[llvm] r297190 - [AArch64] Vulcan is now ThunderXT99
Joel Jones via llvm-commits
llvm-commits at lists.llvm.org
Tue Mar 7 15:44:14 PST 2017
Yes, it seems like for whatever reason the “svn rm” of AArch64SchedVulcan.td didn’t trigger a new cmake, The subsequent clang patch that fixed the errors from clang’s “Driver/aarch64-cpus.c, Frontend/aarch64-taget-cpu.c, and Preprocessor/aarch64-target-features.c”. should fix the pronblem, unless you are using ninja. For those systems, I think forcing a cmake would fix the problem.
Joel
> On Mar 7, 2017, at 3:29 PM, Galina Kistanova <gkistanova at gmail.com> wrote:
>
> Hello Joel,
>
> It look like this commit broke one of our builders:
>
> http://lab.llvm.org:8011/builders/clang-with-thin-lto-ubuntu/builds/2169 <http://lab.llvm.org:8011/builders/clang-with-thin-lto-ubuntu/builds/2169>
>
> Please have a look at this?
>
> Thanks
>
> Galina
>
> On Tue, Mar 7, 2017 at 11:42 AM, Joel Jones via llvm-commits <llvm-commits at lists.llvm.org <mailto:llvm-commits at lists.llvm.org>> wrote:
> Author: joel_k_jones
> Date: Tue Mar 7 13:42:40 2017
> New Revision: 297190
>
> URL: http://llvm.org/viewvc/llvm-project?rev=297190&view=rev <http://llvm.org/viewvc/llvm-project?rev=297190&view=rev>
> Log:
> [AArch64] Vulcan is now ThunderXT99
>
> Broadcom Vulcan is now Cavium ThunderX2T99.
>
> LLVM Bugzilla: http://bugs.llvm.org/show_bug.cgi?id=32113 <http://bugs.llvm.org/show_bug.cgi?id=32113>
>
> Minor fixes for the alignments of loops and functions for
> ThunderX T81/T83/T88 (better performance).
>
> Patch was tested with SpecCPU2006.
>
> Patch by Stefan Teleman
>
> Differential Revision: https://reviews.llvm.org/D30510 <https://reviews.llvm.org/D30510>
>
> Added:
> llvm/trunk/lib/Target/AArch64/AArch64SchedThunderX2T99.td
> Removed:
> llvm/trunk/lib/Target/AArch64/AArch64SchedVulcan.td
> Modified:
> llvm/trunk/include/llvm/Support/AArch64TargetParser.def
> llvm/trunk/lib/Target/AArch64/AArch64.td
> llvm/trunk/lib/Target/AArch64/AArch64SchedThunderX.td
> llvm/trunk/lib/Target/AArch64/AArch64Subtarget.cpp
> llvm/trunk/lib/Target/AArch64/AArch64Subtarget.h
> llvm/trunk/test/CodeGen/AArch64/cpus.ll
> llvm/trunk/test/CodeGen/AArch64/machine-combiner-madd.ll
> llvm/trunk/test/CodeGen/AArch64/remat.ll
> llvm/trunk/unittests/Support/TargetParserTest.cpp
>
> Modified: llvm/trunk/include/llvm/Support/AArch64TargetParser.def
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/include/llvm/Support/AArch64TargetParser.def?rev=297190&r1=297189&r2=297190&view=diff <http://llvm.org/viewvc/llvm-project/llvm/trunk/include/llvm/Support/AArch64TargetParser.def?rev=297190&r1=297189&r2=297190&view=diff>
> ==============================================================================
> --- llvm/trunk/include/llvm/Support/AArch64TargetParser.def (original)
> +++ llvm/trunk/include/llvm/Support/AArch64TargetParser.def Tue Mar 7 13:42:40 2017
> @@ -73,8 +73,9 @@ AARCH64_CPU_NAME("falkor", AK_ARMV8A, FK
> (AArch64::AEK_SIMD | AArch64::AEK_CRC | AArch64::AEK_CRYPTO))
> AARCH64_CPU_NAME("kryo", AK_ARMV8A, FK_CRYPTO_NEON_FP_ARMV8, false,
> (AArch64::AEK_SIMD | AArch64::AEK_CRC | AArch64::AEK_CRYPTO))
> -AARCH64_CPU_NAME("vulcan", AK_ARMV8_1A, FK_CRYPTO_NEON_FP_ARMV8, false,
> - (AArch64::AEK_SIMD | AArch64::AEK_CRC | AArch64::AEK_CRYPTO))
> +AARCH64_CPU_NAME("thunderx2t99", AK_ARMV8_1A, FK_CRYPTO_NEON_FP_ARMV8, false,
> + (AArch64::AEK_SIMD | AArch64::AEK_LSE | AArch64::AEK_CRC |
> + AArch64::AEK_CRYPTO))
> AARCH64_CPU_NAME("thunderx", AK_ARMV8A, FK_CRYPTO_NEON_FP_ARMV8, false,
> (AArch64::AEK_SIMD | AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_FP | AArch64::AEK_PROFILE))
> AARCH64_CPU_NAME("thunderxt88", AK_ARMV8A, FK_CRYPTO_NEON_FP_ARMV8, false,
>
> Modified: llvm/trunk/lib/Target/AArch64/AArch64.td
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AArch64/AArch64.td?rev=297190&r1=297189&r2=297190&view=diff <http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AArch64/AArch64.td?rev=297190&r1=297189&r2=297190&view=diff>
> ==============================================================================
> --- llvm/trunk/lib/Target/AArch64/AArch64.td (original)
> +++ llvm/trunk/lib/Target/AArch64/AArch64.td Tue Mar 7 13:42:40 2017
> @@ -161,7 +161,7 @@ include "AArch64SchedFalkor.td"
> include "AArch64SchedKryo.td"
> include "AArch64SchedM1.td"
> include "AArch64SchedThunderX.td"
> -include "AArch64SchedVulcan.td"
> +include "AArch64SchedThunderX2T99.td"
>
> def ProcA35 : SubtargetFeature<"a35", "ARMProcFamily", "CortexA35",
> "Cortex-A35 ARM processors", [
> @@ -288,16 +288,18 @@ def ProcFalkor : SubtargetFeature<"falk
> FeatureZCZeroing
> ]>;
>
> -def ProcVulcan : SubtargetFeature<"vulcan", "ARMProcFamily", "Vulcan",
> - "Broadcom Vulcan processors", [
> - FeatureCRC,
> - FeatureCrypto,
> - FeatureFPARMv8,
> - FeatureArithmeticBccFusion,
> - FeatureNEON,
> - FeaturePostRAScheduler,
> - FeaturePredictableSelectIsExpensive,
> - HasV8_1aOps]>;
> +def ProcThunderX2T99 : SubtargetFeature<"thunderx2t99", "ARMProcFamily",
> + "ThunderX2T99",
> + "Cavium ThunderX2 processors", [
> + FeatureCRC,
> + FeatureCrypto,
> + FeatureFPARMv8,
> + FeatureArithmeticBccFusion,
> + FeatureNEON,
> + FeaturePostRAScheduler,
> + FeaturePredictableSelectIsExpensive,
> + FeatureLSE,
> + HasV8_1aOps]>;
>
> def ProcThunderX : SubtargetFeature<"thunderx", "ARMProcFamily", "ThunderX",
> "Cavium ThunderX processors", [
> @@ -363,12 +365,13 @@ def : ProcessorModel<"exynos-m2", Exynos
> def : ProcessorModel<"exynos-m3", ExynosM1Model, [ProcExynosM2]>;
> def : ProcessorModel<"falkor", FalkorModel, [ProcFalkor]>;
> def : ProcessorModel<"kryo", KryoModel, [ProcKryo]>;
> -def : ProcessorModel<"vulcan", VulcanModel, [ProcVulcan]>;
> // Cavium ThunderX/ThunderX T8X Processors
> def : ProcessorModel<"thunderx", ThunderXT8XModel, [ProcThunderX]>;
> def : ProcessorModel<"thunderxt88", ThunderXT8XModel, [ProcThunderXT88]>;
> def : ProcessorModel<"thunderxt81", ThunderXT8XModel, [ProcThunderXT81]>;
> def : ProcessorModel<"thunderxt83", ThunderXT8XModel, [ProcThunderXT83]>;
> +// Cavium ThunderX2T9X Processors. Formerly Broadcom Vulcan.
> +def : ProcessorModel<"thunderx2t99", ThunderX2T99Model, [ProcThunderX2T99]>;
>
> //===----------------------------------------------------------------------===//
> // Assembly parser
>
> Modified: llvm/trunk/lib/Target/AArch64/AArch64SchedThunderX.td
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AArch64/AArch64SchedThunderX.td?rev=297190&r1=297189&r2=297190&view=diff <http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AArch64/AArch64SchedThunderX.td?rev=297190&r1=297189&r2=297190&view=diff>
> ==============================================================================
> --- llvm/trunk/lib/Target/AArch64/AArch64SchedThunderX.td (original)
> +++ llvm/trunk/lib/Target/AArch64/AArch64SchedThunderX.td Tue Mar 7 13:42:40 2017
> @@ -23,6 +23,7 @@ def ThunderXT8XModel : SchedMachineModel
> let MicroOpBufferSize = 0; // ThunderX T88/T81/T83 are in-order.
> let LoadLatency = 3; // Optimistic load latency.
> let MispredictPenalty = 8; // Branch mispredict penalty.
> + let PostRAScheduler = 1; // Use PostRA scheduler.
> let CompleteModel = 1;
> }
>
>
> Added: llvm/trunk/lib/Target/AArch64/AArch64SchedThunderX2T99.td
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AArch64/AArch64SchedThunderX2T99.td?rev=297190&view=auto <http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AArch64/AArch64SchedThunderX2T99.td?rev=297190&view=auto>
> ==============================================================================
> --- llvm/trunk/lib/Target/AArch64/AArch64SchedThunderX2T99.td (added)
> +++ llvm/trunk/lib/Target/AArch64/AArch64SchedThunderX2T99.td Tue Mar 7 13:42:40 2017
> @@ -0,0 +1,852 @@
> +//=- AArch64SchedThunderX2T99.td - Cavium ThunderX T99 Scheduling ---*- tablegen -*-=//
> +//
> +// The LLVM Compiler Infrastructure
> +//
> +// This file is distributed under the University of Illinois Open Source
> +// License. See LICENSE.TXT for details.
> +//
> +//===----------------------------------------------------------------------===//
> +//
> +// This file defines the scheduling model for Cavium ThunderX2T99
> +// processors.
> +// Based on Broadcom Vulcan.
> +//
> +//===----------------------------------------------------------------------===//
> +
> +//===----------------------------------------------------------------------===//
> +// 2. Pipeline Description.
> +
> +def ThunderX2T99Model : SchedMachineModel {
> + let IssueWidth = 4; // 4 micro-ops dispatched at a time.
> + let MicroOpBufferSize = 180; // 180 entries in micro-op re-order buffer.
> + let LoadLatency = 4; // Optimistic load latency.
> + let MispredictPenalty = 12; // Extra cycles for mispredicted branch.
> + // Determined via a mix of micro-arch details and experimentation.
> + let LoopMicroOpBufferSize = 32;
> + let PostRAScheduler = 1; // Using PostRA sched.
> + let CompleteModel = 1;
> +}
> +
> +// Define the issue ports.
> +
> +// Port 0: ALU, FP/SIMD.
> +def THX2T99P0 : ProcResource<1>;
> +
> +// Port 1: ALU, FP/SIMD, integer mul/div.
> +def THX2T99P1 : ProcResource<1>;
> +
> +// Port 2: ALU, Branch.
> +def THX2T99P2 : ProcResource<1>;
> +
> +// Port 3: Store data.
> +def THX2T99P3 : ProcResource<1>;
> +
> +// Port 4: Load/store.
> +def THX2T99P4 : ProcResource<1>;
> +
> +// Port 5: Load/store.
> +def THX2T99P5 : ProcResource<1>;
> +
> +let SchedModel = ThunderX2T99Model in {
> +
> +// Define groups for the functional units on each issue port. Each group
> +// created will be used by a WriteRes later on.
