Philip Zeyliger has uploaded this change for review. ( 
http://gerrit.cloudera.org:8080/8211


Change subject: IMPALA-5243: Speed up code gen for wide Avro tables. (wip)
......................................................................

IMPALA-5243: Speed up code gen for wide Avro tables. (wip)

HdfsAvroScanner::CodegenMaterializeTuple generates a function
linear in size to the number of columns. On 1000 column tables,
codegen time is significant.

This change breaks up MaterializeTuple() into multiple
smaller functions, and then calls them in order. When
breaking up into 200-column chunks, there is a noticeable
speed-up.

TODO:
* Run tests.
* Check by inspection that the base case of < 200 columns gets
  inlined into not having a helper function. Or, if it doesn't
  get inlined, convince LLVM to inline it. Or, skip the helper
  functions in that case.
* Compare query performance on a many-column table.
* Remove XXX blocks. There's a tiny bit of code marked XXX
  which I'm using for debugging and iteration. I've left it in
  in case folks have suggestions for how to better do the things
  I'm doing.

I measured codegen time for various "step sizes." The case where there are no 
helper
functions is about 2.6s. The best case was about a step size of 200,
with timings of 1.35s.

$(for step in 10 50 75 100 125 150 175 200 225 250 300 400 500 1000; do for try 
in $(seq 5); do echo $step > /tmp/step_size.txt; echo -n "$step "; 
impala-shell.sh -q "select count(int_col16) from 
functional_avro.widetable_1000_cols; profile;" 2> /dev/null | grep -A9 
'CodeGen:(Total: [0-9]*s' -m 1 | sed -e 's/ - / /' |
sed -e 's/([0-9]*)//' | tr -d '\n' | tr -s ' ' ' '; echo; done; done) | tee 
out.txt
10  CodeGen:(Total: 2s389ms, non-child: 2s389ms, % non-child: 100.00%) 
CodegenTime: 476.490us CompileTime: 1s260ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 133 NumInstructions: 9.01K 
OptimizationTime: 1s100ms PeakMemoryUsage: 4.40 MB PrepareTime: 10.489ms
10  CodeGen:(Total: 2s402ms, non-child: 2s402ms, % non-child: 100.00%) 
CodegenTime: 467.903us CompileTime: 1s264ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 133 NumInstructions: 9.01K 
OptimizationTime: 1s108ms PeakMemoryUsage: 4.40 MB PrepareTime: 10.183ms
10  CodeGen:(Total: 2s405ms, non-child: 2s405ms, % non-child: 100.00%) 
CodegenTime: 470.827us CompileTime: 1s266ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 133 NumInstructions: 9.01K 
OptimizationTime: 1s110ms PeakMemoryUsage: 4.40 MB PrepareTime: 10.378ms
10  CodeGen:(Total: 2s431ms, non-child: 2s431ms, % non-child: 100.00%) 
CodegenTime: 479.621us CompileTime: 1s284ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 133 NumInstructions: 9.01K 
OptimizationTime: 1s117ms PeakMemoryUsage: 4.40 MB PrepareTime: 10.355ms
10  CodeGen:(Total: 2s390ms, non-child: 2s390ms, % non-child: 100.00%) 
CodegenTime: 541.217us CompileTime: 1s260ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 133 NumInstructions: 9.01K 
OptimizationTime: 1s101ms PeakMemoryUsage: 4.40 MB PrepareTime: 10.186ms
50  CodeGen:(Total: 2s497ms, non-child: 2s497ms, % non-child: 100.00%) 
CodegenTime: 469.008us CompileTime: 1s267ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 53 NumInstructions: 8.53K 
OptimizationTime: 1s200ms PeakMemoryUsage: 4.16 MB PrepareTime: 10.131ms
50  CodeGen:(Total: 2s468ms, non-child: 2s468ms, % non-child: 100.00%) 
CodegenTime: 479.658us CompileTime: 1s268ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 53 NumInstructions: 8.53K 
OptimizationTime: 1s170ms PeakMemoryUsage: 4.16 MB PrepareTime: 10.332ms
