pound-emu/archived-ballistic
1
0
Fork 0
mirror of https://github.com/pound-emu/ballistic.git synced 2026-01-31 01:15:21 +01:00
Code Issues Packages Projects Releases Wiki Activity
Files
main
archived-ballistic/spec/arm64_xml/bmops_za_pp_zz.xml
Ronald Caesar 26a677f8b4 decoder: Add ARM specification docs 
Signed-off-by: Ronald Caesar <github43132@proton.me>
2025-12-12 18:11:36 -04:00

171 lines
10 KiB
XML
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.

<?xml version="1.0" encoding="utf-8"?>
<?xml-stylesheet type="text/xsl" encoding="UTF-8" href="iform.xsl" version="1.0"?>
<!DOCTYPE instructionsection PUBLIC "-//ARM//DTD instructionsection //EN" "iform-p.dtd">
<!-- Copyright (c) 2010-2022 Arm Limited or its affiliates. All rights reserved. -->
<!-- This document is Non-Confidential. This document may only be used and distributed in accordance with the terms of the agreement entered into by Arm and the party that Arm delivered this document to. -->
<instructionsection id="bmops_za_pp_zz" title="BMOPS" type="instruction">
<docvars>
<docvar key="instr-class" value="mortlach2" />
<docvar key="isa" value="A64" />
<docvar key="mnemonic" value="BMOPS" />
</docvars>
<heading>BMOPS</heading>
<desc>
<brief>Bitwise exclusive NOR population count outer product and subtract</brief>
<description>
<para>This instruction works with 32-bit element ZA tile. This instruction generates an outer product of the first source SVL<sub>S</sub>×1 vector and the second source 1×SVL<sub>S</sub> vector. Each outer product element is obtained as population count of the bitwise XNOR result of the corresponding 32-bit elements of the first source vector and the second source vector. Each source vector is independently predicated by a corresponding governing predicate. When either source vector element is inactive the corresponding destination tile element remains unmodified. The resulting SVL<sub>S</sub>×SVL<sub>S</sub> product is then destructively subtracted from the destination tile.</para>
</description>
<status>Green</status>
<predicated>True</predicated>
<uses_dit condition="FEAT_SVE2 is implemented or FEAT_SME is implemented">True</uses_dit>
<sm_policy>SM_1_only</sm_policy>
<is_gov_pred_pair>True</is_gov_pred_pair>
</desc>
<alias_list howmany="0"></alias_list>
<classes>
<iclass name="SME2" oneof="1" id="iclass_mortlach2" no_encodings="1" isa="A64">
<docvars>
<docvar key="instr-class" value="mortlach2" />
<docvar key="isa" value="A64" />
<docvar key="mnemonic" value="BMOPS" />
</docvars>
<iclassintro count="1"></iclassintro>
<arch_variants>
<arch_variant name="FEAT_SME2" feature="FEAT_SME2" />
</arch_variants>
<regdiagram form="32" psname="BMOPS-ZA.PP.ZZ-32" tworows="1">
<box hibit="31" width="2" settings="2">
<c>1</c>
<c>0</c>
</box>
<box hibit="29" width="9" settings="9">
<c>0</c>
<c>0</c>
<c>0</c>
<c>0</c>
<c>0</c>
<c>0</c>
<c>1</c>
<c>0</c>
<c>0</c>
</box>
<box hibit="20" width="5" name="Zm" usename="1">
<c colspan="5"></c>
</box>
<box hibit="15" width="3" name="Pm" usename="1">
<c colspan="3"></c>
</box>
<box hibit="12" width="3" name="Pn" usename="1">
<c colspan="3"></c>
</box>
<box hibit="9" width="5" name="Zn" usename="1">
<c colspan="5"></c>
</box>
<box hibit="4" name="S" usename="1" settings="1">
<c>1</c>
</box>
<box hibit="3" settings="1">
<c>1</c>
</box>
<box hibit="2" settings="1">
<c>0</c>
</box>
<box hibit="1" width="2" name="ZAda" usename="1">
