View Issue Details

Related ChangesetsJump to NotesJump to History
IDProjectCategoryView StatusDate SubmittedLast Update
0024285Open CASCADEOCCT:Foundation Classespublic2013-10-24 09:472015-10-16 17:01
ReporterazvAssigned Tobugmaster  
PriorityhighSeverityminor 
Status closedResolutionfixed 
Product Version6.6.0 
Target Version6.9.0Fixed in Version6.9.0 
Summary0024285: Updates of PLib::EvalPolynomial for code acceleration
DescriptionSome changes on PLib::EvalPolynomial related to loop nesting may perform code execution faster.
Steps To Reproduceperf bspline intersect
boolean bcut_complex Q9
boolean bsection P4
TagsNo tags attached.
Test case numberNot needed

Attached Files

  • eval_poly_template_metaprogram.zip (4,544 bytes)

Relationships

related to 0026110 closedbugmaster Open CASCADE Restoring perf bspline test subgroup 

Activities

azv

2013-10-24 09:51

administrator   ~0026248

Branch CR24285 added

abv

2013-10-28 18:07

manager   ~0026308

No remarks, please test

mkv

2013-10-30 08:15

tester   ~0026345

Dear BugMaster,

Branch CR24285 (and products from GIT master) was compiled on Linux and Windows platforms and tested.
SHA-1: 1f4674920d52faedb25eb1dce34facd985de5e2e

Number of compiler warnings:

occt component :
Linux: 319 (323 on master)
Windows: 0 (0 on master)

products component :
Linux: 189 (189 on master)
Windows: 287 (287 on master)

Regressions/Differences:
No regressions/differences

Testing cases:
Not needed

Testing on Linux:
Total MEMORY difference: 353301752 / 356295124
Total CPU difference: 42535.01000000177 / 40672.21000000072

Testing on Windows:
Total MEMORY difference: 407352400 / 406499704
Total CPU difference: 27747.078125 / 34773.5

There are not differences in images found by testdiff.

abv

2013-12-19 11:50

manager   ~0027258

Waiting for OLA

git

2015-01-20 17:12

administrator   ~0036272

Branch CR24285_3 has been created by azv.

SHA-1: e1eaf5e85f3443e7e9338d7b4f48fe45cf731a2b


Detailed log of new commits:

Author: azv
Date: Tue Jan 20 17:06:03 2015 +0300

    0024285: Updates of PLib::EvalPolynomial for code acceleration
    
    1. Loop nesting in functions PLib::EvalPolynomial and PLib::NoDerivativeEvalPolynomial was changed
    2. Avoided pointer arithmetic
    3. There is an afford for automatic generation of SSE instructions by modern compilers

azv

2015-01-20 17:37

administrator   ~0036274

Last edited: 2015-01-20 17:41

 Results of performance test (perf/bspline/intersect) for different parameters of compilation, please, find below.

=====
Machine configuration:

Hardware: Intel Core i5-3470 3.2GHz, 8GB RAM
OS: Windows 7 Professional x64 SP1
Compiler: Microsoft Visual C++ 17.00.50727.1 x64 (Visual Studio 2012)

=====
Abbreviations:

Build No1 (BM1): master (version Dec, 26, 2014), disabled optimization (/Od) for project TKMath
Build No2 (BM2): master (version Dec, 26, 2014), maximize speed (/O2) for TKMath
Build No3 (BM3): master (version Dec, 26, 2014), maximize speed (/O2) and AVX enabled (/arch:AVX) for TKMath

Build No4 (BO1): branch CR24285_3, disabled optimization (/Od) for project TKMath
Build No5 (BO2): branch CR24285_3, maximize speed (/O2) for TKMath
Build No6 (BO3): branch CR24285_3, maximize speed (/O2) and AVX enabled (/arch:AVX) for TKMath

=====
Results (sec.): there were made 5 runs of perf/bspline/intersect test case and calculated average time

    | BM1 | BM2 | BM3
 1 | 91.7753883 | 38.9690498 | 36.6290348
 2 | 91.0421836 | 37.4558401 | 36.6446349
 3 | 91.0421836 | 37.3622395 | 36.5354342
 4 | 91.0889839 | 37.7210418 | 36.7226354
 5 | 91.1825845 | 37.6898416 | 36.8006359
Avg | 91.22626478 | 37.83960256 | 36.66647504

    | BO1 | BO2 | BO3
 1 | 78.936506 | 36.8474362 | 34.3982205
 2 | 79.1393073 | 36.8318361 | 33.2438131
 3 | 79.7789114 | 36.7538356 | 33.1190123
 4 | 79.092507 | 36.9566369 | 33.4622145
 5 | 79.0613068 | 36.5822345 | 33.2126129
Avg | 79.2017077 | 36.79439586 | 33.48717466