50  CodeGen:(Total: 2s475ms, non-child: 2s475ms, % non-child: 100.00%) 
CodegenTime: 488.243us CompileTime: 1s273ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 53 NumInstructions: 8.53K 
OptimizationTime: 1s173ms PeakMemoryUsage: 4.16 MB PrepareTime: 10.248ms
50  CodeGen:(Total: 2s488ms, non-child: 2s488ms, % non-child: 100.00%) 
CodegenTime: 492.705us CompileTime: 1s277ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 53 NumInstructions: 8.53K 
OptimizationTime: 1s181ms PeakMemoryUsage: 4.16 MB PrepareTime: 10.859ms
50  CodeGen:(Total: 2s456ms, non-child: 2s456ms, % non-child: 100.00%) 
CodegenTime: 466.180us CompileTime: 1s262ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 53 NumInstructions: 8.53K 
OptimizationTime: 1s165ms PeakMemoryUsage: 4.16 MB PrepareTime: 10.114ms
75  CodeGen:(Total: 2s875ms, non-child: 2s875ms, % non-child: 100.00%) 
CodegenTime: 474.302us CompileTime: 1s277ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 47 NumInstructions: 8.49K 
OptimizationTime: 1s568ms PeakMemoryUsage: 4.15 MB PrepareTime: 10.905ms
75  CodeGen:(Total: 2s856ms, non-child: 2s856ms, % non-child: 100.00%) 
CodegenTime: 510.086us CompileTime: 1s263ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 47 NumInstructions: 8.49K 
OptimizationTime: 1s563ms PeakMemoryUsage: 4.15 MB PrepareTime: 10.430ms
75  CodeGen:(Total: 2s877ms, non-child: 2s877ms, % non-child: 100.00%) 
CodegenTime: 507.069us CompileTime: 1s269ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 47 NumInstructions: 8.49K 
OptimizationTime: 1s578ms PeakMemoryUsage: 4.15 MB PrepareTime: 10.227ms
75  CodeGen:(Total: 2s875ms, non-child: 2s875ms, % non-child: 100.00%) 
CodegenTime: 504.735us CompileTime: 1s273ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 47 NumInstructions: 8.49K 
OptimizationTime: 1s573ms PeakMemoryUsage: 4.15 MB PrepareTime: 10.293ms
75  CodeGen:(Total: 2s867ms, non-child: 2s867ms, % non-child: 100.00%) 
CodegenTime: 480.478us CompileTime: 1s274ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 47 NumInstructions: 8.49K 
OptimizationTime: 1s563ms PeakMemoryUsage: 4.15 MB PrepareTime: 10.499ms
100  CodeGen:(Total: 2s960ms, non-child: 2s960ms, % non-child: 100.00%) 
CodegenTime: 513.029us CompileTime: 1s259ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 43 NumInstructions: 8.47K 
OptimizationTime: 1s671ms PeakMemoryUsage: 4.13 MB PrepareTime: 10.181ms
100  CodeGen:(Total: 2s959ms, non-child: 2s959ms, % non-child: 100.00%) 
CodegenTime: 515.783us CompileTime: 1s260ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 43 NumInstructions: 8.47K 
OptimizationTime: 1s670ms PeakMemoryUsage: 4.13 MB PrepareTime: 10.245ms
100  CodeGen:(Total: 3s001ms, non-child: 3s001ms, % non-child: 100.00%) 
CodegenTime: 481.243us CompileTime: 1s272ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 43 NumInstructions: 8.47K 
OptimizationTime: 1s699ms PeakMemoryUsage: 4.13 MB PrepareTime: 10.308ms
100  CodeGen:(Total: 2s958ms, non-child: 2s958ms, % non-child: 100.00%) 
CodegenTime: 497.638us CompileTime: 1s258ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 43 NumInstructions: 8.47K 
OptimizationTime: 1s670ms PeakMemoryUsage: 4.13 MB PrepareTime: 10.485ms
100  CodeGen:(Total: 2s998ms, non-child: 2s998ms, % non-child: 100.00%) 
CodegenTime: 524.061us CompileTime: 1s272ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 43 NumInstructions: 8.47K 
OptimizationTime: 1s697ms PeakMemoryUsage: 4.13 MB PrepareTime: 10.156ms
125  CodeGen:(Total: 2s974ms, non-child: 2s974ms, % non-child: 100.00%) 
