Core i9 9900K LLVM LVI Mitigation

LLVM LVI testing by Michael Larabel for a future article.

HTML result view exported from: https://openbenchmarking.org/result/2004036-NI-COREI999088&grr&rdt.

Core i9 9900K LLVM LVI MitigationProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriverCompilerFile-SystemScreen ResolutionLVI MitigatedDefaultIntel Core i9-9900K @ 5.00GHz (8 Cores / 16 Threads)ASRock Z390M Pro4 (P4.20 BIOS)Intel Cannon Lake PCH16GB240GB Force MP510Intel UHD 630 (1200MHz)Realtek ALC892G237HLIntel I219-VUbuntu 20.045.4.0-9-generic (x86_64)GNOME Shell 3.34.1X Server 1.20.5modesetting 1.20.5Clang 11.0.0ext41920x1080OpenBenchmarking.orgEnvironment Details- LVI Mitigated: CXXFLAGS="-O3 -mlvi-cfi" CFLAGS="-O3 -mlvi-cfi"- Default: CXXFLAGS=-O3 CFLAGS=-O3Compiler Details- Optimized build; Default target: x86_64-unknown-linux-gnu; Host CPU: skylake Processor Details- Scaling Governor: intel_pstate powersave - CPU Microcode: 0xcaSecurity Details- itlb_multihit: KVM: Vulnerable + l1tf: Not affected + mds: Vulnerable; SMT vulnerable + meltdown: Not affected + spec_store_bypass: Vulnerable + spectre_v1: Vulnerable: __user pointer sanitization and usercopy barriers only; no swapgs barriers + spectre_v2: Vulnerable IBPB: disabled STIBP: disabled + tsx_async_abort: Vulnerable

Core i9 9900K LLVM LVI Mitigationjohn-the-ripper: MD5himeno: Poisson Pressure Solversqlite-speedtest: Timed Time - Size 1,000nginx: Static Web Page Servingdav1d: Summer Nature 4Kvpxenc: Speed 5leveldb: Seq Fillleveldb: Seq Fillleveldb: Rand Deleteencode-flac: WAV To FLACx264: H.264 Video Encodingsmhasher: fasthash32smhasher: fasthash32smhasher: wyhashleveldb: Seek Randencode-mp3: WAV To MP3leveldb: Rand Readleveldb: Hot Readlammps: Rhodopsin Proteinleveldb: Rand Fillleveldb: Rand Fillleveldb: Overwriteleveldb: OverwriteLVI MitigatedDefault1091194171.76051860.55451338.94146.6128.6436.70048.237.3519.36872.7944.4677468.3834.6779.3389.2507.6807.6324.49140.64843.540.41843.81105594183.81481849.33651953.36150.3633.0634.84750.833.7499.33390.4932.3037461.9923.8259.1388.5557.5027.4574.56735.99149.136.16848.9OpenBenchmarking.org

John The Ripper

Test: MD5

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 1.9.0-jumbo-1Test: MD5LVI MitigatedDefault20K40K60K80K100KSE +/- 10.44, N = 3SE +/- 38.68, N = 3109119110559-mlvi-cfi1. (CC) gcc options: -m64 -lssl -lcrypto -lgmp -pthread -lm -lz -ldl -lcrypt -lbz2 -mavx2 -O3 -march=native -std=gnu89 -Qunused-arguments -funroll-loops

Himeno Benchmark

Poisson Pressure Solver

OpenBenchmarking.orgMFLOPS, More Is BetterHimeno Benchmark 3.0Poisson Pressure SolverLVI MitigatedDefault9001800270036004500SE +/- 5.15, N = 3SE +/- 1.77, N = 34171.764183.81-mlvi-cfi1. (CC) gcc options: -O3 -mavx2

SQLite Speedtest

Timed Time - Size 1,000

OpenBenchmarking.orgSeconds, Fewer Is BetterSQLite Speedtest 3.30Timed Time - Size 1,000LVI MitigatedDefault1428425670SE +/- 0.09, N = 3SE +/- 0.22, N = 360.5549.34-mlvi-cfi1. (CC) gcc options: -O3 -ldl -lz -lpthread

NGINX Benchmark

Static Web Page Serving

OpenBenchmarking.orgRequests Per Second, More Is BetterNGINX Benchmark 1.9.9Static Web Page ServingLVI MitigatedDefault11K22K33K44K55KSE +/- 190.13, N = 3SE +/- 163.47, N = 351338.9451953.36-mlvi-cfi1. (CC) gcc options: -lpthread -lcrypt -lcrypto -lz -O3 -march=native

dav1d

Video Input: Summer Nature 4K

OpenBenchmarking.orgFPS, More Is Betterdav1d 0.6.0Video Input: Summer Nature 4KLVI MitigatedDefault306090120150SE +/- 0.17, N = 3SE +/- 0.85, N = 3146.61150.36-mlvi-cfi - MIN: 135.28 / MAX: 161.58MIN: 139.18 / MAX: 164.011. (CC) gcc options: -O3 -pthread

VP9 libvpx Encoding

Speed: Speed 5

OpenBenchmarking.orgFrames Per Second, More Is BetterVP9 libvpx Encoding 1.8.2Speed: Speed 5LVI MitigatedDefault816243240SE +/- 0.17, N = 3SE +/- 0.08, N = 328.6433.06-mlvi-cfi1. (CXX) g++ options: -m64 -lm -lpthread -O3 -fPIC -U_FORTIFY_SOURCE -std=c++11

LevelDB

Benchmark: Sequential Fill

OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Sequential FillLVI MitigatedDefault816243240SE +/- 0.08, N = 3SE +/- 0.28, N = 336.7034.85-mlvi-cfi1. (CXX) g++ options: -O3 -lsnappy -lpthread

