powerpc/mm: Update page table format for book3s 64
v6 cover powerpc/mm: Update page table format for book3s 64
Hi All,
This patch series attempt to update book3s 64 linux page table format to make it more flexible. Our current pte format is very restrictive and we overload multiple pte bits. This is due to the non-availability of free bits in pte_t. We use pte_t to track the validity of 4K subpages. This patch series free up pte_t of 11 bits by tracking 4 4k subpage using one bit. The pte format is updated such that we have a better method for identifying a pte entry at pmd level. This will also enable us to implement hugetlb migration(not yet done in this series). Before making the changes to the pte format, I am splitting the pte header definition such that we now have the below layout for headers
book3s
32
hash.h pgtable.h
64
hash.h pgtable.h hash-4k.h hash-64k.h
booke
32
pgtable.h pte-40x.h pte-44x.h pte-8xx.h pte-fsl-booke.h
64
pgtable-4k.h pgtable-64k.h pgtable.h
I have done the header split such that booke headers and modified to the minimum so as to avoid causing breakage in booke.
The patch series can also be found at
https://github.com/kvaneesh/linux.git book3s-pte-format-v2
https://github.com/kvaneesh/linux/commits/book3s-pte-format-v2
Changes from V5:
* We need a min pte frag size to be 4096 bytes, This is needed to track base page information with a THP config with 4K hash pte. Hence drop the changes to free up second half of pgtable_t. This implies we don't make any change to the size of pte fragment in this series.
Changes from V4:
* rebase to latest linus
* Add mmtest numbers
Changes from V3:
* Add missing #define pgprot_*
* Add Acked-by
Changes from V2:
* rebase to -next for powerpc tree
Changes from V1:
1) Build fix with STRICT_MM_TYPES enabled
2) pte_mkwrite fix for nohash
3) rebase to latest linus tree.
Performance numbers with and without patch series.
-------------- 4k hash pte guest config start-------------------- aim9
4.3.0 4.3.0
13672-gc669d35-dirty 13708-g8e488c2-dirty
Min page_test 460586.67 ( 0.00%) 435200.00 ( -5.51%)
Min brk_test 4155429.71 ( 0.00%) 4231312.46 ( 1.83%)
Min exec_test 960.69 ( 0.00%) 904.33 ( -5.87%)
Min fork_test 2142.38 ( 0.00%) 2059.80 ( -3.85%)
Hmean page_test 467906.57 ( 0.00%) 445186.92 ( -4.86%)
Hmean brk_test 4260526.41 ( 0.00%) 4257237.07 ( -0.08%)
Hmean exec_test 1013.71 ( 0.00%) 966.27 ( -4.68%)
Hmean fork_test 2292.90 ( 0.00%) 2257.22 ( -1.56%)
Stddev page_test 4416.83 ( 0.00%) 4694.17 ( -6.28%)
Stddev brk_test 72844.41 ( 0.00%) 12520.81 ( 82.81%)
Stddev exec_test 27.23 ( 0.00%) 30.88 (-13.41%)
Stddev fork_test 76.29 ( 0.00%) 106.89 (-40.10%)
CoeffVar page_test 0.94 ( 0.00%) 1.05 (-11.70%)
CoeffVar brk_test 1.71 ( 0.00%) 0.29 ( 82.79%)
CoeffVar exec_test 2.68 ( 0.00%) 3.19 (-18.94%)
CoeffVar fork_test 3.32 ( 0.00%) 4.72 (-42.16%)
Max page_test 473053.33 ( 0.00%) 452426.67 ( -4.36%)
Max brk_test 4366733.33 ( 0.00%) 4272666.67 ( -2.15%)
Max exec_test 1056.67 ( 0.00%) 1014.67 ( -3.97%)
Max fork_test 2396.80 ( 0.00%) 2441.78 ( 1.88%) 4.3.0 4.3.0
13672-gc669d35-dirty13708-g8e488c2-dirty
User 0.24 0.23
System 0.47 0.51Elapsed 723.43 723.04
4.3.0 4.3.0
13672-gc669d35-dirty13708-g8e488c2-dirty
Minor Faults 37700464 36522147
Major Faults 25 25
Swap Ins 0 0
Swap Outs 0 0
Allocation stalls 0 0
DMA allocs 12485233 11996182
DMA32 allocs 0 0
Normal allocs 0 0
Movable allocs 0 0
Direct pages scanned 0 0
Kswapd pages scanned 0 0
Kswapd pages reclaimed 0 0
Direct pages reclaimed 0 0Kswapd efficiency 100% 100%
Kswapd velocity 0.000 0.000
Direct efficiency 100% 100%
Direct velocity 0.000 0.000
Percentage direct scans 0% 0%
Zone normal velocity 0.000 0.000 Zone dma32 velocity 0.000 0.000 Zone dma velocity 0.000 0.000 Page writes by reclaim 0.000 0.000 Page writes file 0 0 Page writes anon 0 0 Page reclaim immediate 0 0 Sector Reads 4224 4224 Sector Writes 69944 69296 Page rescued immediate 0 0 Slabs scanned 0 0 Direct inode steals 0 0 Kswapd inode steals 0 0 Kswapd skipped wait 0 0 THP fault alloc 0 0 THP collapse alloc 0 0 THP splits 0 0 THP fault fallback 0 0 THP collapse fail 0 0 Compaction stalls 0 0 Compaction success 0 0 Compaction failures 0 0 Page migrate success 0 0 Page migrate failure 0 0 Compaction pages isolated 0 0 Compaction migrate scanned 0 0 Compaction free scanned 0 0 Compaction cost 0 0 NUMA alloc hit 12482507 11993566 NUMA alloc miss 0 0 NUMA interleave hit 0 0
NUMA alloc local 12482507 11993566
NUMA base PTE updates 647 643
NUMA huge PMD updates 0 0
NUMA page range updates 647 643
NUMA hint faults 844060 873604 NUMA hint local faults 844060 873604
NUMA hint local percent 100 100
NUMA pages migrated 0 0
AutoNUMA cost 4220% 4368%
pft timings
4.3.0 4.3.0
13672-gc669d35-dirty 13708-g8e488c2-dirty
Min system-1 0.0700 ( 0.00%) 0.0700 ( 0.00%)
Min system-3 0.1400 ( 0.00%) 0.1300 ( 7.14%)
Min system-5 0.2400 ( 0.00%) 0.2300 ( 4.17%)
Min system-7 0.2500 ( 0.00%) 0.2800 (-12.00%)
Min system-8 0.2800 ( 0.00%) 0.2900 ( -3.57%)
Min elapsed-1 0.0900 ( 0.00%) 0.0800 ( 11.11%)
Min elapsed-3 0.0600 ( 0.00%) 0.0600 ( 0.00%)
Min elapsed-5 0.0600 ( 0.00%) 0.0600 ( 0.00%)
Min elapsed-7 0.0600 ( 0.00%) 0.0600 ( 0.00%)
Min elapsed-8 0.0600 ( 0.00%) 0.0600 ( 0.00%)
Amean system-1 0.1677 ( 0.00%) 0.1574 ( 6.18%)
Amean system-3 0.1714 ( 0.00%) 0.1575 ( 8.10%)
Amean system-5 0.3611 ( 0.00%) 0.2919 ( 19.18%)
Amean system-7 0.3364 ( 0.00%) 0.3396 ( -0.97%)
Amean system-8 0.4108 ( 0.00%) 0.3704 ( 9.83%)
Amean elapsed-1 0.2040 ( 0.00%) 0.1929 ( 5.45%)
Amean elapsed-3 0.0733 ( 0.00%) 0.0686 ( 6.31%)
Amean elapsed-5 0.0968 ( 0.00%) 0.0779 ( 19.51%)
Amean elapsed-7 0.0656 ( 0.00%) 0.0665 ( -1.33%)
Amean elapsed-8 0.0690 ( 0.00%) 0.0622 ( 9.78%)
Stddev system-1 0.3338 ( 0.00%) 0.3156 ( 5.45%)
Stddev system-3 0.0203 ( 0.00%) 0.0150 ( 26.29%)
Stddev system-5 0.2636 ( 0.00%) 0.0340 ( 87.11%)
Stddev system-7 0.0343 ( 0.00%) 0.0316 ( 7.83%)
Stddev system-8 0.2066 ( 0.00%) 0.0386 ( 81.31%)
Stddev elapsed-1 0.4417 ( 0.00%) 0.4287 ( 2.95%)
Stddev elapsed-3 0.0085 ( 0.00%) 0.0072 ( 15.09%)
Stddev elapsed-5 0.0706 ( 0.00%) 0.0089 ( 87.39%)
Stddev elapsed-7 0.0063 ( 0.00%) 0.0071 (-12.63%)
Stddev elapsed-8 0.0335 ( 0.00%) 0.0042 ( 87.54%)
CoeffVar system-1 198.9598 ( 0.00%) 200.5134 ( -0.78%)
CoeffVar system-3 11.8421 ( 0.00%) 9.4973 ( 19.80%)
CoeffVar system-5 72.9902 ( 0.00%) 11.6450 ( 84.05%)
CoeffVar system-7 10.1931 ( 0.00%) 9.3046 ( 8.72%)
CoeffVar system-8 50.2929 ( 0.00%) 10.4259 ( 79.27%)
CoeffVar elapsed-1 216.5224 ( 0.00%) 222.2603 ( -2.65%)
CoeffVar elapsed-3 11.5790 ( 0.00%) 10.4938 ( 9.37%)
CoeffVar elapsed-5 72.9538 ( 0.00%) 11.4303 ( 84.33%)
CoeffVar elapsed-7 9.5902 ( 0.00%) 10.6597 (-11.15%)
CoeffVar elapsed-8 48.5886 ( 0.00%) 6.7081 ( 86.19%)
Max system-1 2.6900 ( 0.00%) 2.7500 ( -2.23%)
Max system-3 0.2600 ( 0.00%) 0.2300 ( 11.54%)
Max system-5 1.4800 ( 0.00%) 0.3800 ( 74.32%)
Max system-7 0.4300 ( 0.00%) 0.4400 ( -2.33%)
