CK Browser

Distinct solutions after online classification (auto/crowd-tune GCC compiler flags (minimize execution time))

Scenario UID	8289e0cf24346aa7 (experiment.tune.compiler.flags.gcc.e)
Data UID	788a2c125044df4c

Discuss (optimizations to improve compilers, semantic/data set/hardware features to improve predictions, etc):	GitHub wiki, Google group

Download:	[ All solutions in JSON ], [ Solutions' classification in JSON ]
Reproduce all (with reactions):	ck replay 8289e0cf24346aa7:788a2c125044df4c

Compiler	GCC 4.9
CPU	sun7i
Objective	min

Improvement key IK1	Main kernel execution time speedup [min]
Improvement key IK2	Code size improvement

Improvements (<4% variation)

Distinct workload for highest improvement

Solution UID

IK1

IK2

New distinct optimization choices

Ref

Best species

Worst species

Touched

Iters

Program

CMD

Dataset

Dataset file

CPU freq (MHz)

Cores

Platform

Replay

ada16d565e55e3f4

1.17

0.98

-O3 -ffunction-sections -fno-gcse-lm -fno-indirect-inlining -fno-inline-functions -fivopts -fno-loop-interchange -fno-modulo-sched-allow-regmoves -fmove-loop-invariants -fpeephole -ftoplevel-reorder -fno-prefetch-loop-arrays -fno-rename-registers -fsched-spec-insn-heuristic -fno-sched-dep-count-heuristic -fno-selective-scheduling -fno-tracer -fno-tree-coalesce-vars -fno-tree-forwprop -ftree-loop-optimize -fno-tree-loop-vectorize -fno-tree-partial-pre -fno-tree-switch-conversion -fno-tree-tail-merge --param min-inline-recursive-probability=0 --param max-modulo-backtrack-attempts=53 --param large-function-growth=198 --param max-unswitch-insns=7 --param max-iterations-computation-cost=4 --param hot-bb-count-ws-permille=282 --param max-predicted-iterations=121 --param builtin-expect-probability=46 --param max-grow-copy-bb-insns=4 --param vect-max-peeling-for-alignment=35 --param sched-spec-prob-cutoff=81 --param sched-mem-true-dep-cost=0 --param max-last-value-rtl=15268 --param simultaneous-prefetches=3 --param use-canonical-types=0 --param ipa-cp-array-index-hint-bonus=56

-O3

milepost-codelet-mibench-automotive-susan-e-src-susan-codelet-2-1

default

912, 912

UNKNOWN MID0927

Android 4.2.2

992744c19ad5e9bb

1.14

0.98

-O3 -fno-branch-probabilities -fno-conserve-stack -fno-cx-fortran-rules -fno-data-sections -fgcse-las -fno-indirect-inlining -finline-functions-called-once -fno-keep-inline-functions -fno-peel-loops -frounding-math -fsched2-use-superblocks -fno-schedule-insns -fno-tree-ch -ftree-coalesce-vars -ftree-loop-if-convert -ftree-sink -fno-tree-slsr -funswitch-loops -fno-variable-expansion-in-unroller -fwhole-program -falign-jumps=0 --param max-inline-recursive-depth-auto=11 --param early-inlining-insns=17 --param max-unrolled-insns=187 --param align-threshold=69 --param omega-max-eqs=238 --param selsched-max-lookahead=53 --param selsched-insns-to-rename=0 --param ssp-buffer-size=4 --param max-sched-ready-insns=89 --param prefetch-latency=15 --param lra-max-considered-reload-pseudos=610 --param graphite-max-nb-scop-params=11 --param max-vartrack-reverse-op-size=50 --param allow-load-data-races=1

-O3

cbench-network-dijkstra

default

cdataset-dijkstra-0018

data.txt

912, 912

UNKNOWN MID0927

Android 4.2.2

[ Participated users, platforms, OS, CPU, GPU, GPGPU, NN, NPU ] [ How to participate ] [ Motivation (PPT) (PDF) ] [ Papers 1 , 2 , 3] [ Android app ] [ Collective training set ] [ Unified AI ]

View entry in raw format