search for: ecbd

...olly Dependence analysis. "Polly Dependence Pass" for flop is still high for some benchmarks as we have discussed before. >> SingleSource/Benchmarks/Misc/flops 28.57% >> MultiSource/Benchmarks/MiBench/security-sha/security-sha 22.22% >> MultiSource/Benchmarks/VersaBench/ecbdes/ecbdes 21.05% >> When I look into the compile-time for the flop benchmark using "-ftime-report", I find the extra compile-time overhead mainly comes from the "Combine redundant instructions" pass. >> the top 5 passes when compiled with Polly canonicalization passes...

...simply removing the first "createInstructionCombiningPass". The top 5 benchmarks are listed as follows: SingleSource/Regression/C++/2003-09-29-NonPODsByValue-38.46% SingleSource/Benchmarks/Misc/flops-19.30% SingleSource/Benchmarks/Misc/himenobmtxpa-12.94% MultiSource/Benchmarks/VersaBench/ecbdes/ecbdes-12.68% MultiSource/Benchmarks/ASCI_Purple/SMG2000/smg2000-10.68% Unfortunately, there are also two serious execution-time performance regressions: SingleSource/Benchmarks/Adobe-C++/simple_types_constant_folding204.19% SingleSource/Benchmarks/Polybench/linear-algebra/solvers/dynprog/dynprog...

...egen-scev", only three benchmark shows >20% extra compile-time overhead: I believe so to, but please verify with -debug-pass=Structure > SingleSource/Benchmarks/Misc/flops 28.57% > MultiSource/Benchmarks/MiBench/security-sha/security-sha 22.22% > MultiSource/Benchmarks/VersaBench/ecbdes/ecbdes 21.05% > When I look into the compile-time for the flop benchmark using "-ftime-report", I find the extra compile-time overhead mainly comes from the "Combine redundant instructions" pass. > the top 5 passes when compiled with Polly canonicalization passes: >...

...=none -mllvm -polly-codegen-scev Fortunately, with the option "-polly-codegen-scev", only three benchmark shows >20% extra compile-time overhead: SingleSource/Benchmarks/Misc/flops 28.57% MultiSource/Benchmarks/MiBench/security-sha/security-sha 22.22% MultiSource/Benchmarks/VersaBench/ecbdes/ecbdes 21.05% When I look into the compile-time for the flop benchmark using "-ftime-report", I find the extra compile-time overhead mainly comes from the "Combine redundant instructions" pass. the top 5 passes when compiled with Polly canonicalization passes: ---User Time-...

...gt; > enc-3des: 66.31% > - > > StatementReordering-dbl: 51.45% > - > > Symbolics-dbl: 51.42% > - > > Recurrences-dbl: 51.38% > - > > Packing-dbl: 51.33% > > LO: > > - > > enc-3des: 50.7% > - > > ecbdes: 46.27% > - > > security-rjindael:45.13% > - > > ControlFlow-flt: 25.79% > - > > ControlFlow-dbl: 25.74% > > MO: > > - > > ecbdes: 28.22% > - > > Expansion-flt: 22.56% > - > > Recurrences-flt: 22.19...

...simply removing the first "createInstructionCombiningPass". The top 5 benchmarks are listed as follows: SingleSource/Regression/C++/2003-09-29-NonPODsByValue-38.46% SingleSource/Benchmarks/Misc/flops-19.30% SingleSource/Benchmarks/Misc/himenobmtxpa-12.94% MultiSource/Benchmarks/VersaBench/ecbdes/ecbdes-12.68% MultiSource/Benchmarks/ASCI_Purple/SMG2000/smg2000-10.68% Unfortunately, there are also two serious execution-time performance regressions: SingleSource/Benchmarks/Adobe-C++/simple_types_constant_folding204.19% SingleSource/Benchmarks/Polyb! ench/linear-algebra/solvers/dynprog/dynpr...