Max system-8 2.2000 ( 0.00%) 0.4300 ( 80.45%)
Max elapsed-1 3.4200 ( 0.00%) 3.7200 ( -8.77%)
Max elapsed-3 0.1100 ( 0.00%) 0.0900 ( 18.18%)
Max elapsed-5 0.4200 ( 0.00%) 0.1000 ( 76.19%)
Max elapsed-7 0.0800 ( 0.00%) 0.0800 ( 0.00%)
Max elapsed-8 0.3600 ( 0.00%) 0.0700 ( 80.56%)pft faults
4.3.0 4.3.0
13672-gc669d35-dirty 13708-g8e488c2-dirty
Min faults/cpu-1 4826.3370 ( 0.00%) 4724.0920 ( -2.12%)
Min faults/cpu-3 48054.3130 ( 0.00%) 52057.8790 ( 8.33%)
Min faults/cpu-5 8657.3530 ( 0.00%) 32656.8230 (277.21%)
Min faults/cpu-7 28499.3350 ( 0.00%) 28117.8470 ( -1.34%)
Min faults/cpu-8 5785.6660 ( 0.00%) 28468.9430 (392.06%)
Min faults/sec-1 3818.0350 ( 0.00%) 3503.7170 ( -8.23%)
Min faults/sec-3 115163.0990 ( 0.00%) 142861.2660 ( 24.05%)
Min faults/sec-5 31307.4700 ( 0.00%) 126644.8950 (304.52%)
Min faults/sec-7 161275.0230 ( 0.00%) 154987.8940 ( -3.90%)
Min faults/sec-8 36175.2080 ( 0.00%) 181500.8080 (401.73%)
Hmean faults/cpu-1 73887.2577 ( 0.00%) 78026.1722 ( 5.60%)
Hmean faults/cpu-3 71522.1756 ( 0.00%) 77089.5360 ( 7.78%)
Hmean faults/cpu-5 34277.6489 ( 0.00%) 42103.7728 ( 22.83%)
Hmean faults/cpu-7 36418.2317 ( 0.00%) 36054.2105 ( -1.00%)
Hmean faults/cpu-8 29741.2065 ( 0.00%) 33195.0525 ( 11.61%)
Hmean faults/sec-1 63979.9276 ( 0.00%) 67449.3892 ( 5.42%)
Hmean faults/sec-3 175991.8241 ( 0.00%) 189556.4239 ( 7.71%)
Hmean faults/sec-5 134694.5971 ( 0.00%) 167086.7980 ( 24.05%)
Hmean faults/sec-7 196906.4585 ( 0.00%) 194505.0606 ( -1.22%)
Hmean faults/sec-8 190635.0992 ( 0.00%) 210558.9780 ( 10.45%)
Stddev faults/cpu-1 28421.3218 ( 0.00%) 31119.8455 ( -9.49%)
Stddev faults/cpu-3 7433.8195 ( 0.00%) 7024.7950 ( 5.50%)
Stddev faults/cpu-5 10013.2881 ( 0.00%) 4757.1716 ( 52.49%)
Stddev faults/cpu-7 3856.6692 ( 0.00%) 3128.6064 ( 18.88%)
Stddev faults/cpu-8 4506.9641 ( 0.00%) 3753.0081 ( 16.73%)
Stddev faults/sec-1 27405.3207 ( 0.00%) 29714.3764 ( -8.43%)
Stddev faults/sec-3 16581.7614 ( 0.00%) 16406.9053 ( 1.05%)
Stddev faults/sec-5 41735.1309 ( 0.00%) 18896.6954 ( 54.72%)
Stddev faults/sec-7 13126.2068 ( 0.00%) 16315.6539 (-24.30%)
Stddev faults/sec-8 26939.6466 ( 0.00%) 11738.7089 ( 56.43%)
CoeffVar faults/cpu-1 25.6278 ( 0.00%) 26.5666 ( -3.66%)
CoeffVar faults/cpu-3 10.2707 ( 0.00%) 9.0358 ( 12.02%)
CoeffVar faults/cpu-5 24.5397 ( 0.00%) 11.1592 ( 54.53%)
CoeffVar faults/cpu-7 10.4800 ( 0.00%) 8.6127 ( 17.82%)
CoeffVar faults/cpu-8 14.3209 ( 0.00%) 11.1754 ( 21.96%)
CoeffVar faults/sec-1 26.0287 ( 0.00%) 26.7416 ( -2.74%)
CoeffVar faults/sec-3 9.3294 ( 0.00%) 8.5905 ( 7.92%)
CoeffVar faults/sec-5 25.7575 ( 0.00%) 11.1747 ( 56.62%)
CoeffVar faults/sec-7 6.6347 ( 0.00%) 8.3276 (-25.52%)
CoeffVar faults/sec-8 13.3650 ( 0.00%) 5.5575 ( 58.42%)
Max faults/cpu-1 169773.9820 ( 0.00%) 173628.2170 ( 2.27%)
Max faults/cpu-3 87876.0920 ( 0.00%) 95342.5620 ( 8.50%)
Max faults/cpu-5 52178.9930 ( 0.00%) 53589.5550 ( 2.70%)
Max faults/cpu-7 49544.2420 ( 0.00%) 43912.2830 (-11.37%)
Max faults/cpu-8 42351.8250 ( 0.00%) 42866.1120 ( 1.21%)
Max faults/sec-1 153006.2870 ( 0.00%) 159219.3870 ( 4.06%)
Max faults/sec-3 213038.1300 ( 0.00%) 222970.7520 ( 4.66%)
Max faults/sec-5 216193.7690 ( 0.00%) 216955.0470 ( 0.35%)
Max faults/sec-7 219260.6940 ( 0.00%) 229234.7670 ( 4.55%)
Max faults/sec-8 232367.7020 ( 0.00%) 229220.0190 ( -1.35%) 4.3.0 4.3.0
13672-gc669d35-dirty13708-g8e488c2-dirtyUser 13.47 10.80
System 138.78 134.16
Elapsed 48.68 46.84
4.3.0 4.3.0
13672-gc669d35-dirty13708-g8e488c2-dirty
Minor Faults 5478166 5478364
Major Faults 0 0
Swap Ins 0 0
Swap Outs 0 0
Allocation stalls 0 0
DMA allocs 5269869 5270049
DMA32 allocs 0 0
Normal allocs 0 0
Movable allocs 0 0
Direct pages scanned 0 0
Kswapd pages scanned 0 0
Kswapd pages reclaimed 0 0
Direct pages reclaimed 0 0Kswapd efficiency 100% 100%
Kswapd velocity 0.000 0.000
Direct efficiency 100% 100%
Direct velocity 0.000 0.000
Percentage direct scans 0% 0%
Zone normal velocity 0.000 0.000 Zone dma32 velocity 0.000 0.000 Zone dma velocity 0.000 0.000 Page writes by reclaim 0.000 0.000 Page writes file 0 0 Page writes anon 0 0 Page reclaim immediate 0 0 Sector Reads 72 72 Sector Writes 560 500 Page rescued immediate 0 0 Slabs scanned 0 0 Direct inode steals 0 0 Kswapd inode steals 0 0 Kswapd skipped wait 0 0 THP fault alloc 0 0 THP collapse alloc 0 0 THP splits 0 0 THP fault fallback 0 0 THP collapse fail 0 0 Compaction stalls 0 0 Compaction success 0 0 Compaction failures 0 0 Page migrate success 0 0 Page migrate failure 0 0 Compaction pages isolated 0 0 Compaction migrate scanned 0 0 Compaction free scanned 0 0 Compaction cost 0 0 NUMA alloc hit 5269751 5269915 NUMA alloc miss 0 0 NUMA interleave hit 0 0
NUMA alloc local 5269751 5269915
NUMA base PTE updates 23119 11610
NUMA huge PMD updates 0 0
NUMA page range updates 23119 11610
NUMA hint faults 16 7 NUMA hint local faults 16 7
NUMA hint local percent 100 100
NUMA pages migrated 0 0
AutoNUMA cost 0% 0%
ebizzy Overall Throughput
4.3.0 4.3.0
13672-gc669d35-dirty 13708-g8e488c2-dirty
Min Rsec-1 6486.00 ( 0.00%) 6153.00 ( -5.13%)
Min Rsec-3 9821.00 ( 0.00%) 10083.00 ( 2.67%)
Min Rsec-5 10688.00 ( 0.00%) 10576.00 ( -1.05%)
Min Rsec-7 11359.00 ( 0.00%) 10870.00 ( -4.30%)
Min Rsec-12 14925.00 ( 0.00%) 13443.00 ( -9.93%)
Min Rsec-18 12493.00 ( 0.00%) 11553.00 ( -7.52%)
Min Rsec-24 13696.00 ( 0.00%) 12764.00 ( -6.80%)
Min Rsec-30 13669.00 ( 0.00%) 13299.00 ( -2.71%)
Min Rsec-32 14852.00 ( 0.00%) 13008.00 (-12.42%)
Hmean Rsec-1 6574.40 ( 0.00%) 6230.95 ( -5.22%)
Hmean Rsec-3 10005.71 ( 0.00%) 10403.25 ( 3.97%)
Hmean Rsec-5 11064.39 ( 0.00%) 10999.12 ( -0.59%)
Hmean Rsec-7 11793.27 ( 0.00%) 11044.37 ( -6.35%)
Hmean Rsec-12 15130.50 ( 0.00%) 13929.85 ( -7.94%)
Hmean Rsec-18 13396.79 ( 0.00%) 12284.36 ( -8.30%)
Hmean Rsec-24 13981.41 ( 0.00%) 13493.87 ( -3.49%)
Hmean Rsec-30 14127.08 ( 0.00%) 13482.65 ( -4.56%)
Hmean Rsec-32 15170.35 ( 0.00%) 13493.89 (-11.05%)
Stddev Rsec-1 62.81 ( 0.00%) 52.93 ( 15.73%)
Stddev Rsec-3 152.04 ( 0.00%) 267.01 (-75.62%)
Stddev Rsec-5 260.81 ( 0.00%) 290.26 (-11.29%)
Stddev Rsec-7 540.61 ( 0.00%) 266.96 ( 50.62%)
Stddev Rsec-12 203.42 ( 0.00%) 314.95 (-54.83%)
Stddev Rsec-18 607.51 ( 0.00%) 456.19 ( 24.91%)
Stddev Rsec-24 182.90 ( 0.00%) 698.10 (-281.69%)
Stddev Rsec-30 500.28 ( 0.00%) 194.07 ( 61.21%)
Stddev Rsec-32 247.90 ( 0.00%) 405.56 (-63.60%)
CoeffVar Rsec-1 0.96 ( 0.00%) 0.85 ( 11.08%)
CoeffVar Rsec-3 1.52 ( 0.00%) 2.56 (-68.84%)
CoeffVar Rsec-5 2.36 ( 0.00%) 2.64 (-11.94%)
CoeffVar Rsec-7 4.57 ( 0.00%) 2.42 ( 47.19%)
CoeffVar Rsec-12 1.34 ( 0.00%) 2.26 (-68.11%)
CoeffVar Rsec-18 4.53 ( 0.00%) 3.71 ( 18.05%)
CoeffVar Rsec-24 1.31 ( 0.00%) 5.16 (-294.54%)
CoeffVar Rsec-30 3.54 ( 0.00%) 1.44 ( 59.31%)
CoeffVar Rsec-32 1.63 ( 0.00%) 3.00 (-83.81%)
Max Rsec-1 6679.00 ( 0.00%) 6284.00 ( -5.91%)
Max Rsec-3 10252.00 ( 0.00%) 10865.00 ( 5.98%)
Max Rsec-5 11307.00 ( 0.00%) 11337.00 ( 0.27%)
Max Rsec-7 12841.00 ( 0.00%) 11582.00 ( -9.80%)
