Re: [RFC PATCH] crypto: crc32c-pclmul - Use pmovzxdq to shrink K_table
From: George Spelvin
Date: Fri May 30 2014 - 01:25:23 EST
Olay, recompiled with the acpi-cpufreq driver, so the performance governor
actually works, pegging the frequency at 3900 MHz.
Existing (old) code:
[ 455.641397]
[ 455.641397] testing speed of crc32c
[ 455.641403] test 0 ( 16 byte blocks, 16 bytes per update, 1 updates): 73 cycles/operation, 4 cycles/byte
[ 455.641406] test 1 ( 64 byte blocks, 16 bytes per update, 4 updates): 418 cycles/operation, 6 cycles/byte
[ 455.641409] test 2 ( 64 byte blocks, 64 bytes per update, 1 updates): 89 cycles/operation, 1 cycles/byte
[ 455.641411] test 3 ( 256 byte blocks, 16 bytes per update, 16 updates): 1330 cycles/operation, 5 cycles/byte
[ 455.641417] test 4 ( 256 byte blocks, 64 bytes per update, 4 updates): 502 cycles/operation, 1 cycles/byte
[ 455.641420] test 5 ( 256 byte blocks, 256 bytes per update, 1 updates): 170 cycles/operation, 0 cycles/byte
[ 455.641422] test 6 ( 1024 byte blocks, 16 bytes per update, 64 updates): 4971 cycles/operation, 4 cycles/byte
[ 455.641440] test 7 ( 1024 byte blocks, 256 bytes per update, 4 updates): 805 cycles/operation, 0 cycles/byte
[ 455.641445] test 8 ( 1024 byte blocks, 1024 bytes per update, 1 updates): 371 cycles/operation, 0 cycles/byte
[ 455.641448] test 9 ( 2048 byte blocks, 16 bytes per update, 128 updates): 9839 cycles/operation, 4 cycles/byte
[ 455.641484] test 10 ( 2048 byte blocks, 256 bytes per update, 8 updates): 1436 cycles/operation, 0 cycles/byte
[ 455.641490] test 11 ( 2048 byte blocks, 1024 bytes per update, 2 updates): 824 cycles/operation, 0 cycles/byte
[ 455.641494] test 12 ( 2048 byte blocks, 2048 bytes per update, 1 updates): 494 cycles/operation, 0 cycles/byte
[ 455.641498] test 13 ( 4096 byte blocks, 16 bytes per update, 256 updates): 19561 cycles/operation, 4 cycles/byte
[ 455.641568] test 14 ( 4096 byte blocks, 256 bytes per update, 16 updates): 2757 cycles/operation, 0 cycles/byte
[ 455.641579] test 15 ( 4096 byte blocks, 1024 bytes per update, 4 updates): 1633 cycles/operation, 0 cycles/byte
[ 455.641586] test 16 ( 4096 byte blocks, 4096 bytes per update, 1 updates): 861 cycles/operation, 0 cycles/byte
[ 455.641590] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39015 cycles/operation, 4 cycles/byte
[ 455.641729] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5412 cycles/operation, 0 cycles/byte
[ 455.641749] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3106 cycles/operation, 0 cycles/byte
[ 455.641762] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1656 cycles/operation, 0 cycles/byte
[ 455.641769] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1639 cycles/operation, 0 cycles/byte
[ 480.885336]
[ 480.885336] testing speed of crc32c
[ 480.885342] test 0 ( 16 byte blocks, 16 bytes per update, 1 updates): 81 cycles/operation, 5 cycles/byte
[ 480.885345] test 1 ( 64 byte blocks, 16 bytes per update, 4 updates): 426 cycles/operation, 6 cycles/byte
[ 480.885348] test 2 ( 64 byte blocks, 64 bytes per update, 1 updates): 96 cycles/operation, 1 cycles/byte
[ 480.885350] test 3 ( 256 byte blocks, 16 bytes per update, 16 updates): 1331 cycles/operation, 5 cycles/byte
[ 480.885356] test 4 ( 256 byte blocks, 64 bytes per update, 4 updates): 497 cycles/operation, 1 cycles/byte
[ 480.885359] test 5 ( 256 byte blocks, 256 bytes per update, 1 updates): 179 cycles/operation, 0 cycles/byte
[ 480.885361] test 6 ( 1024 byte blocks, 16 bytes per update, 64 updates): 4961 cycles/operation, 4 cycles/byte
[ 480.885380] test 7 ( 1024 byte blocks, 256 bytes per update, 4 updates): 795 cycles/operation, 0 cycles/byte
[ 480.885384] test 8 ( 1024 byte blocks, 1024 bytes per update, 1 updates): 366 cycles/operation, 0 cycles/byte
[ 480.885387] test 9 ( 2048 byte blocks, 16 bytes per update, 128 updates): 9827 cycles/operation, 4 cycles/byte
[ 480.885423] test 10 ( 2048 byte blocks, 256 bytes per update, 8 updates): 1445 cycles/operation, 0 cycles/byte
[ 480.885430] test 11 ( 2048 byte blocks, 1024 bytes per update, 2 updates): 834 cycles/operation, 0 cycles/byte
[ 480.885434] test 12 ( 2048 byte blocks, 2048 bytes per update, 1 updates): 495 cycles/operation, 0 cycles/byte