CodegenTime: 472.814us CompileTime: 1s256ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 41 NumInstructions: 8.46K 
OptimizationTime: 1s688ms PeakMemoryUsage: 4.13 MB PrepareTime: 10.731ms
125  CodeGen:(Total: 2s997ms, non-child: 2s997ms, % non-child: 100.00%) 
CodegenTime: 486.340us CompileTime: 1s273ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 41 NumInstructions: 8.46K 
OptimizationTime: 1s695ms PeakMemoryUsage: 4.13 MB PrepareTime: 10.807ms
125  CodeGen:(Total: 2s998ms, non-child: 2s998ms, % non-child: 100.00%) 
CodegenTime: 475.155us CompileTime: 1s278ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 41 NumInstructions: 8.46K 
OptimizationTime: 1s688ms PeakMemoryUsage: 4.13 MB PrepareTime: 10.469ms
125  CodeGen:(Total: 2s978ms, non-child: 2s978ms, % non-child: 100.00%) 
CodegenTime: 475.038us CompileTime: 1s261ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 41 NumInstructions: 8.46K 
OptimizationTime: 1s688ms PeakMemoryUsage: 4.13 MB PrepareTime: 10.213ms
125  CodeGen:(Total: 2s976ms, non-child: 2s976ms, % non-child: 100.00%) 
CodegenTime: 542.806us CompileTime: 1s269ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 41 NumInstructions: 8.46K 
OptimizationTime: 1s677ms PeakMemoryUsage: 4.13 MB PrepareTime: 10.300ms
150  CodeGen:(Total: 1s388ms, non-child: 1s388ms, % non-child: 100.00%) 
CodegenTime: 532.355us CompileTime: 559.165ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 40 NumInstructions: 8.45K 
OptimizationTime: 802.083ms PeakMemoryUsage: 4.13 MB PrepareTime: 10.468ms
150  CodeGen:(Total: 1s389ms, non-child: 1s389ms, % non-child: 100.00%) 
CodegenTime: 489.706us CompileTime: 565.535ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 40 NumInstructions: 8.45K 
OptimizationTime: 797.119ms PeakMemoryUsage: 4.13 MB PrepareTime: 10.301ms
150  CodeGen:(Total: 1s385ms, non-child: 1s385ms, % non-child: 100.00%) 
CodegenTime: 564.083us CompileTime: 560.444ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 40 NumInstructions: 8.45K 
OptimizationTime: 797.872ms PeakMemoryUsage: 4.13 MB PrepareTime: 10.224ms
150  CodeGen:(Total: 1s377ms, non-child: 1s377ms, % non-child: 100.00%) 
CodegenTime: 482.622us CompileTime: 559.696ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 40 NumInstructions: 8.45K 
OptimizationTime: 789.466ms PeakMemoryUsage: 4.13 MB PrepareTime: 11.258ms
150  CodeGen:(Total: 1s383ms, non-child: 1s383ms, % non-child: 100.00%) 
CodegenTime: 488.556us CompileTime: 560.105ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 40 NumInstructions: 8.45K 
OptimizationTime: 796.790ms PeakMemoryUsage: 4.13 MB PrepareTime: 10.104ms
175  CodeGen:(Total: 1s342ms, non-child: 1s342ms, % non-child: 100.00%) 
CodegenTime: 664.551us CompileTime: 581.642ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 39 NumInstructions: 8.44K 
OptimizationTime: 733.573ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.681ms
175  CodeGen:(Total: 1s341ms, non-child: 1s341ms, % non-child: 100.00%) 
CodegenTime: 527.015us CompileTime: 571.381ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 39 NumInstructions: 8.44K 
OptimizationTime: 743.133ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.406ms
175  CodeGen:(Total: 1s326ms, non-child: 1s326ms, % non-child: 100.00%) 
CodegenTime: 620.484us CompileTime: 571.566ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 39 NumInstructions: 8.44K 
OptimizationTime: 727.645ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.765ms
175  CodeGen:(Total: 1s340ms, non-child: 1s340ms, % non-child: 100.00%) 
CodegenTime: 640.895us CompileTime: 579.613ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 39 NumInstructions: 8.44K 