Max Rsec-12 15509.00 ( 0.00%) 14361.00 ( -7.40%)
Max Rsec-18 14229.00 ( 0.00%) 12792.00 (-10.10%)
Max Rsec-24 14266.00 ( 0.00%) 14826.00 ( 3.93%)
Max Rsec-30 15055.00 ( 0.00%) 13840.00 ( -8.07%)
Max Rsec-32 15508.00 ( 0.00%) 14237.00 ( -8.20%)ebizzy Per-thread
4.3.0 4.3.0
13672-gc669d35-dirty 13708-g8e488c2-dirty
Min Rsec-1 6486.00 ( 0.00%) 6153.00 ( -5.13%)
Min Rsec-3 3100.00 ( 0.00%) 3232.00 ( 4.26%)
Min Rsec-5 1940.00 ( 0.00%) 2018.00 ( 4.02%)
Min Rsec-7 1532.00 ( 0.00%) 1465.00 ( -4.37%)
Min Rsec-12 1107.00 ( 0.00%) 1012.00 ( -8.58%)
Min Rsec-18 585.00 ( 0.00%) 538.00 ( -8.03%)
Min Rsec-24 459.00 ( 0.00%) 431.00 ( -6.10%)
Min Rsec-30 373.00 ( 0.00%) 358.00 ( -4.02%)
Min Rsec-32 372.00 ( 0.00%) 332.00 (-10.75%)
Hmean Rsec-1 6574.40 ( 0.00%) 6230.95 ( -5.22%)
Hmean Rsec-3 3328.16 ( 0.00%) 3457.35 ( 3.88%)
Hmean Rsec-5 2209.26 ( 0.00%) 2197.43 ( -0.54%)
Hmean Rsec-7 1661.38 ( 0.00%) 1576.23 ( -5.13%)
Hmean Rsec-12 1248.27 ( 0.00%) 1148.41 ( -8.00%)
Hmean Rsec-18 692.79 ( 0.00%) 639.55 ( -7.68%)
Hmean Rsec-24 570.63 ( 0.00%) 545.79 ( -4.35%)
Hmean Rsec-30 461.30 ( 0.00%) 439.45 ( -4.74%)
Hmean Rsec-32 462.88 ( 0.00%) 413.36 (-10.70%)
Stddev Rsec-1 62.81 ( 0.00%) 52.93 (-15.73%)
Stddev Rsec-3 158.59 ( 0.00%) 211.53 ( 33.38%)
Stddev Rsec-5 96.83 ( 0.00%) 88.33 ( -8.78%)
Stddev Rsec-7 290.12 ( 0.00%) 57.71 (-80.11%)
Stddev Rsec-12 134.37 ( 0.00%) 136.35 ( 1.47%)
Stddev Rsec-18 253.56 ( 0.00%) 220.27 (-13.13%)
Stddev Rsec-24 85.84 ( 0.00%) 106.71 ( 24.32%)
Stddev Rsec-30 76.73 ( 0.00%) 72.81 ( -5.11%)
Stddev Rsec-32 75.83 ( 0.00%) 62.71 (-17.30%)
CoeffVar Rsec-1 0.96 ( 0.00%) 0.85 ( 11.08%)
CoeffVar Rsec-3 4.75 ( 0.00%) 6.10 (-28.23%)
CoeffVar Rsec-5 4.37 ( 0.00%) 4.01 ( 8.25%)
CoeffVar Rsec-7 17.19 ( 0.00%) 3.66 ( 78.73%)
CoeffVar Rsec-12 10.66 ( 0.00%) 11.75 (-10.19%)
CoeffVar Rsec-18 34.02 ( 0.00%) 32.25 ( 5.19%)
CoeffVar Rsec-24 14.74 ( 0.00%) 18.95 (-28.52%)
CoeffVar Rsec-30 16.29 ( 0.00%) 16.22 ( 0.45%)
CoeffVar Rsec-32 16.01 ( 0.00%) 14.88 ( 7.07%)
Max Rsec-1 6679.00 ( 0.00%) 6284.00 ( -5.91%)
Max Rsec-3 3602.00 ( 0.00%) 3967.00 ( 10.13%)
Max Rsec-5 2351.00 ( 0.00%) 2433.00 ( 3.49%)
Max Rsec-7 3355.00 ( 0.00%) 1737.00 (-48.23%)
Max Rsec-12 1890.00 ( 0.00%) 1787.00 ( -5.45%)
Max Rsec-18 1448.00 ( 0.00%) 1295.00 (-10.57%)
Max Rsec-24 818.00 ( 0.00%) 905.00 ( 10.64%)
Max Rsec-30 771.00 ( 0.00%) 724.00 ( -6.10%)
Max Rsec-32 676.00 ( 0.00%) 648.00 ( -4.14%)ebizzy Thread spread
4.3.0 4.3.0
13672-gc669d35-dirty 13708-g8e488c2-dirty
Min spread-1 0.00 ( 0.00%) 0.00 ( 0.00%)
Min spread-3 272.00 ( 0.00%) 342.00 (-25.74%)
Min spread-5 119.00 ( 0.00%) 132.00 (-10.92%)
Min spread-7 90.00 ( 0.00%) 101.00 (-12.22%)
Min spread-12 278.00 ( 0.00%) 112.00 ( 59.71%)
Min spread-18 751.00 ( 0.00%) 638.00 ( 15.05%)
Min spread-24 192.00 ( 0.00%) 260.00 (-35.42%)
Min spread-30 307.00 ( 0.00%) 260.00 ( 15.31%)
Min spread-32 240.00 ( 0.00%) 179.00 ( 25.42%)
Hmean spread-1 0.00 ( 0.00%) 0.00 ( 0.00%)
Hmean spread-3 315.55 ( 0.00%) 411.86 (-30.52%)
Hmean spread-5 179.34 ( 0.00%) 172.47 ( 3.83%)
Hmean spread-7 121.46 ( 0.00%) 134.77 (-10.96%)
Hmean spread-12 331.88 ( 0.00%) 263.00 ( 20.75%)
Hmean spread-18 775.27 ( 0.00%) 692.94 ( 10.62%)
Hmean spread-24 267.67 ( 0.00%) 351.87 (-31.46%)
Hmean spread-30 339.24 ( 0.00%) 304.50 ( 10.24%)
Hmean spread-32 267.37 ( 0.00%) 220.17 ( 17.65%)
Stddev spread-1 0.00 ( 0.00%) 0.00 ( 0.00%)
Stddev spread-3 76.63 ( 0.00%) 63.98 ( 16.50%)
Stddev spread-5 77.12 ( 0.00%) 42.96 ( 44.30%)
Stddev spread-7 689.22 ( 0.00%) 24.63 ( 96.43%)
Stddev spread-12 192.78 ( 0.00%) 258.17 (-33.92%)
Stddev spread-18 27.29 ( 0.00%) 37.57 (-37.67%)
Stddev spread-24 56.34 ( 0.00%) 69.16 (-22.75%)
Stddev spread-30 34.27 ( 0.00%) 40.12 (-17.07%)
Stddev spread-32 24.82 ( 0.00%) 41.70 (-67.98%)
CoeffVar spread-1 0.00 ( 0.00%) 0.00 ( 0.00%)
CoeffVar spread-3 23.26 ( 0.00%) 15.20 (-34.64%)
CoeffVar spread-5 38.03 ( 0.00%) 23.71 (-37.66%)
CoeffVar spread-7 154.95 ( 0.00%) 17.67 (-88.60%)
CoeffVar spread-12 49.76 ( 0.00%) 60.89 ( 22.36%)
CoeffVar spread-18 3.52 ( 0.00%) 5.41 ( 53.76%)
CoeffVar spread-24 20.12 ( 0.00%) 18.90 ( -6.09%)
CoeffVar spread-30 10.00 ( 0.00%) 12.96 ( 29.55%)
CoeffVar spread-32 9.21 ( 0.00%) 18.31 ( 98.81%)
Max spread-1 0.00 ( 0.00%) 0.00 ( 0.00%)
Max spread-3 478.00 ( 0.00%) 535.00 (-11.92%)
Max spread-5 345.00 ( 0.00%) 260.00 ( 24.64%)
Max spread-7 1823.00 ( 0.00%) 175.00 ( 90.40%)
Max spread-12 772.00 ( 0.00%) 751.00 ( 2.72%)
Max spread-18 827.00 ( 0.00%) 746.00 ( 9.79%)
Max spread-24 351.00 ( 0.00%) 448.00 (-27.64%)
Max spread-30 387.00 ( 0.00%) 366.00 ( 5.43%)
Max spread-32 304.00 ( 0.00%) 278.00 ( 8.55%) 4.3.0 4.3.0
13672-gc669d35-dirty13708-g8e488c2-dirty
User 1577.04 1520.68
System 6809.94 6864.08
Elapsed 1353.57 1353.80 4.3.0 4.3.0
13672-gc669d35-dirty13708-g8e488c2-dirty
Minor Faults 100301324 95014933
Major Faults 0 0
Swap Ins 0 0
Swap Outs 0 0
Allocation stalls 0 0
DMA allocs 100222214 94935232
DMA32 allocs 0 0
Normal allocs 0 0
Movable allocs 0 0
Direct pages scanned 0 0
Kswapd pages scanned 0 0
Kswapd pages reclaimed 0 0
Direct pages reclaimed 0 0Kswapd efficiency 100% 100%
Kswapd velocity 0.000 0.000
Direct efficiency 100% 100%
Direct velocity 0.000 0.000
Percentage direct scans 0% 0%
Zone normal velocity 0.000 0.000 Zone dma32 velocity 0.000 0.000 Zone dma velocity 0.000 0.000 Page writes by reclaim 0.000 0.000 Page writes file 0 0 Page writes anon 0 0 Page reclaim immediate 0 0 Sector Reads 440 560 Sector Writes 4324 4156 Page rescued immediate 0 0 Slabs scanned 0 0 Direct inode steals 0 0 Kswapd inode steals 0 0 Kswapd skipped wait 0 0 THP fault alloc 0 0 THP collapse alloc 0 0 THP splits 0 0 THP fault fallback 0 0 THP collapse fail 0 0 Compaction stalls 0 0 Compaction success 0 0 Compaction failures 0 0 Page migrate success 0 0 Page migrate failure 0 0 Compaction pages isolated 0 0 Compaction migrate scanned 0 0 Compaction free scanned 0 0 Compaction cost 0 0 NUMA alloc hit 100222248 94935323 NUMA alloc miss 0 0 NUMA interleave hit 0 0 NUMA alloc local 100222248 94935323 NUMA base PTE updates 43936 45000 NUMA huge PMD updates 0 0 NUMA page range updates 43936 45000 NUMA hint faults 35295 36155 NUMA hint local faults 35295 36155
NUMA hint local percent 100 100
NUMA pages migrated 0 0
AutoNUMA cost 176% 181%
-------------- 4k hash pte guest config stop-------------------- -------------- 64k hash pte powernv config start--------------------
aim9
4.3.0 4.3.0
13672-gc669d35 13708-g8e488c2
Min page_test 2256353.33 ( 0.00%) 2225073.33 ( -1.39%)
Min brk_test 4131000.00 ( 0.00%) 4153666.67 ( 0.55%)
Min exec_test 1639.67 ( 0.00%) 1635.00 ( -0.28%)
Min fork_test 4636.91 ( 0.00%) 4633.82 ( -0.07%)
Hmean page_test 2281108.70 ( 0.00%) 2282854.75 ( 0.08%)
Hmean brk_test 4178161.02 ( 0.00%) 4190571.57 ( 0.30%)
Hmean exec_test 1666.34 ( 0.00%) 1663.18 ( -0.19%)
Hmean fork_test 4659.91 ( 0.00%) 4665.26 ( 0.11%)
