<?xml version='1.0' encoding='utf-8'?>
<!DOCTYPE register_page SYSTEM "registers.dtd">
<!-- Copyright (c) 2010-2026 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. -->
<!--The data contained in this document is preliminary and subject to change or correction following further review. -->
<?xml-stylesheet href="one_register.xsl" type="text/xsl" ?>







<register_page>
  <registers>
  
    <register execution_state="AArch64" is_register="False" is_internal="True" is_stub_entry="False">
      <reg_short_name>DC GZVA</reg_short_name>
        
        <reg_long_name>Data Cache set Allocation Tags and Zero by VA</reg_long_name>



      
            <reg_condition otherwise="UNDEFINED">when FEAT_MTE is implemented</reg_condition>
      




          <reg_reset_value></reg_reset_value>

      <reg_mappings>
        




      </reg_mappings>

        <reg_purpose>
          
    
      <purpose_text>
        <para>Zero data and write a value to the Allocation Tags of a naturally aligned block of N bytes, where the size of N is identified in <register_link state="AArch64" id="AArch64-dczid_el0.xml">DCZID_EL0</register_link>.  The Allocation Tag used is determined by the input address.</para>

      </purpose_text>
      <purpose_text>
        <para>This instruction might have the poison-atomic property for <arm-defined-word>IMPLEMENTATION DEFINED</arm-defined-word> regions of physical memory.</para>
      </purpose_text>

        </reg_purpose>

      <reg_groups>
            <reg_group>Cache</reg_group>
      </reg_groups>
      
      
        
      <reg_attributes>
          
    
      <attributes_text>
        <para>DC GZVA is a 64-bit System instruction.</para>
      </attributes_text>

      </reg_attributes>
      <reg_fieldsets>
        






<fields id="fieldset_0" length="64">
  <text_before_fields/>
  <field id="fieldset_0-63_0" has_partial_fieldset="False" is_linked_to_partial_fieldset="False" is_access_restriction_possible="False" is_variable_length="False" is_constant_value="False" is_partial_field="False" is_conditional_field_name="False">
    <field_name>VA</field_name>
    <field_shortdesc>Virtual address to use</field_shortdesc>
    <field_msb>63</field_msb>
    <field_lsb>0</field_lsb>
    <rel_range>63:0</rel_range>
    <field_description order="before">
      <para>Virtual address to use. There is no alignment restriction on the address within the block of N bytes that is used.</para>
    </field_description>
  </field>
  <text_after_fields/>
</fields>




<reg_fieldset length="64">
  <fieldat id="fieldset_0-63_0" msb="63" lsb="0"/>
</reg_fieldset>


      </reg_fieldsets>

      <access_mechanisms>
          


  
    
      <access_permission_text>
        <para>When this instruction is executed, it can generate memory faults or watchpoints which are prioritized in the same way as other memory-related faults or watchpoints. If a synchronous Data Abort fault or a watchpoint is generated, the CM bit in the <xref linkend="#ESR_ELx">ESR_ELx</xref>.ISS field is not set.</para>

      </access_permission_text>
      <access_permission_text>
        <para>If the memory region being zeroed is any type of Device memory that does not support unaligned accesses, this instruction generates an alignment fault which is prioritized in the same way as other alignment faults that are determined by the memory type.</para>

      </access_permission_text>
      <access_permission_text>
        <para>This instruction applies to Normal memory regardless of cacheability attributes.</para>

      </access_permission_text>
      <access_permission_text>
        <para>This instruction behaves as a set of Stores to each byte and Allocation tag within the block being accessed, and so it:</para>

      </access_permission_text>
      <access_permission_text>
        <list type="unordered">
<listitem><content>
<para>Generates a Permission fault if the translation system does not permit writes to the locations.</para>
</content>
</listitem><listitem><content>
<para>Requires the same considerations for ordering and the management of coherency as any other store instructions.</para>
</content>
</listitem></list>

      </access_permission_text>
      <access_permission_text>
        <para>This system instruction is an alias of the SYS instruction.</para>
      </access_permission_text>





    
        
        <access_mechanism accessor="DC GZVA" type="SystemAccessor">
            <encoding>
            <access_instruction>DC GZVA, &lt;Xt&gt;</access_instruction>
                
                <enc n="op0" v="0b01"/>
                
                <enc n="op1" v="0b011"/>
                
                <enc n="CRn" v="0b0111"/>
                
                <enc n="CRm" v="0b0100"/>
                
                <enc n="op2" v="0b100"/>
            </encoding>
            <access_permission>
                <ps name="DC" sections="1" secttype="access_permission">
                <pstext>
if !IsFeatureImplemented(FEAT_MTE) then
    Undefined();
elsif PSTATE.EL == EL0 then
    if !ELIsInHost(EL0) &amp;&amp; SCTLR_EL1().DZE == '0' then
        if EL2Enabled() &amp;&amp; HCR_EL2().TGE == '1' then
            AArch64_SystemAccessTrap(EL2, 0x18);
        else
            AArch64_SystemAccessTrap(EL1, 0x18);
        end;
    elsif EL2Enabled() &amp;&amp; !ELIsInHost(EL0) &amp;&amp; HCR_EL2().TDZ == '1' then
        AArch64_SystemAccessTrap(EL2, 0x18);
    elsif EL2Enabled() &amp;&amp; !ELIsInHost(EL0) &amp;&amp; IsFeatureImplemented(FEAT_FGT) &amp;&amp; (!HaveEL(EL3) || SCR_EL3().FGTEn == '1') &amp;&amp; HFGITR_EL2().DCZVA == '1' then
        AArch64_SystemAccessTrap(EL2, 0x18);
    elsif ELIsInHost(EL0) &amp;&amp; SCTLR_EL2().DZE == '0' then
        AArch64_SystemAccessTrap(EL2, 0x18);
    else
        AArch64_MemZero(X{64}(t), CacheType_Data_Tag);
    end;
elsif PSTATE.EL == EL1 then
    if EL2Enabled() &amp;&amp; HCR_EL2().TDZ == '1' then
        AArch64_SystemAccessTrap(EL2, 0x18);
    elsif EL2Enabled() &amp;&amp; IsFeatureImplemented(FEAT_FGT) &amp;&amp; (!HaveEL(EL3) || SCR_EL3().FGTEn == '1') &amp;&amp; HFGITR_EL2().DCZVA == '1' then
        AArch64_SystemAccessTrap(EL2, 0x18);
    else
        AArch64_MemZero(X{64}(t), CacheType_Data_Tag);
    end;
elsif PSTATE.EL == EL2 then
    AArch64_MemZero(X{64}(t), CacheType_Data_Tag);
elsif PSTATE.EL == EL3 then
    AArch64_MemZero(X{64}(t), CacheType_Data_Tag);
end;
                </pstext>
                </ps>
            </access_permission>
        </access_mechanism>

      </access_mechanisms>

      <arch_variants>
      </arch_variants>
  </register>
</registers>
<timestamp>2026-03-26 20:27:25</timestamp>
<commit_id>2026-03_rel</commit_id>
</register_page>