OptimizationTime: 734.347ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.252ms
175  CodeGen:(Total: 1s340ms, non-child: 1s340ms, % non-child: 100.00%) 
CodegenTime: 507.545us CompileTime: 577.371ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 39 NumInstructions: 8.44K 
OptimizationTime: 735.613ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.303ms
200  CodeGen:(Total: 1s333ms, non-child: 1s333ms, % non-child: 100.00%) 
CodegenTime: 613.562us CompileTime: 605.320ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 38 NumInstructions: 8.44K 
OptimizationTime: 701.276ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.014ms
200  CodeGen:(Total: 1s323ms, non-child: 1s323ms, % non-child: 100.00%) 
CodegenTime: 532.751us CompileTime: 594.476ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 38 NumInstructions: 8.44K 
OptimizationTime: 701.189ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.897ms
200  CodeGen:(Total: 1s330ms, non-child: 1s330ms, % non-child: 100.00%) 
CodegenTime: 519.793us CompileTime: 599.211ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 38 NumInstructions: 8.44K 
OptimizationTime: 704.668ms PeakMemoryUsage: 4.12 MB PrepareTime: 9.934ms
200  CodeGen:(Total: 1s330ms, non-child: 1s330ms, % non-child: 100.00%) 
CodegenTime: 574.106us CompileTime: 591.840ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 38 NumInstructions: 8.44K 
OptimizationTime: 710.732ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.619ms
200  CodeGen:(Total: 1s343ms, non-child: 1s343ms, % non-child: 100.00%) 
CodegenTime: 531.922us CompileTime: 601.756ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 38 NumInstructions: 8.44K 
OptimizationTime: 714.359ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.674ms
225  CodeGen:(Total: 1s477ms, non-child: 1s477ms, % non-child: 100.00%) 
CodegenTime: 524.866us CompileTime: 607.973ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 38 NumInstructions: 8.44K 
OptimizationTime: 841.991ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.386ms
225  CodeGen:(Total: 1s490ms, non-child: 1s490ms, % non-child: 100.00%) 
CodegenTime: 653.539us CompileTime: 617.348ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 38 NumInstructions: 8.44K 
OptimizationTime: 846.569ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.193ms
225  CodeGen:(Total: 1s482ms, non-child: 1s482ms, % non-child: 100.00%) 
CodegenTime: 607.267us CompileTime: 615.646ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 38 NumInstructions: 8.44K 
OptimizationTime: 839.834ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.399ms
225  CodeGen:(Total: 1s491ms, non-child: 1s491ms, % non-child: 100.00%) 
CodegenTime: 600.223us CompileTime: 620.448ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 38 NumInstructions: 8.44K 
OptimizationTime: 843.065ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.288ms
225  CodeGen:(Total: 1s490ms, non-child: 1s490ms, % non-child: 100.00%) 
CodegenTime: 534.518us CompileTime: 617.852ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 38 NumInstructions: 8.44K 
OptimizationTime: 845.317ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.453ms
250  CodeGen:(Total: 1s415ms, non-child: 1s415ms, % non-child: 100.00%) 
CodegenTime: 538.088us CompileTime: 643.237ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 37 NumInstructions: 8.43K 
OptimizationTime: 744.778ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.123ms
250  CodeGen:(Total: 1s425ms, non-child: 1s425ms, % non-child: 100.00%) 