Stddev page_test 19681.94 ( 0.00%) 23269.98 (-18.23%)
Stddev brk_test 62136.33 ( 0.00%) 41482.83 ( 33.24%)
Stddev exec_test 14.09 ( 0.00%) 11.74 ( 16.69%)
Stddev fork_test 9.66 ( 0.00%) 18.39 (-90.33%)
CoeffVar page_test 0.86 ( 0.00%) 1.02 (-18.14%)
CoeffVar brk_test 1.49 ( 0.00%) 0.99 ( 33.43%)
CoeffVar exec_test 0.85 ( 0.00%) 0.71 ( 16.53%)
CoeffVar fork_test 0.21 ( 0.00%) 0.39 (-90.11%)
Max page_test 2323106.67 ( 0.00%) 2307466.67 ( -0.67%)
Max brk_test 4285133.33 ( 0.00%) 4321400.00 ( 0.85%)
Max exec_test 1684.33 ( 0.00%) 1678.33 ( -0.36%)
Max fork_test 4676.88 ( 0.00%) 4690.21 ( 0.29%) 4.3.0 4.3.0
13672-gc669d3513708-g8e488c2
User 0.09 2.37
System 0.13 0.53Elapsed 722.20 726.32
4.3.0 4.3.0
13672-gc669d3513708-g8e488c2
Minor Faults 84853683 84777713
Major Faults 25 328
Swap Ins 0 0
Swap Outs 0 0
Allocation stalls 0 0
DMA allocs 38237729 38213150
DMA32 allocs 0 0
Normal allocs 0 0
Movable allocs 0 0
Direct pages scanned 0 0
Kswapd pages scanned 0 0
Kswapd pages reclaimed 0 0
Direct pages reclaimed 0 0Kswapd efficiency 100% 100%
Kswapd velocity 0.000 0.000
Direct efficiency 100% 100%
Direct velocity 0.000 0.000
Percentage direct scans 0% 0%
Zone normal velocity 0.000 0.000 Zone dma32 velocity 0.000 0.000 Zone dma velocity 0.000 0.000 Page writes by reclaim 0.000 0.000 Page writes file 0 0 Page writes anon 0 0 Page reclaim immediate 0 0 Sector Reads 4744 22336 Sector Writes 71712 71744 Page rescued immediate 0 0 Slabs scanned 0 0 Direct inode steals 0 0 Kswapd inode steals 0 0 Kswapd skipped wait 0 0 THP fault alloc 0 2 THP collapse alloc 0 0 THP splits 0 0 THP fault fallback 0 3 THP collapse fail 0 0 Compaction stalls 0 0 Compaction success 0 0 Compaction failures 0 0 Page migrate success 0 0 Page migrate failure 0 0 Compaction pages isolated 0 0 Compaction migrate scanned 0 0 Compaction free scanned 0 0 Compaction cost 0 0 NUMA alloc hit 38231124 38206624 NUMA alloc miss 0 0 NUMA interleave hit 0 0
NUMA alloc local 38229247 38204174
NUMA base PTE updates 646 637
NUMA huge PMD updates 0 0
NUMA page range updates 646 637
NUMA hint faults 1803811 1756586 NUMA hint local faults 1803625 1756366
NUMA hint local percent 99 99
NUMA pages migrated 0 0
AutoNUMA cost 9019% 8782%
pft timings
4.3.0 4.3.0
13672-gc669d35 13708-g8e488c2
Min system-1 1.8600 ( 0.00%) 1.8700 ( -0.54%)
Min system-4 4.1900 ( 0.00%) 4.1900 ( 0.00%)
Min system-7 6.5800 ( 0.00%) 5.8500 ( 11.09%)
Min system-12 9.3600 ( 0.00%) 11.0400 (-17.95%)
Min system-21 15.4700 ( 0.00%) 16.2900 ( -5.30%)
Min system-30 22.2600 ( 0.00%) 21.4400 ( 3.68%)
Min system-48 30.7000 ( 0.00%) 31.4000 ( -2.28%)
Min system-79 55.4600 ( 0.00%) 54.2300 ( 2.22%)
Min system-110 63.6400 ( 0.00%) 61.8200 ( 2.86%)
Min system-141 66.8900 ( 0.00%) 67.3400 ( -0.67%)
Min system-160 69.5500 ( 0.00%) 70.4900 ( -1.35%)
Min elapsed-1 2.0300 ( 0.00%) 2.0400 ( -0.49%)
Min elapsed-4 1.3000 ( 0.00%) 1.3100 ( -0.77%)
Min elapsed-7 1.2300 ( 0.00%) 0.9900 ( 19.51%)
Min elapsed-12 0.8700 ( 0.00%) 1.1500 (-32.18%)
Min elapsed-21 0.9100 ( 0.00%) 1.0100 (-10.99%)
Min elapsed-30 0.8700 ( 0.00%) 0.8700 ( 0.00%)
Min elapsed-48 0.8700 ( 0.00%) 0.8800 ( -1.15%)
Min elapsed-79 0.9000 ( 0.00%) 0.8800 ( 2.22%)
Min elapsed-110 0.9500 ( 0.00%) 0.9000 ( 5.26%)
Min elapsed-141 0.8800 ( 0.00%) 0.9300 ( -5.68%)
Min elapsed-160 0.8800 ( 0.00%) 0.9200 ( -4.55%)
Amean system-1 1.8881 ( 0.00%) 1.9090 ( -1.11%)
Amean system-4 6.5492 ( 0.00%) 6.5650 ( -0.24%)
Amean system-7 10.0890 ( 0.00%) 10.1636 ( -0.74%)
Amean system-12 14.7406 ( 0.00%) 15.3927 ( -4.42%)
Amean system-21 22.2554 ( 0.00%) 25.3304 (-13.82%)
Amean system-30 35.2691 ( 0.00%) 29.8990 ( 15.23%)
Amean system-48 45.8848 ( 0.00%) 42.5765 ( 7.21%)
Amean system-79 65.6720 ( 0.00%) 63.4805 ( 3.34%)
Amean system-110 77.0405 ( 0.00%) 75.3361 ( 2.21%)
Amean system-141 83.8607 ( 0.00%) 83.1904 ( 0.80%)
Amean system-160 84.8629 ( 0.00%) 83.5491 ( 1.55%)
Amean elapsed-1 2.0836 ( 0.00%) 2.1036 ( -0.96%)
Amean elapsed-4 1.7204 ( 0.00%) 1.7264 ( -0.35%)
Amean elapsed-7 1.6176 ( 0.00%) 1.6254 ( -0.48%)
Amean elapsed-12 1.5135 ( 0.00%) 1.5618 ( -3.19%)
Amean elapsed-21 1.3556 ( 0.00%) 1.5172 (-11.92%)
Amean elapsed-30 1.5014 ( 0.00%) 1.2959 ( 13.69%)
Amean elapsed-48 1.2863 ( 0.00%) 1.2243 ( 4.82%)
Amean elapsed-79 1.2315 ( 0.00%) 1.1953 ( 2.94%)
Amean elapsed-110 1.1055 ( 0.00%) 1.0705 ( 3.17%)
Amean elapsed-141 1.0494 ( 0.00%) 1.0268 ( 2.16%)
Amean elapsed-160 1.0396 ( 0.00%) 1.0375 ( 0.20%)
Stddev system-1 0.0758 ( 0.00%) 0.0766 ( -1.03%)
Stddev system-4 0.3702 ( 0.00%) 0.2734 ( 26.16%)
Stddev system-7 1.4758 ( 0.00%) 1.4623 ( 0.91%)
Stddev system-12 2.1752 ( 0.00%) 1.7466 ( 19.70%)
Stddev system-21 5.2401 ( 0.00%) 3.9628 ( 24.38%)
Stddev system-30 6.4944 ( 0.00%) 7.3898 (-13.79%)
Stddev system-48 11.4237 ( 0.00%) 9.5028 ( 16.81%)
Stddev system-79 9.4485 ( 0.00%) 6.2530 ( 33.82%)
Stddev system-110 7.8213 ( 0.00%) 6.6034 ( 15.57%)
Stddev system-141 7.2452 ( 0.00%) 6.5108 ( 10.14%)
Stddev system-160 6.9865 ( 0.00%) 6.5068 ( 6.87%)
Stddev elapsed-1 0.0778 ( 0.00%) 0.0819 ( -5.36%)
Stddev elapsed-4 0.0666 ( 0.00%) 0.0470 ( 29.39%)
Stddev elapsed-7 0.1404 ( 0.00%) 0.1481 ( -5.50%)
Stddev elapsed-12 0.1691 ( 0.00%) 0.1135 ( 32.90%)
Stddev elapsed-21 0.2723 ( 0.00%) 0.1641 ( 39.75%)
Stddev elapsed-30 0.2001 ( 0.00%) 0.2761 (-37.99%)
Stddev elapsed-48 0.2472 ( 0.00%) 0.2112 ( 14.55%)
Stddev elapsed-79 0.1377 ( 0.00%) 0.1097 ( 20.36%)
Stddev elapsed-110 0.0812 ( 0.00%) 0.0808 ( 0.46%)
Stddev elapsed-141 0.0544 ( 0.00%) 0.0557 ( -2.27%)
Stddev elapsed-160 0.0637 ( 0.00%) 0.0577 ( 9.37%)
CoeffVar system-1 4.0135 ( 0.00%) 4.0105 ( 0.08%)
CoeffVar system-4 5.6527 ( 0.00%) 4.1640 ( 26.34%)
CoeffVar system-7 14.6274 ( 0.00%) 14.3878 ( 1.64%)
CoeffVar system-12 14.7568 ( 0.00%) 11.3470 ( 23.11%)
CoeffVar system-21 23.5452 ( 0.00%) 15.6444 ( 33.56%)
CoeffVar system-30 18.4137 ( 0.00%) 24.7158 (-34.22%)
CoeffVar system-48 24.8964 ( 0.00%) 22.3194 ( 10.35%)
CoeffVar system-79 14.3874 ( 0.00%) 9.8502 ( 31.54%)
CoeffVar system-110 10.1521 ( 0.00%) 8.7653 ( 13.66%)
CoeffVar system-141 8.6396 ( 0.00%) 7.8264 ( 9.41%)
CoeffVar system-160 8.2327 ( 0.00%) 7.7880 ( 5.40%)
CoeffVar elapsed-1 3.7316 ( 0.00%) 3.8941 ( -4.35%)
CoeffVar elapsed-4 3.8704 ( 0.00%) 2.7235 ( 29.63%)
CoeffVar elapsed-7 8.6807 ( 0.00%) 9.1146 ( -5.00%)
CoeffVar elapsed-12 11.1714 ( 0.00%) 7.2646 ( 34.97%)
CoeffVar elapsed-21 20.0897 ( 0.00%) 10.8143 ( 46.17%)
CoeffVar elapsed-30 13.3289 ( 0.00%) 21.3093 (-59.87%)
CoeffVar elapsed-48 19.2176 ( 0.00%) 17.2536 ( 10.22%)
CoeffVar elapsed-79 11.1827 ( 0.00%) 9.1764 ( 17.94%)
CoeffVar elapsed-110 7.3416 ( 0.00%) 7.5471 ( -2.80%)
CoeffVar elapsed-141 5.1854 ( 0.00%) 5.4200 ( -4.53%)
CoeffVar elapsed-160 6.1260 ( 0.00%) 5.5631 ( 9.19%)
Max system-1 2.3000 ( 0.00%) 2.2800 ( 0.87%)
Max system-4 6.6900 ( 0.00%) 6.7000 ( -0.15%)
Max system-7 11.6800 ( 0.00%) 11.7500 ( -0.60%)
Max system-12 17.9100 ( 0.00%) 18.4500 ( -3.02%)