=====
Relative average speedup:

BM2/BM1 58.52%
BM3/BM2 3.1 %
BO2/BO1 53.54%
BO3/BO2 8.99%

BO1/BM1 13.18%
BO2/BM2 2.76%
BO3/BM3 8.67%

azv

2015-01-20 17:43

administrator   ~0036276

Dear Andrey, Mikhail,

Please review branch CR24285_3.

abv

2015-01-21 07:14

manager   ~0036283

Have you been able to observe some acceleration on this branch? As for me, first test shown degradation of performance (59 min vs. 56 on master). Will try more...

azv

2015-01-21 07:33

administrator   ~0036284

Last edited: 2015-01-21 07:38

 I suppose, you compiled x86 version, didn't you?

I had observed that the performance of x86 compilation (with SSE2 enabled) is too much floating. During testing you may get different results: on my machine I have average speedup 0.42% only (on perf/bspline/intersect test case). So, I propose to use x64 compilation for performance checking because it gives more stable results.

Results of full non-regression testing both x86 and x64 builds did not change too much on my machine.

abv

2015-01-21 08:17

manager   ~0036285

No, I have built x64 version naturally, vc10 compiler. Sorry I have overlooked your report on performance above... I will check results on my system more carefully and report here this evening.

abv

2015-01-23 06:51

manager   ~0036424

Sorry I was wrong saying about downgrade: that looks like a time variation on one test run, and consequent runs showed almost the same time as master. That is, I see no variation of time in general. On test perf bspline intersect, the times are:
master: 28.8 sec
CR24285_2: 25.8 sec
CR24285_3: 26.5 sec

msv

2015-01-23 09:23

developer   ~0036428

I would like to know what part of time of that test is occupied by run of PLib::EvalPolynomial.

azv

2015-01-27 07:11

administrator   ~0036682

According to Intel Amplifier the running PLib::EvalPolynomial takes 25% of full time of testing perf/bspline/intersect on the current master.

msv

2015-01-27 10:47

developer   ~0036688

Well, 25% of 28.8 = 7.2 sec.
Gain is 28.8-26.5 = 2.3 sec.
So, the method was speed-ed up on 2.3/7.2=32%
According to the statements of Roman and Istvan, they speed-ed up it on 1000% and more. So, the target of this improvement was not hit.

abv

2015-01-27 11:28

manager   ~0036692

Last edited: 2015-01-27 11:30

 Mikhail, your calculations are very strange for me.

If you calculate acceleration as ratio to original value (as you did), then you get 32% - the fraction of time spend by accelerated algorithm from original one. That is, you have three times less time spent (68% acceleration). Note that with that approach you cannot get more than 100% of acceleration.

If you calculate acceleration as ratio to new value (where you can have 1000% or more), you should calculate it as 7.2/2.3 = 310%

msv

2015-01-27 13:23

developer   ~0036703

You are right. All in all, 310% is not 1000%.

abv

2015-02-14 19:28

manager  

eval_poly_template_metaprogram.zip (4,544 bytes)

abv

2015-02-14 19:32

manager   ~0037543

Last edited: 2015-02-17 12:04

 I suggest we shall try variant of evaluator implemented by Istvan Csanady (see http://dev.opencascade.org/index.php?q=node/1043 , archive with his sources is attached, eval_poly_template_metaprogram.zip). It uses templates metaprogramming and according to Istvan yields higher acceleration than plain C++ code. In order to try this on desktop, the intrinsics used in the code need to be adopted for the platform.

git

2015-02-20 15:02

administrator   ~0037748

Branch CR24285_azn has been created by azn.

SHA-1: 6b46853f59c2907d55bfa825bdd4ca9297a4cf41


Detailed log of new commits:

Author: azn
Date: Tue Feb 17 14:39:04 2015 +0300

    0024285: Updates of PLib::EvalPolynomial for code acceleration
    
    SSE2 & temaplte implementation have been integrated to PLib::EvalPolynomial.

git

2015-03-05 14:24

administrator   ~0038151

Branch CR24285_azn has been updated by azn.