On 09/08/2013 08:03 PM, Star Tan wrote: > Hello all, > > > I have done some basic experiments about Polly canonicalization passes and I found the SCEV canonicalization has significant impact on both compile-time and execution-time performance. Interesting. > Detailed results for SCEV and default canonicalization can be viewed on: http://188.40.87.11:8000/db_default/v4/nts/32 (or

Hello all, I have done some basic experiments about Polly canonicalization passes and I found the SCEV canonicalization has significant impact on both compile-time and execution-time performance. Detailed results for SCEV and default canonicalization can be viewed on: http://188.40.87.11:8000/db_default/v4/nts/32 (or 33, 34) *pNoGen with SCEV canonicalization (run 32): -O3 -Xclang -load

Hi all, I've started looking at the GlobalMerge pass, enabled by default on ARM and AArch64. I think we should reconsider that, at least for AArch64. As is, the pass just merges all globals together, in groups of 4KB (AArch64, 128B on ARM). At the time it was enabled, the general thinking was "it's almost free, it doesn't affect performance much, we might as well use it".

.... #ls -al #cd / #zpool status zpool scrub tank zpool status Console output is: SUNW-MSG-ID: ZFS-8000-D3, TYPE: Fault, VER: 1, SEVERITY: Major EVENT-TIME: Thu Nov 2 16:20:36 EST 2006 PLATFORM: SUNW,Sun-Fire-V210, CSN: -, HOSTNAME: jsclient1 SOURCE: zfs-diagnosis, REV: 1.0 EVENT-ID: 7bac21b6-76e7-ecbd-a63a-982be2230f9d DESC: A ZFS device failed. Refer to http://sun.com/msg/ZFS-8000-D3 for more information. AUTO-RESPONSE: No automated response will occur. IMPACT: Fault tolerance of the pool may be compromised. REC-ACTION: Run ''zpool status -x'' and replace the bad device. panic...

...e/Applications/SPASS/SPASS | * | * | MultiSource/Applications/oggenc/oggenc | * | * | MultiSource/Benchmarks/VersaBench/bmm/bmm | * | * | MultiSource/Benchmarks/VersaBench/ecbdes/ecbdes | * | * | MultiSource/Benchmarks/VersaBench/beamformer/beamformer | * | * | MultiSource/Benchmarks/VersaBench/dbms/dbms | * | * | MultiSource/Benchmarks/VersaBench/8b10b/8b10b...

I’m River and I’m a compiler engineer at PlayStation. Recently, I’ve been working on an interprocedural outlining (code folding) pass for code size improvement at the IR level. We hit a couple of use cases that the current code size solutions didn’t handle well enough. Outlining is one of the avenues that seemed potentially beneficial. -- Algorithmic Approach -- The general implementation can be

...-0.3612158257126 Benchmarks/FreeBench/neural/neural 0.1771 0.1765 -0.3387916431394 Benchmarks/Ptrdist/ft/ft 0.6569 0.6549 -0.3044603440401 Benchmarks/Trimaran/enc-3des/enc-3des 1.3386 1.3354 -0.2390557298670 Benchmarks/VersaBench/ecbdes/ecbdes 1.5638 1.5623 -0.0959201943982 Benchmarks/TSVC/Recurrences-dbl/Recurrences 2.8128 2.8102 -0.0924345847554 Benchmarks/Trimaran/netbench-crc/netbench-c 0.5665 0.566 -0.0882612533098 Benchmarks/Prolangs-C++/life/life 1.826 1.8244...

...ench/gs/gs -10.2% SingleSource/Benchmarks/Misc/ReedSolomon -9.8% MultiSource/Applications/siod/siod -9.7% MultiSource/Benchmarks/Prolangs-C/gnugo/gnugo -9.1% MultiSource/Benchmarks/llubenchmark/llu -9.1% MultiSource/Benchmarks/FreeBench/analyzer/analyzer -8.9% MultiSource/Benchmarks/VersaBench/ecbdes/ecbdes -8.8% SingleSource/Benchmarks/Shootout/sieve -8.6% MultiSource/Benchmarks/MiBench/security-rijndael/security-rijndael -7.9% SingleSource/Benchmarks/Misc/mandel-2 -6.2% MultiSource/Benchmarks/MiBench/network-dijkstra/network-dijkstra -6.2% External/SPEC/CFP2006/447.dealII/447.dealII -...