Max system-21 31.9700 ( 0.00%) 31.7300 ( 0.75%)
Max system-30 43.3700 ( 0.00%) 43.2700 ( 0.23%)
Max system-48 70.3200 ( 0.00%) 70.1900 ( 0.18%)
Max system-79 106.7500 ( 0.00%) 90.1600 ( 15.54%)
Max system-110 102.1600 ( 0.00%) 101.2100 ( 0.93%)
Max system-141 103.0400 ( 0.00%) 96.7000 ( 6.15%)
Max system-160 98.4800 ( 0.00%) 97.5000 ( 1.00%)
Max elapsed-1 2.4900 ( 0.00%) 2.4800 ( 0.40%)
Max elapsed-4 1.7400 ( 0.00%) 1.7500 ( -0.57%)
Max elapsed-7 1.7500 ( 0.00%) 1.7400 ( 0.57%)
Max elapsed-12 1.7400 ( 0.00%) 1.7500 ( -0.57%)
Max elapsed-21 1.7500 ( 0.00%) 1.7000 ( 2.86%)
Max elapsed-30 1.7300 ( 0.00%) 1.7100 ( 1.16%)
Max elapsed-48 1.7100 ( 0.00%) 1.7000 ( 0.58%)
Max elapsed-79 1.7000 ( 0.00%) 1.5300 ( 10.00%)
Max elapsed-110 1.3500 ( 0.00%) 1.3600 ( -0.74%)
Max elapsed-141 1.1900 ( 0.00%) 1.1800 ( 0.84%)
Max elapsed-160 1.2300 ( 0.00%) 1.2300 ( 0.00%)pft faults
4.3.0 4.3.0
13672-gc669d35 13708-g8e488c2
Min faults/cpu-1 167962.7340 ( 0.00%) 169258.4000 ( 0.77%)
Min faults/cpu-4 60713.3780 ( 0.00%) 60626.9530 ( -0.14%)
Min faults/cpu-7 35103.6480 ( 0.00%) 34927.7700 ( -0.50%)
Min faults/cpu-12 22960.1440 ( 0.00%) 22309.4670 ( -2.83%)
Min faults/cpu-21 12900.6920 ( 0.00%) 13012.9840 ( 0.87%)
Min faults/cpu-30 9525.1380 ( 0.00%) 9555.3660 ( 0.32%)
Min faults/cpu-48 5893.4010 ( 0.00%) 5906.2790 ( 0.22%)
Min faults/cpu-79 3888.9070 ( 0.00%) 4603.1070 ( 18.37%)
Min faults/cpu-110 4062.4060 ( 0.00%) 4100.8300 ( 0.95%)
Min faults/cpu-141 4028.4030 ( 0.00%) 4292.1770 ( 6.55%)
Min faults/cpu-160 4212.2100 ( 0.00%) 4256.9010 ( 1.06%)
Min faults/sec-1 167817.6050 ( 0.00%) 168734.8460 ( 0.55%)
Min faults/sec-4 239459.5830 ( 0.00%) 238660.6900 ( -0.33%)
Min faults/sec-7 238787.6430 ( 0.00%) 239458.0500 ( 0.28%)
Min faults/sec-12 240464.0970 ( 0.00%) 238214.1180 ( -0.94%)
Min faults/sec-21 238517.9660 ( 0.00%) 245078.0380 ( 2.75%)
Min faults/sec-30 241525.0900 ( 0.00%) 244625.2980 ( 1.28%)
Min faults/sec-48 244673.7080 ( 0.00%) 245524.4950 ( 0.35%)
Min faults/sec-79 246298.3920 ( 0.00%) 273087.1550 ( 10.88%)
Min faults/sec-110 309127.0460 ( 0.00%) 306026.2640 ( -1.00%)
Min faults/sec-141 351861.9470 ( 0.00%) 353729.4440 ( 0.53%)
Min faults/sec-160 338721.8730 ( 0.00%) 339156.8390 ( 0.13%)
Hmean faults/cpu-1 201043.7845 ( 0.00%) 198997.0168 ( -1.02%)
Hmean faults/cpu-4 61948.9837 ( 0.00%) 61760.6204 ( -0.30%)
Hmean faults/cpu-7 40535.5843 ( 0.00%) 40239.4941 ( -0.73%)
Hmean faults/cpu-12 27836.2410 ( 0.00%) 26689.3711 ( -4.12%)
Hmean faults/cpu-21 18488.2448 ( 0.00%) 16271.0796 (-11.99%)
Hmean faults/cpu-30 11692.2366 ( 0.00%) 13790.2478 ( 17.94%)
Hmean faults/cpu-48 8998.8424 ( 0.00%) 9697.2403 ( 7.76%)
Hmean faults/cpu-79 6298.5336 ( 0.00%) 6516.3831 ( 3.46%)
Hmean faults/cpu-110 5375.1056 ( 0.00%) 5497.3005 ( 2.27%)
Hmean faults/cpu-141 4942.4573 ( 0.00%) 4983.7162 ( 0.83%)
Hmean faults/cpu-160 4884.7553 ( 0.00%) 4962.4938 ( 1.59%)
Hmean faults/sec-1 200497.7126 ( 0.00%) 198516.6748 ( -0.99%)
Hmean faults/sec-4 242883.1475 ( 0.00%) 241973.6927 ( -0.37%)
Hmean faults/sec-7 258232.4571 ( 0.00%) 256973.3980 ( -0.49%)
Hmean faults/sec-12 275982.6468 ( 0.00%) 267524.8543 ( -3.06%)
Hmean faults/sec-21 308067.0715 ( 0.00%) 275282.8535 (-10.64%)
Hmean faults/sec-30 278272.5295 ( 0.00%) 322318.2277 ( 15.83%)
Hmean faults/sec-48 324712.2770 ( 0.00%) 341047.4148 ( 5.03%)
Hmean faults/sec-79 339270.7201 ( 0.00%) 349595.9658 ( 3.04%)
Hmean faults/sec-110 377555.4863 ( 0.00%) 389982.2628 ( 3.29%)
Hmean faults/sec-141 398205.3191 ( 0.00%) 407199.2710 ( 2.26%)
Hmean faults/sec-160 402260.8158 ( 0.00%) 402662.2649 ( 0.10%)
Stddev faults/cpu-1 6579.0174 ( 0.00%) 6939.5357 ( -5.48%)
Stddev faults/cpu-4 5081.5381 ( 0.00%) 3729.6536 ( 26.60%)
Stddev faults/cpu-7 6863.0849 ( 0.00%) 6915.7844 ( -0.77%)
Stddev faults/cpu-12 4949.4091 ( 0.00%) 3301.5541 ( 33.29%)
Stddev faults/cpu-21 4556.4181 ( 0.00%) 3091.1372 ( 32.16%)
Stddev faults/cpu-30 2605.7291 ( 0.00%) 3308.9108 (-26.99%)
Stddev faults/cpu-48 2074.9654 ( 0.00%) 1755.5029 ( 15.40%)
Stddev faults/cpu-79 755.4743 ( 0.00%) 570.7272 ( 24.45%)
Stddev faults/cpu-110 518.6887 ( 0.00%) 457.2601 ( 11.84%)
Stddev faults/cpu-141 420.8814 ( 0.00%) 399.1634 ( 5.16%)
Stddev faults/cpu-160 409.1068 ( 0.00%) 389.8045 ( 4.72%)
Stddev faults/sec-1 6547.5672 ( 0.00%) 6924.4985 ( -5.76%)
Stddev faults/sec-4 12454.6505 ( 0.00%) 8746.6168 ( 29.77%)
Stddev faults/sec-7 25389.6889 ( 0.00%) 29485.1401 (-16.13%)
Stddev faults/sec-12 40255.2835 ( 0.00%) 21610.2497 ( 46.32%)
Stddev faults/sec-21 71662.3000 ( 0.00%) 36438.3667 ( 49.15%)
Stddev faults/sec-30 47102.8263 ( 0.00%) 74717.8822 (-58.63%)
Stddev faults/sec-48 66397.4011 ( 0.00%) 58784.5253 ( 11.47%)
Stddev faults/sec-79 37604.4314 ( 0.00%) 33859.6768 ( 9.96%)
Stddev faults/sec-110 27245.1710 ( 0.00%) 28753.8621 ( -5.54%)
Stddev faults/sec-141 20904.0801 ( 0.00%) 21653.4136 ( -3.58%)
Stddev faults/sec-160 25264.6583 ( 0.00%) 22119.1868 ( 12.45%)
CoeffVar faults/cpu-1 3.2684 ( 0.00%) 3.4826 ( -6.55%)
CoeffVar faults/cpu-4 8.1660 ( 0.00%) 6.0243 ( 26.23%)
CoeffVar faults/cpu-7 16.5350 ( 0.00%) 16.7871 ( -1.52%)
CoeffVar faults/cpu-12 17.3398 ( 0.00%) 12.2030 ( 29.62%)
CoeffVar faults/cpu-21 23.3188 ( 0.00%) 18.4598 ( 20.84%)
CoeffVar faults/cpu-30 21.4245 ( 0.00%) 22.6766 ( -5.84%)
CoeffVar faults/cpu-48 21.8405 ( 0.00%) 17.4212 ( 20.23%)
CoeffVar faults/cpu-79 11.7968 ( 0.00%) 8.6851 ( 26.38%)
CoeffVar faults/cpu-110 9.5579 ( 0.00%) 8.2588 ( 13.59%)
CoeffVar faults/cpu-141 8.4542 ( 0.00%) 7.9601 ( 5.85%)
CoeffVar faults/cpu-160 8.3184 ( 0.00%) 7.8078 ( 6.14%)
CoeffVar faults/sec-1 3.2617 ( 0.00%) 3.4834 ( -6.80%)
CoeffVar faults/sec-4 5.1176 ( 0.00%) 3.6111 ( 29.44%)
CoeffVar faults/sec-7 9.7495 ( 0.00%) 11.3571 (-16.49%)
CoeffVar faults/sec-12 14.3571 ( 0.00%) 8.0310 ( 44.06%)
CoeffVar faults/sec-21 22.2373 ( 0.00%) 13.0512 ( 41.31%)
CoeffVar faults/sec-30 16.5586 ( 0.00%) 22.1067 (-33.51%)
CoeffVar faults/sec-48 19.6920 ( 0.00%) 16.7508 ( 14.94%)
CoeffVar faults/sec-79 10.9508 ( 0.00%) 9.6012 ( 12.32%)
CoeffVar faults/sec-110 7.1785 ( 0.00%) 7.3326 ( -2.15%)
CoeffVar faults/sec-141 5.2355 ( 0.00%) 5.3024 ( -1.28%)
CoeffVar faults/sec-160 6.2566 ( 0.00%) 5.4766 ( 12.47%)
Max faults/cpu-1 206718.3130 ( 0.00%) 205567.6440 ( -0.56%)
Max faults/cpu-4 95165.9670 ( 0.00%) 94731.2700 ( -0.46%)
Max faults/cpu-7 61266.2550 ( 0.00%) 68833.4340 ( 12.35%)
Max faults/cpu-12 43476.9210 ( 0.00%) 37082.6580 (-14.71%)
Max faults/cpu-21 26492.9770 ( 0.00%) 25181.4640 ( -4.95%)
Max faults/cpu-30 18445.4910 ( 0.00%) 19158.5420 ( 3.87%)
Max faults/cpu-48 13377.7470 ( 0.00%) 13082.6810 ( -2.21%)
Max faults/cpu-79 7442.6860 ( 0.00%) 7615.1420 ( 2.32%)
Max faults/cpu-110 6503.6080 ( 0.00%) 6687.2100 ( 2.82%)
Max faults/cpu-141 6188.6650 ( 0.00%) 6146.9070 ( -0.67%)
Max faults/cpu-160 5950.8970 ( 0.00%) 5871.3010 ( -1.34%)
Max faults/sec-1 205979.4530 ( 0.00%) 204798.6030 ( -0.57%)
Max faults/sec-4 322267.9360 ( 0.00%) 319804.0130 ( -0.76%)
