>B</td>
|
||
<td>V/V</td>
|
||
<td>AVX512_VBMI2 AVX512VL</td>
|
||
<td>Concatenate xmm1 and xmm2, extract result shifted to the right by value in xmm3/m128 into xmm1.</td></tr>
|
||
<tr>
|
||
<td>EVEX.256.66.0F38.W0 73 /r VPSHRDVD ymm1{k1}{z}, ymm2, ymm3/m256/m32bcst</td>
|
||
<td>B</td>
|
||
<td>V/V</td>
|
||
<td>AVX512_VBMI2 AVX512VL</td>
|
||
<td>Concatenate ymm1 and ymm2, extract result shifted to the right by value in xmm3/m256 into ymm1.</td></tr>
|
||
<tr>
|
||
<td>EVEX.512.66.0F38.W0 73 /r VPSHRDVD zmm1{k1}{z}, zmm2, zmm3/m512/m32bcst</td>
|
||
<td>B</td>
|
||
<td>V/V</td>
|
||
<td>AVX512_VBMI2</td>
|
||
<td>Concatenate zmm1 and zmm2, extract result shifted to the right by value in zmm3/m512 into zmm1.</td></tr>
|
||
<tr>
|
||
<td>EVEX.128.66.0F38.W1 73 /r VPSHRDVQ xmm1{k1}{z}, xmm2, xmm3/m128/m64bcst</td>
|
||
<td>B</td>
|
||
<td>V/V</td>
|
||
<td>AVX512_VBMI2 AVX512VL</td>
|
||
<td>Concatenate xmm1 and xmm2, extract result shifted to the right by value in xmm3/m128 into xmm1.</td></tr>
|
||
<tr>
|
||
<td>EVEX.256.66.0F38.W1 73 /r VPSHRDVQ ymm1{k1}{z}, ymm2, ymm3/m256/m64bcst</td>
|
||
<td>B</td>
|
||
<td>V/V</td>
|
||
<td>AVX512_VBMI2 AVX512VL</td>
|
||
<td>Concatenate ymm1 and ymm2, extract result shifted to the right by value in xmm3/m256 into ymm1.</td></tr>
|
||
<tr>
|
||
<td>EVEX.512.66.0F38.W1 73 /r VPSHRDVQ zmm1{k1}{z}, zmm2, zmm3/m512/m64bcst</td>
|
||
<td>B</td>
|
||
<td>V/V</td>
|
||
<td>AVX512_VBMI2</td>
|
||
<td>Concatenate zmm1 and zmm2, extract result shifted to the right by value in zmm3/m512 into zmm1.</td></tr></table>
|
||
<h2 id="instruction-operand-encoding">Instruction Operand Encoding<a class="anchor" href="#instruction-operand-encoding">
|
||
¶
|
||
</a></h2>
|
||
<table>
|
||
<tr>
|
||
<th>Op/En</th>
|
||
<th>Tuple</th>
|
||
<th>Operand 1</th>
|
||
<th>Operand 2</th>
|
||
<th>Operand 3</th>
|
||
<th>Operand 4</th></tr>
|
||
<tr>
|
||
<td>A</td>
|
||
<td>Full Mem</td>
|
||
<td>ModRM:reg (r, w)</td>
|
||
<td>EVEX.vvvv (r)</td>
|
||
<td>ModRM:r/m (r)</td>
|
||
<td>N/A</td></tr>
|
||
<tr>
|
||
<td>B</td>
|
||
<td>Full</td>
|
||
<td>ModRM:reg (r, w)</td>
|
||
<td>EVEX.vvvv (r)</td>
|
||
<td>ModRM:r/m (r)</td>
|
||
<td>N/A</td></tr></table>
|
||
<h3 id="description">Description<a class="anchor" href="#description">
|
||
¶
|
||
</a></h3>
|
||
<p>Concatenate packed data, extract result shifted to the right by variable value.</p>
|
||
<p>This instruction supports memory fault suppression.</p>
|
||
<h3 id="operation">Operation<a class="anchor" href="#operation">
|
||
¶
|
||
</a></h3>
|
||
<h4 id="vpshrdvw-dest--src2--src3">VPSHRDVW DEST, SRC2, SRC3<a class="anchor" href="#vpshrdvw-dest--src2--src3">