> +//
> +// NOTE: Some groups only contain one member. This is a way to create names for
> +// the various functional units that share a single issue port. For example,
> +// "THX2T99I1" for ALU ops on port 1 and "THX2T99F1" for FP ops on port 1.
> +
> +// Integer divide and multiply micro-ops only on port 1.
> +def THX2T99I1 : ProcResGroup<[THX2T99P1]>;
> +
> +// Branch micro-ops only on port 2.
> +def THX2T99I2 : ProcResGroup<[THX2T99P2]>;
> +
> +// ALU micro-ops on ports 0, 1, and 2.
> +def THX2T99I012 : ProcResGroup<[THX2T99P0, THX2T99P1, THX2T99P2]>;
> +
> +// Crypto FP/SIMD micro-ops only on port 1.
> +def THX2T99F1 : ProcResGroup<[THX2T99P1]>;
> +
> +// FP/SIMD micro-ops on ports 0 and 1.
> +def THX2T99F01 : ProcResGroup<[THX2T99P0, THX2T99P1]>;
> +
> +// Store data micro-ops only on port 3.
> +def THX2T99SD : ProcResGroup<[THX2T99P3]>;
> +
> +// Load/store micro-ops on ports 4 and 5.
> +def THX2T99LS01 : ProcResGroup<[THX2T99P4, THX2T99P5]>;
> +
> +// 60 entry unified scheduler.
> +def THX2T99Any : ProcResGroup<[THX2T99P0, THX2T99P1, THX2T99P2,
> + THX2T99P3, THX2T99P4, THX2T99P5]> {
> + let BufferSize=60;
> +}
> +
> +// Define commonly used write types for InstRW specializations.
> +// All definitions follow the format: THX2T99Write_<NumCycles>Cyc_<Resources>.
> +
> +// 3 cycles on I1.
> +def THX2T99Write_3Cyc_I1 : SchedWriteRes<[THX2T99I1]> { let Latency = 3; }
> +
> +// 4 cycles on I1.
> +def THX2T99Write_4Cyc_I1 : SchedWriteRes<[THX2T99I1]> { let Latency = 4; }
> +
> +// 1 cycle on I0, I1, or I2.
> +def THX2T99Write_1Cyc_I012 : SchedWriteRes<[THX2T99I012]> { let Latency = 1; }
> +
> +// 5 cycles on F1.
> +def THX2T99Write_5Cyc_F1 : SchedWriteRes<[THX2T99F1]> { let Latency = 5; }
> +
> +// 7 cycles on F1.
> +def THX2T99Write_7Cyc_F1 : SchedWriteRes<[THX2T99F1]> { let Latency = 7; }
> +
> +// 4 cycles on F0 or F1.
> +def THX2T99Write_4Cyc_F01 : SchedWriteRes<[THX2T99F01]> { let Latency = 4; }
> +
> +// 5 cycles on F0 or F1.
> +def THX2T99Write_5Cyc_F01 : SchedWriteRes<[THX2T99F01]> { let Latency = 5; }
> +
> +// 6 cycles on F0 or F1.
> +def THX2T99Write_6Cyc_F01 : SchedWriteRes<[THX2T99F01]> { let Latency = 6; }
> +
> +// 7 cycles on F0 or F1.
> +def THX2T99Write_7Cyc_F01 : SchedWriteRes<[THX2T99F01]> { let Latency = 7; }
> +
> +// 8 cycles on F0 or F1.
> +def THX2T99Write_8Cyc_F01 : SchedWriteRes<[THX2T99F01]> { let Latency = 8; }
> +
> +// 16 cycles on F0 or F1.
> +def THX2T99Write_16Cyc_F01 : SchedWriteRes<[THX2T99F01]> {
> + let Latency = 16;
> + let ResourceCycles = [8];
> +}
> +
> +// 23 cycles on F0 or F1.
> +def THX2T99Write_23Cyc_F01 : SchedWriteRes<[THX2T99F01]> {
> + let Latency = 23;
> + let ResourceCycles = [11];
> +}
> +
> +// 1 cycles on LS0 or LS1.
> +def THX2T99Write_1Cyc_LS01 : SchedWriteRes<[THX2T99LS01]> { let Latency = 1; }
> +
> +// 4 cycles on LS0 or LS1.
> +def THX2T99Write_4Cyc_LS01 : SchedWriteRes<[THX2T99LS01]> { let Latency = 4; }
> +
> +// 5 cycles on LS0 or LS1.
> +def THX2T99Write_5Cyc_LS01 : SchedWriteRes<[THX2T99LS01]> { let Latency = 5; }
> +
> +// 6 cycles on LS0 or LS1.
> +def THX2T99Write_6Cyc_LS01 : SchedWriteRes<[THX2T99LS01]> { let Latency = 6; }
> +
> +// 5 cycles on LS0 or LS1 and I0, I1, or I2.
> +def THX2T99Write_5Cyc_LS01_I012 : SchedWriteRes<[THX2T99LS01, THX2T99I012]> {
> + let Latency = 5;
> + let NumMicroOps = 2;
> +}
> +
> +// 5 cycles on LS0 or LS1 and 2 of I0, I1, or I2.
> +def THX2T99Write_6Cyc_LS01_I012_I012 :
> + SchedWriteRes<[THX2T99LS01, THX2T99I012, THX2T99I012]> {
> + let Latency = 6;
> + let NumMicroOps = 3;
> +}
> +
> +// 1 cycles on LS0 or LS1 and F0 or F1.
> +def THX2T99Write_1Cyc_LS01_F01 : SchedWriteRes<[THX2T99LS01, THX2T99F01]> {
> + let Latency = 1;
> + let NumMicroOps = 2;
> +}
> +
> +// 5 cycles on LS0 or LS1 and F0 or F1.
> +def THX2T99Write_5Cyc_LS01_F01 : SchedWriteRes<[THX2T99LS01, THX2T99F01]> {
> + let Latency = 5;
> + let NumMicroOps = 2;
> +}
> +
> +// 6 cycles on LS0 or LS1 and F0 or F1.
> +def THX2T99Write_6Cyc_LS01_F01 : SchedWriteRes<[THX2T99LS01, THX2T99F01]> {
> + let Latency = 6;
> + let NumMicroOps = 2;
> +}
> +
> +// 7 cycles on LS0 or LS1 and F0 or F1.
> +def THX2T99Write_7Cyc_LS01_F01 : SchedWriteRes<[THX2T99LS01, THX2T99F01]> {
> + let Latency = 7;
> + let NumMicroOps = 2;
> +}
> +
> +// 8 cycles on LS0 or LS1 and F0 or F1.
> +def THX2T99Write_8Cyc_LS01_F01 : SchedWriteRes<[THX2T99LS01, THX2T99F01]> {
> + let Latency = 8;
> + let NumMicroOps = 2;
> +}
> +
> +// Define commonly used read types.
> +
> +// No forwarding is provided for these types.
> +def : ReadAdvance<ReadI, 0>;
> +def : ReadAdvance<ReadISReg, 0>;
> +def : ReadAdvance<ReadIEReg, 0>;
> +def : ReadAdvance<ReadIM, 0>;
> +def : ReadAdvance<ReadIMA, 0>;
> +def : ReadAdvance<ReadID, 0>;
> +def : ReadAdvance<ReadExtrHi, 0>;
> +def : ReadAdvance<ReadAdrBase, 0>;
> +def : ReadAdvance<ReadVLD, 0>;
> +
> +}
> +
> +
> +//===----------------------------------------------------------------------===//
> +// 3. Instruction Tables.
> +
> +let SchedModel = ThunderX2T99Model in {
> +
> +//---
> +// 3.1 Branch Instructions
> +//---
> +
> +// Branch, immed
> +// Branch and link, immed
> +// Compare and branch
> +def : WriteRes<WriteBr, [THX2T99I2]> { let Latency = 1; }
> +
> +def : WriteRes<WriteSys, []> { let Latency = 1; }
> +def : WriteRes<WriteBarrier, []> { let Latency = 1; }
> +def : WriteRes<WriteHint, []> { let Latency = 1; }
> +
> +def : WriteRes<WriteAtomic, []> { let Unsupported = 1; }
> +
> +// Branch, register
> +// Branch and link, register != LR
> +// Branch and link, register = LR
> +def : WriteRes<WriteBrReg, [THX2T99I2]> { let Latency = 1; }
> +
> +//---
> +// 3.2 Arithmetic and Logical Instructions
> +// 3.3 Move and Shift Instructions
> +//---
> +
> +// ALU, basic
> +// Conditional compare
> +// Conditional select
> +// Address generation
> +def : WriteRes<WriteI, [THX2T99I012]> { let Latency = 1; }
> +def : InstRW<[WriteI], (instrs COPY)>;
> +
> +// ALU, extend and/or shift
> +def : WriteRes<WriteISReg, [THX2T99I012]> {
> + let Latency = 2;
> + let ResourceCycles = [2];
> +}
> +
> +def : WriteRes<WriteIEReg, [THX2T99I012]> {
> + let Latency = 2;
> + let ResourceCycles = [2];
> +}
> +
> +// Move immed
> +def : WriteRes<WriteImm, [THX2T99I012]> { let Latency = 1; }
> +
> +// Variable shift
> +def : WriteRes<WriteIS, [THX2T99I012]> { let Latency = 1; }
> +
> +//---
> +// 3.4 Divide and Multiply Instructions
> +//---
> +
> +// Divide, W-form
> +// Latency range of 13-23. Take the average.
> +def : WriteRes<WriteID32, [THX2T99I1]> {
> + let Latency = 18;
> + let ResourceCycles = [18];
> +}
> +
> +// Divide, X-form
> +// Latency range of 13-39. Take the average.
> +def : WriteRes<WriteID64, [THX2T99I1]> {
> + let Latency = 26;
> + let ResourceCycles = [26];
> +}
> +
> +// Multiply accumulate, W-form
> +def : WriteRes<WriteIM32, [THX2T99I012]> { let Latency = 5; }
> +
> +// Multiply accumulate, X-form
> +def : WriteRes<WriteIM64, [THX2T99I012]> { let Latency = 5; }
> +
> +// Bitfield extract, two reg
> +def : WriteRes<WriteExtr, [THX2T99I012]> { let Latency = 1; }
> +
> +// Bitfield move, basic
> +// Bitfield move, insert
> +// NOTE: Handled by WriteIS.
> +
> +// Count leading
> +def : InstRW<[THX2T99Write_3Cyc_I1], (instregex "^CLS(W|X)r$",
> + "^CLZ(W|X)r$")>;
> +
> +// Reverse bits/bytes
> +// NOTE: Handled by WriteI.
> +
> +//---
> +// 3.6 Load Instructions
> +// 3.10 FP Load Instructions
> +//---
> +
> +// Load register, literal
> +// Load register, unscaled immed
> +// Load register, immed unprivileged
> +// Load register, unsigned immed
> +def : WriteRes<WriteLD, [THX2T99LS01]> { let Latency = 4; }
> +
> +// Load register, immed post-index
> +// NOTE: Handled by WriteLD, WriteI.
> +// Load register, immed pre-index
> +// NOTE: Handled by WriteLD, WriteAdr.
> +def : WriteRes<WriteAdr, [THX2T99I012]> { let Latency = 1; }
> +
> +// Load register offset, basic
> +// Load register, register offset, scale by 4/8
> +// Load register, register offset, scale by 2
> +// Load register offset, extend
> +// Load register, register offset, extend, scale by 4/8
> +// Load register, register offset, extend, scale by 2
> +def THX2T99WriteLDIdx : SchedWriteVariant<[
> + SchedVar<ScaledIdxPred, [THX2T99Write_6Cyc_LS01_I012_I012]>,
> + SchedVar<NoSchedPred, [THX2T99Write_5Cyc_LS01_I012]>]>;
> +def : SchedAlias<WriteLDIdx, THX2T99WriteLDIdx>;
> +
> +def THX2T99ReadAdrBase : SchedReadVariant<[
> + SchedVar<ScaledIdxPred, [ReadDefault]>,
> + SchedVar<NoSchedPred, [ReadDefault]>]>;
> +def : SchedAlias<ReadAdrBase, THX2T99ReadAdrBase>;
> +
> +// Load pair, immed offset, normal
> +// Load pair, immed offset, signed words, base != SP
> +// Load pair, immed offset signed words, base = SP
> +// LDP only breaks into *one* LS micro-op. Thus
> +// the resources are handling by WriteLD.