Max faults/sec-7 338997.6600 ( 0.00%) 422940.9730 ( 24.76%)
Max faults/sec-12 477630.1990 ( 0.00%) 361992.2860 (-24.21%)
Max faults/sec-21 460652.0560 ( 0.00%) 414704.3010 ( -9.97%)
Max faults/sec-30 481274.4730 ( 0.00%) 479694.9450 ( -0.33%)
Max faults/sec-48 478355.8440 ( 0.00%) 475307.0260 ( -0.64%)
Max faults/sec-79 461884.5660 ( 0.00%) 473182.5810 ( 2.45%)
Max faults/sec-110 441502.3200 ( 0.00%) 462415.9300 ( 4.74%)
Max faults/sec-141 473293.5230 ( 0.00%) 447330.8000 ( -5.49%)
Max faults/sec-160 475931.1320 ( 0.00%) 453387.9520 ( -4.74%) 4.3.0 4.3.0
13672-gc669d3513708-g8e488c2User 2202.96 2188.14
System 43841.77 43231.29
Elapsed 1363.37 1355.71
4.3.0 4.3.0
13672-gc669d3513708-g8e488c2
Minor Faults 370404167 370419290
Major Faults 0 32
Swap Ins 0 0
Swap Outs 0 0
Allocation stalls 0 0
DMA allocs 368416400 368424912
DMA32 allocs 0 0
Normal allocs 0 0
Movable allocs 0 0
Direct pages scanned 0 0
Kswapd pages scanned 0 0
Kswapd pages reclaimed 0 0
Direct pages reclaimed 0 0Kswapd efficiency 100% 100%
Kswapd velocity 0.000 0.000
Direct efficiency 100% 100%
Direct velocity 0.000 0.000
Percentage direct scans 0% 0%
Zone normal velocity 0.000 0.000 Zone dma32 velocity 0.000 0.000 Zone dma velocity 0.000 0.000 Page writes by reclaim 0.000 0.000 Page writes file 0 0 Page writes anon 0 0 Page reclaim immediate 0 0 Sector Reads 168 3696 Sector Writes 7396 9136 Page rescued immediate 0 0 Slabs scanned 0 0 Direct inode steals 0 0 Kswapd inode steals 0 0 Kswapd skipped wait 0 0 THP fault alloc 0 0 THP collapse alloc 0 0 THP splits 0 0 THP fault fallback 0 0 THP collapse fail 0 0 Compaction stalls 0 0 Compaction success 0 0 Compaction failures 0 0 Page migrate success 452 442 Page migrate failure 0 0 Compaction pages isolated 0 0 Compaction migrate scanned 0 0 Compaction free scanned 0 0 Compaction cost 0 0 NUMA alloc hit 368352412 368362898 NUMA alloc miss 0 0 NUMA interleave hit 0 0
NUMA alloc local 249887344 239441708
NUMA base PTE updates 30421298 30252358
NUMA huge PMD updates 0 0
NUMA page range updates 30421298 30252358
NUMA hint faults 39412 34328 NUMA hint local faults 28543 25263
NUMA hint local percent 72 73
NUMA pages migrated 452 442
AutoNUMA cost 410% 383%
ebizzy Overall Throughput
4.3.0 4.3.0
13672-gc669d35 13708-g8e488c2
Min Rsec-1 30631.00 ( 0.00%) 27088.00 (-11.57%)
Min Rsec-4 57821.00 ( 0.00%) 47893.00 (-17.17%)
Min Rsec-7 64229.00 ( 0.00%) 62025.00 ( -3.43%)
Min Rsec-12 75825.00 ( 0.00%) 83296.00 ( 9.85%)
Min Rsec-21 81774.00 ( 0.00%) 88602.00 ( 8.35%)
Min Rsec-30 92366.00 ( 0.00%) 96488.00 ( 4.46%)
Min Rsec-48 98346.00 ( 0.00%) 102081.00 ( 3.80%)
Min Rsec-79 107800.00 ( 0.00%) 107806.00 ( 0.01%)
Min Rsec-110 115001.00 ( 0.00%) 114767.00 ( -0.20%)
Min Rsec-141 118423.00 ( 0.00%) 120113.00 ( 1.43%)
Min Rsec-172 128361.00 ( 0.00%) 126839.00 ( -1.19%)
Min Rsec-203 128782.00 ( 0.00%) 124506.00 ( -3.32%)
Min Rsec-234 129973.00 ( 0.00%) 125048.00 ( -3.79%)
Min Rsec-265 129718.00 ( 0.00%) 124645.00 ( -3.91%)
Min Rsec-296 131203.00 ( 0.00%) 124085.00 ( -5.43%)
Min Rsec-327 136167.00 ( 0.00%) 93782.00 (-31.13%)
Min Rsec-358 132593.00 ( 0.00%) 126162.00 ( -4.85%)
Min Rsec-389 133791.00 ( 0.00%) 125713.00 ( -6.04%)
Min Rsec-420 132244.00 ( 0.00%) 125662.00 ( -4.98%)
Min Rsec-451 130741.00 ( 0.00%) 131577.00 ( 0.64%)
Min Rsec-482 132991.00 ( 0.00%) 133038.00 ( 0.04%)
Min Rsec-513 135247.00 ( 0.00%) 132752.00 ( -1.84%)
Min Rsec-544 133541.00 ( 0.00%) 138052.00 ( 3.38%)
Min Rsec-575 134361.00 ( 0.00%) 140407.00 ( 4.50%)
Min Rsec-606 136566.00 ( 0.00%) 140320.00 ( 2.75%)
Min Rsec-637 136596.00 ( 0.00%) 99349.00 (-27.27%)
Min Rsec-640 101422.00 ( 0.00%) 99561.00 ( -1.83%)
Hmean Rsec-1 30824.32 ( 0.00%) 28676.70 ( -6.97%)
Hmean Rsec-4 64157.06 ( 0.00%) 62294.92 ( -2.90%)
Hmean Rsec-7 65651.01 ( 0.00%) 69718.38 ( 6.20%)
Hmean Rsec-12 78873.91 ( 0.00%) 91860.80 ( 16.47%)
Hmean Rsec-21 86468.13 ( 0.00%) 96639.75 ( 11.76%)
Hmean Rsec-30 95530.97 ( 0.00%) 103998.21 ( 8.86%)
Hmean Rsec-48 103374.48 ( 0.00%) 108060.90 ( 4.53%)
Hmean Rsec-79 110468.40 ( 0.00%) 111625.07 ( 1.05%)
Hmean Rsec-110 117068.12 ( 0.00%) 117277.05 ( 0.18%)
Hmean Rsec-141 122303.78 ( 0.00%) 123704.79 ( 1.15%)
Hmean Rsec-172 129919.04 ( 0.00%) 127485.39 ( -1.87%)
Hmean Rsec-203 129848.20 ( 0.00%) 126836.53 ( -2.32%)
Hmean Rsec-234 131271.19 ( 0.00%) 126981.96 ( -3.27%)
Hmean Rsec-265 130786.79 ( 0.00%) 126195.37 ( -3.51%)
Hmean Rsec-296 136679.24 ( 0.00%) 125519.14 ( -8.17%)
Hmean Rsec-327 136987.94 ( 0.00%) 118013.15 (-13.85%)
Hmean Rsec-358 134069.52 ( 0.00%) 126610.36 ( -5.56%)
Hmean Rsec-389 134492.50 ( 0.00%) 126879.43 ( -5.66%)
Hmean Rsec-420 134001.91 ( 0.00%) 127656.60 ( -4.74%)
Hmean Rsec-451 133772.89 ( 0.00%) 133757.64 ( -0.01%)
Hmean Rsec-482 135542.90 ( 0.00%) 135620.98 ( 0.06%)
Hmean Rsec-513 137292.43 ( 0.00%) 135231.99 ( -1.50%)
Hmean Rsec-544 135299.06 ( 0.00%) 141228.12 ( 4.38%)
Hmean Rsec-575 137473.02 ( 0.00%) 140835.74 ( 2.45%)
Hmean Rsec-606 137456.90 ( 0.00%) 143092.13 ( 4.10%)
Hmean Rsec-637 137146.14 ( 0.00%) 129751.36 ( -5.39%)
Hmean Rsec-640 131512.67 ( 0.00%) 129329.67 ( -1.66%)
Stddev Rsec-1 144.49 ( 0.00%) 1292.15 (-794.30%)
Stddev Rsec-4 4005.88 ( 0.00%) 8292.35 (-107.00%)
Stddev Rsec-7 959.09 ( 0.00%) 5545.25 (-478.18%)
Stddev Rsec-12 2728.12 ( 0.00%) 7819.40 (-186.62%)
Stddev Rsec-21 5449.55 ( 0.00%) 8175.21 (-50.02%)
Stddev Rsec-30 2577.87 ( 0.00%) 4946.96 (-91.90%)
Stddev Rsec-48 4972.04 ( 0.00%) 4285.92 ( 13.80%)
Stddev Rsec-79 1860.85 ( 0.00%) 3294.02 (-77.02%)
Stddev Rsec-110 2042.24 ( 0.00%) 1385.94 ( 32.14%)
Stddev Rsec-141 2275.70 ( 0.00%) 2258.88 ( 0.74%)
Stddev Rsec-172 1003.21 ( 0.00%) 679.57 ( 32.26%)
Stddev Rsec-203 623.13 ( 0.00%) 1220.15 (-95.81%)
Stddev Rsec-234 723.22 ( 0.00%) 1138.08 (-57.36%)
Stddev Rsec-265 742.86 ( 0.00%) 1321.18 (-77.85%)
Stddev Rsec-296 2939.87 ( 0.00%) 993.39 ( 66.21%)
Stddev Rsec-327 604.06 ( 0.00%) 12956.36 (-2044.87%)
Stddev Rsec-358 1390.12 ( 0.00%) 285.62 ( 79.45%)
Stddev Rsec-389 847.67 ( 0.00%) 855.57 ( -0.93%)
Stddev Rsec-420 1151.99 ( 0.00%) 1619.63 (-40.59%)
Stddev Rsec-451 1570.06 ( 0.00%) 1573.54 ( -0.22%)
Stddev Rsec-482 1509.33 ( 0.00%) 1984.46 (-31.48%)
Stddev Rsec-513 1480.83 ( 0.00%) 1873.63 (-26.53%)
Stddev Rsec-544 1470.26 ( 0.00%) 3914.73 (-166.26%)
Stddev Rsec-575 1898.92 ( 0.00%) 613.86 ( 67.67%)
Stddev Rsec-606 694.98 ( 0.00%) 2642.72 (-280.26%)
Stddev Rsec-637 670.80 ( 0.00%) 16535.67 (-2365.07%)
Stddev Rsec-640 16352.20 ( 0.00%) 16160.95 ( 1.17%)
CoeffVar Rsec-1 0.47 ( 0.00%) 4.50 (-859.37%)
CoeffVar Rsec-4 6.22 ( 0.00%) 13.04 (-109.75%)
CoeffVar Rsec-7 1.46 ( 0.00%) 7.90 (-441.04%)
CoeffVar Rsec-12 3.45 ( 0.00%) 8.45 (-144.72%)
CoeffVar Rsec-21 6.28 ( 0.00%) 8.40 (-33.83%)
CoeffVar Rsec-30 2.70 ( 0.00%) 4.75 (-76.00%)
CoeffVar Rsec-48 4.80 ( 0.00%) 3.96 ( 17.49%)
CoeffVar Rsec-79 1.68 ( 0.00%) 2.95 (-75.08%)
CoeffVar Rsec-110 1.74 ( 0.00%) 1.18 ( 32.25%)
CoeffVar Rsec-141 1.86 ( 0.00%) 1.83 ( 1.86%)