|
||
¶
|
||
</a></h4>
|
||
<pre>(KL, VL) = (8, 128), (16, 256), (32, 512)
|
||
FOR j := 0 TO KL-1:
|
||
IF MaskBit(j) OR *no writemask*:
|
||
DEST.word[j] := concat(SRC2.word[j], DEST.word[j]) >> (SRC3.word[j] & 15)
|
||
ELSE IF *zeroing*:
|
||
DEST.word[j] := 0
|
||
*ELSE DEST.word[j] remains unchanged*
|
||
DEST[MAX_VL-1:VL] := 0
|
||
</pre>
|
||
<h4 id="vpshrdvd-dest--src2--src3">VPSHRDVD DEST, SRC2, SRC3<a class="anchor" href="#vpshrdvd-dest--src2--src3">
|
||
¶
|
||
</a></h4>
|
||
<pre>(KL, VL) = (4, 128), (8, 256), (16, 512)
|
||
FOR j := 0 TO KL-1:
|
||
IF SRC3 is broadcast memop:
|
||
tsrc3 := SRC3.dword[0]
|
||
ELSE:
|
||
tsrc3 := SRC3.dword[j]
|
||
IF MaskBit(j) OR *no writemask*:
|
||
DEST.dword[j] := concat(SRC2.dword[j], DEST.dword[j]) >> (tsrc3 & 31)
|
||
ELSE IF *zeroing*:
|
||
DEST.dword[j] := 0
|
||
*ELSE DEST.dword[j] remains unchanged*
|
||
DEST[MAX_VL-1:VL] := 0
|
||
</pre>
|
||
<h4 id="vpshrdvq-dest--src2--src3">VPSHRDVQ DEST, SRC2, SRC3<a class="anchor" href="#vpshrdvq-dest--src2--src3">
|
||
¶
|
||
</a></h4>
|
||
<pre>(KL, VL) = (2, 128), (4, 256), (8, 512)
|
||
FOR j := 0 TO KL-1:
|
||
IF SRC3 is broadcast memop:
|
||
tsrc3 := SRC3.qword[0]
|
||
ELSE:
|
||
tsrc3 := SRC3.qword[j]
|
||
IF MaskBit(j) OR *no writemask*:
|
||
DEST.qword[j] := concat(SRC2.qword[j], DEST.qword[j]) >> (tsrc3 & 63)
|
||
ELSE IF *zeroing*:
|
||
DEST.qword[j] := 0
|
||
*ELSE DEST.qword[j] remains unchanged*
|
||
DEST[MAX_VL-1:VL] := 0
|
||
</pre>
|
||
<h3 id="intel-c-c++-compiler-intrinsic-equivalent">Intel C/C++ Compiler Intrinsic Equivalent<a class="anchor" href="#intel-c-c++-compiler-intrinsic-equivalent">
|
||
¶
|
||
</a></h3>
|
||
<pre>VPSHRDVQ __m128i _mm_shrdv_epi64(__m128i, __m128i, __m128i);
|
||
</pre>
|
||
<pre>VPSHRDVQ __m128i _mm_mask_shrdv_epi64(__m128i, __mmask8, __m128i, __m128i);
|
||
</pre>
|
||
<pre>VPSHRDVQ __m128i _mm_maskz_shrdv_epi64(__mmask8, __m128i, __m128i, __m128i);
|
||
</pre>
|
||
<pre>VPSHRDVQ __m256i _mm256_shrdv_epi64(__m256i, __m256i, __m256i);
|
||
</pre>
|
||
<pre>VPSHRDVQ __m256i _mm256_mask_shrdv_epi64(__m256i, __mmask8, __m256i, __m256i);
|
||
</pre>
|
||
<pre>VPSHRDVQ __m256i _mm256_maskz_shrdv_epi64(__mmask8, __m256i, __m256i, __m256i);
|
||
</pre>
|
||
<pre>VPSHRDVQ __m512i _mm512_shrdv_epi64(__m512i, __m512i, __m512i);
|
||
</pre>
|
||
<pre>VPSHRDVQ __m512i _mm512_mask_shrdv_epi64(__m512i, __mmask8, __m512i, __m512i);
|
||
</pre>
|
||
<pre>VPSHRDVQ __m512i _mm512_maskz_shrdv_epi64(__mmask8, __m512i, __m512i, __m512i);
|
||
</pre>
|
||
<pre>VPSHRDVD __m128i _mm_shrdv_epi32(__m128i, __m128i, __m128i);
