289 lines
12 KiB
HTML
289 lines
12 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>VPMOVQB/VPMOVSQB/VPMOVUSQB
|
||
— Down Convert QWord to Byte</title></head><body><header><nav><ul><li><a href='index.html'>Index</a></li><li>December 2023</li></ul></nav></header><h1>VPMOVQB/VPMOVSQB/VPMOVUSQB
|
||
— Down Convert QWord to Byte</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.F3.0F38.W0 32 /r VPMOVQB xmm1/m16 {k1}{z}, xmm2</td>
|
||
<td>A</td>
|
||
<td>V/V</td>
|
||
<td>AVX512VL AVX512F</td>
|
||
<td>Converts 2 packed quad-word integers from xmm2 into 2 packed byte integers in xmm1/m16 with truncation under writemask k1.</td></tr>
|
||
<tr>
|
||
<td>EVEX.128.F3.0F38.W0 22 /r VPMOVSQB xmm1/m16 {k1}{z}, xmm2</td>
|
||
<td>A</td>
|
||
<td>V/V</td>
|
||
<td>AVX512VL AVX512F</td>
|
||
<td>Converts 2 packed signed quad-word integers from xmm2 into 2 packed signed byte integers in xmm1/m16 using signed saturation under writemask k1.</td></tr>
|
||
<tr>
|
||
<td>EVEX.128.F3.0F38.W0 12 /r VPMOVUSQB xmm1/m16 {k1}{z}, xmm2</td>
|
||
<td>A</td>
|
||
<td>V/V</td>
|
||
<td>AVX512VL AVX512F</td>
|
||
<td>Converts 2 packed unsigned quad-word integers from xmm2 into 2 packed unsigned byte integers in xmm1/m16 using unsigned saturation under writemask k1.</td></tr>
|
||
<tr>
|
||
<td>EVEX.256.F3.0F38.W0 32 /r VPMOVQB xmm1/m32 {k1}{z}, ymm2</td>
|
||
<td>A</td>
|
||
<td>V/V</td>
|
||
<td>AVX512VL AVX512F</td>
|
||
<td>Converts 4 packed quad-word integers from ymm2 into 4 packed byte integers in xmm1/m32 with truncation under writemask k1.</td></tr>
|
||
<tr>
|
||
<td>EVEX.256.F3.0F38.W0 22 /r VPMOVSQB xmm1/m32 {k1}{z}, ymm2</td>
|
||
<td>A</td>
|
||
<td>V/V</td>
|
||
<td>AVX512VL AVX512F</td>
|
||
<td>Converts 4 packed signed quad-word integers from ymm2 into 4 packed signed byte integers in xmm1/m32 using signed saturation under writemask k1.</td></tr>
|
||
<tr>
|
||
<td>EVEX.256.F3.0F38.W0 12 /r VPMOVUSQB xmm1/m32 {k1}{z}, ymm2</td>
|
||
<td>A</td>
|
||
<td>V/V</td>
|
||
<td>AVX512VL AVX512F</td>
|
||
<td>Converts 4 packed unsigned quad-word integers from ymm2 into 4 packed unsigned byte integers in xmm1/m32 using unsigned saturation under writemask k1.</td></tr>
|
||
<tr>
|
||
<td>EVEX.512.F3.0F38.W0 32 /r VPMOVQB xmm1/m64 {k1}{z}, zmm2</td>
|
||
<td>A</td>
|
||
<td>V/V</td>
|
||
<td>AVX512F</td>
|
||
<td>Converts 8 packed quad-word integers from zmm2 into 8 packed byte integers in xmm1/m64 with truncation under writemask k1.</td></tr>
|
||
<tr>
|
||
<td>EVEX.512.F3.0F38.W0 22 /r VPMOVSQB xmm1/m64 {k1}{z}, zmm2</td>
|
||
<td>A</td>
|
||
<td>V/V</td>
|
||
<td>AVX512F</td>
|
||
<td>Converts 8 packed signed quad-word integers from zmm2 into 8 packed signed byte integers in xmm1/m64 using signed saturation under writemask k1.</td></tr>
|
||
<tr>
|
||
<td>EVEX.512.F3.0F38.W0 12 /r VPMOVUSQB xmm1/m64 {k1}{z}, zmm2</td>
|
||
<td>A</td>
|
||
<td>V/V</td>
|
||
<td>AVX512F</td>
|
||