> +def : WriteRes<WriteLDHi, []> {
> + let Latency = 5;
> +}
> +
> +// Load pair, immed pre-index, normal
> +// Load pair, immed pre-index, signed words
> +// Load pair, immed post-index, normal
> +// Load pair, immed post-index, signed words
> +// NOTE: Handled by WriteLD, WriteLDHi, WriteAdr.
> +
> +//--
> +// 3.7 Store Instructions
> +// 3.11 FP Store Instructions
> +//--
> +
> +// Store register, unscaled immed
> +// Store register, immed unprivileged
> +// Store register, unsigned immed
> +def : WriteRes<WriteST, [THX2T99LS01, THX2T99SD]> {
> + let Latency = 1;
> + let NumMicroOps = 2;
> +}
> +
> +// Store register, immed post-index
> +// NOTE: Handled by WriteAdr, WriteST, ReadAdrBase
> +
> +// Store register, immed pre-index
> +// NOTE: Handled by WriteAdr, WriteST
> +
> +// Store register, register offset, basic
> +// Store register, register offset, scaled by 4/8
> +// Store register, register offset, scaled by 2
> +// Store register, register offset, extend
> +// Store register, register offset, extend, scale by 4/8
> +// Store register, register offset, extend, scale by 1
> +def : WriteRes<WriteSTIdx, [THX2T99LS01, THX2T99SD, THX2T99I012]> {
> + let Latency = 1;
> + let NumMicroOps = 3;
> +}
> +
> +// Store pair, immed offset, W-form
> +// Store pair, immed offset, X-form
> +def : WriteRes<WriteSTP, [THX2T99LS01, THX2T99SD]> {
> + let Latency = 1;
> + let NumMicroOps = 2;
> +}
> +
> +// Store pair, immed post-index, W-form
> +// Store pair, immed post-index, X-form
> +// Store pair, immed pre-index, W-form
> +// Store pair, immed pre-index, X-form
> +// NOTE: Handled by WriteAdr, WriteSTP.
> +
> +//---
> +// 3.8 FP Data Processing Instructions
> +//---
> +
> +// FP absolute value
> +// FP min/max
> +// FP negate
> +def : WriteRes<WriteF, [THX2T99F01]> { let Latency = 5; }
> +
> +// FP arithmetic
> +def : InstRW<[THX2T99Write_6Cyc_F01], (instregex "^FADD", "^FSUB")>;
> +
> +// FP compare
> +def : WriteRes<WriteFCmp, [THX2T99F01]> { let Latency = 5; }
> +
> +// FP divide, S-form
> +// FP square root, S-form
> +def : WriteRes<WriteFDiv, [THX2T99F01]> {
> + let Latency = 16;
> + let ResourceCycles = [8];
> +}
> +
> +// FP divide, D-form
> +// FP square root, D-form
> +def : InstRW<[THX2T99Write_23Cyc_F01], (instrs FDIVDrr, FSQRTDr)>;
> +
> +// FP multiply
> +// FP multiply accumulate
> +def : WriteRes<WriteFMul, [THX2T99F01]> { let Latency = 6; }
> +
> +// FP round to integral
> +def : InstRW<[THX2T99Write_7Cyc_F01],
> + (instregex "^FRINT(A|I|M|N|P|X|Z)(Sr|Dr)")>;
> +
> +// FP select
> +def : InstRW<[THX2T99Write_4Cyc_F01], (instregex "^FCSEL")>;
> +
> +//---
> +// 3.9 FP Miscellaneous Instructions
> +//---
> +
> +// FP convert, from vec to vec reg
> +// FP convert, from gen to vec reg
> +// FP convert, from vec to gen reg
> +def : WriteRes<WriteFCvt, [THX2T99F01]> { let Latency = 7; }
> +
> +// FP move, immed
> +// FP move, register
> +def : WriteRes<WriteFImm, [THX2T99F01]> { let Latency = 4; }
> +
> +// FP transfer, from gen to vec reg
> +// FP transfer, from vec to gen reg
> +def : WriteRes<WriteFCopy, [THX2T99F01]> { let Latency = 4; }
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instrs FMOVXDHighr, FMOVDXHighr)>;
> +
> +//---
> +// 3.12 ASIMD Integer Instructions
> +//---
> +
> +// ASIMD absolute diff, D-form
> +// ASIMD absolute diff, Q-form
> +// ASIMD absolute diff accum, D-form
> +// ASIMD absolute diff accum, Q-form
> +// ASIMD absolute diff accum long
> +// ASIMD absolute diff long
> +// ASIMD arith, basic
> +// ASIMD arith, complex
> +// ASIMD compare
> +// ASIMD logical (AND, BIC, EOR)
> +// ASIMD max/min, basic
> +// ASIMD max/min, reduce, 4H/4S
> +// ASIMD max/min, reduce, 8B/8H
> +// ASIMD max/min, reduce, 16B
> +// ASIMD multiply, D-form
> +// ASIMD multiply, Q-form
> +// ASIMD multiply accumulate long
> +// ASIMD multiply accumulate saturating long
> +// ASIMD multiply long
> +// ASIMD pairwise add and accumulate
> +// ASIMD shift accumulate
> +// ASIMD shift by immed, basic
> +// ASIMD shift by immed and insert, basic, D-form
> +// ASIMD shift by immed and insert, basic, Q-form
> +// ASIMD shift by immed, complex
> +// ASIMD shift by register, basic, D-form
> +// ASIMD shift by register, basic, Q-form
> +// ASIMD shift by register, complex, D-form
> +// ASIMD shift by register, complex, Q-form
> +def : WriteRes<WriteV, [THX2T99F01]> { let Latency = 7; }
> +
> +// ASIMD arith, reduce, 4H/4S
> +// ASIMD arith, reduce, 8B/8H
> +// ASIMD arith, reduce, 16B
> +def : InstRW<[THX2T99Write_5Cyc_F01],
> + (instregex "^ADDVv", "^SADDLVv", "^UADDLVv")>;
> +
> +// ASIMD logical (MOV, MVN, ORN, ORR)
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^ORRv", "^ORNv", "^NOTv")>;
> +
> +// ASIMD polynomial (8x8) multiply long
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instrs PMULLv8i8, PMULLv16i8)>;
> +
> +//---
> +// 3.13 ASIMD Floating-point Instructions
> +//---
> +
> +// ASIMD FP absolute value
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^FABSv")>;
> +
> +// ASIMD FP arith, normal, D-form
> +// ASIMD FP arith, normal, Q-form
> +def : InstRW<[THX2T99Write_6Cyc_F01], (instregex "^FABDv", "^FADDv", "^FSUBv")>;
> +
> +// ASIMD FP arith,pairwise, D-form
> +// ASIMD FP arith, pairwise, Q-form
> +def : InstRW<[THX2T99Write_6Cyc_F01], (instregex "^FADDPv")>;
> +
> +// ASIMD FP compare, D-form
> +// ASIMD FP compare, Q-form
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^FACGEv", "^FACGTv")>;
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^FCMEQv", "^FCMGEv",
> + "^FCMGTv", "^FCMLEv",
> + "^FCMLTv")>;
> +
> +// ASIMD FP convert, long
> +// ASIMD FP convert, narrow
> +// ASIMD FP convert, other, D-form
> +// ASIMD FP convert, other, Q-form
> +// NOTE: Handled by WriteV.
> +
> +// ASIMD FP divide, D-form, F32
> +def : InstRW<[THX2T99Write_16Cyc_F01], (instrs FDIVv2f32)>;
> +
> +// ASIMD FP divide, Q-form, F32
> +def : InstRW<[THX2T99Write_16Cyc_F01], (instrs FDIVv4f32)>;
> +
> +// ASIMD FP divide, Q-form, F64
> +def : InstRW<[THX2T99Write_23Cyc_F01], (instrs FDIVv2f64)>;
> +
> +// ASIMD FP max/min, normal, D-form
> +// ASIMD FP max/min, normal, Q-form
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^FMAXv", "^FMAXNMv",
> + "^FMINv", "^FMINNMv")>;
> +
> +// ASIMD FP max/min, pairwise, D-form
> +// ASIMD FP max/min, pairwise, Q-form
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^FMAXPv", "^FMAXNMPv",
> + "^FMINPv", "^FMINNMPv")>;
> +
> +// ASIMD FP max/min, reduce
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^FMAXVv", "^FMAXNMVv",
> + "^FMINVv", "^FMINNMVv")>;
> +
> +// ASIMD FP multiply, D-form, FZ
> +// ASIMD FP multiply, D-form, no FZ
> +// ASIMD FP multiply, Q-form, FZ
> +// ASIMD FP multiply, Q-form, no FZ
> +def : InstRW<[THX2T99Write_6Cyc_F01], (instregex "^FMULv", "^FMULXv")>;
> +
> +// ASIMD FP multiply accumulate, Dform, FZ
> +// ASIMD FP multiply accumulate, Dform, no FZ
> +// ASIMD FP multiply accumulate, Qform, FZ
> +// ASIMD FP multiply accumulate, Qform, no FZ
> +def : InstRW<[THX2T99Write_6Cyc_F01], (instregex "^FMLAv", "^FMLSv")>;
> +
> +// ASIMD FP negate
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^FNEGv")>;
> +
> +// ASIMD FP round, D-form
> +// ASIMD FP round, Q-form
> +// NOTE: Handled by WriteV.
> +
> +//--
> +// 3.14 ASIMD Miscellaneous Instructions
> +//--
> +
> +// ASIMD bit reverse
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^RBITv")>;
> +
> +// ASIMD bitwise insert, D-form
> +// ASIMD bitwise insert, Q-form
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^BIFv", "^BITv", "^BSLv")>;
> +
> +// ASIMD count, D-form
> +// ASIMD count, Q-form
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^CLSv", "^CLZv", "^CNTv")>;
> +
> +// ASIMD duplicate, gen reg
> +// ASIMD duplicate, element
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^DUPv")>;
> +
> +// ASIMD extract
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^EXTv")>;
> +
> +// ASIMD extract narrow
> +// ASIMD extract narrow, saturating
> +// NOTE: Handled by WriteV.
