ia32-64/x86/vcmpsh.html
2025-07-08 02:23:29 -03:00

112 lines
4.8 KiB
HTML
Raw Permalink Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

<!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>VCMPSH
— Compare Scalar FP16 Values</title></head><body><header><nav><ul><li><a href='index.html'>Index</a></li><li>December 2023</li></ul></nav></header><h1>VCMPSH
— Compare Scalar FP16 Values</h1>
<table>
<tr>
<th> Instruction En Bit Mode Flag
Support Instruction En Bit Mode Flag
Support 64/32 CPUID Feature Instruction En Bit Mode Flag CPUID Feature Instruction En Bit Mode Flag Op/ 64/32 CPUID Feature Instruction En Bit Mode Flag 64/32 CPUID Feature Instruction En Bit Mode Flag CPUID Feature Instruction En Bit Mode Flag Op/ 64/32 CPUID Feature </th>
<th></th>
<th>Support</th>
<th></th>
<th>Description</th></tr>
<tr>
<td>EVEX.LLIG.F3.0F3A.W0 C2 /r /ib VCMPSH k1{k2}, xmm2, xmm3/m16 {sae}, imm8</td>
<td>A</td>
<td>V/V</td>
<td>AVX512-FP16</td>
<td>Compare low FP16 values in xmm3/m16 and xmm2 using bits 4:0 of imm8 as a comparison predicate subject to writemask k2, and store the result in mask register 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</th>
<th>Operand 1</th>
<th>Operand 2</th>
<th>Operand 3</th>
<th>Operand 4</th></tr>
<tr>
<td>A</td>
<td>Scalar</td>
<td>ModRM:reg (w)</td>
<td>VEX.vvvv (r)</td>
<td>ModRM:r/m (r)</td>
<td>imm8 (r)</td></tr></table>
<h3 id="description">Description<a class="anchor" href="#description">
</a></h3>
<p>This instruction compares the FP16 values from the lowest element of the source operands and stores the result in the destination mask operand. The comparison predicate operand (immediate byte bits 4:0) specifies the type of comparison performed on the pair of packed FP16 values. The low destination bit is updated according to the writemask. Bits MAXKL-1:1 of the destination operand are zeroed.</p>
<h3 id="operation">Operation<a class="anchor" href="#operation">
</a></h3>
<pre>CASE (imm8 &amp; 0x1F) OF
0: CMP_OPERATOR := EQ_OQ;
1: CMP_OPERATOR := LT_OS;
2: CMP_OPERATOR := LE_OS;
3: CMP_OPERATOR := UNORD_Q;
4: CMP_OPERATOR := NEQ_UQ;
5: CMP_OPERATOR := NLT_US;
6: CMP_OPERATOR := NLE_US;
7: CMP_OPERATOR := ORD_Q;
8: CMP_OPERATOR := EQ_UQ;
9: CMP_OPERATOR := NGE_US;
10: CMP_OPERATOR := NGT_US;
11: CMP_OPERATOR := FALSE_OQ;
12: CMP_OPERATOR := NEQ_OQ;
13: CMP_OPERATOR := GE_OS;
14: CMP_OPERATOR := GT_OS;
15: CMP_OPERATOR := TRUE_UQ;
16: CMP_OPERATOR := EQ_OS;
17: CMP_OPERATOR := LT_OQ;
18: CMP_OPERATOR := LE_OQ;
19: CMP_OPERATOR := UNORD_S;
20: CMP_OPERATOR := NEQ_US;
21: CMP_OPERATOR := NLT_UQ;
22: CMP_OPERATOR := NLE_UQ;
23: CMP_OPERATOR := ORD_S;
24: CMP_OPERATOR := EQ_US;
25: CMP_OPERATOR := NGE_UQ;
26: CMP_OPERATOR := NGT_UQ;
27: CMP_OPERATOR := FALSE_OS;
28: CMP_OPERATOR := NEQ_OS;
29: CMP_OPERATOR := GE_OQ;
30: CMP_OPERATOR := GT_OQ;
31: CMP_OPERATOR := TRUE_US;
ESAC
</pre>
<h4 id="vcmpsh--evex-encoded-versions-">VCMPSH (EVEX Encoded Versions)<a class="anchor" href="#vcmpsh--evex-encoded-versions-">
</a></h4>
<pre>IF k2[0] OR *no writemask*:
DEST.bit[0] := SRC1.fp16[0] CMP_OPERATOR SRC2.fp16[0]
ELSE
DEST.bit[0] := 0
DEST[MAXKL-1:1] := 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>VCMPSH __mmask8 _mm_cmp_round_sh_mask (__m128h a, __m128h b, const int imm8, const int sae);
</pre>
<pre>VCMPSH __mmask8 _mm_mask_cmp_round_sh_mask (__mmask8 k1, __m128h a, __m128h b, const int imm8, const int sae);
</pre>
<pre>VCMPSH __mmask8 _mm_cmp_sh_mask (__m128h a, __m128h b, const int imm8);
</pre>
<pre>VCMPSH __mmask8 _mm_mask_cmp_sh_mask (__mmask8 k1, __m128h a, __m128h b, const int imm8);
</pre>
<h3 class="exceptions" id="simd-floating-point-exceptions">SIMD Floating-Point Exceptions<a class="anchor" href="#simd-floating-point-exceptions">
</a></h3>
<p>Invalid, Denormal.</p>
<h3 class="exceptions" id="other-exceptions">Other Exceptions<a class="anchor" href="#other-exceptions">
</a></h3>
<p>EVEX-encoded instructions, see <span class="not-imported">Table 2-47</span>, “Type E3 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 Developers Manual</a> for anything serious.
</p></footer></body></html>