CodegenTime: 515.369us CompileTime: 647.450ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 37 NumInstructions: 8.43K 
OptimizationTime: 749.996ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.509ms
250  CodeGen:(Total: 1s442ms, non-child: 1s442ms, % non-child: 100.00%) 
CodegenTime: 547.479us CompileTime: 643.160ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 37 NumInstructions: 8.43K 
OptimizationTime: 770.421ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.672ms
250  CodeGen:(Total: 1s411ms, non-child: 1s411ms, % non-child: 100.00%) 
CodegenTime: 557.810us CompileTime: 636.188ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 37 NumInstructions: 8.43K 
OptimizationTime: 746.959ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.448ms
250  CodeGen:(Total: 1s437ms, non-child: 1s437ms, % non-child: 100.00%) 
CodegenTime: 538.374us CompileTime: 653.973ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 37 NumInstructions: 8.43K 
OptimizationTime: 748.269ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.362ms
300  CodeGen:(Total: 1s606ms, non-child: 1s606ms, % non-child: 100.00%) 
CodegenTime: 521.837us CompileTime: 677.628ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 37 NumInstructions: 8.43K 
OptimizationTime: 901.113ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.261ms
300  CodeGen:(Total: 1s584ms, non-child: 1s584ms, % non-child: 100.00%) 
CodegenTime: 560.073us CompileTime: 664.160ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 37 NumInstructions: 8.43K 
OptimizationTime: 891.712ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.287ms
300  CodeGen:(Total: 1s596ms, non-child: 1s596ms, % non-child: 100.00%) 
CodegenTime: 556.069us CompileTime: 669.866ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 37 NumInstructions: 8.43K 
OptimizationTime: 898.832ms PeakMemoryUsage: 4.12 MB PrepareTime: 9.956ms
300  CodeGen:(Total: 1s588ms, non-child: 1s588ms, % non-child: 100.00%) 
CodegenTime: 603.589us CompileTime: 668.717ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 37 NumInstructions: 8.43K 
OptimizationTime: 892.021ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.260ms
300  CodeGen:(Total: 1s596ms, non-child: 1s596ms, % non-child: 100.00%) 
CodegenTime: 556.296us CompileTime: 670.226ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 37 NumInstructions: 8.43K 
OptimizationTime: 898.943ms PeakMemoryUsage: 4.12 MB PrepareTime: 10.306ms
400  CodeGen:(Total: 1s604ms, non-child: 1s604ms, % non-child: 100.00%) 
CodegenTime: 565.012us CompileTime: 739.865ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 36 NumInstructions: 8.43K 
OptimizationTime: 836.981ms PeakMemoryUsage: 4.11 MB PrepareTime: 10.174ms
400  CodeGen:(Total: 1s584ms, non-child: 1s584ms, % non-child: 100.00%) 
CodegenTime: 575.709us CompileTime: 736.661ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 36 NumInstructions: 8.43K 
OptimizationTime: 820.141ms PeakMemoryUsage: 4.11 MB PrepareTime: 10.431ms
400  CodeGen:(Total: 1s603ms, non-child: 1s603ms, % non-child: 100.00%) 
CodegenTime: 578.476us CompileTime: 738.403ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 36 NumInstructions: 8.43K 
OptimizationTime: 836.339ms PeakMemoryUsage: 4.11 MB PrepareTime: 11.222ms
400  CodeGen:(Total: 1s601ms, non-child: 1s601ms, % non-child: 100.00%) 
CodegenTime: 565.338us CompileTime: 739.958ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 36 NumInstructions: 8.43K 
OptimizationTime: 834.638ms PeakMemoryUsage: 4.11 MB PrepareTime: 10.093ms
400  CodeGen:(Total: 1s586ms, non-child: 1s586ms, % non-child: 100.00%) 