[ 480.885437] test 13 ( 4096 byte blocks, 16 bytes per update, 256 updates): 19560 cycles/operation, 4 cycles/byte
[ 480.885507] test 14 ( 4096 byte blocks, 256 bytes per update, 16 updates): 2767 cycles/operation, 0 cycles/byte
[ 480.885518] test 15 ( 4096 byte blocks, 1024 bytes per update, 4 updates): 1643 cycles/operation, 0 cycles/byte
[ 480.885525] test 16 ( 4096 byte blocks, 4096 bytes per update, 1 updates): 862 cycles/operation, 0 cycles/byte
[ 480.885530] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39013 cycles/operation, 4 cycles/byte
[ 480.885669] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5417 cycles/operation, 0 cycles/byte
[ 480.885689] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3113 cycles/operation, 0 cycles/byte
[ 480.885701] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1665 cycles/operation, 0 cycles/byte
[ 480.885708] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1646 cycles/operation, 0 cycles/byte
Proposed (new) code:
[ 800.253907]
[ 800.253907] testing speed of crc32c
[ 800.253913] test 0 ( 16 byte blocks, 16 bytes per update, 1 updates): 75 cycles/operation, 4 cycles/byte
[ 800.253915] test 1 ( 64 byte blocks, 16 bytes per update, 4 updates): 421 cycles/operation, 6 cycles/byte
[ 800.253919] test 2 ( 64 byte blocks, 64 bytes per update, 1 updates): 88 cycles/operation, 1 cycles/byte
[ 800.253920] test 3 ( 256 byte blocks, 16 bytes per update, 16 updates): 1339 cycles/operation, 5 cycles/byte
[ 800.253942] test 4 ( 256 byte blocks, 64 bytes per update, 4 updates): 511 cycles/operation, 1 cycles/byte
[ 800.253945] test 5 ( 256 byte blocks, 256 bytes per update, 1 updates): 180 cycles/operation, 0 cycles/byte
[ 800.253947] test 6 ( 1024 byte blocks, 16 bytes per update, 64 updates): 4972 cycles/operation, 4 cycles/byte
[ 800.253966] test 7 ( 1024 byte blocks, 256 bytes per update, 4 updates): 789 cycles/operation, 0 cycles/byte
[ 800.253970] test 8 ( 1024 byte blocks, 1024 bytes per update, 1 updates): 371 cycles/operation, 0 cycles/byte
[ 800.253973] test 9 ( 2048 byte blocks, 16 bytes per update, 128 updates): 10093 cycles/operation, 4 cycles/byte
[ 800.254010] test 10 ( 2048 byte blocks, 256 bytes per update, 8 updates): 1443 cycles/operation, 0 cycles/byte
[ 800.254017] test 11 ( 2048 byte blocks, 1024 bytes per update, 2 updates): 829 cycles/operation, 0 cycles/byte
[ 800.254021] test 12 ( 2048 byte blocks, 2048 bytes per update, 1 updates): 495 cycles/operation, 0 cycles/byte
[ 800.254024] test 13 ( 4096 byte blocks, 16 bytes per update, 256 updates): 19556 cycles/operation, 4 cycles/byte
[ 800.254094] test 14 ( 4096 byte blocks, 256 bytes per update, 16 updates): 2762 cycles/operation, 0 cycles/byte
[ 800.254105] test 15 ( 4096 byte blocks, 1024 bytes per update, 4 updates): 1640 cycles/operation, 0 cycles/byte
[ 800.254113] test 16 ( 4096 byte blocks, 4096 bytes per update, 1 updates): 854 cycles/operation, 0 cycles/byte
[ 800.254117] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39015 cycles/operation, 4 cycles/byte
[ 800.254256] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5415 cycles/operation, 0 cycles/byte
[ 800.254276] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3113 cycles/operation, 0 cycles/byte
[ 800.254288] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1666 cycles/operation, 0 cycles/byte
[ 800.254295] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1638 cycles/operation, 0 cycles/byte
[ 808.113346]
[ 808.113346] testing speed of crc32c
[ 808.113353] test 0 ( 16 byte blocks, 16 bytes per update, 1 updates): 70 cycles/operation, 4 cycles/byte
[ 808.113355] test 1 ( 64 byte blocks, 16 bytes per update, 4 updates): 432 cycles/operation, 6 cycles/byte
[ 808.113359] test 2 ( 64 byte blocks, 64 bytes per update, 1 updates): 89 cycles/operation, 1 cycles/byte
[ 808.113360] test 3 ( 256 byte blocks, 16 bytes per update, 16 updates): 1330 cycles/operation, 5 cycles/byte
[ 808.113366] test 4 ( 256 byte blocks, 64 bytes per update, 4 updates): 514 cycles/operation, 2 cycles/byte
[ 808.113369] test 5 ( 256 byte blocks, 256 bytes per update, 1 updates): 171 cycles/operation, 0 cycles/byte
[ 808.113371] test 6 ( 1024 byte blocks, 16 bytes per update, 64 updates): 4968 cycles/operation, 4 cycles/byte