LevelDB

Benchmark: Sequential Fill

OpenBenchmarking.orgMB/s, More Is BetterLevelDB 1.22Benchmark: Sequential FillLVI MitigatedDefault1122334455SE +/- 0.10, N = 3SE +/- 0.39, N = 348.250.8-mlvi-cfi1. (CXX) g++ options: -O3 -lsnappy -lpthread

LevelDB

Benchmark: Random Delete

OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Random DeleteLVI MitigatedDefault918273645SE +/- 0.10, N = 3SE +/- 0.06, N = 337.3533.75-mlvi-cfi1. (CXX) g++ options: -O3 -lsnappy -lpthread

FLAC Audio Encoding

WAV To FLAC

OpenBenchmarking.orgSeconds, Fewer Is BetterFLAC Audio Encoding 1.3.2WAV To FLACLVI MitigatedDefault3691215SE +/- 0.011, N = 5SE +/- 0.007, N = 59.3689.333-mlvi-cfi1. (CXX) g++ options: -O3 -logg -lm

x264

H.264 Video Encoding

OpenBenchmarking.orgFrames Per Second, More Is Betterx264 2019-12-17H.264 Video EncodingLVI MitigatedDefault20406080100SE +/- 0.65, N = 3SE +/- 0.83, N = 1072.7990.49-mlvi-cfi1. (CC) gcc options: -ldl -m64 -lm -lpthread -O3 -ffast-math -std=gnu99 -mstack-alignment=64 -fPIC -fomit-frame-pointer -fno-tree-vectorize

SMHasher

Hash: fasthash32

OpenBenchmarking.orgcycles/hash, Fewer Is BetterSMHasher 2020-02-29Hash: fasthash32LVI MitigatedDefault1020304050SE +/- 0.01, N = 3SE +/- 0.00, N = 344.4732.30-mlvi-cfi1. (CXX) g++ options: -O3 -march=native -lpthread

SMHasher

Hash: fasthash32

OpenBenchmarking.orgMiB/sec, More Is BetterSMHasher 2020-02-29Hash: fasthash32LVI MitigatedDefault16003200480064008000SE +/- 0.53, N = 3SE +/- 2.84, N = 37468.387461.99-mlvi-cfi1. (CXX) g++ options: -O3 -march=native -lpthread

SMHasher

Hash: wyhash

OpenBenchmarking.orgcycles/hash, Fewer Is BetterSMHasher 2020-02-29Hash: wyhashLVI MitigatedDefault816243240SE +/- 0.00, N = 3SE +/- 0.02, N = 334.6823.83-mlvi-cfi1. (CXX) g++ options: -O3 -march=native -lpthread

LevelDB

Benchmark: Seek Random

OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Seek RandomLVI MitigatedDefault3691215SE +/- 0.018, N = 3SE +/- 0.080, N = 39.3389.138-mlvi-cfi1. (CXX) g++ options: -O3 -lsnappy -lpthread

LAME MP3 Encoding

WAV To MP3

OpenBenchmarking.orgSeconds, Fewer Is BetterLAME MP3 Encoding 3.100WAV To MP3LVI MitigatedDefault3691215SE +/- 0.022, N = 3SE +/- 0.004, N = 39.2508.555-mlvi-cfi1. (CC) gcc options: -O3 -pipe -lncurses -lm

LevelDB

Benchmark: Random Read

OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Random ReadLVI MitigatedDefault246810SE +/- 0.060, N = 3SE +/- 0.040, N = 37.6807.502-mlvi-cfi1. (CXX) g++ options: -O3 -lsnappy -lpthread

LevelDB

Benchmark: Hot Read

OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Hot ReadLVI MitigatedDefault246810SE +/- 0.121, N = 3SE +/- 0.099, N = 37.6327.457-mlvi-cfi1. (CXX) g++ options: -O3 -lsnappy -lpthread

LAMMPS Molecular Dynamics Simulator

Model: Rhodopsin Protein

OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 9Jan2020Model: Rhodopsin ProteinLVI MitigatedDefault1.02762.05523.08284.11045.138SE +/- 0.008, N = 3SE +/- 0.040, N = 34.4914.567-mlvi-cfi1. (CXX) g++ options: -O3 -rdynamic -ljpeg -lpng -lz -lfftw3 -lm

LevelDB

Benchmark: Random Fill

OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Random FillLVI MitigatedDefault918273645SE +/- 0.18, N = 3SE +/- 0.08, N = 340.6535.99-mlvi-cfi1. (CXX) g++ options: -O3 -lsnappy -lpthread

LevelDB

Benchmark: Random Fill

OpenBenchmarking.orgMB/s, More Is BetterLevelDB 1.22Benchmark: Random FillLVI MitigatedDefault1122334455SE +/- 0.22, N = 3SE +/- 0.12, N = 343.549.1-mlvi-cfi1. (CXX) g++ options: -O3 -lsnappy -lpthread

LevelDB

Benchmark: Overwrite

OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: OverwriteLVI MitigatedDefault918273645SE +/- 0.14, N = 3SE +/- 0.19, N = 340.4236.17-mlvi-cfi1. (CXX) g++ options: -O3 -lsnappy -lpthread

LevelDB

Benchmark: Overwrite

OpenBenchmarking.orgMB/s, More Is BetterLevelDB 1.22Benchmark: OverwriteLVI MitigatedDefault1122334455SE +/- 0.15, N = 3SE +/- 0.26, N = 343.848.9-mlvi-cfi1. (CXX) g++ options: -O3 -lsnappy -lpthread

Phoronix Test Suite v10.8.4