<c colspan="2"></c>
</box>
</regdiagram>
<encoding name="bmops_za_pp_zz_32" oneofinclass="1" oneof="1" label="">
<docvars>
<docvar key="instr-class" value="mortlach2" />
<docvar key="isa" value="A64" />
<docvar key="mnemonic" value="BMOPS" />
</docvars>
<asmtemplate><text>BMOPS </text><a link="sa_zada" hover="ZA tile ZA0-ZA3 (field &quot;ZAda&quot;)">&lt;ZAda&gt;</a><text>.S, </text><a link="sa_pn" hover="First governing scalable predicate register P0-P7 (field &quot;Pn&quot;)">&lt;Pn&gt;</a><text>/M, </text><a link="sa_pm" hover="Second governing scalable predicate register P0-P7 (field &quot;Pm&quot;)">&lt;Pm&gt;</a><text>/M, </text><a link="sa_zn" hover="First source scalable vector register (field &quot;Zn&quot;)">&lt;Zn&gt;</a><text>.S, </text><a link="sa_zm" hover="Second source scalable vector register (field &quot;Zm&quot;)">&lt;Zm&gt;</a><text>.S</text></asmtemplate>
</encoding>
<ps_section howmany="1">
<ps name="BMOPS-ZA.PP.ZZ-32" mylink="BMOPS-ZA.PP.ZZ-32" enclabels="" sections="1" secttype="noheading">
<pstext mayhavelinks="1" section="Decode" rep_section="decode">if !<a link="impl-aarch64.HaveSME2.0" file="shared_pseudocode.xml" hover="function: boolean HaveSME2()">HaveSME2</a>() then UNDEFINED;
constant integer esize = 32;
integer a = <a link="impl-shared.UInt.1" file="shared_pseudocode.xml" hover="function: integer UInt(bits(N) x)">UInt</a>(Pn);
integer b = <a link="impl-shared.UInt.1" file="shared_pseudocode.xml" hover="function: integer UInt(bits(N) x)">UInt</a>(Pm);
integer n = <a link="impl-shared.UInt.1" file="shared_pseudocode.xml" hover="function: integer UInt(bits(N) x)">UInt</a>(Zn);
integer m = <a link="impl-shared.UInt.1" file="shared_pseudocode.xml" hover="function: integer UInt(bits(N) x)">UInt</a>(Zm);
integer da = <a link="impl-shared.UInt.1" file="shared_pseudocode.xml" hover="function: integer UInt(bits(N) x)">UInt</a>(ZAda);
boolean sub_op = TRUE;</pstext>
</ps>
</ps_section>
</iclass>
</classes>
<explanations scope="all">
<explanation enclist="bmops_za_pp_zz_32" symboldefcount="1">
<symbol link="sa_zada">&lt;ZAda&gt;</symbol>
<account encodedin="ZAda">
<intro>
<para>Is the name of the ZA tile ZA0-ZA3, encoded in the "ZAda" field.</para>
</intro>
</account>
</explanation>
<explanation enclist="bmops_za_pp_zz_32" symboldefcount="1">
<symbol link="sa_pn">&lt;Pn&gt;</symbol>
<account encodedin="Pn">
<intro>
<para>Is the name of the first governing scalable predicate register P0-P7, encoded in the "Pn" field.</para>
</intro>
</account>
</explanation>
<explanation enclist="bmops_za_pp_zz_32" symboldefcount="1">
<symbol link="sa_pm">&lt;Pm&gt;</symbol>
<account encodedin="Pm">
<intro>
<para>Is the name of the second governing scalable predicate register P0-P7, encoded in the "Pm" field.</para>
</intro>
</account>
</explanation>
<explanation enclist="bmops_za_pp_zz_32" symboldefcount="1">
<symbol link="sa_zn">&lt;Zn&gt;</symbol>
<account encodedin="Zn">
<intro>
<para>Is the name of the first source scalable vector register, encoded in the "Zn" field.</para>
</intro>
</account>
</explanation>
<explanation enclist="bmops_za_pp_zz_32" symboldefcount="1">
<symbol link="sa_zm">&lt;Zm&gt;</symbol>
<account encodedin="Zm">
<intro>
<para>Is the name of the second source scalable vector register, encoded in the "Zm" field.</para>
</intro>
</account>
</explanation>
</explanations>
<ps_section howmany="1">