[ 808.113390] test 7 ( 1024 byte blocks, 256 bytes per update, 4 updates): 833 cycles/operation, 0 cycles/byte
[ 808.113394] test 8 ( 1024 byte blocks, 1024 bytes per update, 1 updates): 368 cycles/operation, 0 cycles/byte
[ 808.113398] test 9 ( 2048 byte blocks, 16 bytes per update, 128 updates): 9842 cycles/operation, 4 cycles/byte
[ 808.113434] test 10 ( 2048 byte blocks, 256 bytes per update, 8 updates): 1462 cycles/operation, 0 cycles/byte
[ 808.113440] test 11 ( 2048 byte blocks, 1024 bytes per update, 2 updates): 827 cycles/operation, 0 cycles/byte
[ 808.113444] test 12 ( 2048 byte blocks, 2048 bytes per update, 1 updates): 494 cycles/operation, 0 cycles/byte
[ 808.113448] test 13 ( 4096 byte blocks, 16 bytes per update, 256 updates): 19556 cycles/operation, 4 cycles/byte
[ 808.113518] test 14 ( 4096 byte blocks, 256 bytes per update, 16 updates): 2783 cycles/operation, 0 cycles/byte
[ 808.113529] test 15 ( 4096 byte blocks, 1024 bytes per update, 4 updates): 1645 cycles/operation, 0 cycles/byte
[ 808.113536] test 16 ( 4096 byte blocks, 4096 bytes per update, 1 updates): 853 cycles/operation, 0 cycles/byte
[ 808.113540] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39019 cycles/operation, 4 cycles/byte
[ 808.113679] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5437 cycles/operation, 0 cycles/byte
[ 808.113700] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3118 cycles/operation, 0 cycles/byte
[ 808.113712] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1652 cycles/operation, 0 cycles/byte
[ 808.113719] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1639 cycles/operation, 0 cycles/byte
As you can see, the differences look comparable to the spread in the
measured values. Considering just the last 4 tests (since only blocks
of at least 200 bytes are affected by the change), here are 10 more runs
of each:
Existing (old) code:
[ 2168.651975] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39013 cycles/operation, 4 cycles/byte
[ 2168.652114] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5425 cycles/operation, 0 cycles/byte
[ 2168.652134] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3121 cycles/operation, 0 cycles/byte
[ 2168.652146] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1654 cycles/operation, 0 cycles/byte
[ 2168.652153] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1644 cycles/operation, 0 cycles/byte
[ 2168.672956] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39025 cycles/operation, 4 cycles/byte
[ 2168.673095] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5494 cycles/operation, 0 cycles/byte
[ 2168.673116] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3113 cycles/operation, 0 cycles/byte
[ 2168.673157] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1674 cycles/operation, 0 cycles/byte
[ 2168.673169] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1636 cycles/operation, 0 cycles/byte
[ 2168.696197] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39012 cycles/operation, 4 cycles/byte
[ 2168.696336] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5410 cycles/operation, 0 cycles/byte
[ 2168.696356] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3119 cycles/operation, 0 cycles/byte
[ 2168.696368] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1667 cycles/operation, 0 cycles/byte
[ 2168.696375] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1635 cycles/operation, 0 cycles/byte
[ 2168.716198] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39015 cycles/operation, 4 cycles/byte
[ 2168.716337] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5543 cycles/operation, 0 cycles/byte
[ 2168.716358] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3111 cycles/operation, 0 cycles/byte
[ 2168.716370] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1673 cycles/operation, 0 cycles/byte
[ 2168.716377] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1636 cycles/operation, 0 cycles/byte
[ 2168.739520] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39022 cycles/operation, 4 cycles/byte
[ 2168.739659] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5484 cycles/operation, 0 cycles/byte