SHA-1: ecfb84ee8479ee8bed8b787a78e7eba5c178132f


Detailed log of new commits:

Author: azn
Date: Thu Mar 5 14:23:15 2015 +0300

    0024285: Updates of PLib::EvalPolynomial for code acceleration
    
     - Full template refactoring has been performed.
     - AVX instructions has been integrated.

git

2015-03-12 16:26

administrator   ~0038299

Branch CR24285_azn has been updated by azn.

SHA-1: fc22d2906a133ba08d76fa4db51def5458d5d31e


Detailed log of new commits:

Author: azn
Date: Wed Mar 11 09:00:14 2015 +0300

    0024285: Updates of PLib::EvalPolynomial for code acceleration
    
     - Change order of loops.
     - Using SSE2 instructions.

abv

2015-04-18 22:04

manager   ~0040028

I have compared three versions of the improvement (from branches CR24285, CR24285_3, and CR24285_azn) on current master, running 3 times test bspline intersect on each build. Here are CPU times, sec:

master: 29.; 29.2; 29.1
CR24285: 25.6, 25.8, 25.7
CR24285_3: 25.8; 25.7; 25.8
CR24285_azn: 27.; 26.75;

Then I additionally used OSD_PerfMeter to see how many calls are made to EvalPolynomial vs. NoDerivEvalPolynomial:

NoDerivEvalPolynomial : 102788324 4.35 0.04
EvalPolynomial : 129180814 10.36 0.08

Note that (a) this measurement is made on version _azn (based on templates made by Istvan Csanady), and (b) measurement severely affects performance (total time grows up to 42 sec). Nevertheless, it shows that both functions are called with almost equal frequency.

Finally I propose version 3 should be taken.

abv

2015-04-19 07:39

manager   ~0040029

Statistics on calls of EvalPolynomial with different values of DerivativeRequest and Dimension, on test bspline intersect:

          Perf meter results : enters seconds microsec/enter
EP:1,g : 39003481 2.84 0.07
EP:1,3 : 47607306 1.78 0.04
EP:1,1 : 25502944 0.83 0.03
EP:1,9 : 8730145 0.47 0.05
EP:1,6 : 9200905 0.47 0.05
EP:2,g : 5435 0.02 2.87
EP:2,3 : 4868 0.00 0.00
EP:2,1 : 2370 0.00 0.00
EP:2,9 : 973 0.00 0.00
EP:2,6 : 900 0.00 0.00

and for test boolean bsection P4 which is accelerated by 25% by this fix:

          Perf meter results : enters seconds microsec/enter
EP:1,g : 4924084 1.62 0.33
EP:1,3 : 7812651 0.42 0.05
EP:1,6 : 189456 0.00 0.00
EP:2,g : 2149101 1.06 0.49
EP:2,3 : 2168582 0.16 0.07
EP:2,6 : 38663 0.00 0.00

git

2015-04-19 18:35

administrator   ~0040030

Branch CR24285_4 has been created by abv.

SHA-1: 7c38fa1081ab47fba268f0006e114ebc4bf3bf65


Detailed log of new commits:

Author: azv
Date: Tue Jan 20 17:06:03 2015 +0300

    0024285: Updates of PLib::EvalPolynomial for code acceleration
    
    Functions PLib::EvalPolynomial and PLib::NoDerivativeEvalPolynomial are refactored to allow generation of faster code:
    1. Iteration by degree is made in outer loop
    2. Avoided pointer arithmetic
    3. Recursive templates are used to expand loop by dimension in specific cases (1-15)

abv

2015-04-19 19:02

manager   ~0040031

I have pushed new version of the fix to branch CR24285_4, please review.

This version is based on CR24285_3, with fragments specialized for particular dimension of the vector implemented by recursive template. Use of templates allows covering more special cases, and to optimize also function NoDerivativeEvalPolynomial() in the same way, keeping the same overall amount of code.

Result of my testing vs. master is (for most notable cases):

CPU boolean bcut_complex Q9: 6.6768428 / 10.1244649 [-34.05%]
CPU boolean bsection P4: 6.3648408 / 9.9684639 [-36.15%]
CPU bugs moddata_2 bug6862_3: 10.3116661 / 12.1212777 [-14.93%]
CPU bugs moddata_2 bug6862_4: 10.2648658 / 12.0588773 [-14.88%]
CPU bugs moddata_2 bug6862_6: 24.6793582 / 30.0769928 [-17.95%]
CPU de step_3 F2: 313.3748088 / 361.5791178 [-13.33%]
CPU de step_4 I1: 217.8865967 / 253.1428227 [-13.93%]
...