> +
> +// ASIMD insert, element to element
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^INSv")>;
> +
> +// ASIMD move, integer immed
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^MOVIv", "^MOVIDv")>;
> +
> +// ASIMD move, FP immed
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^FMOVv")>;
> +
> +// ASIMD reciprocal estimate, D-form
> +// ASIMD reciprocal estimate, Q-form
> +def : InstRW<[THX2T99Write_5Cyc_F01],
> + (instregex "^FRECPEv", "^FRECPXv", "^URECPEv",
> + "^FRSQRTEv", "^URSQRTEv")>;
> +
> +// ASIMD reciprocal step, D-form, FZ
> +// ASIMD reciprocal step, D-form, no FZ
> +// ASIMD reciprocal step, Q-form, FZ
> +// ASIMD reciprocal step, Q-form, no FZ
> +def : InstRW<[THX2T99Write_6Cyc_F01], (instregex "^FRECPSv", "^FRSQRTSv")>;
> +
> +// ASIMD reverse
> +def : InstRW<[THX2T99Write_5Cyc_F01],
> + (instregex "^REV16v", "^REV32v", "^REV64v")>;
> +
> +// ASIMD table lookup, D-form
> +// ASIMD table lookup, Q-form
> +def : InstRW<[THX2T99Write_8Cyc_F01], (instregex "^TBLv", "^TBXv")>;
> +
> +// ASIMD transfer, element to word or word
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^UMOVv")>;
> +
> +// ASIMD transfer, element to gen reg
> +def : InstRW<[THX2T99Write_6Cyc_F01], (instregex "^SMOVv", "^UMOVv")>;
> +
> +// ASIMD transfer gen reg to element
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^INSv")>;
> +
> +// ASIMD transpose
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^TRN1v", "^TRN2v",
> + "^UZP1v", "^UZP2v")>;
> +
> +// ASIMD unzip/zip
> +def : InstRW<[THX2T99Write_5Cyc_F01], (instregex "^ZIP1v", "^ZIP2v")>;
> +
> +//--
> +// 3.15 ASIMD Load Instructions
> +//--
> +
> +// ASIMD load, 1 element, multiple, 1 reg, D-form
> +// ASIMD load, 1 element, multiple, 1 reg, Q-form
> +def : InstRW<[THX2T99Write_4Cyc_LS01],
> + (instregex "^LD1Onev(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_4Cyc_LS01, WriteAdr],
> + (instregex "^LD1Onev(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> +
> +// ASIMD load, 1 element, multiple, 2 reg, D-form
> +// ASIMD load, 1 element, multiple, 2 reg, Q-form
> +def : InstRW<[THX2T99Write_4Cyc_LS01],
> + (instregex "^LD1Twov(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_4Cyc_LS01, WriteAdr],
> + (instregex "^LD1Twov(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> +
> +// ASIMD load, 1 element, multiple, 3 reg, D-form
> +// ASIMD load, 1 element, multiple, 3 reg, Q-form
> +def : InstRW<[THX2T99Write_5Cyc_LS01],
> + (instregex "^LD1Threev(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_5Cyc_LS01, WriteAdr],
> + (instregex "^LD1Threev(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> +
> +// ASIMD load, 1 element, multiple, 4 reg, D-form
> +// ASIMD load, 1 element, multiple, 4 reg, Q-form
> +def : InstRW<[THX2T99Write_6Cyc_LS01],
> + (instregex "^LD1Fourv(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_6Cyc_LS01, WriteAdr],
> + (instregex "^LD1Fourv(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> +
> +// ASIMD load, 1 element, one lane, B/H/S
> +// ASIMD load, 1 element, one lane, D
> +def : InstRW<[THX2T99Write_5Cyc_LS01_F01], (instregex "^LD1i(8|16|32|64)$")>;
> +def : InstRW<[THX2T99Write_5Cyc_LS01_F01, WriteAdr],
> + (instregex "^LD1i(8|16|32|64)_POST$")>;
> +
> +// ASIMD load, 1 element, all lanes, D-form, B/H/S
> +// ASIMD load, 1 element, all lanes, D-form, D
> +// ASIMD load, 1 element, all lanes, Q-form
> +def : InstRW<[THX2T99Write_5Cyc_LS01_F01],
> + (instregex "^LD1Rv(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_5Cyc_LS01_F01, WriteAdr],
> + (instregex "^LD1Rv(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> +
> +// ASIMD load, 2 element, multiple, D-form, B/H/S
> +// ASIMD load, 2 element, multiple, Q-form, D
> +def : InstRW<[THX2T99Write_5Cyc_LS01_F01],
> + (instregex "^LD2Twov(8b|4h|2s|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_5Cyc_LS01_F01, WriteAdr],
> + (instregex "^LD2Twov(8b|4h|2s|16b|8h|4s|2d)_POST$")>;
> +
> +// ASIMD load, 2 element, one lane, B/H
> +// ASIMD load, 2 element, one lane, S
> +// ASIMD load, 2 element, one lane, D
> +def : InstRW<[THX2T99Write_5Cyc_LS01_F01], (instregex "^LD2i(8|16|32|64)$")>;
> +def : InstRW<[THX2T99Write_5Cyc_LS01_F01, WriteAdr],
> + (instregex "^LD2i(8|16|32|64)_POST$")>;
> +
> +// ASIMD load, 2 element, all lanes, D-form, B/H/S
> +// ASIMD load, 2 element, all lanes, D-form, D
> +// ASIMD load, 2 element, all lanes, Q-form
> +def : InstRW<[THX2T99Write_5Cyc_LS01_F01],
> + (instregex "^LD2Rv(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_5Cyc_LS01_F01, WriteAdr],
> + (instregex "^LD2Rv(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> +
> +// ASIMD load, 3 element, multiple, D-form, B/H/S
> +// ASIMD load, 3 element, multiple, Q-form, B/H/S
> +// ASIMD load, 3 element, multiple, Q-form, D
> +def : InstRW<[THX2T99Write_8Cyc_LS01_F01],
> + (instregex "^LD3Threev(8b|4h|2s|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_8Cyc_LS01_F01, WriteAdr],
> + (instregex "^LD3Threev(8b|4h|2s|16b|8h|4s|2d)_POST$")>;
> +
> +// ASIMD load, 3 element, one lone, B/H
> +// ASIMD load, 3 element, one lane, S
> +// ASIMD load, 3 element, one lane, D
> +def : InstRW<[THX2T99Write_7Cyc_LS01_F01], (instregex "^LD3i(8|16|32|64)$")>;
> +def : InstRW<[THX2T99Write_7Cyc_LS01_F01, WriteAdr],
> + (instregex "^LD3i(8|16|32|64)_POST$")>;
> +
> +// ASIMD load, 3 element, all lanes, D-form, B/H/S
> +// ASIMD load, 3 element, all lanes, D-form, D
> +// ASIMD load, 3 element, all lanes, Q-form, B/H/S
> +// ASIMD load, 3 element, all lanes, Q-form, D
> +def : InstRW<[THX2T99Write_7Cyc_LS01_F01],
> + (instregex "^LD3Rv(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_7Cyc_LS01_F01, WriteAdr],
> + (instregex "^LD3Rv(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> +
> +// ASIMD load, 4 element, multiple, D-form, B/H/S
> +// ASIMD load, 4 element, multiple, Q-form, B/H/S
> +// ASIMD load, 4 element, multiple, Q-form, D
> +def : InstRW<[THX2T99Write_8Cyc_LS01_F01],
> + (instregex "^LD4Fourv(8b|4h|2s|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_8Cyc_LS01_F01, WriteAdr],
> + (instregex "^LD4Fourv(8b|4h|2s|16b|8h|4s|2d)_POST$")>;
> +
> +// ASIMD load, 4 element, one lane, B/H
> +// ASIMD load, 4 element, one lane, S
> +// ASIMD load, 4 element, one lane, D
> +def : InstRW<[THX2T99Write_6Cyc_LS01_F01], (instregex "^LD4i(8|16|32|64)$")>;
> +def : InstRW<[THX2T99Write_6Cyc_LS01_F01, WriteAdr],
> + (instregex "^LD4i(8|16|32|64)_POST$")>;
> +
> +// ASIMD load, 4 element, all lanes, D-form, B/H/S
> +// ASIMD load, 4 element, all lanes, D-form, D
> +// ASIMD load, 4 element, all lanes, Q-form, B/H/S
> +// ASIMD load, 4 element, all lanes, Q-form, D
> +def : InstRW<[THX2T99Write_6Cyc_LS01_F01],
> + (instregex "^LD4Rv(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_6Cyc_LS01_F01, WriteAdr],
> + (instregex "^LD4Rv(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> +
> +//--
> +// 3.16 ASIMD Store Instructions
> +//--
> +
> +// ASIMD store, 1 element, multiple, 1 reg, D-form
> +// ASIMD store, 1 element, multiple, 1 reg, Q-form
> +def : InstRW<[THX2T99Write_1Cyc_LS01],
> + (instregex "^ST1Onev(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_1Cyc_LS01, WriteAdr],
> + (instregex "^ST1Onev(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> +
> +// ASIMD store, 1 element, multiple, 2 reg, D-form
> +// ASIMD store, 1 element, multiple, 2 reg, Q-form
> +def : InstRW<[THX2T99Write_1Cyc_LS01],
> + (instregex "^ST1Twov(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_1Cyc_LS01, WriteAdr],
> + (instregex "^ST1Twov(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> +
> +// ASIMD store, 1 element, multiple, 3 reg, D-form
> +// ASIMD store, 1 element, multiple, 3 reg, Q-form
> +def : InstRW<[THX2T99Write_1Cyc_LS01],
> + (instregex "^ST1Threev(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_1Cyc_LS01, WriteAdr],
> + (instregex "^ST1Threev(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> +
> +// ASIMD store, 1 element, multiple, 4 reg, D-form
> +// ASIMD store, 1 element, multiple, 4 reg, Q-form
> +def : InstRW<[THX2T99Write_1Cyc_LS01],
> + (instregex "^ST1Fourv(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_1Cyc_LS01, WriteAdr],
> + (instregex "^ST1Fourv(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> +
> +// ASIMD store, 1 element, one lane, B/H/S
> +// ASIMD store, 1 element, one lane, D
> +def : InstRW<[THX2T99Write_1Cyc_LS01_F01],
> + (instregex "^ST1i(8|16|32|64)$")>;
> +def : InstRW<[THX2T99Write_1Cyc_LS01_F01, WriteAdr],
> + (instregex "^ST1i(8|16|32|64)_POST$")>;
> +
> +// ASIMD store, 2 element, multiple, D-form, B/H/S
> +// ASIMD store, 2 element, multiple, Q-form, B/H/S
> +// ASIMD store, 2 element, multiple, Q-form, D
> +def : InstRW<[THX2T99Write_1Cyc_LS01_F01],
> + (instregex "^ST2Twov(8b|4h|2s|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_1Cyc_LS01_F01, WriteAdr],
> + (instregex "^ST2Twov(8b|4h|2s|16b|8h|4s|2d)_POST$")>;
> +
> +// ASIMD store, 2 element, one lane, B/H/S
> +// ASIMD store, 2 element, one lane, D
> +def : InstRW<[THX2T99Write_1Cyc_LS01_F01],
> + (instregex "^ST2i(8|16|32|64)$")>;
> +def : InstRW<[THX2T99Write_1Cyc_LS01_F01, WriteAdr],
> + (instregex "^ST2i(8|16|32|64)_POST$")>;
> +
> +// ASIMD store, 3 element, multiple, D-form, B/H/S
> +// ASIMD store, 3 element, multiple, Q-form, B/H/S
> +// ASIMD store, 3 element, multiple, Q-form, D
> +def : InstRW<[THX2T99Write_1Cyc_LS01_F01],
> + (instregex "^ST3Threev(8b|4h|2s|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_1Cyc_LS01_F01, WriteAdr],
> + (instregex "^ST3Threev(8b|4h|2s|16b|8h|4s|2d)_POST$")>;
> +
> +// ASIMD store, 3 element, one lane, B/H
> +// ASIMD store, 3 element, one lane, S
> +// ASIMD store, 3 element, one lane, D
> +def : InstRW<[THX2T99Write_1Cyc_LS01_F01], (instregex "^ST3i(8|16|32|64)$")>;
> +def : InstRW<[THX2T99Write_1Cyc_LS01_F01, WriteAdr],
> + (instregex "^ST3i(8|16|32|64)_POST$")>;
> +
> +// ASIMD store, 4 element, multiple, D-form, B/H/S
> +// ASIMD store, 4 element, multiple, Q-form, B/H/S
> +// ASIMD store, 4 element, multiple, Q-form, D
> +def : InstRW<[THX2T99Write_1Cyc_LS01_F01],
> + (instregex "^ST4Fourv(8b|4h|2s|16b|8h|4s|2d)$")>;
> +def : InstRW<[THX2T99Write_1Cyc_LS01_F01, WriteAdr],
> + (instregex "^ST4Fourv(8b|4h|2s|16b|8h|4s|2d)_POST$")>;
> +
> +// ASIMD store, 4 element, one lane, B/H
> +// ASIMD store, 4 element, one lane, S
> +// ASIMD store, 4 element, one lane, D
> +def : InstRW<[THX2T99Write_1Cyc_LS01_F01], (instregex "^ST4i(8|16|32|64)$")>;
> +def : InstRW<[THX2T99Write_1Cyc_LS01_F01, WriteAdr],
> + (instregex "^ST4i(8|16|32|64)_POST$")>;
> +
> +//--
> +// 3.17 Cryptography Extensions
> +//--
> +
> +// Crypto AES ops
> +def : InstRW<[THX2T99Write_5Cyc_F1], (instregex "^AES")>;
> +
> +// Crypto polynomial (64x64) multiply long
> +def : InstRW<[THX2T99Write_5Cyc_F1], (instrs PMULLv1i64, PMULLv2i64)>;
> +
> +// Crypto SHA1 xor ops
> +// Crypto SHA1 schedule acceleration ops
> +// Crypto SHA256 schedule acceleration op (1 u-op)
> +// Crypto SHA256 schedule acceleration op (2 u-ops)
> +// Crypto SHA256 hash acceleration ops
> +def : InstRW<[THX2T99Write_7Cyc_F1], (instregex "^SHA")>;
> +
> +//--
> +// 3.18 CRC
> +//--
> +
> +// CRC checksum ops
> +def : InstRW<[THX2T99Write_4Cyc_I1], (instregex "^CRC32")>;
> +
> +} // SchedModel = ThunderX2T99Model
>
> Removed: llvm/trunk/lib/Target/AArch64/AArch64SchedVulcan.td
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AArch64/AArch64SchedVulcan.td?rev=297189&view=auto <http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AArch64/AArch64SchedVulcan.td?rev=297189&view=auto>
> ==============================================================================
> --- llvm/trunk/lib/Target/AArch64/AArch64SchedVulcan.td (original)
> +++ llvm/trunk/lib/Target/AArch64/AArch64SchedVulcan.td (removed)
> @@ -1,852 +0,0 @@
> -//=- AArch64SchedVulcan.td - Vulcan Scheduling Defs ----------*- tablegen -*-=//
> -//
> -// The LLVM Compiler Infrastructure
> -//
> -// This file is distributed under the University of Illinois Open Source
> -// License. See LICENSE.TXT for details.