<td>Converts 8 packed unsigned quad-word integers from zmm2 into 8 packed unsigned byte integers in xmm1/m64 using unsigned saturation under writemask k1.</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 Type</th>
|
||
<th>Operand 1</th>
|
||
<th>Operand 2</th>
|
||
<th>Operand 3</th>
|
||
<th>Operand 4</th></tr>
|
||
<tr>
|
||
<td>A</td>
|
||
<td>Eighth Mem</td>
|
||
<td>ModRM:r/m (w)</td>
|
||
<td>ModRM:reg (r)</td>
|
||
<td>N/A</td>
|
||
<td>N/A</td></tr></table>
|
||
<h3 id="description">Description<a class="anchor" href="#description">
|
||
¶
|
||
</a></h3>
|
||
<p>VPMOVQB down converts 64-bit integer elements in the source operand (the second operand) into packed byte elements using truncation. VPMOVSQB converts signed 64-bit integers into packed signed bytes using signed saturation. VPMOVUSQB convert unsigned quad-word values into unsigned byte values using unsigned saturation. The source operand is a vector register. The destination operand is an XMM register or a memory location.</p>
|
||
<p>Down-converted byte elements are written to the destination operand (the first operand) from the least-significant byte. Byte elements of the destination operand are updated according to the writemask. Bits (MAXVL-1:64) of the destination are zeroed.</p>
|
||
<p>EVEX.vvvv is reserved and must be 1111b otherwise instructions will #UD.</p>
|
||
<h3 id="operation">Operation<a class="anchor" href="#operation">
|
||
¶
|
||
</a></h3>
|
||
<h4 id="vpmovqb-instruction--evex-encoded-versions--when-dest-is-a-register">VPMOVQB instruction (EVEX encoded versions) when dest is a register<a class="anchor" href="#vpmovqb-instruction--evex-encoded-versions--when-dest-is-a-register">
|
||
¶
|
||
</a></h4>
|
||
<pre>(KL, VL) = (2, 128), (4, 256), (8, 512)
|
||
FOR j := 0 TO KL-1
|
||
i := j * 8
|
||
m := j * 64
|
||
IF k1[j] OR *no writemask*
|
||
THEN DEST[i+7:i] := TruncateQuadWordToByte (SRC[m+63:m])
|
||
ELSE
|
||
IF *merging-masking* ; merging-masking
|
||
THEN *DEST[i+7:i] remains unchanged*
|
||
ELSE *zeroing-masking*
|
||
; zeroing-masking
|
||
DEST[i+7:i] := 0
|
||
FI
|
||
FI;
|
||
ENDFOR
|
||
DEST[MAXVL-1:VL/8] := 0;
|
||
</pre>
|
||
<h4 id="vpmovqb-instruction--evex-encoded-versions--when-dest-is-memory">VPMOVQB instruction (EVEX encoded versions) when dest is memory<a class="anchor" href="#vpmovqb-instruction--evex-encoded-versions--when-dest-is-memory">
|
||
¶
|
||
</a></h4>
|
||
<pre>(KL, VL) = (2, 128), (4, 256), (8, 512)
|
||
FOR j := 0 TO KL-1
|
||
i := j * 8
|
||
m := j * 64
|
||
IF k1[j] OR *no writemask*
|
||
THEN DEST[i+7:i] := TruncateQuadWordToByte (SRC[m+63:m])
|
||
ELSE
|
||
*DEST[i+7:i] remains unchanged* ; merging-masking
|
||
FI;
|
||
ENDFOR
|
||
</pre>
|
||
<h4 id="vpmovsqb-instruction--evex-encoded-versions--when-dest-is-a-register">VPMOVSQB instruction (EVEX encoded versions) when dest is a register<a class="anchor" href="#vpmovsqb-instruction--evex-encoded-versions--when-dest-is-a-register">
|
||
¶
|
||
</a></h4>
|
||
<pre>(KL, VL) = (2, 128), (4, 256), (8, 512)
|
||
FOR j := 0 TO KL-1
|
||
i := j * 8
|
||
m := j * 64
|
