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

135 lines
9.4 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>ADDSUBPD
— Packed Double Precision Floating-Point Add/Subtract</title></head><body><header><nav><ul><li><a href='index.html'>Index</a></li><li>December 2023</li></ul></nav></header><h1>ADDSUBPD
— Packed Double Precision Floating-Point Add/Subtract</h1>
<table>
<tr>
<th>Opcode/Instruction</th>
<th>Op/En</th>
<th>64/32-bit Mode</th>
<th>CPUID Feature Flag</th>
<th>Description</th></tr>
<tr>
<td>66 0F D0 /r ADDSUBPD xmm1, xmm2/m128</td>
<td>RM</td>
<td>V/V</td>
<td>SSE3</td>
<td>Add/subtract double precision floating-point values from xmm2/m128 to xmm1.</td></tr>
<tr>
<td>VEX.128.66.0F.WIG D0 /r VADDSUBPD xmm1, xmm2, xmm3/m128</td>
<td>RVM</td>
<td>V/V</td>
<td>AVX</td>
<td>Add/subtract packed double precision floating-point values from xmm3/mem to xmm2 and stores result in xmm1.</td></tr>
<tr>
<td>VEX.256.66.0F.WIG D0 /r VADDSUBPD ymm1, ymm2, ymm3/m256</td>
<td>RVM</td>
<td>V/V</td>
<td>AVX</td>
<td>Add / subtract packed double precision floating-point values from ymm3/mem to ymm2 and stores result in ymm1.</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>Operand 1</th>
<th>Operand 2</th>
<th>Operand 3</th>
<th>Operand 4</th></tr>
<tr>
<td>RM</td>
<td>ModRM:reg (r, w)</td>
<td>ModRM:r/m (r)</td>
<td>N/A</td>
<td>N/A</td></tr>
<tr>
<td>RVM</td>
<td>ModRM:reg (w)</td>
<td>VEX.vvvv (r)</td>
<td>ModRM:r/m (r)</td>
<td>N/A</td></tr></table>
<h2 id="description">Description<a class="anchor" href="#description">
</a></h2>
<p>Adds odd-numbered double precision floating-point values of the first source operand (second operand) with the corresponding double precision floating-point values from the second source operand (third operand); stores the result in the odd-numbered values of the destination operand (first operand). Subtracts the even-numbered double precision floating-point values from the second source operand from the corresponding double precision floating values in the first source operand; stores the result into the even-numbered values of the destination operand.</p>
<p>In 64-bit mode, using a REX prefix in the form of REX.R permits this instruction to access additional registers (XMM8-XMM15).</p>
<p>128-bit Legacy SSE version: The second source can be an XMM register or an 128-bit memory location. The destination is not distinct from the first source XMM register and the upper bits (MAXVL-1:128) of the corresponding YMM register destination are unmodified. See <a href='addsubpd.html#fig-3-3'>Figure 3-3</a>.</p>
<p>VEX.128 encoded version: the first source operand is an XMM register or 128-bit memory location. The destination operand is an XMM register. The upper bits (MAXVL-1:128) of the corresponding YMM register destination are zeroed.</p>
<p>VEX.256 encoded version: The first source 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.</p>
<figure id="fig-3-3">
<svg style="width: 432.018pt; height: 151.206pt" viewBox="128.824 0.0 365.015 131.005">
<g xmlns="http://www.w3.org/2000/svg" style="stroke: none; fill: none">
<rect height="126.00500000000001" style="stroke: rgb(0%, 0%, 0%)" width="360.015" x="131.324" y="0.0"></rect>
<rect height="27.001" style="fill: rgb(0%, 0%, 0%)" width="144.006" x="144.82500000000002" y="25.875999999999976"></rect>
<rect height="27.001" style="stroke: rgb(0%, 0%, 0%)" width="144.006" x="144.82500000000002" y="25.875999999999976"></rect>
<rect height="27.001" style="fill: rgb(0%, 0%, 0%)" width="144.006" x="144.82500000000002" y="74.25300000000004"></rect>
<rect height="27.001" style="stroke: rgb(0%, 0%, 0%)" width="144.006" x="144.82500000000002" y="74.25300000000004"></rect>
<rect height="27.001" style="fill: rgb(0%, 0%, 0%)" width="144.006" x="288.831" y="74.25300000000004"></rect>
<rect height="27.001" style="stroke: rgb(0%, 0%, 0%)" width="144.006" x="288.831" y="74.25300000000004"></rect>
<rect height="27.001" style="fill: rgb(0%, 0%, 0%)" width="144.006" x="288.831" y="25.875999999999976"></rect>
<rect height="27.001" style="stroke: rgb(0%, 0%, 0%)" width="144.006" x="288.831" y="25.875999999999976"></rect>
<path d="M 216.828 69.09300000000007 L 216.828 52.88200000000006" style="fill-rule: nonzero; stroke: rgb(0%, 0%, 0%)"></path>