[ 2168.739680] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3110 cycles/operation, 0 cycles/byte
[ 2168.739692] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1687 cycles/operation, 0 cycles/byte
[ 2168.739699] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1640 cycles/operation, 0 cycles/byte
[ 2168.762814] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39015 cycles/operation, 4 cycles/byte
[ 2168.762953] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5412 cycles/operation, 0 cycles/byte
[ 2168.762973] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3109 cycles/operation, 0 cycles/byte
[ 2168.762985] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1675 cycles/operation, 0 cycles/byte
[ 2168.762992] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1634 cycles/operation, 0 cycles/byte
[ 2168.796244] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39008 cycles/operation, 4 cycles/byte
[ 2168.796383] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5413 cycles/operation, 0 cycles/byte
[ 2168.796403] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3106 cycles/operation, 0 cycles/byte
[ 2168.796415] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1658 cycles/operation, 0 cycles/byte
[ 2168.796422] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1637 cycles/operation, 0 cycles/byte
[ 2168.819616] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39040 cycles/operation, 4 cycles/byte
[ 2168.819757] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5416 cycles/operation, 0 cycles/byte
[ 2168.819777] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3109 cycles/operation, 0 cycles/byte
[ 2168.819814] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1674 cycles/operation, 0 cycles/byte
[ 2168.819823] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1649 cycles/operation, 0 cycles/byte
[ 2168.859652] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39011 cycles/operation, 4 cycles/byte
[ 2168.859806] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5412 cycles/operation, 0 cycles/byte
[ 2168.859826] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3117 cycles/operation, 0 cycles/byte
[ 2168.859841] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1665 cycles/operation, 0 cycles/byte
[ 2168.859850] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1639 cycles/operation, 0 cycles/byte
[ 2168.896378] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39023 cycles/operation, 4 cycles/byte
[ 2168.896532] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5424 cycles/operation, 0 cycles/byte
[ 2168.896554] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3126 cycles/operation, 0 cycles/byte
[ 2168.896567] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1664 cycles/operation, 0 cycles/byte
[ 2168.896574] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1634 cycles/operation, 0 cycles/byte
Proposed (new) code:
[ 2061.715381] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39018 cycles/operation, 4 cycles/byte
[ 2061.715520] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5420 cycles/operation, 0 cycles/byte
[ 2061.715540] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3101 cycles/operation, 0 cycles/byte
[ 2061.715552] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1662 cycles/operation, 0 cycles/byte
[ 2061.715559] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1647 cycles/operation, 0 cycles/byte
[ 2061.734935] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39029 cycles/operation, 4 cycles/byte
[ 2061.735074] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5416 cycles/operation, 0 cycles/byte
[ 2061.735094] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3253 cycles/operation, 0 cycles/byte
[ 2061.735107] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1685 cycles/operation, 0 cycles/byte
[ 2061.735114] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1642 cycles/operation, 0 cycles/byte
[ 2061.761667] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39027 cycles/operation, 4 cycles/byte
[ 2061.761806] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5415 cycles/operation, 0 cycles/byte
[ 2061.761826] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3112 cycles/operation, 0 cycles/byte