> -//
> -//===----------------------------------------------------------------------===//
> -// 1. Introduction
> -//
> -// This file defines the machine model for Broadcom Vulcan to support
> -// instruction scheduling and other instruction cost heuristics.
> -//
> -//===----------------------------------------------------------------------===//
> -
> -//===----------------------------------------------------------------------===//
> -// 2. Pipeline Description.
> -
> -def VulcanModel : SchedMachineModel {
> - let IssueWidth = 4; // 4 micro-ops dispatched at a time.
> - let MicroOpBufferSize = 180; // 180 entries in micro-op re-order buffer.
> - let LoadLatency = 4; // Optimistic load latency.
> - let MispredictPenalty = 12; // Extra cycles for mispredicted branch.
> - // Determined via a mix of micro-arch details and experimentation.
> - let LoopMicroOpBufferSize = 32;
> - let PostRAScheduler = 1; // Using PostRA sched.
> - let CompleteModel = 1;
> -}
> -
> -// Define the issue ports.
> -
> -// Port 0: ALU, FP/SIMD.
> -def VulcanP0 : ProcResource<1>;
> -
> -// Port 1: ALU, FP/SIMD, integer mul/div.
> -def VulcanP1 : ProcResource<1>;
> -
> -// Port 2: ALU, Branch.
> -def VulcanP2 : ProcResource<1>;
> -
> -// Port 3: Store data.
> -def VulcanP3 : ProcResource<1>;
> -
> -// Port 4: Load/store.
> -def VulcanP4 : ProcResource<1>;
> -
> -// Port 5: Load/store.
> -def VulcanP5 : ProcResource<1>;
> -
> -let SchedModel = VulcanModel in {
> -
> -// Define groups for the functional units on each issue port. Each group
> -// created will be used by a WriteRes later on.
> -//
> -// NOTE: Some groups only contain one member. This is a way to create names for
> -// the various functional units that share a single issue port. For example,
> -// "VulcanI1" for ALU ops on port 1 and "VulcanF1" for FP ops on port 1.
> -
> -// Integer divide and multiply micro-ops only on port 1.
> -def VulcanI1 : ProcResGroup<[VulcanP1]>;
> -
> -// Branch micro-ops only on port 2.
> -def VulcanI2 : ProcResGroup<[VulcanP2]>;
> -
> -// ALU micro-ops on ports 0, 1, and 2.
> -def VulcanI012 : ProcResGroup<[VulcanP0, VulcanP1, VulcanP2]>;
> -
> -// Crypto FP/SIMD micro-ops only on port 1.
> -def VulcanF1 : ProcResGroup<[VulcanP1]>;
> -
> -// FP/SIMD micro-ops on ports 0 and 1.
> -def VulcanF01 : ProcResGroup<[VulcanP0, VulcanP1]>;
> -
> -// Store data micro-ops only on port 3.
> -def VulcanSD : ProcResGroup<[VulcanP3]>;
> -
> -// Load/store micro-ops on ports 4 and 5.
> -def VulcanLS01 : ProcResGroup<[VulcanP4, VulcanP5]>;
> -
> -// 60 entry unified scheduler.
> -def VulcanAny : ProcResGroup<[VulcanP0, VulcanP1, VulcanP2,
> - VulcanP3, VulcanP4, VulcanP5]> {
> - let BufferSize=60;
> -}
> -
> -// Define commonly used write types for InstRW specializations.
> -// All definitions follow the format: VulcanWrite_<NumCycles>Cyc_<Resources>.
> -
> -// 3 cycles on I1.
> -def VulcanWrite_3Cyc_I1 : SchedWriteRes<[VulcanI1]> { let Latency = 3; }
> -
> -// 4 cycles on I1.
> -def VulcanWrite_4Cyc_I1 : SchedWriteRes<[VulcanI1]> { let Latency = 4; }
> -
> -// 1 cycle on I0, I1, or I2.
> -def VulcanWrite_1Cyc_I012 : SchedWriteRes<[VulcanI012]> { let Latency = 1; }
> -
> -// 5 cycles on F1.
> -def VulcanWrite_5Cyc_F1 : SchedWriteRes<[VulcanF1]> { let Latency = 5; }
> -
> -// 7 cycles on F1.
> -def VulcanWrite_7Cyc_F1 : SchedWriteRes<[VulcanF1]> { let Latency = 7; }
> -
> -// 4 cycles on F0 or F1.
> -def VulcanWrite_4Cyc_F01 : SchedWriteRes<[VulcanF01]> { let Latency = 4; }
> -
> -// 5 cycles on F0 or F1.
> -def VulcanWrite_5Cyc_F01 : SchedWriteRes<[VulcanF01]> { let Latency = 5; }
> -
> -// 6 cycles on F0 or F1.
> -def VulcanWrite_6Cyc_F01 : SchedWriteRes<[VulcanF01]> { let Latency = 6; }
> -
> -// 7 cycles on F0 or F1.
> -def VulcanWrite_7Cyc_F01 : SchedWriteRes<[VulcanF01]> { let Latency = 7; }
> -
> -// 8 cycles on F0 or F1.
> -def VulcanWrite_8Cyc_F01 : SchedWriteRes<[VulcanF01]> { let Latency = 8; }
> -
> -// 16 cycles on F0 or F1.
> -def VulcanWrite_16Cyc_F01 : SchedWriteRes<[VulcanF01]> {
> - let Latency = 16;
> - let ResourceCycles = [8];
> -}
> -
> -// 23 cycles on F0 or F1.
> -def VulcanWrite_23Cyc_F01 : SchedWriteRes<[VulcanF01]> {
> - let Latency = 23;
> - let ResourceCycles = [11];
> -}
> -
> -// 1 cycles on LS0 or LS1.
> -def VulcanWrite_1Cyc_LS01 : SchedWriteRes<[VulcanLS01]> { let Latency = 1; }
> -
> -// 4 cycles on LS0 or LS1.
> -def VulcanWrite_4Cyc_LS01 : SchedWriteRes<[VulcanLS01]> { let Latency = 4; }
> -
> -// 5 cycles on LS0 or LS1.
> -def VulcanWrite_5Cyc_LS01 : SchedWriteRes<[VulcanLS01]> { let Latency = 5; }
> -
> -// 6 cycles on LS0 or LS1.
> -def VulcanWrite_6Cyc_LS01 : SchedWriteRes<[VulcanLS01]> { let Latency = 6; }
> -
> -// 5 cycles on LS0 or LS1 and I0, I1, or I2.
> -def VulcanWrite_5Cyc_LS01_I012 : SchedWriteRes<[VulcanLS01, VulcanI012]> {
> - let Latency = 5;
> - let NumMicroOps = 2;
> -}
> -
> -// 5 cycles on LS0 or LS1 and 2 of I0, I1, or I2.
> -def VulcanWrite_6Cyc_LS01_I012_I012 :
> - SchedWriteRes<[VulcanLS01, VulcanI012, VulcanI012]> {
> - let Latency = 6;
> - let NumMicroOps = 3;
> -}
> -
> -// 1 cycles on LS0 or LS1 and F0 or F1.
> -def VulcanWrite_1Cyc_LS01_F01 : SchedWriteRes<[VulcanLS01, VulcanF01]> {
> - let Latency = 1;
> - let NumMicroOps = 2;
> -}
> -
> -// 5 cycles on LS0 or LS1 and F0 or F1.
> -def VulcanWrite_5Cyc_LS01_F01 : SchedWriteRes<[VulcanLS01, VulcanF01]> {
> - let Latency = 5;
> - let NumMicroOps = 2;
> -}
> -
> -// 6 cycles on LS0 or LS1 and F0 or F1.
> -def VulcanWrite_6Cyc_LS01_F01 : SchedWriteRes<[VulcanLS01, VulcanF01]> {
> - let Latency = 6;
> - let NumMicroOps = 2;
> -}
> -
> -// 7 cycles on LS0 or LS1 and F0 or F1.
> -def VulcanWrite_7Cyc_LS01_F01 : SchedWriteRes<[VulcanLS01, VulcanF01]> {
> - let Latency = 7;
> - let NumMicroOps = 2;
> -}
> -
> -// 8 cycles on LS0 or LS1 and F0 or F1.
> -def VulcanWrite_8Cyc_LS01_F01 : SchedWriteRes<[VulcanLS01, VulcanF01]> {
> - let Latency = 8;
> - let NumMicroOps = 2;
> -}
> -
> -// Define commonly used read types.
> -
> -// No forwarding is provided for these types.
> -def : ReadAdvance<ReadI, 0>;
> -def : ReadAdvance<ReadISReg, 0>;
> -def : ReadAdvance<ReadIEReg, 0>;
> -def : ReadAdvance<ReadIM, 0>;
> -def : ReadAdvance<ReadIMA, 0>;
> -def : ReadAdvance<ReadID, 0>;
> -def : ReadAdvance<ReadExtrHi, 0>;
> -def : ReadAdvance<ReadAdrBase, 0>;
> -def : ReadAdvance<ReadVLD, 0>;
> -
> -}
> -
> -
> -//===----------------------------------------------------------------------===//
> -// 3. Instruction Tables.
> -
> -let SchedModel = VulcanModel in {
> -
> -//---
> -// 3.1 Branch Instructions
> -//---
> -
> -// Branch, immed
> -// Branch and link, immed
> -// Compare and branch
> -def : WriteRes<WriteBr, [VulcanI2]> { let Latency = 1; }
> -
> -def : WriteRes<WriteSys, []> { let Latency = 1; }
> -def : WriteRes<WriteBarrier, []> { let Latency = 1; }
> -def : WriteRes<WriteHint, []> { let Latency = 1; }
> -
> -def : WriteRes<WriteAtomic, []> { let Unsupported = 1; }
> -
> -// Branch, register
> -// Branch and link, register != LR
> -// Branch and link, register = LR
> -def : WriteRes<WriteBrReg, [VulcanI2]> { let Latency = 1; }
> -
> -//---
> -// 3.2 Arithmetic and Logical Instructions
> -// 3.3 Move and Shift Instructions
> -//---
> -
> -// ALU, basic
> -// Conditional compare
> -// Conditional select
> -// Address generation
> -def : WriteRes<WriteI, [VulcanI012]> { let Latency = 1; }
> -def : InstRW<[WriteI], (instrs COPY)>;
> -
> -// ALU, extend and/or shift
> -def : WriteRes<WriteISReg, [VulcanI012]> {
> - let Latency = 2;
> - let ResourceCycles = [2];
> -}
> -
> -def : WriteRes<WriteIEReg, [VulcanI012]> {
> - let Latency = 2;
> - let ResourceCycles = [2];
> -}
> -
> -// Move immed
> -def : WriteRes<WriteImm, [VulcanI012]> { let Latency = 1; }
> -
> -// Variable shift
> -def : WriteRes<WriteIS, [VulcanI012]> { let Latency = 1; }
> -
> -//---
> -// 3.4 Divide and Multiply Instructions
> -//---
> -
> -// Divide, W-form
> -// Latency range of 13-23. Take the average.