Andy and I briefly discussed this the other day, we have not yet got chance to list a detailed pass order for the pre- and post- IPO scalar optimizations. This is wish-list in our mind: pre-IPO: based on the ordering he propose, get rid of the inlining (or just inline tiny func), get rid of all loop xforms... post-IPO: get rid of inlining, or maybe we still need it, only

...1.305024772| -0.98| 9932045354| 0|1.292741542| -0.04| 9932045350| 0| |MultiSource/Benchmarks/VersaBench/dbms/dbms.test | 45|0.810661753| 3391009910|0.802206556| 1.05| 3324663069| 2|0.800454835| 1.28| 3324663069| 2| |MultiSource/Benchmarks/VersaBench/ecbdes/ecbdes.test | 40|1.457606778|11993065051|1.457628021| 0|11993065048| 0|1.458776986| -0.08|11993065048| 0| |MultiSource/Benchmarks/llubenchmark/llu.test | 40|5.202703599| 2493038287| 5.21286191| -0.19| 2493038293| 0|5....

...39 | 1.00    0.99    n/a          n/a > MultiSource/Benchmarks/VersaBench/dbms/dbms                              | > 0.1200 36468    0.2500      *                0.2400      |    2.34    2.23 >  2.17 *           2.35 | 1.05    1.08    n/a          n/a > MultiSource/Benchmarks/VersaBench/ecbdes/ecbdes                          | > 0.0599 9740     0.0200      *                0.0200      |    2.87    2.91 >  2.77 *           2.76 | 0.99    1.04    n/a          n/a > MultiSource/Benchmarks/llubenchmark/llu                                  | > 0.0000 3896     0.0300      *      ...

Hi Tanya, > 1) Compile llvm from source and untar the llvm-test in the projects > directory (name it llvm-test or test-suite). Choose to use a > pre-compiled llvm-gcc or re-compile it yourself. I compiled llvm and llvm-gcc with separate objects directories. Platform is x86_64-linux-gnu. > 2) Run make check, report any failures (FAIL or unexpected pass). Note > that you need to

...0.99 n/a n/a > MultiSource/Benchmarks/VersaBench/dbms/ > dbms | 0.1200 36468 0.2500 > * 0.2400 | 2.34 2.23 2.17 * > 2.35 | 1.05 1.08 n/a n/a > MultiSource/Benchmarks/VersaBench/ecbdes/ > ecbdes | 0.0599 9740 0.0200 > * 0.0200 | 2.87 2.91 2.77 * > 2.76 | 0.99 1.04 n/a n/a > MultiSource/Benchmarks/llubenchmark/ > llu | 0.0000 3896 0.03...

...ultiSource/Benchmarks/MiBench/consumer-jpeg/consumer-jpeg.compile > <http://llvm.org/perf/db_default/simple/nts/380/graph?pset.0=on&test.MultiSource%2FBenchmarks%2FMiBench%2Fconsumer-jpeg%2Fconsumer-jpeg.compile=on> > 1.37% 4.4549 4.5158 0.0077 > MultiSource/Benchmarks/VersaBench/ecbdes/ecbdes.compile > <http://llvm.org/perf/db_default/simple/nts/380/graph?pset.0=on&test.MultiSource%2FBenchmarks%2FVersaBench%2Fecbdes%2Fecbdes.compile=on> > 1.34% 0.2461 0.2494 0.0003 > MultiSource/Applications/hbd/hbd.compile > <http://llvm.org/perf/db_default/simple/nts/...