PLI (register)

Preload Instruction signals the memory system that instruction memory accesses from a specified address are likely in the near future. The memory system can respond by taking actions that are expected to speed up the memory accesses when they do occur, such as pre-loading the cache line containing the specified address into the instruction cache.

The effect of a PLI instruction is implementation defined. For more information, see Preloading caches.

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

A1

313029282726252423222120191817161514131211109876543210
11110110U101Rn(1)(1)(1)(1)imm5stype0Rm

Rotate right with extend (imm5 == 00000 && stype == 11)

PLI{<c>}{<q>} [<Rn>, {+/-}<Rm> , RRX]

Shift or rotate by value (!(imm5 == 00000 && stype == 11))

PLI{<c>}{<q>} [<Rn>, {+/-}<Rm> {, <shift> #<amount>}]

n = UInt(Rn); m = UInt(Rm); add = (U == '1'); (shift_t, shift_n) = DecodeImmShift(stype, imm5); if m == 15 then UNPREDICTABLE;

T1

15141312111098765432101514131211109876543210
111110010001!= 11111111000000imm2Rm
Rn

T1

PLI{<c>}{<q>} [<Rn>, {+}<Rm> {, LSL #<amount>}]

if Rn == '1111' then SEE "PLI (immediate, literal)"; n = UInt(Rn); m = UInt(Rm); add = TRUE; (shift_t, shift_n) = (SRType_LSL, UInt(imm2)); if m == 15 then UNPREDICTABLE; // Armv8-A removes UNPREDICTABLE for R13

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

Assembler Symbols

<c>

For encoding A1: see Standard assembler syntax fields. <c> must be AL or omitted.

For encoding T1: see Standard assembler syntax fields.

<q>

See Standard assembler syntax fields.

<Rn>

Is the general-purpose base register, encoded in the "Rn" field.

+/-

Specifies the index register is added to or subtracted from the base register, defaulting to + if omitted and encoded in U:

U +/-
0 -
1 +
+

Specifies the index register is added to the base register.

<Rm>

Is the general-purpose index register, encoded in the "Rm" field.

<shift>

Is the type of shift to be applied to the index register, encoded in stype:

stype <shift>
00 LSL
01 LSR
10 ASR
11 ROR
<amount>

For encoding A1: is the shift amount, in the range 1 to 31 (when <shift> = LSL or ROR) or 1 to 32 (when <shift> = LSR or ASR) encoded in the "imm5" field as <amount> modulo 32.

For encoding T1: is the shift amount, in the range 0 to 3, defaulting to 0 and encoded in the "imm2" field.

Operation

if ConditionPassed() then EncodingSpecificOperations(); offset = Shift(R[m], shift_t, shift_n, PSTATE.C); address = if add then (R[n] + offset) else (R[n] - offset); Hint_PreloadInstr(address);


Internal version only: isa v01_31, pseudocode v2023-06_rel, sve v2023-06_rel ; Build timestamp: 2023-07-04T18:06

Copyright © 2010-2023 Arm Limited or its affiliates. All rights reserved. This document is Non-Confidential.