<path d="M 219.588 68.73300000000006 L 216.828 74.25300000000004 L 214.06799999999998 68.73300000000006 L 219.588 68.73300000000006" style="fill: rgb(0%, 0%, 0%); fill-rule: evenodd"></path>
<path d="M 360.834 69.09300000000007 L 360.834 52.88200000000006" style="fill-rule: nonzero; stroke: rgb(0%, 0%, 0%)"></path>
<path d="M 363.594 68.73300000000006 L 360.834 74.25300000000004 L 358.074 68.73300000000006 L 363.594 68.73300000000006" style="fill: rgb(0%, 0%, 0%); fill-rule: evenodd"></path>
<text lengthAdjust="spacingAndGlyphs" style="font-size: 7.408277799999951pt; fill: #000" textLength="116.90838390000019" x="230.38335018" y="18.14341284000011">ADDSUBPD xmm1, xmm2/m128</text>
<text lengthAdjust="spacingAndGlyphs" style="font-size: 7.408277799999951pt; fill: #000" textLength="44.009650300000146" x="436.83669189000005" y="43.456362040000045">xmm2/m128</text>
<text lengthAdjust="spacingAndGlyphs" style="font-size: 7.408277799999951pt; fill: #000" textLength="28.913084200000014" x="202.3676" y="43.456863">[127:64]</text>
<text lengthAdjust="spacingAndGlyphs" style="font-size: 7.408277799999951pt; fill: #000" textLength="20.016750600000023" x="350.82370000000003" y="43.456863">[63:0]</text>
<text lengthAdjust="spacingAndGlyphs" style="font-size: 7.408277799999951pt; fill: #000" textLength="32.30521139999996" x="436.83669189000005" y="87.03319611000006">RESULT:</text>
<text lengthAdjust="spacingAndGlyphs" style="font-size: 7.408277799999951pt; fill: #000" textLength="132.73297730000004" x="150.4553" y="91.83386300000006">xmm1[127:64] + xmm2/m128[127:64]</text>
<text lengthAdjust="spacingAndGlyphs" style="font-size: 7.408277799999951pt; fill: #000" textLength="112.93223480000023" x="304.3718" y="91.83386300000006">xmm1[63:0] - xmm2/m128[63:0]</text>
<text lengthAdjust="spacingAndGlyphs" style="font-size: 7.408277799999951pt; fill: #000" textLength="21.776816600000075" x="436.83669189000005" y="96.63355611000009">xmm1</text>
<text lengthAdjust="spacingAndGlyphs" style="font-size: 7.408277799999951pt; fill: #000" textLength="28.913084200000014" x="202.3775" y="115.48466300000007">[127:64]</text>
<text lengthAdjust="spacingAndGlyphs" style="font-size: 7.408277799999951pt; fill: #000" textLength="20.016750600000023" x="350.83346689" y="115.48466300000007">[63:0]</text></g></svg>
<figcaption><a href='addsubpd.html#fig-3-3'>Figure 3-3</a>. ADDSUBPD—Packed Double Precision Floating-Point Add/Subtract</figcaption></figure>
<h2 id="operation">Operation<a class="anchor" href="#operation">
</a></h2>
<h3 id="addsubpd--128-bit-legacy-sse-version-">ADDSUBPD (128-bit Legacy SSE Version)<a class="anchor" href="#addsubpd--128-bit-legacy-sse-version-">
</a></h3>
<pre>DEST[63:0] := DEST[63:0] - SRC[63:0]
DEST[127:64] := DEST[127:64] + SRC[127:64]
DEST[MAXVL-1:128] (Unmodified)
</pre>
<h3 id="vaddsubpd--vex-128-encoded-version-">VADDSUBPD (VEX.128 Encoded Version)<a class="anchor" href="#vaddsubpd--vex-128-encoded-version-">
</a></h3>
<pre>DEST[63:0] := SRC1[63:0] - SRC2[63:0]
DEST[127:64] := SRC1[127:64] + SRC2[127:64]
DEST[MAXVL-1:128] := 0
</pre>
<h3 id="vaddsubpd--vex-256-encoded-version-">VADDSUBPD (VEX.256 Encoded Version)<a class="anchor" href="#vaddsubpd--vex-256-encoded-version-">
</a></h3>
<pre>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]
</pre>
<h2 id="intel-c-c++-compiler-intrinsic-equivalent">Intel C/C++ Compiler Intrinsic Equivalent<a class="anchor" href="#intel-c-c++-compiler-intrinsic-equivalent">
</a></h2>
<pre>ADDSUBPD __m128d _mm_addsub_pd(__m128d a, __m128d b)
</pre>
<pre>VADDSUBPD __m256d _mm256_addsub_pd (__m256d a, __m256d b)
</pre>
<h2 class="exceptions" id="exceptions">Exceptions<a class="anchor" href="#exceptions">
</a></h2>
<p>When the source operand is a memory operand, it must be aligned on a 16-byte boundary or a general-protection exception (#GP) will be generated.</p>
<h2 class="exceptions" id="simd-floating-point-exceptions">SIMD Floating-Point Exceptions<a class="anchor" href="#simd-floating-point-exceptions">
</a></h2>
<p>Overflow, Underflow, Invalid, Precision, Denormal.</p>
<h2 class="exceptions" id="other-exceptions">Other Exceptions<a class="anchor" href="#other-exceptions">
</a></h2>
<p>See <span class="not-imported">Table 2-19</span>, “Type 2 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>