This is a collection of Intel®’ IA32® Software Developer's Manuals (URL of the day) and AMD' AMD64 Architecture Programmer's Manual together with the related specifications, application notes, white papers, and change logs. The collection aims to keep all available revisions. It was originally created by Michal Necasek, see OS/2 Museum.

If you have a public document, related to the IA32® specifications and missing from the collection, please mail it to me. The content of this URL and all sub-ULRs is available for convenient bulk download by rsync x86docs password "" (empty).

DC ZVA

DC ZVA, Data Cache Zero by VA

The DC ZVA characteristics are:

Purpose

Zero data cache by address. Zeroes a naturally aligned block of N bytes, where the size of N is identified in DCZID_EL0.

This instruction might have the poison-atomic property for IMPLEMENTATION DEFINED regions of physical memory.

Configuration

This instruction is present only when FEAT_AA64 is implemented. Otherwise, direct accesses to DC ZVA are UNDEFINED.

Attributes

DC ZVA is a 64-bit System instruction.

Field descriptions

6362616059585756555453525150494847464544434241403938373635343332
313029282726252423222120191817161514131211109876543210
VA
VA

VA, bits [63:0]

Virtual address to use. There is no alignment restriction on the address within the block of N bytes that is used.

Executing DC ZVA

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 ESR_ELx.ISS field is set to 0.

If the memory region being zeroed is any type of Device memory, this instruction generates an Alignment fault which is prioritized in the same way as other Alignment faults that are determined by the memory type.

This instruction applies to Normal memory regardless of cacheability attributes.

This instruction behaves as a set of Stores to each byte within the block being accessed, and so it:

This system instruction is an alias of the SYS instruction.

Accesses to this instruction use the following encodings in the System instruction encoding space:

DC ZVA, <Xt>

op0op1CRnCRmop2
0b010b0110b01110b01000b001

if !IsFeatureImplemented(FEAT_AA64) then Undefined(); elsif PSTATE.EL == EL0 then if !ELIsInHost(EL0) && SCTLR_EL1().DZE == '0' then if EL2Enabled() && HCR_EL2().TGE == '1' then AArch64_SystemAccessTrap(EL2, 0x18); else AArch64_SystemAccessTrap(EL1, 0x18); end; elsif EL2Enabled() && !ELIsInHost(EL0) && HCR_EL2().TDZ == '1' then AArch64_SystemAccessTrap(EL2, 0x18); elsif EL2Enabled() && !ELIsInHost(EL0) && IsFeatureImplemented(FEAT_FGT) && (!HaveEL(EL3) || SCR_EL3().FGTEn == '1') && HFGITR_EL2().DCZVA == '1' then AArch64_SystemAccessTrap(EL2, 0x18); elsif ELIsInHost(EL0) && SCTLR_EL2().DZE == '0' then AArch64_SystemAccessTrap(EL2, 0x18); else AArch64_MemZero(X{64}(t), CacheType_Data); end; elsif PSTATE.EL == EL1 then if EL2Enabled() && HCR_EL2().TDZ == '1' then AArch64_SystemAccessTrap(EL2, 0x18); elsif EL2Enabled() && IsFeatureImplemented(FEAT_FGT) && (!HaveEL(EL3) || SCR_EL3().FGTEn == '1') && HFGITR_EL2().DCZVA == '1' then AArch64_SystemAccessTrap(EL2, 0x18); else AArch64_MemZero(X{64}(t), CacheType_Data); end; elsif PSTATE.EL == EL2 then AArch64_MemZero(X{64}(t), CacheType_Data); elsif PSTATE.EL == EL3 then AArch64_MemZero(X{64}(t), CacheType_Data); end;


2026-03-12 12:23:09, 2025-09_rel_asl1

Copyright © 2010-2025 Arm Limited or its affiliates. All rights reserved. This document is Non-Confidential.