CPU of test perf bspline intersect is 25.1-25.4 sec, i.e. better than other versions. Overall CPU difference by all cases vs. master is reported as [-2%]

azv

2015-04-20 08:16

administrator   ~0040034

Reviewed without remarks. Please test.

abv

2015-04-20 08:52

manager   ~0040035

When reporting results of testing, please include results of testdiff on the test cases listed in 0024285:0040031

abv

2015-04-20 13:51

manager   ~0040046

I have tested the effect of the fix in Debug mode. Surprisingly, performance also has improved:

| Test case | master | CR24285_4 |
| boolean bcut_complex Q9 | 56.4 sec | 44.5 sec |
| boolean bsection P4 | 54.5 sec | 12.9 sec |
| bugs moddata_2 bug6862_6 | 236 sec | 38 sec |

mkv

2015-04-22 17:08

tester   ~0040113

Dear BugMaster,
Branch CR24285_4 from occt git-repository (and master from products git-repository) was compiled on Linux, MacOS and Windows platforms and tested on Release mode.
SHA-1: 7c38fa1081ab47fba268f0006e114ebc4bf3bf65

Number of compiler warnings:

occt component :
Linux: 18 (18 on master)
Windows: 0 (0 on master)

products component :
Linux: 4 (4 on master)
Windows: 0 (0 on master)

Regressions/Differences:
No regressions/differences

Testing cases:
Not needed

Testing on Linux:
occt component :
Total MEMORY difference: 95201625 / 94523880 [+0.72%]
Total CPU difference: 52416.46999999925 / 52229.55999999946 [+0.36%]

CPU boolean bcut_complex Q9: 24.92 / 31.98 [-22.08%]
CPU boolean bsection P4: 24.69 / 28.68 [-13.91%]
CPU de step_3 F2: 1282.14 / 1454.54 [-11.85%]
CPU de step_4 I1: 887.69 / 982.29 [-9.63%]

products component :
Total MEMORY difference: 23754993 / 23685903 [+0.29%]
Total CPU difference: 17585.62 / 17782.939999999973 [-1.11%]

Testing on Windows:
occt component :
Total MEMORY difference: 57155174 / 57165566 [-0.02%]
Total CPU difference: 15625.855765098884 / 16283.477980599077 [-4.04%]

CPU boolean bcut_complex Q9: 7.9092507 / 12.2616786 [-35.50%]
CPU boolean bsection P4: 7.4412477 / 11.8248758 [-37.07%]
CPU bugs moddata_2 bug6862_3: 12.2148783 / 14.2584914 [-14.33%]
CPU bugs moddata_2 bug6862_4: 12.1836781 / 14.3364919 [-15.02%]
CPU bugs moddata_2 bug6862_6: 29.2033872 / 36.0050308 [-18.89%]
CPU de step_3 F2: 372.6551888 / 433.4019782 [-14.02%]
CPU de step_4 I1: 258.6652581 / 304.6387528 [-15.09%]

products component :
Total MEMORY difference: 15561014 / 15564896 [-0.02%]
Total CPU difference: 6210.836612799964 / 6585.207812599981 [-5.69%]

There are no differences in images found by testdiff.

Istvan Csanady

2015-04-28 12:01

reporter   ~0040321

Some remarks: I had biggest performance gains in cases when the polynomails that I was evaluating were low order (degree 4-6). I believe that was due to the compiler was able to optimize and place most of/all the data to the SIMD units/registers, thus eliminating almost every memory access.
Also try to compile with optimization set to -fast (fastest, aggressive optimizations) with Clang, in my case that allowed the compiler to optimize a lot more compared to -o3.

git

2015-05-14 16:25

administrator   ~0040941

Branch CR24285 has been deleted by inv.

SHA-1: 1f4674920d52faedb25eb1dce34facd985de5e2e

git

2015-05-14 16:25

administrator   ~0040942

Branch CR24285_1 has been deleted by inv.

SHA-1: 751c824ac9c6055a38f7e67f8e01298ae4e809fe

git

2015-05-14 16:25

administrator   ~0040943

Branch CR24285_2 has been deleted by inv.