<ps name="BMOPS-ZA.PP.ZZ-32" mylink="execute" enclabels="" sections="1" secttype="Operation">
<pstext mayhavelinks="1" section="Execute" rep_section="execute"><a link="impl-aarch64.CheckStreamingSVEAndZAEnabled.0" file="shared_pseudocode.xml" hover="function: CheckStreamingSVEAndZAEnabled()">CheckStreamingSVEAndZAEnabled</a>();
constant integer VL = <a link="impl-aarch64.CurrentVL.read.none" file="shared_pseudocode.xml" hover="accessor: integer CurrentVL">CurrentVL</a>;
constant integer PL = VL DIV 8;
constant integer dim = VL DIV esize;
bits(PL) mask1 = <a link="impl-aarch64.P.read.2" file="shared_pseudocode.xml" hover="accessor: bits(width) P[integer n, integer width]">P</a>[a, PL];
bits(PL) mask2 = <a link="impl-aarch64.P.read.2" file="shared_pseudocode.xml" hover="accessor: bits(width) P[integer n, integer width]">P</a>[b, PL];
bits(VL) operand1 = <a link="impl-aarch64.Z.read.2" file="shared_pseudocode.xml" hover="accessor: bits(width) Z[integer n, integer width]">Z</a>[n, VL];
bits(VL) operand2 = <a link="impl-aarch64.Z.read.2" file="shared_pseudocode.xml" hover="accessor: bits(width) Z[integer n, integer width]">Z</a>[m, VL];
bits(dim*dim*esize) operand3 = <a link="impl-aarch64.ZAtile.read.3" file="shared_pseudocode.xml" hover="accessor: bits(width) ZAtile[integer tile, integer esize, integer width]">ZAtile</a>[da, esize, dim*dim*esize];
bits(dim*dim*esize) result;
for row = 0 to dim-1
bits(esize) element1 = <a link="impl-shared.Elem.read.3" file="shared_pseudocode.xml" hover="accessor: bits(size) Elem[bits(N) vector, integer e, integer size]">Elem</a>[operand1, row, esize];
for col = 0 to dim-1
bits(esize) element2 = <a link="impl-shared.Elem.read.3" file="shared_pseudocode.xml" hover="accessor: bits(size) Elem[bits(N) vector, integer e, integer size]">Elem</a>[operand2, col, esize];
bits(esize) element3 = <a link="impl-shared.Elem.read.3" file="shared_pseudocode.xml" hover="accessor: bits(size) Elem[bits(N) vector, integer e, integer size]">Elem</a>[operand3, row*dim + col, esize];
if (<a link="impl-aarch64.ActivePredicateElement.3" file="shared_pseudocode.xml" hover="function: boolean ActivePredicateElement(bits(N) pred, integer e, integer esize)">ActivePredicateElement</a>(mask1, row, esize) &amp;&amp;
<a link="impl-aarch64.ActivePredicateElement.3" file="shared_pseudocode.xml" hover="function: boolean ActivePredicateElement(bits(N) pred, integer e, integer esize)">ActivePredicateElement</a>(mask2, col, esize)) then
integer res = <a link="impl-shared.BitCount.1" file="shared_pseudocode.xml" hover="function: integer BitCount(bits(N) x)">BitCount</a>(NOT(element1 EOR element2));
if sub_op then res = -res;
<a link="impl-shared.Elem.write.3" file="shared_pseudocode.xml" hover="accessor: Elem[bits(N) &amp;vector, integer e, integer size] = bits(size) value">Elem</a>[result, row*dim + col, esize] = element3 + res;
else
<a link="impl-shared.Elem.write.3" file="shared_pseudocode.xml" hover="accessor: Elem[bits(N) &amp;vector, integer e, integer size] = bits(size) value">Elem</a>[result, row*dim + col, esize] = element3;
<a link="impl-aarch64.ZAtile.write.3" file="shared_pseudocode.xml" hover="accessor: ZAtile[integer tile, integer esize, integer width] = bits(width) value">ZAtile</a>[da, esize, dim*dim*esize] = result;</pstext>
</ps>
</ps_section>
</instructionsection>
Reference in New Issue View Git Blame Copy Permalink