||
IF k1[j] OR *no writemask*
|
||
THEN DEST[i+7:i] := SaturateSignedQuadWordToByte (SRC[m+63:m])
|
||
ELSE
|
||
IF *merging-masking* ; merging-masking
|
||
THEN *DEST[i+7:i] remains unchanged*
|
||
ELSE *zeroing-masking*
|
||
; zeroing-masking
|
||
DEST[i+7:i] := 0
|
||
FI
|
||
FI;
|
||
ENDFOR
|
||
DEST[MAXVL-1:VL/8] := 0;
|
||
</pre>
|
||
<h4 id="vpmovsqb-instruction--evex-encoded-versions--when-dest-is-memory">VPMOVSQB instruction (EVEX encoded versions) when dest is memory<a class="anchor" href="#vpmovsqb-instruction--evex-encoded-versions--when-dest-is-memory">
|
||
¶
|
||
</a></h4>
|
||
<pre>(KL, VL) = (2, 128), (4, 256), (8, 512)
|
||
FOR j := 0 TO KL-1
|
||
i := j * 8
|
||
m := j * 64
|
||
IF k1[j] OR *no writemask*
|
||
THEN DEST[i+7:i] := SaturateSignedQuadWordToByte (SRC[m+63:m])
|
||
ELSE
|
||
*DEST[i+7:i] remains unchanged* ; merging-masking
|
||
FI;
|
||
ENDFOR
|
||
</pre>
|
||
<h4 id="vpmovusqb-instruction--evex-encoded-versions--when-dest-is-a-register">VPMOVUSQB instruction (EVEX encoded versions) when dest is a register<a class="anchor" href="#vpmovusqb-instruction--evex-encoded-versions--when-dest-is-a-register">
|
||
¶
|
||
</a></h4>
|
||
<pre>(KL, VL) = (2, 128), (4, 256), (8, 512)
|
||
FOR j := 0 TO KL-1
|
||
i := j * 8
|
||
m := j * 64
|
||
IF k1[j] OR *no writemask*
|
||
THEN DEST[i+7:i] := SaturateUnsignedQuadWordToByte (SRC[m+63:m])
|
||
ELSE
|
||
IF *merging-masking* ; merging-masking
|
||
THEN *DEST[i+7:i] remains unchanged*
|
||
ELSE *zeroing-masking*
|
||
; zeroing-masking
|
||
DEST[i+7:i] := 0
|
||
FI
|
||
FI;
|
||
ENDFOR
|
||
DEST[MAXVL-1:VL/8] := 0;
|
||
</pre>
|
||
<h4 id="vpmovusqb-instruction--evex-encoded-versions--when-dest-is-memory">VPMOVUSQB instruction (EVEX encoded versions) when dest is memory<a class="anchor" href="#vpmovusqb-instruction--evex-encoded-versions--when-dest-is-memory">
|
||
¶
|
||
</a></h4>
|
||
<pre>(KL, VL) = (2, 128), (4, 256), (8, 512)
|
||
FOR j := 0 TO KL-1
|
||
i := j * 8
|
||
m := j * 64
|
||
IF k1[j] OR *no writemask*
|
||
THEN DEST[i+7:i] := SaturateUnsignedQuadWordToByte (SRC[m+63:m])
|
||
ELSE
|
||
*DEST[i+7:i] remains unchanged* ; merging-masking
|
||
FI;
|
||
ENDFOR
|
||
</pre>
|
||
<h3 id="intel-c-c++-compiler-intrinsic-equivalents">Intel C/C++ Compiler Intrinsic Equivalents<a class="anchor" href="#intel-c-c++-compiler-intrinsic-equivalents">
|
||
¶
|
||
</a></h3>
|
||
<pre>VPMOVQB __m128i _mm512_cvtepi64_epi8( __m512i a);
|
||
</pre>
|
||
<pre>VPMOVQB __m128i _mm512_mask_cvtepi64_epi8(__m128i s, __mmask8 k, __m512i a);
|
||
</pre>
|
||
<pre>VPMOVQB __m128i _mm512_maskz_cvtepi64_epi8( __mmask8 k, __m512i a);
|
||
</pre>
|
||
<pre>VPMOVQB void _mm512_mask_cvtepi64_storeu_epi8(void * d, __mmask8 k, __m512i a);
|
||
</pre>
|
||
<pre>VPMOVSQB __m128i _mm512_cvtsepi64_epi8( __m512i a);
|
||
</pre>
|
||
<pre>VPMOVSQB __m128i _mm512_mask_cvtsepi64_epi8(__m128i s, __mmask8 k, __m512i a);
|
||
</pre>
|
||
<pre>VPMOVSQB __m128i _mm512_maskz_cvtsepi64_epi8( __mmask8 k, __m512i a);
|
||