SHA-1: 09aaf89911592e15edd6b72d57a05b6c1d58d27c

git

2015-05-14 16:25

administrator   ~0040944

Branch CR24285_3 has been deleted by inv.

SHA-1: e1eaf5e85f3443e7e9338d7b4f48fe45cf731a2b

git

2015-05-14 16:25

administrator   ~0040945

Branch CR24285_4 has been deleted by inv.

SHA-1: 7c38fa1081ab47fba268f0006e114ebc4bf3bf65

git

2015-05-14 16:25

administrator   ~0040946

Branch CR24285_azn has been deleted by inv.

SHA-1: fc22d2906a133ba08d76fa4db51def5458d5d31e

git

2015-05-20 11:04

administrator   ~0041300

Branch CR24285_azn_test has been created by azn.

SHA-1: 3ab7cd619bb5af261d05a21cbacdaf4336ca0ef9


Detailed log of new commits:

Author: azn
Date: Tue Feb 17 14:39:04 2015 +0300

    0024285: Updates of PLib::EvalPolynomial for code acceleration
    
    SSE2 & temaplte implementation (by Istvan Csanady) have been integrated to PLib::EvalPolynomial.

git

2015-05-20 11:05

administrator   ~0041302

Branch CR24285_Istvan has been created by azn.

SHA-1: 3ab7cd619bb5af261d05a21cbacdaf4336ca0ef9


No new revisions were added by this update.

git

2015-05-20 11:40

administrator   ~0041305

Branch CR24285_azn has been created by azn.

SHA-1: 881bc61f39e7bac99b76e36898283b2db38363ba


Detailed log of new commits:

Author: azn
Date: Tue Feb 17 14:39:04 2015 +0300

    0024285: Updates of PLib::EvalPolynomial for code acceleration
    
     - Temaplte implementation.
     - Using SSE2 instructions.

git

2015-10-16 17:01

administrator   ~0047109

Branch CR24285_Istvan has been deleted by kgv.

SHA-1: 3ab7cd619bb5af261d05a21cbacdaf4336ca0ef9

git

2015-10-16 17:01

administrator   ~0047110

Branch CR24285_azn has been deleted by kgv.

SHA-1: 881bc61f39e7bac99b76e36898283b2db38363ba

git

2015-10-16 17:01

administrator   ~0047111

Branch CR24285_azn_test has been deleted by kgv.

SHA-1: 3ab7cd619bb5af261d05a21cbacdaf4336ca0ef9

Related Changesets

occt: master d721c8eb

2015-01-20 14:06:03

azv


Committer: bugmaster Details Diff		 0024285: Updates of PLib::EvalPolynomial for code acceleration

Functions PLib::EvalPolynomial and PLib::NoDerivativeEvalPolynomial are refactored to allow generation of faster code:
1. Iteration by degree is made in outer loop
2. Avoided pointer arithmetic
3. Recursive templates are used to expand loop by dimension in specific cases (1-15)		 Affected Issues
0024285
mod - src/PLib/PLib.cxx Diff File
add - tests/perf/bspline/intersect Diff File
mod - tests/perf/grids.list Diff File