CodegenTime: 581.171us CompileTime: 732.219ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 36 NumInstructions: 8.43K 
OptimizationTime: 827.611ms PeakMemoryUsage: 4.11 MB PrepareTime: 10.068ms
500  CodeGen:(Total: 1s826ms, non-child: 1s826ms, % non-child: 100.00%) 
CodegenTime: 608.492us CompileTime: 849.240ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 35 NumInstructions: 8.42K 
OptimizationTime: 949.177ms PeakMemoryUsage: 4.11 MB PrepareTime: 10.111ms
500  CodeGen:(Total: 1s818ms, non-child: 1s818ms, % non-child: 100.00%) 
CodegenTime: 575.550us CompileTime: 852.164ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 35 NumInstructions: 8.42K 
OptimizationTime: 937.886ms PeakMemoryUsage: 4.11 MB PrepareTime: 10.390ms
500  CodeGen:(Total: 1s820ms, non-child: 1s820ms, % non-child: 100.00%) 
CodegenTime: 736.558us CompileTime: 844.141ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 35 NumInstructions: 8.42K 
OptimizationTime: 947.839ms PeakMemoryUsage: 4.11 MB PrepareTime: 10.301ms
500  CodeGen:(Total: 1s824ms, non-child: 1s824ms, % non-child: 100.00%) 
CodegenTime: 594.016us CompileTime: 848.599ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 35 NumInstructions: 8.42K 
OptimizationTime: 943.845ms PeakMemoryUsage: 4.11 MB PrepareTime: 14.425ms
500  CodeGen:(Total: 1s836ms, non-child: 1s836ms, % non-child: 100.00%) 
CodegenTime: 681.831us CompileTime: 856.213ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 35 NumInstructions: 8.42K 
OptimizationTime: 952.317ms PeakMemoryUsage: 4.11 MB PrepareTime: 10.635ms
1000  CodeGen:(Total: 2s657ms, non-child: 2s657ms, % non-child: 100.00%) 
CodegenTime: 651.648us CompileTime: 1s267ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 34 NumInstructions: 8.41K 
OptimizationTime: 1s361ms PeakMemoryUsage: 4.11 MB PrepareTime: 10.251ms
1000  CodeGen:(Total: 2s635ms, non-child: 2s635ms, % non-child: 100.00%) 
CodegenTime: 581.623us CompileTime: 1s252ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 34 NumInstructions: 8.41K 
OptimizationTime: 1s353ms PeakMemoryUsage: 4.11 MB PrepareTime: 10.277ms
1000  CodeGen:(Total: 2s679ms, non-child: 2s679ms, % non-child: 100.00%) 
CodegenTime: 551.666us CompileTime: 1s270ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 34 NumInstructions: 8.41K 
OptimizationTime: 1s379ms PeakMemoryUsage: 4.11 MB PrepareTime: 10.257ms
1000  CodeGen:(Total: 2s676ms, non-child: 2s676ms, % non-child: 100.00%) 
CodegenTime: 623.111us CompileTime: 1s275ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 34 NumInstructions: 8.41K 
OptimizationTime: 1s372ms PeakMemoryUsage: 4.11 MB PrepareTime: 10.357ms
1000  CodeGen:(Total: 2s659ms, non-child: 2s659ms, % non-child: 100.00%) 
CodegenTime: 558.860us CompileTime: 1s267ms LoadTime: 0.000ns 
ModuleBitcodeSize: 1.95 MB NumFunctions: 34 NumInstructions: 8.41K 
OptimizationTime: 1s362ms PeakMemoryUsage: 4.11 MB PrepareTime: 10.574ms

Change-Id: I7f1b390be4adf6e6699a18344234f8ff7ee74476
(cherry picked from commit 7fa2443b436d274f92c6d163ecb554bc0ead6725)
---
M be/src/codegen/llvm-codegen.cc
M be/src/exec/hdfs-avro-scanner.cc
M be/src/exec/hdfs-avro-scanner.h
3 files changed, 134 insertions(+), 40 deletions(-)



  git pull ssh://gerrit.cloudera.org:29418/Impala-ASF refs/changes/11/8211/1
--
To view, visit http://gerrit.cloudera.org:8080/8211
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: newchange
Gerrit-Change-Id: I7f1b390be4adf6e6699a18344234f8ff7ee74476
Gerrit-Change-Number: 8211
Gerrit-PatchSet: 1
Gerrit-Owner: Philip Zeyliger <[email protected]>

Reply via email to