[ 2061.761838] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1673 cycles/operation, 0 cycles/byte
[ 2061.761845] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1644 cycles/operation, 0 cycles/byte
[ 2061.781846] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39010 cycles/operation, 4 cycles/byte
[ 2061.781985] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5424 cycles/operation, 0 cycles/byte
[ 2061.782005] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3242 cycles/operation, 0 cycles/byte
[ 2061.782018] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1687 cycles/operation, 0 cycles/byte
[ 2061.782025] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1641 cycles/operation, 0 cycles/byte
[ 2061.801881] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39020 cycles/operation, 4 cycles/byte
[ 2061.802020] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5424 cycles/operation, 0 cycles/byte
[ 2061.802041] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3113 cycles/operation, 0 cycles/byte
[ 2061.802053] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1673 cycles/operation, 0 cycles/byte
[ 2061.802060] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1638 cycles/operation, 0 cycles/byte
[ 2061.822194] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39034 cycles/operation, 4 cycles/byte
[ 2061.822333] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5414 cycles/operation, 0 cycles/byte
[ 2061.822353] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3246 cycles/operation, 0 cycles/byte
[ 2061.822366] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1703 cycles/operation, 0 cycles/byte
[ 2061.822373] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1643 cycles/operation, 0 cycles/byte
[ 2061.842361] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39017 cycles/operation, 4 cycles/byte
[ 2061.842500] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5411 cycles/operation, 0 cycles/byte
[ 2061.842520] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3115 cycles/operation, 0 cycles/byte
[ 2061.842532] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1671 cycles/operation, 0 cycles/byte
[ 2061.842539] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1641 cycles/operation, 0 cycles/byte
[ 2061.875909] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39020 cycles/operation, 4 cycles/byte
[ 2061.876048] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5414 cycles/operation, 0 cycles/byte
[ 2061.876068] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3250 cycles/operation, 0 cycles/byte
[ 2061.876081] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1699 cycles/operation, 0 cycles/byte
[ 2061.876088] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1640 cycles/operation, 0 cycles/byte
[ 2061.899397] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39017 cycles/operation, 4 cycles/byte
[ 2061.899536] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5420 cycles/operation, 0 cycles/byte
[ 2061.899556] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3109 cycles/operation, 0 cycles/byte
[ 2061.899568] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1683 cycles/operation, 0 cycles/byte
[ 2061.899576] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1640 cycles/operation, 0 cycles/byte
[ 2061.922872] test 17 ( 8192 byte blocks, 16 bytes per update, 512 updates): 39016 cycles/operation, 4 cycles/byte
[ 2061.923010] test 18 ( 8192 byte blocks, 256 bytes per update, 32 updates): 5538 cycles/operation, 0 cycles/byte
[ 2061.923032] test 19 ( 8192 byte blocks, 1024 bytes per update, 8 updates): 3248 cycles/operation, 0 cycles/byte
[ 2061.923044] test 20 ( 8192 byte blocks, 4096 bytes per update, 2 updates): 1683 cycles/operation, 0 cycles/byte
[ 2061.923052] test 21 ( 8192 byte blocks, 8192 bytes per update, 1 updates): 1640 cycles/operation, 0 cycles/byte
Averaging the 8K bytes per update, I do see an average of 3.2 cycles per
operation (that is, per 8K of data processed) lost, or about 1 cycle per
(3K or less) block processed. I'm hoping the reduced D-cache polution
makes it up somewhere else.
Comments?
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/