NRZCode/ia32-64
1
0
Fork 1
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
ia32-64/x86/vpshldv.html

230 lines
8.8 KiB
HTML
Raw Normal View History

Initial commit
2025-07-08 02:23:29 -03:00
<!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>VPSHLDV
— Concatenate and Variable Shift Packed Data Left Logical</title></head><body><header><nav><ul><li><a href='index.html'>Index</a></li><li>December 2023</li></ul></nav></header><h1>VPSHLDV
— Concatenate and Variable Shift Packed Data Left 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 70 /r VPSHLDVW 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 left by value in xmm3/m128 into xmm1.</td></tr>
<tr>
<td>EVEX.256.66.0F38.W1 70 /r VPSHLDVW 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 left by value in xmm3/m256 into ymm1.</td></tr>
<tr>
<td>EVEX.512.66.0F38.W1 70 /r VPSHLDVW 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 left by value in zmm3/m512 into zmm1.</td></tr>
<tr>
<td>EVEX.128.66.0F38.W0 71 /r VPSHLDVD 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 left by value in xmm3/m128 into xmm1.</td></tr>
<tr>
<td>EVEX.256.66.0F38.W0 71 /r VPSHLDVD 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 left by value in xmm3/m256 into ymm1.</td></tr>
<tr>
<td>EVEX.512.66.0F38.W0 71 /r VPSHLDVD 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 left by value in zmm3/m512 into zmm1.</td></tr>
<tr>
<td>EVEX.128.66.0F38.W1 71 /r VPSHLDVQ 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 left by value in xmm3/m128 into xmm1.</td></tr>
<tr>
<td>EVEX.256.66.0F38.W1 71 /r VPSHLDVQ 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 left by value in xmm3/m256 into ymm1.</td></tr>
<tr>
<td>EVEX.512.66.0F38.W1 71 /r VPSHLDVQ 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 left 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 left by variable value.</p>
<p>This instruction supports memory fault suppression.</p>
<h3 id="operation">Operation<a class="anchor" href="#operation">
</a></h3>
<pre>FUNCTION concat(a,b):
IF words:
d.word[1] := a
d.word[0] := b
return d
ELSE IF dwords:
q.dword[1] := a
q.dword[0] := b
return q
ELSE IF qwords:
o.qword[1] := a
o.qword[0] := b
return o
</pre>
<h4 id="vpshldvw-dest--src2--src3">VPSHLDVW DEST, SRC2, SRC3<a class="anchor" href="#vpshldvw-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*:
tmp := concat(DEST.word[j], SRC2.word[j]) &lt;&lt; (SRC3.word[j] &amp; 15)
DEST.word[j] := tmp.word[1]
ELSE IF *zeroing*:
DEST.word[j] := 0
*ELSE DEST.word[j] remains unchanged*
DEST[MAX_VL-1:VL] := 0
</pre>
<h4 id="vpshldvd-dest--src2--src3">VPSHLDVD DEST, SRC2, SRC3<a class="anchor" href="#vpshldvd-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*:
tmp := concat(DEST.dword[j], SRC2.dword[j]) &lt;&lt; (tsrc3 &amp; 31)
DEST.dword[j] := tmp.dword[1]
ELSE IF *zeroing*:
DEST.dword[j] := 0
*ELSE DEST.dword[j] remains unchanged*
DEST[MAX_VL-1:VL] := 0
</pre>
<h4 id="vpshldvq-dest--src2--src3">VPSHLDVQ DEST, SRC2, SRC3<a class="anchor" href="#vpshldvq-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*:
tmp := concat(DEST.qword[j], SRC2.qword[j]) &lt;&lt; (tsrc3 &amp; 63)
DEST.qword[j] := tmp.qword[1]
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>VPSHLDVW __m128i _mm_shldv_epi16(__m128i, __m128i, __m128i);
</pre>
<pre>VPSHLDVW __m128i _mm_mask_shldv_epi16(__m128i, __mmask8, __m128i, __m128i);
</pre>
<pre>VPSHLDVW __m128i _mm_maskz_shldv_epi16(__mmask8, __m128i, __m128i, __m128i);
</pre>
<pre>VPSHLDVW __m256i _mm256_shldv_epi16(__m256i, __m256i, __m256i);
</pre>
<pre>VPSHLDVW __m256i _mm256_mask_shldv_epi16(__m256i, __mmask16, __m256i, __m256i);
</pre>
<pre>VPSHLDVW __m256i _mm256_maskz_shldv_epi16(__mmask16, __m256i, __m256i, __m256i);
</pre>
<pre>VPSHLDVQ __m512i _mm512_shldv_epi64(__m512i, __m512i, __m512i);
</pre>
<pre>VPSHLDVQ __m512i _mm512_mask_shldv_epi64(__m512i, __mmask8, __m512i, __m512i);
</pre>
<pre>VPSHLDVQ __m512i _mm512_maskz_shldv_epi64(__mmask8, __m512i, __m512i, __m512i);
</pre>
<pre>VPSHLDVW __m128i _mm_shldv_epi16(__m128i, __m128i, __m128i);
</pre>
<pre>VPSHLDVW __m128i _mm_mask_shldv_epi16(__m128i, __mmask8, __m128i, __m128i);
</pre>
<pre>VPSHLDVW __m128i _mm_maskz_shldv_epi16(__mmask8, __m128i, __m128i, __m128i);
</pre>
<pre>VPSHLDVW __m256i _mm256_shldv_epi16(__m256i, __m256i, __m256i);
</pre>
<pre>VPSHLDVW __m256i _mm256_mask_shldv_epi16(__m256i, __mmask16, __m256i, __m256i);
</pre>
<pre>VPSHLDVW __m256i _mm256_maskz_shldv_epi16(__mmask16, __m256i, __m256i, __m256i);
</pre>
<pre>VPSHLDVW __m512i _mm512_shldv_epi16(__m512i, __m512i, __m512i);
</pre>
<pre>VPSHLDVW __m512i _mm512_mask_shldv_epi16(__m512i, __mmask32, __m512i, __m512i);
</pre>
<pre>VPSHLDVW __m512i _mm512_maskz_shldv_epi16(__mmask32, __m512i, __m512i, __m512i);
</pre>
<pre>VPSHLDVD __m128i _mm_shldv_epi32(__m128i, __m128i, __m128i);
</pre>
<pre>VPSHLDVD __m128i _mm_mask_shldv_epi32(__m128i, __mmask8, __m128i, __m128i);
</pre>
<pre>VPSHLDVD __m128i _mm_maskz_shldv_epi32(__mmask8, __m128i, __m128i, __m128i);
</pre>
<pre>VPSHLDVD __m256i _mm256_shldv_epi32(__m256i, __m256i, __m256i);
</pre>
<pre>VPSHLDVD __m256i _mm256_mask_shldv_epi32(__m256i, __mmask8, __m256i, __m256i);
</pre>
<pre>VPSHLDVD __m256i _mm256_maskz_shldv_epi32(__mmask8, __m256i, __m256i, __m256i);
</pre>
<pre>VPSHLDVD __m512i _mm512_shldv_epi32(__m512i, __m512i, __m512i);
</pre>
<pre>VPSHLDVD __m512i _mm512_mask_shldv_epi32(__m512i, __mmask16, __m512i, __m512i);
</pre>
<pre>VPSHLDVD __m512i _mm512_maskz_shldv_epi32(__mmask16, __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>
Reference in a new issue Copy permalink