</pre>
|
||
<pre>VPMOVSQB void _mm512_mask_cvtsepi64_storeu_epi8(void * d, __mmask8 k, __m512i a);
|
||
</pre>
|
||
<pre>VPMOVUSQB __m128i _mm512_cvtusepi64_epi8( __m512i a);
|
||
</pre>
|
||
<pre>VPMOVUSQB __m128i _mm512_mask_cvtusepi64_epi8(__m128i s, __mmask8 k, __m512i a);
|
||
</pre>
|
||
<pre>VPMOVUSQB __m128i _mm512_maskz_cvtusepi64_epi8( __mmask8 k, __m512i a);
|
||
</pre>
|
||
<pre>VPMOVUSQB void _mm512_mask_cvtusepi64_storeu_epi8(void * d, __mmask8 k, __m512i a);
|
||
</pre>
|
||
<pre>VPMOVUSQB __m128i _mm256_cvtusepi64_epi8(__m256i a);
|
||
</pre>
|
||
<pre>VPMOVUSQB __m128i _mm256_mask_cvtusepi64_epi8(__m128i a, __mmask8 k, __m256i b);
|
||
</pre>
|
||
<pre>VPMOVUSQB __m128i _mm256_maskz_cvtusepi64_epi8( __mmask8 k, __m256i b);
|
||
</pre>
|
||
<pre>VPMOVUSQB void _mm256_mask_cvtusepi64_storeu_epi8(void * , __mmask8 k, __m256i b);
|
||
</pre>
|
||
<pre>VPMOVUSQB __m128i _mm_cvtusepi64_epi8(__m128i a);
|
||
</pre>
|
||
<pre>VPMOVUSQB __m128i _mm_mask_cvtusepi64_epi8(__m128i a, __mmask8 k, __m128i b);
|
||
</pre>
|
||
<pre>VPMOVUSQB __m128i _mm_maskz_cvtusepi64_epi8( __mmask8 k, __m128i b);
|
||
</pre>
|
||
<pre>VPMOVUSQB void _mm_mask_cvtusepi64_storeu_epi8(void * , __mmask8 k, __m128i b);
|
||
</pre>
|
||
<pre>VPMOVSQB __m128i _mm256_cvtsepi64_epi8(__m256i a);
|
||
</pre>
|
||
<pre>VPMOVSQB __m128i _mm256_mask_cvtsepi64_epi8(__m128i a, __mmask8 k, __m256i b);
|
||
</pre>
|
||
<pre>VPMOVSQB __m128i _mm256_maskz_cvtsepi64_epi8( __mmask8 k, __m256i b);
|
||
</pre>
|
||
<pre>VPMOVSQB void _mm256_mask_cvtsepi64_storeu_epi8(void * , __mmask8 k, __m256i b);
|
||
</pre>
|
||
<pre>VPMOVSQB __m128i _mm_cvtsepi64_epi8(__m128i a);
|
||
</pre>
|
||
<pre>VPMOVSQB __m128i _mm_mask_cvtsepi64_epi8(__m128i a, __mmask8 k, __m128i b);
|
||
</pre>
|
||
<pre>VPMOVSQB __m128i _mm_maskz_cvtsepi64_epi8( __mmask8 k, __m128i b);
|
||
</pre>
|
||
<pre>VPMOVSQB void _mm_mask_cvtsepi64_storeu_epi8(void * , __mmask8 k, __m128i b);
|
||
</pre>
|
||
<pre>VPMOVQB __m128i _mm256_cvtepi64_epi8(__m256i a);
|
||
</pre>
|
||
<pre>VPMOVQB __m128i _mm256_mask_cvtepi64_epi8(__m128i a, __mmask8 k, __m256i b);
|
||
</pre>
|
||
<pre>VPMOVQB __m128i _mm256_maskz_cvtepi64_epi8( __mmask8 k, __m256i b);
|
||
</pre>
|
||
<pre>VPMOVQB void _mm256_mask_cvtepi64_storeu_epi8(void * , __mmask8 k, __m256i b);
|
||
</pre>
|
||
<pre>VPMOVQB __m128i _mm_cvtepi64_epi8(__m128i a);
|
||
</pre>
|
||
<pre>VPMOVQB __m128i _mm_mask_cvtepi64_epi8(__m128i a, __mmask8 k, __m128i b);
|
||
</pre>
|
||
<pre>VPMOVQB __m128i _mm_maskz_cvtepi64_epi8( __mmask8 k, __m128i b);
|
||
</pre>
|
||
<pre>VPMOVQB void _mm_mask_cvtepi64_storeu_epi8(void * , __mmask8 k, __m128i b);
|
||
</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>EVEX-encoded instruction, see <span class="not-imported">Table 2-53</span>, “Type E6 Class Exception Conditions.”</p>
|
||
<p>Additionally:</p>
|
||
<table>
|
||
<tr>
|
||
<td>#UD</td>
|
||
<td>If EVEX.vvvv != 1111B.</td></tr></table><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>
|