> -def : WriteRes<WriteID32, [VulcanI1]> {
> - let Latency = 18;
> - let ResourceCycles = [18];
> -}
> -
> -// Divide, X-form
> -// Latency range of 13-39. Take the average.
> -def : WriteRes<WriteID64, [VulcanI1]> {
> - let Latency = 26;
> - let ResourceCycles = [26];
> -}
> -
> -// Multiply accumulate, W-form
> -def : WriteRes<WriteIM32, [VulcanI012]> { let Latency = 5; }
> -
> -// Multiply accumulate, X-form
> -def : WriteRes<WriteIM64, [VulcanI012]> { let Latency = 5; }
> -
> -// Bitfield extract, two reg
> -def : WriteRes<WriteExtr, [VulcanI012]> { let Latency = 1; }
> -
> -// Bitfield move, basic
> -// Bitfield move, insert
> -// NOTE: Handled by WriteIS.
> -
> -// Count leading
> -def : InstRW<[VulcanWrite_3Cyc_I1], (instregex "^CLS(W|X)r$",
> - "^CLZ(W|X)r$")>;
> -
> -// Reverse bits/bytes
> -// NOTE: Handled by WriteI.
> -
> -//---
> -// 3.6 Load Instructions
> -// 3.10 FP Load Instructions
> -//---
> -
> -// Load register, literal
> -// Load register, unscaled immed
> -// Load register, immed unprivileged
> -// Load register, unsigned immed
> -def : WriteRes<WriteLD, [VulcanLS01]> { let Latency = 4; }
> -
> -// Load register, immed post-index
> -// NOTE: Handled by WriteLD, WriteI.
> -// Load register, immed pre-index
> -// NOTE: Handled by WriteLD, WriteAdr.
> -def : WriteRes<WriteAdr, [VulcanI012]> { let Latency = 1; }
> -
> -// Load register offset, basic
> -// Load register, register offset, scale by 4/8
> -// Load register, register offset, scale by 2
> -// Load register offset, extend
> -// Load register, register offset, extend, scale by 4/8
> -// Load register, register offset, extend, scale by 2
> -def VulcanWriteLDIdx : SchedWriteVariant<[
> - SchedVar<ScaledIdxPred, [VulcanWrite_6Cyc_LS01_I012_I012]>,
> - SchedVar<NoSchedPred, [VulcanWrite_5Cyc_LS01_I012]>]>;
> -def : SchedAlias<WriteLDIdx, VulcanWriteLDIdx>;
> -
> -def VulcanReadAdrBase : SchedReadVariant<[
> - SchedVar<ScaledIdxPred, [ReadDefault]>,
> - SchedVar<NoSchedPred, [ReadDefault]>]>;
> -def : SchedAlias<ReadAdrBase, VulcanReadAdrBase>;
> -
> -// Load pair, immed offset, normal
> -// Load pair, immed offset, signed words, base != SP
> -// Load pair, immed offset signed words, base = SP
> -// LDP only breaks into *one* LS micro-op. Thus
> -// the resources are handling by WriteLD.
> -def : WriteRes<WriteLDHi, []> {
> - let Latency = 5;
> -}
> -
> -// Load pair, immed pre-index, normal
> -// Load pair, immed pre-index, signed words
> -// Load pair, immed post-index, normal
> -// Load pair, immed post-index, signed words
> -// NOTE: Handled by WriteLD, WriteLDHi, WriteAdr.
> -
> -//--
> -// 3.7 Store Instructions
> -// 3.11 FP Store Instructions
> -//--
> -
> -// Store register, unscaled immed
> -// Store register, immed unprivileged
> -// Store register, unsigned immed
> -def : WriteRes<WriteST, [VulcanLS01, VulcanSD]> {
> - let Latency = 1;
> - let NumMicroOps = 2;
> -}
> -
> -// Store register, immed post-index
> -// NOTE: Handled by WriteAdr, WriteST, ReadAdrBase
> -
> -// Store register, immed pre-index
> -// NOTE: Handled by WriteAdr, WriteST
> -
> -// Store register, register offset, basic
> -// Store register, register offset, scaled by 4/8
> -// Store register, register offset, scaled by 2
> -// Store register, register offset, extend
> -// Store register, register offset, extend, scale by 4/8
> -// Store register, register offset, extend, scale by 1
> -def : WriteRes<WriteSTIdx, [VulcanLS01, VulcanSD, VulcanI012]> {
> - let Latency = 1;
> - let NumMicroOps = 3;
> -}
> -
> -// Store pair, immed offset, W-form
> -// Store pair, immed offset, X-form
> -def : WriteRes<WriteSTP, [VulcanLS01, VulcanSD]> {
> - let Latency = 1;
> - let NumMicroOps = 2;
> -}
> -
> -// Store pair, immed post-index, W-form
> -// Store pair, immed post-index, X-form
> -// Store pair, immed pre-index, W-form
> -// Store pair, immed pre-index, X-form
> -// NOTE: Handled by WriteAdr, WriteSTP.
> -
> -//---
> -// 3.8 FP Data Processing Instructions
> -//---
> -
> -// FP absolute value
> -// FP min/max
> -// FP negate
> -def : WriteRes<WriteF, [VulcanF01]> { let Latency = 5; }
> -
> -// FP arithmetic
> -def : InstRW<[VulcanWrite_6Cyc_F01], (instregex "^FADD", "^FSUB")>;
> -
> -// FP compare
> -def : WriteRes<WriteFCmp, [VulcanF01]> { let Latency = 5; }
> -
> -// FP divide, S-form
> -// FP square root, S-form
> -def : WriteRes<WriteFDiv, [VulcanF01]> {
> - let Latency = 16;
> - let ResourceCycles = [8];
> -}
> -
> -// FP divide, D-form
> -// FP square root, D-form
> -def : InstRW<[VulcanWrite_23Cyc_F01], (instrs FDIVDrr, FSQRTDr)>;
> -
> -// FP multiply
> -// FP multiply accumulate
> -def : WriteRes<WriteFMul, [VulcanF01]> { let Latency = 6; }
> -
> -// FP round to integral
> -def : InstRW<[VulcanWrite_7Cyc_F01],
> - (instregex "^FRINT(A|I|M|N|P|X|Z)(Sr|Dr)")>;
> -
> -// FP select
> -def : InstRW<[VulcanWrite_4Cyc_F01], (instregex "^FCSEL")>;
> -
> -//---
> -// 3.9 FP Miscellaneous Instructions
> -//---
> -
> -// FP convert, from vec to vec reg
> -// FP convert, from gen to vec reg
> -// FP convert, from vec to gen reg
> -def : WriteRes<WriteFCvt, [VulcanF01]> { let Latency = 7; }
> -
> -// FP move, immed
> -// FP move, register
> -def : WriteRes<WriteFImm, [VulcanF01]> { let Latency = 4; }
> -
> -// FP transfer, from gen to vec reg
> -// FP transfer, from vec to gen reg
> -def : WriteRes<WriteFCopy, [VulcanF01]> { let Latency = 4; }
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instrs FMOVXDHighr, FMOVDXHighr)>;
> -
> -//---
> -// 3.12 ASIMD Integer Instructions
> -//---
> -
> -// ASIMD absolute diff, D-form
> -// ASIMD absolute diff, Q-form
> -// ASIMD absolute diff accum, D-form
> -// ASIMD absolute diff accum, Q-form
> -// ASIMD absolute diff accum long
> -// ASIMD absolute diff long
> -// ASIMD arith, basic
> -// ASIMD arith, complex
> -// ASIMD compare
> -// ASIMD logical (AND, BIC, EOR)
> -// ASIMD max/min, basic
> -// ASIMD max/min, reduce, 4H/4S
> -// ASIMD max/min, reduce, 8B/8H
> -// ASIMD max/min, reduce, 16B
> -// ASIMD multiply, D-form
> -// ASIMD multiply, Q-form
> -// ASIMD multiply accumulate long
> -// ASIMD multiply accumulate saturating long
> -// ASIMD multiply long
> -// ASIMD pairwise add and accumulate
> -// ASIMD shift accumulate
> -// ASIMD shift by immed, basic
> -// ASIMD shift by immed and insert, basic, D-form
> -// ASIMD shift by immed and insert, basic, Q-form
> -// ASIMD shift by immed, complex
> -// ASIMD shift by register, basic, D-form
> -// ASIMD shift by register, basic, Q-form
> -// ASIMD shift by register, complex, D-form
> -// ASIMD shift by register, complex, Q-form
> -def : WriteRes<WriteV, [VulcanF01]> { let Latency = 7; }
> -
> -// ASIMD arith, reduce, 4H/4S
> -// ASIMD arith, reduce, 8B/8H
> -// ASIMD arith, reduce, 16B
> -def : InstRW<[VulcanWrite_5Cyc_F01],
> - (instregex "^ADDVv", "^SADDLVv", "^UADDLVv")>;
> -
> -// ASIMD logical (MOV, MVN, ORN, ORR)
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^ORRv", "^ORNv", "^NOTv")>;
> -
> -// ASIMD polynomial (8x8) multiply long
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instrs PMULLv8i8, PMULLv16i8)>;
> -
> -//---
> -// 3.13 ASIMD Floating-point Instructions
> -//---
> -
> -// ASIMD FP absolute value
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^FABSv")>;
> -
> -// ASIMD FP arith, normal, D-form
> -// ASIMD FP arith, normal, Q-form
> -def : InstRW<[VulcanWrite_6Cyc_F01], (instregex "^FABDv", "^FADDv", "^FSUBv")>;
> -
> -// ASIMD FP arith,pairwise, D-form
> -// ASIMD FP arith, pairwise, Q-form
> -def : InstRW<[VulcanWrite_6Cyc_F01], (instregex "^FADDPv")>;
> -
> -// ASIMD FP compare, D-form
> -// ASIMD FP compare, Q-form
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^FACGEv", "^FACGTv")>;
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^FCMEQv", "^FCMGEv",
> - "^FCMGTv", "^FCMLEv",
> - "^FCMLTv")>;
> -
> -// ASIMD FP convert, long
> -// ASIMD FP convert, narrow
> -// ASIMD FP convert, other, D-form
> -// ASIMD FP convert, other, Q-form
> -// NOTE: Handled by WriteV.
> -
> -// ASIMD FP divide, D-form, F32
> -def : InstRW<[VulcanWrite_16Cyc_F01], (instrs FDIVv2f32)>;
> -
> -// ASIMD FP divide, Q-form, F32
> -def : InstRW<[VulcanWrite_16Cyc_F01], (instrs FDIVv4f32)>;
> -
> -// ASIMD FP divide, Q-form, F64
> -def : InstRW<[VulcanWrite_23Cyc_F01], (instrs FDIVv2f64)>;
> -
> -// ASIMD FP max/min, normal, D-form
> -// ASIMD FP max/min, normal, Q-form
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^FMAXv", "^FMAXNMv",
> - "^FMINv", "^FMINNMv")>;
> -
> -// ASIMD FP max/min, pairwise, D-form
> -// ASIMD FP max/min, pairwise, Q-form
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^FMAXPv", "^FMAXNMPv",
> - "^FMINPv", "^FMINNMPv")>;
> -
> -// ASIMD FP max/min, reduce
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^FMAXVv", "^FMAXNMVv",
> - "^FMINVv", "^FMINNMVv")>;
> -
> -// ASIMD FP multiply, D-form, FZ
> -// ASIMD FP multiply, D-form, no FZ
> -// ASIMD FP multiply, Q-form, FZ
> -// ASIMD FP multiply, Q-form, no FZ
> -def : InstRW<[VulcanWrite_6Cyc_F01], (instregex "^FMULv", "^FMULXv")>;
> -
> -// ASIMD FP multiply accumulate, Dform, FZ
> -// ASIMD FP multiply accumulate, Dform, no FZ
> -// ASIMD FP multiply accumulate, Qform, FZ
> -// ASIMD FP multiply accumulate, Qform, no FZ
> -def : InstRW<[VulcanWrite_6Cyc_F01], (instregex "^FMLAv", "^FMLSv")>;
> -
> -// ASIMD FP negate
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^FNEGv")>;
> -
> -// ASIMD FP round, D-form
> -// ASIMD FP round, Q-form
> -// NOTE: Handled by WriteV.