CoeffVar Rsec-172 0.77 ( 0.00%) 0.53 ( 30.97%)
CoeffVar Rsec-203 0.48 ( 0.00%) 0.96 (-100.45%)
CoeffVar Rsec-234 0.55 ( 0.00%) 0.90 (-62.67%)
CoeffVar Rsec-265 0.57 ( 0.00%) 1.05 (-84.31%)
CoeffVar Rsec-296 2.15 ( 0.00%) 0.79 ( 63.19%)
CoeffVar Rsec-327 0.44 ( 0.00%) 10.83 (-2354.96%)
CoeffVar Rsec-358 1.04 ( 0.00%) 0.23 ( 78.24%)
CoeffVar Rsec-389 0.63 ( 0.00%) 0.67 ( -6.99%)
CoeffVar Rsec-420 0.86 ( 0.00%) 1.27 (-47.57%)
CoeffVar Rsec-451 1.17 ( 0.00%) 1.18 ( -0.23%)
CoeffVar Rsec-482 1.11 ( 0.00%) 1.46 (-31.39%)
CoeffVar Rsec-513 1.08 ( 0.00%) 1.39 (-28.44%)
CoeffVar Rsec-544 1.09 ( 0.00%) 2.77 (-154.92%)
CoeffVar Rsec-575 1.38 ( 0.00%) 0.44 ( 68.44%)
CoeffVar Rsec-606 0.51 ( 0.00%) 1.85 (-265.17%)
CoeffVar Rsec-637 0.49 ( 0.00%) 12.50 (-2455.70%)
CoeffVar Rsec-640 12.21 ( 0.00%) 12.27 ( -0.48%)
Max Rsec-1 31000.00 ( 0.00%) 30276.00 ( -2.34%)
Max Rsec-4 69454.00 ( 0.00%) 71739.00 ( 3.29%)
Max Rsec-7 67182.00 ( 0.00%) 75573.00 ( 12.49%)
Max Rsec-12 82739.00 ( 0.00%) 106341.00 ( 28.53%)
Max Rsec-21 95156.00 ( 0.00%) 110486.00 ( 16.11%)
Max Rsec-30 98859.00 ( 0.00%) 109381.00 ( 10.64%)
Max Rsec-48 113061.00 ( 0.00%) 114346.00 ( 1.14%)
Max Rsec-79 113578.00 ( 0.00%) 117539.00 ( 3.49%)
Max Rsec-110 119618.00 ( 0.00%) 118806.00 ( -0.68%)
Max Rsec-141 125009.00 ( 0.00%) 126417.00 ( 1.13%)
Max Rsec-172 131250.00 ( 0.00%) 128709.00 ( -1.94%)
Max Rsec-203 130718.00 ( 0.00%) 128025.00 ( -2.06%)
Max Rsec-234 131856.00 ( 0.00%) 128528.00 ( -2.52%)
Max Rsec-265 132036.00 ( 0.00%) 128059.00 ( -3.01%)
Max Rsec-296 139572.00 ( 0.00%) 126819.00 ( -9.14%)
Max Rsec-327 138043.00 ( 0.00%) 126581.00 ( -8.30%)
Max Rsec-358 136616.00 ( 0.00%) 126986.00 ( -7.05%)
Max Rsec-389 136122.00 ( 0.00%) 128050.00 ( -5.93%)
Max Rsec-420 135873.00 ( 0.00%) 129725.00 ( -4.52%)
Max Rsec-451 134918.00 ( 0.00%) 135687.00 ( 0.57%)
Max Rsec-482 137574.00 ( 0.00%) 138546.00 ( 0.71%)
Max Rsec-513 139144.00 ( 0.00%) 138358.00 ( -0.56%)
Max Rsec-544 137415.00 ( 0.00%) 148411.00 ( 8.00%)
Max Rsec-575 140015.00 ( 0.00%) 142036.00 ( 1.44%)
Max Rsec-606 138658.00 ( 0.00%) 148103.00 ( 6.81%)
Max Rsec-637 138429.00 ( 0.00%) 143293.00 ( 3.51%)
Max Rsec-640 144351.00 ( 0.00%) 141507.00 ( -1.97%)ebizzy Per-thread
4.3.0 4.3.0
13672-gc669d35 13708-g8e488c2
Min Rsec-1 30631.00 ( 0.00%) 27088.00 (-11.57%)
Min Rsec-4 12735.00 ( 0.00%) 11127.00 (-12.63%)
Min Rsec-7 7954.00 ( 0.00%) 7116.00 (-10.54%)
Min Rsec-12 4721.00 ( 0.00%) 5315.00 ( 12.58%)
Min Rsec-21 2410.00 ( 0.00%) 2304.00 ( -4.40%)
Min Rsec-30 1931.00 ( 0.00%) 2093.00 ( 8.39%)
Min Rsec-48 1451.00 ( 0.00%) 1403.00 ( -3.31%)
Min Rsec-79 821.00 ( 0.00%) 715.00 (-12.91%)
Min Rsec-110 647.00 ( 0.00%) 668.00 ( 3.25%)
Min Rsec-141 477.00 ( 0.00%) 495.00 ( 3.77%)
Min Rsec-172 463.00 ( 0.00%) 452.00 ( -2.38%)
Min Rsec-203 367.00 ( 0.00%) 314.00 (-14.44%)
Min Rsec-234 304.00 ( 0.00%) 290.00 ( -4.61%)
Min Rsec-265 260.00 ( 0.00%) 244.00 ( -6.15%)
Min Rsec-296 211.00 ( 0.00%) 205.00 ( -2.84%)
Min Rsec-327 185.00 ( 0.00%) 152.00 (-17.84%)
Min Rsec-358 170.00 ( 0.00%) 171.00 ( 0.59%)
Min Rsec-389 134.00 ( 0.00%) 139.00 ( 3.73%)
Min Rsec-420 153.00 ( 0.00%) 159.00 ( 3.92%)
Min Rsec-451 133.00 ( 0.00%) 135.00 ( 1.50%)
Min Rsec-482 127.00 ( 0.00%) 127.00 ( 0.00%)
Min Rsec-513 120.00 ( 0.00%) 130.00 ( 8.33%)
Min Rsec-544 107.00 ( 0.00%) 101.00 ( -5.61%)
Min Rsec-575 94.00 ( 0.00%) 114.00 ( 21.28%)
Min Rsec-606 94.00 ( 0.00%) 127.00 ( 35.11%)
Min Rsec-637 101.00 ( 0.00%) 101.00 ( 0.00%)
Min Rsec-640 81.00 ( 0.00%) 90.00 ( 11.11%)
Hmean Rsec-1 30824.32 ( 0.00%) 28676.70 ( -6.97%)
Hmean Rsec-4 15947.39 ( 0.00%) 15456.86 ( -3.08%)
Hmean Rsec-7 9303.67 ( 0.00%) 9820.37 ( 5.55%)
Hmean Rsec-12 6388.57 ( 0.00%) 7416.20 ( 16.09%)
Hmean Rsec-21 3321.63 ( 0.00%) 3634.83 ( 9.43%)
Hmean Rsec-30 2815.97 ( 0.00%) 3019.42 ( 7.22%)
Hmean Rsec-48 1981.55 ( 0.00%) 1982.90 ( 0.07%)
Hmean Rsec-79 1276.93 ( 0.00%) 1287.34 ( 0.82%)
Hmean Rsec-110 956.47 ( 0.00%) 985.88 ( 3.08%)
Hmean Rsec-141 783.13 ( 0.00%) 783.22 ( 0.01%)
Hmean Rsec-172 689.92 ( 0.00%) 667.13 ( -3.30%)
Hmean Rsec-203 587.77 ( 0.00%) 556.88 ( -5.26%)
Hmean Rsec-234 506.71 ( 0.00%) 504.36 ( -0.46%)
Hmean Rsec-265 446.66 ( 0.00%) 432.97 ( -3.06%)
Hmean Rsec-296 412.77 ( 0.00%) 378.81 ( -8.23%)
Hmean Rsec-327 359.13 ( 0.00%) 315.30 (-12.20%)
Hmean Rsec-358 335.46 ( 0.00%) 312.77 ( -6.76%)
Hmean Rsec-389 291.33 ( 0.00%) 281.59 ( -3.34%)
Hmean Rsec-420 289.74 ( 0.00%) 278.40 ( -3.91%)
Hmean Rsec-451 255.18 ( 0.00%) 263.41 ( 3.22%)
Hmean Rsec-482 241.82 ( 0.00%) 253.38 ( 4.78%)
Hmean Rsec-513 237.84 ( 0.00%) 233.55 ( -1.81%)
Hmean Rsec-544 219.99 ( 0.00%) 216.96 ( -1.38%)
Hmean Rsec-575 205.77 ( 0.00%) 216.15 ( 5.04%)
Hmean Rsec-606 199.08 ( 0.00%) 215.75 ( 8.37%)
Hmean Rsec-637 183.93 ( 0.00%) 190.02 ( 3.31%)
Hmean Rsec-640 183.38 ( 0.00%) 184.17 ( 0.43%)
Stddev Rsec-1 144.49 ( 0.00%) 1292.15 (794.30%)
Stddev Rsec-4 1525.43 ( 0.00%) 2515.03 ( 64.87%)
Stddev Rsec-7 853.17 ( 0.00%) 1373.82 ( 61.02%)
Stddev Rsec-12 1178.82 ( 0.00%) 1543.19 ( 30.91%)
Stddev Rsec-21 2459.94 ( 0.00%) 2892.79 ( 17.60%)
Stddev Rsec-30 1585.28 ( 0.00%) 1903.96 ( 20.10%)
Stddev Rsec-48 972.90 ( 0.00%) 1219.25 ( 25.32%)
Stddev Rsec-79 629.60 ( 0.00%) 637.97 ( 1.33%)
Stddev Rsec-110 529.71 ( 0.00%) 417.64 (-21.16%)
Stddev Rsec-141 421.55 ( 0.00%) 452.80 ( 7.41%)
Stddev Rsec-172 341.05 ( 0.00%) 375.21 ( 10.02%)
Stddev Rsec-203 249.78 ( 0.00%) 281.78 ( 12.81%)
Stddev Rsec-234 244.43 ( 0.00%) 193.14 (-20.98%)
Stddev Rsec-265 186.43 ( 0.00%) 196.87 ( 5.60%)
Stddev Rsec-296 192.08 ( 0.00%) 184.98 ( -3.70%)
Stddev Rsec-327 186.84 ( 0.00%) 161.37 (-13.63%)
Stddev Rsec-358 150.51 ( 0.00%) 147.22 ( -2.19%)
Stddev Rsec-389 163.74 ( 0.00%) 139.43 (-14.85%)
Stddev Rsec-420 123.57 ( 0.00%) 112.75 ( -8.76%)
Stddev Rsec-451 130.29 ( 0.00%) 127.71 ( -1.98%)
Stddev Rsec-482 130.08 ( 0.00%) 109.39 (-15.90%)
Stddev Rsec-513 109.04 ( 0.00%) 105.78 ( -2.99%)
Stddev Rsec-544 100.31 ( 0.00%) 126.31 ( 25.92%)
Stddev Rsec-575 102.79 ( 0.00%) 98.62 ( -4.06%)
Stddev Rsec-606 93.57 ( 0.00%) 82.75 (-11.57%)
Stddev Rsec-637 102.32 ( 0.00%) 67.40 (-34.13%)
Stddev Rsec-640 78.48 ( 0.00%) 89.02 ( 13.43%)
CoeffVar Rsec-1 0.47 ( 0.00%) 4.50 (-859.37%)
CoeffVar Rsec-4 9.47 ( 0.00%) 15.82 (-67.06%)
CoeffVar Rsec-7 9.10 ( 0.00%) 13.71 (-50.68%)
CoeffVar Rsec-12 17.91 ( 0.00%) 20.02 (-11.77%)
CoeffVar Rsec-21 59.52 ( 0.00%) 62.44 ( -4.91%)
CoeffVar Rsec-30 49.75 ( 0.00%) 54.80 (-10.15%)
CoeffVar Rsec-48 45.09 ( 0.00%) 54.08 (-19.96%)
CoeffVar Rsec-79 45.03 ( 0.00%) 45.13 ( -0.22%)
CoeffVar Rsec-110 49.78 ( 0.00%) 39.18 ( 21.28%)
CoeffVar Rsec-141 48.61 ( 0.00%) 51.62 ( -6.20%)
CoeffVar Rsec-172 45.18 ( 0.00%) 50.65 (-12.12%)
CoeffVar Rsec-203 39.08 ( 0.00%) 45.13 (-15.48%)
CoeffVar Rsec-234 43.61 ( 0.00%) 35.62 ( 18.32%)
CoeffVar Rsec-265 37.81 ( 0.00%) 41.38 ( -9.44%)
CoeffVar Rsec-296 41.62 ( 0.00%) 43.67 ( -4.92%)
CoeffVar Rsec-327 44.65 ( 0.00%) 44.15 ( 1.13%)
