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).

FSTMDBX, FSTMIAX -- AArch32

FSTMDBX, FSTMIAX

FSTMX stores multiple SIMD&FP registers from the Advanced SIMD and floating-point register file to consecutive locations in using an address from a general-purpose register.

Arm deprecates use of FSTMDBX and FSTMIAX, except for disassembly purposes, and reassembly of disassembled code.

Depending on settings in the CPACR, NSACR, HCPTR, and FPEXC registers, and the Security state and PE mode in which the instruction is executed, an attempt to execute the instruction might be UNDEFINED, or trapped to Hyp mode. For more information, see Enabling Advanced SIMD and floating-point support.

It has encodings from the following instruction sets: A32 ( A1 ) and T32 ( T1 ) .

A1

313029282726252423222120191817161514131211109876543210
!= 1111110PUDW0RnVd1011xxxxxxx1
condLsizeimm8

Encoding for the Decrement Before variant

Applies when (P == 1 && U == 0 && W == 1)

FSTMDBX{<c>}{<q>} <Rn>!, <dreglist>

Encoding for the Increment After variant

Applies when (P == 0 && U == 1)

FSTMIAX{<c>}{<q>} <Rn>{!}, <dreglist>

Decode for all variants of this encoding

if P == '0' && U == '0' && W == '0' then See("Related encodings"); end; if P == '1' && W == '0' then See("VSTR"); end; if P == U && W == '1' then Undefined(); end; // Remaining combinations are PUW = 010 (IA without !), 011 (IA with !), 101 (DB with !) let single_regs : boolean = FALSE; let add : boolean = (U == '1'); let wback : boolean = (W == '1'); let d : integer = UInt(D::Vd); let n : integer = UInt(Rn); let imm32 : bits(32) = ZeroExtend{}(imm8::'00'); let regs : integer = UInt(imm8) DIVRM 2; // If UInt(imm8) is odd, see "FSTDBMX, FSTMIAX". if n == 15 && (wback || CurrentInstrSet() != InstrSet_A32) then UnpredictableProcedure(); end; if regs == 0 || regs > 16 || (d+regs) > 32 then UnpredictableProcedure(); end; if imm8[0] == '1' && (d+regs) > 16 then UnpredictableProcedure(); end;

CONSTRAINED UNPREDICTABLE behavior

If regs == 0, then one of the following behaviors must occur:

If regs > 16 || (d+regs) > 16, then one of the following behaviors must occur:

T1

15141312111098765432101514131211109876543210
1110110PUDW0RnVd1011xxxxxxx1
Lsizeimm8

Encoding for the Decrement Before variant

Applies when (P == 1 && U == 0 && W == 1)

FSTMDBX{<c>}{<q>} <Rn>!, <dreglist>

Encoding for the Increment After variant

Applies when (P == 0 && U == 1)

FSTMIAX{<c>}{<q>} <Rn>{!}, <dreglist>

Decode for all variants of this encoding

if P == '0' && U == '0' && W == '0' then See("Related encodings"); end; if P == '1' && W == '0' then See("VSTR"); end; if P == U && W == '1' then Undefined(); end; // Remaining combinations are PUW = 010 (IA without !), 011 (IA with !), 101 (DB with !) let single_regs : boolean = FALSE; let add : boolean = (U == '1'); let wback : boolean = (W == '1'); let d : integer = UInt(D::Vd); let n : integer = UInt(Rn); let imm32 : bits(32) = ZeroExtend{}(imm8::'00'); let regs : integer = UInt(imm8) DIVRM 2; // If UInt(imm8) is odd, see "FSTDBMX, FSTMIAX". if n == 15 && (wback || CurrentInstrSet() != InstrSet_A32) then UnpredictableProcedure(); end; if regs == 0 || regs > 16 || (d+regs) > 32 then UnpredictableProcedure(); end; if imm8[0] == '1' && (d+regs) > 16 then UnpredictableProcedure(); end;

CONSTRAINED UNPREDICTABLE behavior

If regs == 0, then one of the following behaviors must occur:

If regs > 16 || (d+regs) > 16, then one of the following behaviors must occur:

For more information about the CONSTRAINED UNPREDICTABLE behavior of this instruction, see Architectural Constraints on UNPREDICTABLE behaviors.

Related encodings: See Advanced SIMD and floating-point 64-bit move for the T32 instruction set, or Advanced SIMD and floating-point 64-bit move for the A32 instruction set.

Assembler Symbols

<c>

See Standard assembler syntax fields.

<q>

See Standard assembler syntax fields.

<Rn>

Is the general-purpose base register, encoded in the "Rn" field. If writeback is not specified, the PC can be used. However, Arm deprecates use of the PC.

<dreglist>

Is the list of consecutively numbered 64-bit SIMD&FP registers to be transferred. The first register in the list is encoded in "D:Vd", and "imm8" is set to twice the number of registers in the list plus one. The list must contain at least one register, all registers must be in the range D0-D15, and must not contain more than 16 registers.

!

Specifies base register writeback, encoded in W:

W !
0 [absent]
1 [present]

Operation

if ConditionPassed() then EncodingSpecificOperations(); CheckVFPEnabled(TRUE); var address : bits(32) = if add then R(n) else R(n)-imm32; for r = 0 to regs-1 do if single_regs then MemA{32}(address) = S(d+r); address = address+4; else // Store as two word-aligned words in the correct order for current endianness. if BigEndian(AccessType_ASIMD) then MemA{32}(address) = D(d+r)[63:32]; MemA{32}(address+4) = D(d+r)[31:0]; else MemA{32}(address) = D(d+r)[31:0]; MemA{32}(address+4) = D(d+r)[63:32]; end; address = address+8; end; end; if wback then R(n) = if add then R(n)+imm32 else R(n)-imm32; end; end;

Operational information

This instruction is a data-independent-time instruction as described in About the DIT bit.


2026-03_rel 2026-03-26 20:48:11

Copyright © 2010-2026 Arm Limited or its affiliates. All rights reserved. This document is Non-Confidential.