|
||
</pre>
|
||
<pre>VPSHRDVD __m128i _mm_mask_shrdv_epi32(__m128i, __mmask8, __m128i, __m128i);
|
||
</pre>
|
||
<pre>VPSHRDVD __m128i _mm_maskz_shrdv_epi32(__mmask8, __m128i, __m128i, __m128i);
|
||
</pre>
|
||
<pre>VPSHRDVD __m256i _mm256_shrdv_epi32(__m256i, __m256i, __m256i);
|
||
</pre>
|
||
<pre>VPSHRDVD __m256i _mm256_mask_shrdv_epi32(__m256i, __mmask8, __m256i, __m256i);
|
||
</pre>
|
||
<pre>VPSHRDVD __m256i _mm256_maskz_shrdv_epi32(__mmask8, __m256i, __m256i, __m256i);
|
||
</pre>
|
||
<pre>VPSHRDVD __m512i _mm512_shrdv_epi32(__m512i, __m512i, __m512i);
|
||
</pre>
|
||
<pre>VPSHRDVD __m512i _mm512_mask_shrdv_epi32(__m512i, __mmask16, __m512i, __m512i);
|
||
</pre>
|
||
<pre>VPSHRDVD __m512i _mm512_maskz_shrdv_epi32(__mmask16, __m512i, __m512i, __m512i);
|
||
</pre>
|
||
<pre>VPSHRDVW __m128i _mm_shrdv_epi16(__m128i, __m128i, __m128i);
|
||
</pre>
|
||
<pre>VPSHRDVW __m128i _mm_mask_shrdv_epi16(__m128i, __mmask8, __m128i, __m128i);
|
||
</pre>
|
||
<pre>VPSHRDVW __m128i _mm_maskz_shrdv_epi16(__mmask8, __m128i, __m128i, __m128i);
|
||
</pre>
|
||
<pre>VPSHRDVW __m256i _mm256_shrdv_epi16(__m256i, __m256i, __m256i);
|
||
</pre>
|
||
<pre>VPSHRDVW __m256i _mm256_mask_shrdv_epi16(__m256i, __mmask16, __m256i, __m256i);
|
||
</pre>
|
||
<pre>VPSHRDVW __m256i _mm256_maskz_shrdv_epi16(__mmask16, __m256i, __m256i, __m256i);
|
||
</pre>
|
||
<pre>VPSHRDVW __m512i _mm512_shrdv_epi16(__m512i, __m512i, __m512i);
|
||
</pre>
|
||
<pre>VPSHRDVW __m512i _mm512_mask_shrdv_epi16(__m512i, __mmask32, __m512i, __m512i);
|
||
</pre>
|
||
<pre>VPSHRDVW __m512i _mm512_maskz_shrdv_epi16(__mmask32, __m512i, __m512i, __m512i);
|
||
</pre>
|
||
<h3 class="exceptions" id="simd-floating-point-exceptions">SIMD Floating-Point Exceptions<a class="anchor" href="#simd-floating-point-exceptions">
|
||
¶
|
||
</a></h3>
|
||
<p>None.</p>
|
||
<h3 class="exceptions" id="other-exceptions">Other Exceptions<a class="anchor" href="#other-exceptions">
|
||
¶
|
||
</a></h3>
|
||
<p>See <span class="not-imported">Table 2-49</span>, “Type E4 Class Exception Conditions.”</p><footer><p>
|
||
This UNOFFICIAL, mechanically-separated, non-verified reference is provided for convenience, but it may be
|
||
inc<span style="opacity: 0.2">omp</span>lete or b<sub>r</sub>oke<sub>n</sub> in various obvious or non-obvious
|
||
ways. Refer to <a href="https://software.intel.com/en-us/download/intel-64-and-ia-32-architectures-sdm-combined-volumes-1-2a-2b-2c-2d-3a-3b-3c-3d-and-4">Intel® 64 and IA-32 Architectures Software Developer’s Manual</a> for anything serious.
|
||
</p></footer></body></html>
|