> -
> -//--
> -// 3.14 ASIMD Miscellaneous Instructions
> -//--
> -
> -// ASIMD bit reverse
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^RBITv")>;
> -
> -// ASIMD bitwise insert, D-form
> -// ASIMD bitwise insert, Q-form
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^BIFv", "^BITv", "^BSLv")>;
> -
> -// ASIMD count, D-form
> -// ASIMD count, Q-form
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^CLSv", "^CLZv", "^CNTv")>;
> -
> -// ASIMD duplicate, gen reg
> -// ASIMD duplicate, element
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^DUPv")>;
> -
> -// ASIMD extract
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^EXTv")>;
> -
> -// ASIMD extract narrow
> -// ASIMD extract narrow, saturating
> -// NOTE: Handled by WriteV.
> -
> -// ASIMD insert, element to element
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^INSv")>;
> -
> -// ASIMD move, integer immed
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^MOVIv", "^MOVIDv")>;
> -
> -// ASIMD move, FP immed
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^FMOVv")>;
> -
> -// ASIMD reciprocal estimate, D-form
> -// ASIMD reciprocal estimate, Q-form
> -def : InstRW<[VulcanWrite_5Cyc_F01],
> - (instregex "^FRECPEv", "^FRECPXv", "^URECPEv",
> - "^FRSQRTEv", "^URSQRTEv")>;
> -
> -// ASIMD reciprocal step, D-form, FZ
> -// ASIMD reciprocal step, D-form, no FZ
> -// ASIMD reciprocal step, Q-form, FZ
> -// ASIMD reciprocal step, Q-form, no FZ
> -def : InstRW<[VulcanWrite_6Cyc_F01], (instregex "^FRECPSv", "^FRSQRTSv")>;
> -
> -// ASIMD reverse
> -def : InstRW<[VulcanWrite_5Cyc_F01],
> - (instregex "^REV16v", "^REV32v", "^REV64v")>;
> -
> -// ASIMD table lookup, D-form
> -// ASIMD table lookup, Q-form
> -def : InstRW<[VulcanWrite_8Cyc_F01], (instregex "^TBLv", "^TBXv")>;
> -
> -// ASIMD transfer, element to word or word
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^UMOVv")>;
> -
> -// ASIMD transfer, element to gen reg
> -def : InstRW<[VulcanWrite_6Cyc_F01], (instregex "^SMOVv", "^UMOVv")>;
> -
> -// ASIMD transfer gen reg to element
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^INSv")>;
> -
> -// ASIMD transpose
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^TRN1v", "^TRN2v",
> - "^UZP1v", "^UZP2v")>;
> -
> -// ASIMD unzip/zip
> -def : InstRW<[VulcanWrite_5Cyc_F01], (instregex "^ZIP1v", "^ZIP2v")>;
> -
> -//--
> -// 3.15 ASIMD Load Instructions
> -//--
> -
> -// ASIMD load, 1 element, multiple, 1 reg, D-form
> -// ASIMD load, 1 element, multiple, 1 reg, Q-form
> -def : InstRW<[VulcanWrite_4Cyc_LS01],
> - (instregex "^LD1Onev(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_4Cyc_LS01, WriteAdr],
> - (instregex "^LD1Onev(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> -
> -// ASIMD load, 1 element, multiple, 2 reg, D-form
> -// ASIMD load, 1 element, multiple, 2 reg, Q-form
> -def : InstRW<[VulcanWrite_4Cyc_LS01],
> - (instregex "^LD1Twov(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_4Cyc_LS01, WriteAdr],
> - (instregex "^LD1Twov(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> -
> -// ASIMD load, 1 element, multiple, 3 reg, D-form
> -// ASIMD load, 1 element, multiple, 3 reg, Q-form
> -def : InstRW<[VulcanWrite_5Cyc_LS01],
> - (instregex "^LD1Threev(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_5Cyc_LS01, WriteAdr],
> - (instregex "^LD1Threev(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> -
> -// ASIMD load, 1 element, multiple, 4 reg, D-form
> -// ASIMD load, 1 element, multiple, 4 reg, Q-form
> -def : InstRW<[VulcanWrite_6Cyc_LS01],
> - (instregex "^LD1Fourv(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_6Cyc_LS01, WriteAdr],
> - (instregex "^LD1Fourv(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> -
> -// ASIMD load, 1 element, one lane, B/H/S
> -// ASIMD load, 1 element, one lane, D
> -def : InstRW<[VulcanWrite_5Cyc_LS01_F01], (instregex "^LD1i(8|16|32|64)$")>;
> -def : InstRW<[VulcanWrite_5Cyc_LS01_F01, WriteAdr],
> - (instregex "^LD1i(8|16|32|64)_POST$")>;
> -
> -// ASIMD load, 1 element, all lanes, D-form, B/H/S
> -// ASIMD load, 1 element, all lanes, D-form, D
> -// ASIMD load, 1 element, all lanes, Q-form
> -def : InstRW<[VulcanWrite_5Cyc_LS01_F01],
> - (instregex "^LD1Rv(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_5Cyc_LS01_F01, WriteAdr],
> - (instregex "^LD1Rv(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> -
> -// ASIMD load, 2 element, multiple, D-form, B/H/S
> -// ASIMD load, 2 element, multiple, Q-form, D
> -def : InstRW<[VulcanWrite_5Cyc_LS01_F01],
> - (instregex "^LD2Twov(8b|4h|2s|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_5Cyc_LS01_F01, WriteAdr],
> - (instregex "^LD2Twov(8b|4h|2s|16b|8h|4s|2d)_POST$")>;
> -
> -// ASIMD load, 2 element, one lane, B/H
> -// ASIMD load, 2 element, one lane, S
> -// ASIMD load, 2 element, one lane, D
> -def : InstRW<[VulcanWrite_5Cyc_LS01_F01], (instregex "^LD2i(8|16|32|64)$")>;
> -def : InstRW<[VulcanWrite_5Cyc_LS01_F01, WriteAdr],
> - (instregex "^LD2i(8|16|32|64)_POST$")>;
> -
> -// ASIMD load, 2 element, all lanes, D-form, B/H/S
> -// ASIMD load, 2 element, all lanes, D-form, D
> -// ASIMD load, 2 element, all lanes, Q-form
> -def : InstRW<[VulcanWrite_5Cyc_LS01_F01],
> - (instregex "^LD2Rv(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_5Cyc_LS01_F01, WriteAdr],
> - (instregex "^LD2Rv(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> -
> -// ASIMD load, 3 element, multiple, D-form, B/H/S
> -// ASIMD load, 3 element, multiple, Q-form, B/H/S
> -// ASIMD load, 3 element, multiple, Q-form, D
> -def : InstRW<[VulcanWrite_8Cyc_LS01_F01],
> - (instregex "^LD3Threev(8b|4h|2s|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_8Cyc_LS01_F01, WriteAdr],
> - (instregex "^LD3Threev(8b|4h|2s|16b|8h|4s|2d)_POST$")>;
> -
> -// ASIMD load, 3 element, one lone, B/H
> -// ASIMD load, 3 element, one lane, S
> -// ASIMD load, 3 element, one lane, D
> -def : InstRW<[VulcanWrite_7Cyc_LS01_F01], (instregex "^LD3i(8|16|32|64)$")>;
> -def : InstRW<[VulcanWrite_7Cyc_LS01_F01, WriteAdr],
> - (instregex "^LD3i(8|16|32|64)_POST$")>;
> -
> -// ASIMD load, 3 element, all lanes, D-form, B/H/S
> -// ASIMD load, 3 element, all lanes, D-form, D
> -// ASIMD load, 3 element, all lanes, Q-form, B/H/S
> -// ASIMD load, 3 element, all lanes, Q-form, D
> -def : InstRW<[VulcanWrite_7Cyc_LS01_F01],
> - (instregex "^LD3Rv(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_7Cyc_LS01_F01, WriteAdr],
> - (instregex "^LD3Rv(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> -
> -// ASIMD load, 4 element, multiple, D-form, B/H/S
> -// ASIMD load, 4 element, multiple, Q-form, B/H/S
> -// ASIMD load, 4 element, multiple, Q-form, D
> -def : InstRW<[VulcanWrite_8Cyc_LS01_F01],
> - (instregex "^LD4Fourv(8b|4h|2s|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_8Cyc_LS01_F01, WriteAdr],
> - (instregex "^LD4Fourv(8b|4h|2s|16b|8h|4s|2d)_POST$")>;
> -
> -// ASIMD load, 4 element, one lane, B/H
> -// ASIMD load, 4 element, one lane, S
> -// ASIMD load, 4 element, one lane, D
> -def : InstRW<[VulcanWrite_6Cyc_LS01_F01], (instregex "^LD4i(8|16|32|64)$")>;
> -def : InstRW<[VulcanWrite_6Cyc_LS01_F01, WriteAdr],
> - (instregex "^LD4i(8|16|32|64)_POST$")>;
> -
> -// ASIMD load, 4 element, all lanes, D-form, B/H/S
> -// ASIMD load, 4 element, all lanes, D-form, D
> -// ASIMD load, 4 element, all lanes, Q-form, B/H/S
> -// ASIMD load, 4 element, all lanes, Q-form, D
> -def : InstRW<[VulcanWrite_6Cyc_LS01_F01],
> - (instregex "^LD4Rv(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_6Cyc_LS01_F01, WriteAdr],
> - (instregex "^LD4Rv(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> -
> -//--
> -// 3.16 ASIMD Store Instructions
> -//--
> -
> -// ASIMD store, 1 element, multiple, 1 reg, D-form
> -// ASIMD store, 1 element, multiple, 1 reg, Q-form
> -def : InstRW<[VulcanWrite_1Cyc_LS01],
> - (instregex "^ST1Onev(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_1Cyc_LS01, WriteAdr],
> - (instregex "^ST1Onev(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> -
> -// ASIMD store, 1 element, multiple, 2 reg, D-form
> -// ASIMD store, 1 element, multiple, 2 reg, Q-form
> -def : InstRW<[VulcanWrite_1Cyc_LS01],
> - (instregex "^ST1Twov(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_1Cyc_LS01, WriteAdr],
> - (instregex "^ST1Twov(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> -
> -// ASIMD store, 1 element, multiple, 3 reg, D-form
> -// ASIMD store, 1 element, multiple, 3 reg, Q-form
> -def : InstRW<[VulcanWrite_1Cyc_LS01],
> - (instregex "^ST1Threev(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_1Cyc_LS01, WriteAdr],
> - (instregex "^ST1Threev(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> -
> -// ASIMD store, 1 element, multiple, 4 reg, D-form
> -// ASIMD store, 1 element, multiple, 4 reg, Q-form
> -def : InstRW<[VulcanWrite_1Cyc_LS01],
> - (instregex "^ST1Fourv(8b|4h|2s|1d|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_1Cyc_LS01, WriteAdr],
> - (instregex "^ST1Fourv(8b|4h|2s|1d|16b|8h|4s|2d)_POST$")>;
> -
> -// ASIMD store, 1 element, one lane, B/H/S
> -// ASIMD store, 1 element, one lane, D
> -def : InstRW<[VulcanWrite_1Cyc_LS01_F01],
> - (instregex "^ST1i(8|16|32|64)$")>;
> -def : InstRW<[VulcanWrite_1Cyc_LS01_F01, WriteAdr],
> - (instregex "^ST1i(8|16|32|64)_POST$")>;
> -
> -// ASIMD store, 2 element, multiple, D-form, B/H/S
> -// ASIMD store, 2 element, multiple, Q-form, B/H/S
> -// ASIMD store, 2 element, multiple, Q-form, D
> -def : InstRW<[VulcanWrite_1Cyc_LS01_F01],
> - (instregex "^ST2Twov(8b|4h|2s|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_1Cyc_LS01_F01, WriteAdr],
> - (instregex "^ST2Twov(8b|4h|2s|16b|8h|4s|2d)_POST$")>;