Issue History

Date Modified Username Field Change
2013-10-24 09:47 azv New Issue
2013-10-24 09:47 azv Assigned To => azv
2013-10-24 09:51 azv Note Added: 0026248
2013-10-24 09:51 azv Assigned To azv => abv
2013-10-24 09:51 azv Status new => resolved
2013-10-28 18:07 abv Note Added: 0026308
2013-10-28 18:07 abv Assigned To abv => bugmaster
2013-10-28 18:07 abv Status resolved => reviewed
2013-10-29 06:51 mkv Assigned To bugmaster => mkv
2013-10-29 11:44 mkv Assigned To mkv => azv
2013-10-29 11:44 mkv Status reviewed => assigned
2013-10-29 11:57 mkv Assigned To azv => mkv
2013-10-29 11:57 mkv Status assigned => resolved
2013-10-29 11:57 mkv Status resolved => reviewed
2013-10-30 08:15 mkv Note Added: 0026345
2013-10-30 08:16 mkv Test case number => Not needed
2013-10-30 08:16 mkv Assigned To mkv => bugmaster
2013-10-30 08:16 mkv Status reviewed => tested
2013-12-19 11:50 abv Note Added: 0027258
2013-12-19 11:50 abv Status tested => feedback
2013-12-19 11:53 abv Assigned To bugmaster => abv
2014-04-04 17:42 abv Target Version => 6.8.0
2014-09-11 09:23 abv Target Version 6.8.0 => 7.1.0
2015-01-20 17:12 git Note Added: 0036272
2015-01-20 17:12 azv Assigned To abv => azv
2015-01-20 17:12 azv Status feedback => assigned
2015-01-20 17:37 azv Note Added: 0036274
2015-01-20 17:39 azv Note Edited: 0036274
2015-01-20 17:41 azv Note Edited: 0036274
2015-01-20 17:43 azv Note Added: 0036276
2015-01-20 17:43 azv Assigned To azv => abv
2015-01-20 17:43 azv Status assigned => resolved
2015-01-20 17:43 azv Steps to Reproduce Updated
2015-01-21 07:14 abv Note Added: 0036283
2015-01-21 07:33 azv Note Added: 0036284
2015-01-21 07:38 azv Note Edited: 0036284
2015-01-21 08:17 abv Note Added: 0036285
2015-01-23 06:51 abv Note Added: 0036424
2015-01-23 09:23 msv Note Added: 0036428
2015-01-27 07:11 azv Note Added: 0036682
2015-01-27 10:47 msv Note Added: 0036688
2015-01-27 11:28 abv Note Added: 0036692
2015-01-27 11:30 abv Note Edited: 0036692
2015-01-27 13:23 msv Note Added: 0036703
2015-02-14 19:28 abv File Added: eval_poly_template_metaprogram.zip
2015-02-14 19:32 abv Note Added: 0037543
2015-02-14 19:32 abv Assigned To abv => azn
2015-02-14 19:32 abv Status resolved => assigned
2015-02-17 12:04 abv Note Edited: 0037543
2015-02-20 15:02 git Note Added: 0037748
2015-03-05 14:24 git Note Added: 0038151
2015-03-12 16:26 git Note Added: 0038299
2015-04-18 22:04 abv Note Added: 0040028
2015-04-19 07:39 abv Note Added: 0040029
2015-04-19 09:44 abv Steps to Reproduce Updated
2015-04-19 18:35 git Note Added: 0040030
2015-04-19 19:02 abv Note Added: 0040031
2015-04-19 19:02 abv Assigned To azn => azv
2015-04-19 19:02 abv Priority low => high
2015-04-19 19:02 abv Status assigned => resolved
2015-04-19 19:02 abv Target Version 7.1.0 => 6.9.0
2015-04-20 08:16 azv Note Added: 0040034
2015-04-20 08:16 azv Assigned To azv => bugmaster
2015-04-20 08:16 azv Status resolved => reviewed
2015-04-20 08:52 abv Note Added: 0040035
2015-04-20 11:58 aiv Project Internal => Open CASCADE
2015-04-20 13:51 abv Note Added: 0040046
2015-04-20 16:16 abv Relationship added related to 0026110
2015-04-21 16:30 mkv Assigned To bugmaster => mkv
2015-04-22 17:08 mkv Note Added: 0040113
2015-04-22 17:08 mkv Assigned To mkv => bugmaster
2015-04-22 17:08 mkv Status reviewed => tested
2015-04-25 17:37 bugmaster Changeset attached => occt master d721c8eb
2015-04-25 17:37 bugmaster Status tested => verified
2015-04-25 17:37 bugmaster Resolution open => fixed
2015-04-28 11:25 abv View Status private => public
2015-04-28 12:01 Istvan Csanady Note Added: 0040321
2015-05-14 15:28 aiv Status verified => closed
2015-05-14 15:30 aiv Fixed in Version => 6.9.0
2015-05-14 16:25 git Note Added: 0040941
2015-05-14 16:25 git Note Added: 0040942
2015-05-14 16:25 git Note Added: 0040943
2015-05-14 16:25 git Note Added: 0040944
2015-05-14 16:25 git Note Added: 0040945
2015-05-14 16:25 git Note Added: 0040946
2015-05-20 11:04 git Note Added: 0041300
2015-05-20 11:05 git Note Added: 0041302
2015-05-20 11:40 git Note Added: 0041305
2015-10-16 17:01 git Note Added: 0047109
2015-10-16 17:01 git Note Added: 0047110
2015-10-16 17:01 git Note Added: 0047111