> -
> -// ASIMD store, 2 element, one lane, B/H/S
> -// ASIMD store, 2 element, one lane, D
> -def : InstRW<[VulcanWrite_1Cyc_LS01_F01],
> - (instregex "^ST2i(8|16|32|64)$")>;
> -def : InstRW<[VulcanWrite_1Cyc_LS01_F01, WriteAdr],
> - (instregex "^ST2i(8|16|32|64)_POST$")>;
> -
> -// ASIMD store, 3 element, multiple, D-form, B/H/S
> -// ASIMD store, 3 element, multiple, Q-form, B/H/S
> -// ASIMD store, 3 element, multiple, Q-form, D
> -def : InstRW<[VulcanWrite_1Cyc_LS01_F01],
> - (instregex "^ST3Threev(8b|4h|2s|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_1Cyc_LS01_F01, WriteAdr],
> - (instregex "^ST3Threev(8b|4h|2s|16b|8h|4s|2d)_POST$")>;
> -
> -// ASIMD store, 3 element, one lane, B/H
> -// ASIMD store, 3 element, one lane, S
> -// ASIMD store, 3 element, one lane, D
> -def : InstRW<[VulcanWrite_1Cyc_LS01_F01], (instregex "^ST3i(8|16|32|64)$")>;
> -def : InstRW<[VulcanWrite_1Cyc_LS01_F01, WriteAdr],
> - (instregex "^ST3i(8|16|32|64)_POST$")>;
> -
> -// ASIMD store, 4 element, multiple, D-form, B/H/S
> -// ASIMD store, 4 element, multiple, Q-form, B/H/S
> -// ASIMD store, 4 element, multiple, Q-form, D
> -def : InstRW<[VulcanWrite_1Cyc_LS01_F01],
> - (instregex "^ST4Fourv(8b|4h|2s|16b|8h|4s|2d)$")>;
> -def : InstRW<[VulcanWrite_1Cyc_LS01_F01, WriteAdr],
> - (instregex "^ST4Fourv(8b|4h|2s|16b|8h|4s|2d)_POST$")>;
> -
> -// ASIMD store, 4 element, one lane, B/H
> -// ASIMD store, 4 element, one lane, S
> -// ASIMD store, 4 element, one lane, D
> -def : InstRW<[VulcanWrite_1Cyc_LS01_F01], (instregex "^ST4i(8|16|32|64)$")>;
> -def : InstRW<[VulcanWrite_1Cyc_LS01_F01, WriteAdr],
> - (instregex "^ST4i(8|16|32|64)_POST$")>;
> -
> -//--
> -// 3.17 Cryptography Extensions
> -//--
> -
> -// Crypto AES ops
> -def : InstRW<[VulcanWrite_5Cyc_F1], (instregex "^AES")>;
> -
> -// Crypto polynomial (64x64) multiply long
> -def : InstRW<[VulcanWrite_5Cyc_F1], (instrs PMULLv1i64, PMULLv2i64)>;
> -
> -// Crypto SHA1 xor ops
> -// Crypto SHA1 schedule acceleration ops
> -// Crypto SHA256 schedule acceleration op (1 u-op)
> -// Crypto SHA256 schedule acceleration op (2 u-ops)
> -// Crypto SHA256 hash acceleration ops
> -def : InstRW<[VulcanWrite_7Cyc_F1], (instregex "^SHA")>;
> -
> -//--
> -// 3.18 CRC
> -//--
> -
> -// CRC checksum ops
> -def : InstRW<[VulcanWrite_4Cyc_I1], (instregex "^CRC32")>;
> -
> -} // SchedModel = VulcanModel
>
> Modified: llvm/trunk/lib/Target/AArch64/AArch64Subtarget.cpp
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AArch64/AArch64Subtarget.cpp?rev=297190&r1=297189&r2=297190&view=diff <http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AArch64/AArch64Subtarget.cpp?rev=297190&r1=297189&r2=297190&view=diff>
> ==============================================================================
> --- llvm/trunk/lib/Target/AArch64/AArch64Subtarget.cpp (original)
> +++ llvm/trunk/lib/Target/AArch64/AArch64Subtarget.cpp Tue Mar 7 13:42:40 2017
> @@ -81,16 +81,22 @@ void AArch64Subtarget::initializePropert
> MinPrefetchStride = 1024;
> MaxPrefetchIterationsAhead = 11;
> break;
> - case Vulcan:
> + case ThunderX2T99:
> + CacheLineSize = 64;
> + PrefFunctionAlignment = 3;
> + PrefLoopAlignment = 2;
> MaxInterleaveFactor = 4;
> + PrefetchDistance = 128;
> + MinPrefetchStride = 1024;
> + MaxPrefetchIterationsAhead = 4;
> break;
> case ThunderX:
> case ThunderXT88:
> case ThunderXT81:
> case ThunderXT83:
> CacheLineSize = 128;
> - PrefFunctionAlignment = 4;
> - PrefLoopAlignment = 4;
> + PrefFunctionAlignment = 3;
> + PrefLoopAlignment = 2;
> break;
> case CortexA35: break;
> case CortexA53: break;
>
> Modified: llvm/trunk/lib/Target/AArch64/AArch64Subtarget.h
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AArch64/AArch64Subtarget.h?rev=297190&r1=297189&r2=297190&view=diff <http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AArch64/AArch64Subtarget.h?rev=297190&r1=297189&r2=297190&view=diff>
> ==============================================================================
> --- llvm/trunk/lib/Target/AArch64/AArch64Subtarget.h (original)
> +++ llvm/trunk/lib/Target/AArch64/AArch64Subtarget.h Tue Mar 7 13:42:40 2017
> @@ -45,7 +45,7 @@ public:
> ExynosM1,
> Falkor,
> Kryo,
> - Vulcan,
> + ThunderX2T99,
> ThunderX,
> ThunderXT81,
> ThunderXT83,
>
> Modified: llvm/trunk/test/CodeGen/AArch64/cpus.ll
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/AArch64/cpus.ll?rev=297190&r1=297189&r2=297190&view=diff <http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/AArch64/cpus.ll?rev=297190&r1=297189&r2=297190&view=diff>
> ==============================================================================
> --- llvm/trunk/test/CodeGen/AArch64/cpus.ll (original)
> +++ llvm/trunk/test/CodeGen/AArch64/cpus.ll Tue Mar 7 13:42:40 2017
> @@ -12,7 +12,7 @@
> ; RUN: llc < %s -mtriple=arm64-unknown-unknown -mcpu=exynos-m3 2>&1 | FileCheck %s
> ; RUN: llc < %s -mtriple=arm64-unknown-unknown -mcpu=falkor 2>&1 | FileCheck %s
> ; RUN: llc < %s -mtriple=arm64-unknown-unknown -mcpu=kryo 2>&1 | FileCheck %s
> -; RUN: llc < %s -mtriple=arm64-unknown-unknown -mcpu=vulcan 2>&1 | FileCheck %s
> +; RUN: llc < %s -mtriple=arm64-unknown-unknown -mcpu=thunderx2t99 2>&1 | FileCheck %s
> ; RUN: llc < %s -mtriple=arm64-unknown-unknown -mcpu=invalidcpu 2>&1 | FileCheck %s --check-prefix=INVALID
>
> ; CHECK-NOT: {{.*}} is not a recognized processor for this target
>
> Modified: llvm/trunk/test/CodeGen/AArch64/machine-combiner-madd.ll
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/AArch64/machine-combiner-madd.ll?rev=297190&r1=297189&r2=297190&view=diff <http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/AArch64/machine-combiner-madd.ll?rev=297190&r1=297189&r2=297190&view=diff>
> ==============================================================================
> --- llvm/trunk/test/CodeGen/AArch64/machine-combiner-madd.ll (original)
> +++ llvm/trunk/test/CodeGen/AArch64/machine-combiner-madd.ll Tue Mar 7 13:42:40 2017
> @@ -6,7 +6,7 @@
> ; RUN: llc -mtriple=aarch64-linux-gnu -mcpu=exynos-m1 < %s | FileCheck %s
> ; RUN: llc -mtriple=aarch64-linux-gnu -mcpu=exynos-m2 < %s | FileCheck %s
> ; RUN: llc -mtriple=aarch64-linux-gnu -mcpu=kryo < %s | FileCheck %s
> -; RUN: llc -mtriple=aarch64-linux-gnu -mcpu=vulcan < %s | FileCheck %s
> +; RUN: llc -mtriple=aarch64-linux-gnu -mcpu=thunderx2t99 < %s | FileCheck %s
>
> ; Make sure that inst-combine fuses the multiply add in the addressing mode of
> ; the load.
>
> Modified: llvm/trunk/test/CodeGen/AArch64/remat.ll
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/AArch64/remat.ll?rev=297190&r1=297189&r2=297190&view=diff <http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/AArch64/remat.ll?rev=297190&r1=297189&r2=297190&view=diff>
> ==============================================================================
> --- llvm/trunk/test/CodeGen/AArch64/remat.ll (original)
> +++ llvm/trunk/test/CodeGen/AArch64/remat.ll Tue Mar 7 13:42:40 2017
> @@ -8,7 +8,7 @@
> ; RUN: llc -mtriple=aarch64-linux-gnuabi -mcpu=exynos-m3 -o - %s | FileCheck %s
> ; RUN: llc -mtriple=aarch64-linux-gnuabi -mcpu=falkor -o - %s | FileCheck %s
> ; RUN: llc -mtriple=aarch64-linux-gnuabi -mcpu=kryo -o - %s | FileCheck %s
> -; RUN: llc -mtriple=aarch64-linux-gnuabi -mcpu=vulcan -o - %s | FileCheck %s
> +; RUN: llc -mtriple=aarch64-linux-gnuabi -mcpu=thunderx2t99 -o - %s | FileCheck %s
> ; RUN: llc -mtriple=aarch64-linux-gnuabi -mattr=+custom-cheap-as-move -o - %s | FileCheck %s
>
> %X = type { i64, i64, i64 }
>
> Modified: llvm/trunk/unittests/Support/TargetParserTest.cpp
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/unittests/Support/TargetParserTest.cpp?rev=297190&r1=297189&r2=297190&view=diff <http://llvm.org/viewvc/llvm-project/llvm/trunk/unittests/Support/TargetParserTest.cpp?rev=297190&r1=297189&r2=297190&view=diff>
> ==============================================================================
> --- llvm/trunk/unittests/Support/TargetParserTest.cpp (original)
> +++ llvm/trunk/unittests/Support/TargetParserTest.cpp Tue Mar 7 13:42:40 2017
> @@ -643,8 +643,9 @@ TEST(TargetParserTest, testAArch64CPU) {
> "kryo", "armv8-a", "crypto-neon-fp-armv8",
> AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_SIMD, "8-A"));
> EXPECT_TRUE(testAArch64CPU(
> - "vulcan", "armv8.1-a", "crypto-neon-fp-armv8",
> - AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_SIMD, "8.1-A"));
> + "thunderx2t99", "armv8.1-a", "crypto-neon-fp-armv8",
> + AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_LSE |
> + AArch64::AEK_SIMD, "8.1-A"));
> EXPECT_TRUE(testAArch64CPU(
> "thunderx", "armv8-a", "crypto-neon-fp-armv8",
> AArch64::AEK_CRC | AArch64::AEK_CRYPTO | AArch64::AEK_SIMD |
> @@ -700,7 +701,7 @@ TEST(TargetParserTest, testAArch64Extens
> EXPECT_FALSE(testAArch64Extension("cyclone", 0, "ras"));
> EXPECT_FALSE(testAArch64Extension("exynos-m1", 0, "ras"));
> EXPECT_FALSE(testAArch64Extension("kryo", 0, "ras"));
> - EXPECT_FALSE(testAArch64Extension("vulcan", 0, "ras"));
> + EXPECT_FALSE(testAArch64Extension("thunderx2t99", 0, "ras"));
> EXPECT_FALSE(testAArch64Extension("thunderx", 0, "lse"));
> EXPECT_FALSE(testAArch64Extension("thunderxt81", 0, "lse"));
> EXPECT_FALSE(testAArch64Extension("thunderxt83", 0, "lse"));
>
>
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at lists.llvm.org <mailto:llvm-commits at lists.llvm.org>
> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits <http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170307/f1a383d8/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 1861 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170307/f1a383d8/attachment-0001.bin>
More information about the llvm-commits
mailing list