CoeffVar Rsec-358 40.24 ( 0.00%) 41.68 ( -3.59%)
CoeffVar Rsec-389 47.42 ( 0.00%) 42.81 ( 9.73%)
CoeffVar Rsec-420 38.79 ( 0.00%) 37.15 ( 4.23%)
CoeffVar Rsec-451 43.99 ( 0.00%) 43.13 ( 1.96%)
CoeffVar Rsec-482 46.33 ( 0.00%) 38.94 ( 15.96%)
CoeffVar Rsec-513 40.81 ( 0.00%) 40.19 ( 1.52%)
CoeffVar Rsec-544 40.41 ( 0.00%) 48.71 (-20.54%)
CoeffVar Rsec-575 43.07 ( 0.00%) 40.34 ( 6.33%)
CoeffVar Rsec-606 41.34 ( 0.00%) 35.10 ( 15.08%)
CoeffVar Rsec-637 47.63 ( 0.00%) 32.53 ( 31.70%)
CoeffVar Rsec-640 37.58 ( 0.00%) 43.34 (-15.33%)
Max Rsec-1 31000.00 ( 0.00%) 30276.00 ( -2.34%)
Max Rsec-4 18406.00 ( 0.00%) 19844.00 ( 7.81%)
Max Rsec-7 11200.00 ( 0.00%) 12913.00 ( 15.29%)
Max Rsec-12 10612.00 ( 0.00%) 11465.00 ( 8.04%)
Max Rsec-21 15254.00 ( 0.00%) 15490.00 ( 1.55%)
Max Rsec-30 10529.00 ( 0.00%) 11822.00 ( 12.28%)
Max Rsec-48 8134.00 ( 0.00%) 8236.00 ( 1.25%)
Max Rsec-79 5105.00 ( 0.00%) 5945.00 ( 16.45%)
Max Rsec-110 4728.00 ( 0.00%) 4371.00 ( -7.55%)
Max Rsec-141 3628.00 ( 0.00%) 3997.00 ( 10.17%)
Max Rsec-172 3261.00 ( 0.00%) 3138.00 ( -3.77%)
Max Rsec-203 2453.00 ( 0.00%) 2995.00 ( 22.10%)
Max Rsec-234 2363.00 ( 0.00%) 2194.00 ( -7.15%)
Max Rsec-265 2038.00 ( 0.00%) 2235.00 ( 9.67%)
Max Rsec-296 2174.00 ( 0.00%) 1942.00 (-10.67%)
Max Rsec-327 1791.00 ( 0.00%) 1826.00 ( 1.95%)
Max Rsec-358 1389.00 ( 0.00%) 1538.00 ( 10.73%)
Max Rsec-389 1630.00 ( 0.00%) 1273.00 (-21.90%)
Max Rsec-420 1249.00 ( 0.00%) 1047.00 (-16.17%)
Max Rsec-451 1179.00 ( 0.00%) 1316.00 ( 11.62%)
Max Rsec-482 1187.00 ( 0.00%) 1211.00 ( 2.02%)
Max Rsec-513 1257.00 ( 0.00%) 1018.00 (-19.01%)
Max Rsec-544 1139.00 ( 0.00%) 1216.00 ( 6.76%)
Max Rsec-575 1089.00 ( 0.00%) 912.00 (-16.25%)
Max Rsec-606 903.00 ( 0.00%) 1014.00 ( 12.29%)
Max Rsec-637 1053.00 ( 0.00%) 660.00 (-37.32%)
Max Rsec-640 749.00 ( 0.00%) 1081.00 ( 44.33%)ebizzy Thread spread
4.3.0 4.3.0
13672-gc669d35 13708-g8e488c2
Min spread-1 0.00 ( 0.00%) 0.00 ( 0.00%)
Min spread-4 1122.00 ( 0.00%) 1464.00 (-30.48%)
Min spread-7 1537.00 ( 0.00%) 1949.00 (-26.81%)
Min spread-12 3466.00 ( 0.00%) 3667.00 ( -5.80%)
Min spread-21 6507.00 ( 0.00%) 6406.00 ( 1.55%)
Min spread-30 5648.00 ( 0.00%) 7543.00 (-33.55%)
Min spread-48 3531.00 ( 0.00%) 4244.00 (-20.19%)
Min spread-79 2933.00 ( 0.00%) 3184.00 ( -8.56%)
Min spread-110 2341.00 ( 0.00%) 1892.00 ( 19.18%)
Min spread-141 2084.00 ( 0.00%) 2641.00 (-26.73%)
Min spread-172 1447.00 ( 0.00%) 2001.00 (-38.29%)
Min spread-203 1268.00 ( 0.00%) 1686.00 (-32.97%)
Min spread-234 1703.00 ( 0.00%) 1157.00 ( 32.06%)
Min spread-265 1385.00 ( 0.00%) 1370.00 ( 1.08%)
Min spread-296 1139.00 ( 0.00%) 1263.00 (-10.89%)
Min spread-327 1259.00 ( 0.00%) 328.00 ( 73.95%)
Min spread-358 781.00 ( 0.00%) 1002.00 (-28.30%)
Min spread-389 746.00 ( 0.00%) 767.00 ( -2.82%)
Min spread-420 703.00 ( 0.00%) 673.00 ( 4.27%)
Min spread-451 812.00 ( 0.00%) 603.00 ( 25.74%)
Min spread-482 530.00 ( 0.00%) 685.00 (-29.25%)
Min spread-513 564.00 ( 0.00%) 456.00 ( 19.15%)
Min spread-544 530.00 ( 0.00%) 664.00 (-25.28%)
Min spread-575 497.00 ( 0.00%) 443.00 ( 10.87%)
Min spread-606 372.00 ( 0.00%) 460.00 (-23.66%)
Min spread-637 377.00 ( 0.00%) 185.00 ( 50.93%)
Min spread-640 295.00 ( 0.00%) 82.00 ( 72.20%)
Hmean spread-1 0.00 ( 0.00%) 0.00 ( 0.00%)
Hmean spread-4 2198.84 ( 0.00%) 2371.33 ( -7.84%)
Hmean spread-7 2165.67 ( 0.00%) 2771.13 (-27.96%)
Hmean spread-12 4198.82 ( 0.00%) 4507.36 ( -7.35%)
Hmean spread-21 7559.98 ( 0.00%) 8509.79 (-12.56%)
Hmean spread-30 6617.29 ( 0.00%) 8461.26 (-27.87%)
Hmean spread-48 4726.30 ( 0.00%) 5794.29 (-22.60%)
Hmean spread-79 3596.75 ( 0.00%) 3913.30 ( -8.80%)
Hmean spread-110 3210.81 ( 0.00%) 2530.22 ( 21.20%)
Hmean spread-141 2642.11 ( 0.00%) 2904.53 ( -9.93%)
Hmean spread-172 2147.18 ( 0.00%) 2374.37 (-10.58%)
Hmean spread-203 1778.54 ( 0.00%) 2008.80 (-12.95%)
Hmean spread-234 1843.86 ( 0.00%) 1385.32 ( 24.87%)
Hmean spread-265 1516.18 ( 0.00%) 1641.22 ( -8.25%)
Hmean spread-296 1396.98 ( 0.00%) 1441.88 ( -3.21%)
Hmean spread-327 1418.21 ( 0.00%) 789.92 ( 44.30%)
Hmean spread-358 940.59 ( 0.00%) 1187.35 (-26.23%)
Hmean spread-389 1050.33 ( 0.00%) 922.54 ( 12.17%)
Hmean spread-420 837.76 ( 0.00%) 772.29 ( 7.82%)
Hmean spread-451 874.68 ( 0.00%) 821.89 ( 6.04%)
Hmean spread-482 755.71 ( 0.00%) 782.04 ( -3.49%)
Hmean spread-513 716.35 ( 0.00%) 682.52 ( 4.72%)
Hmean spread-544 659.82 ( 0.00%) 807.36 (-22.36%)
Hmean spread-575 667.82 ( 0.00%) 584.47 ( 12.48%)
Hmean spread-606 547.04 ( 0.00%) 556.28 ( -1.69%)
Hmean spread-637 571.10 ( 0.00%) 340.03 ( 40.46%)
Hmean spread-640 413.62 ( 0.00%) 247.08 ( 40.26%)
Stddev spread-1 0.00 ( 0.00%) 0.00 ( 0.00%)
Stddev spread-4 1055.03 ( 0.00%) 1918.21 (-81.81%)
Stddev spread-7 675.30 ( 0.00%) 998.13 (-47.80%)
Stddev spread-12 645.46 ( 0.00%) 726.78 (-12.60%)
Stddev spread-21 2351.57 ( 0.00%) 2788.71 (-18.59%)
Stddev spread-30 1044.36 ( 0.00%) 717.65 ( 31.28%)
Stddev spread-48 1259.45 ( 0.00%) 912.70 ( 27.53%)
Stddev spread-79 445.51 ( 0.00%) 833.11 (-87.00%)
Stddev spread-110 595.82 ( 0.00%) 624.63 ( -4.84%)
Stddev spread-141 370.68 ( 0.00%) 285.28 ( 23.04%)
Stddev spread-172 472.11 ( 0.00%) 250.80 ( 46.88%)
Stddev spread-203 288.31 ( 0.00%) 393.40 (-36.45%)
Stddev spread-234 130.86 ( 0.00%) 329.63 (-151.90%)
Stddev spread-265 129.24 ( 0.00%) 196.98 (-52.41%)
Stddev spread-296 292.03 ( 0.00%) 157.32 ( 46.13%)
Stddev spread-327 136.07 ( 0.00%) 443.66 (-226.05%)
Stddev spread-358 168.12 ( 0.00%) 116.22 ( 30.87%)
Stddev spread-389 273.41 ( 0.00%) 120.46 ( 55.94%)
Stddev spread-420 124.12 ( 0.00%) 89.51 ( 27.88%)
Stddev spread-451 80.65 ( 0.00%) 220.33 (-173.18%)
Stddev spread-482 211.68 ( 0.00%) 146.76 ( 30.67%)
Stddev spread-513 202.23 ( 0.00%) 156.38 ( 22.67%)
Stddev spread-544 183.05 ( 0.00%) 149.90 ( 18.11%)
Stddev spread-575 170.16 ( 0.00%) 134.70 ( 20.84%)
Stddev spread-606 145.56 ( 0.00%) 161.55 (-10.99%)
Stddev spread-637 240.03 ( 0.00%) 109.43 ( 54.41%)
Stddev spread-640 119.87 ( 0.00%) 295.05 (-146.15%)
CoeffVar spread-1 0.00 ( 0.00%) 0.00 ( 0.00%)
CoeffVar spread-4 39.42 ( 0.00%) 60.93 ( 54.57%)
CoeffVar spread-7 28.47 ( 0.00%) 32.47 ( 14.07%)
CoeffVar spread-12 15.04 ( 0.00%) 15.75 ( 4.74%)
CoeffVar spread-21 29.16 ( 0.00%) 29.98 ( 2.80%)
CoeffVar spread-30 15.43 ( 0.00%) 8.42 (-45.42%)
CoeffVar spread-48 25.02 ( 0.00%) 15.30 (-38.85%)
CoeffVar spread-79 12.19 ( 0.00%) 20.43 ( 67.55%)
CoeffVar spread-110 17.88 ( 0.00%) 23.42 ( 30.99%)
CoeffVar spread-141 13.74 ( 0.00%) 9.73 (-29.15%)
CoeffVar spread-172 20.83 ( 0.00%) 10.44 (-49.89%)
CoeffVar spread-203 15.70 ( 0.00%) 18.94 ( 20.66%)
CoeffVar spread-234 7.06 ( 0.00%) 22.68 (221.11%)
CoeffVar spread-265 8.47 ( 0.00%) 11.83 ( 39.72%)
CoeffVar spread-296 20.16 ( 0.00%) 10.79 (-46.46%)
CoeffVar spread-327 9.51 ( 0.00%) 41.00 (331.14%)
CoeffVar spread-358 17.36 ( 0.00%) 9.69 (-44.18%)
CoeffVar spread-389 24.42 ( 0.00%) 12.84 (-47.41%)
CoeffVar spread-420 14.52 ( 0.00%) 11.43 (-21.27%)
CoeffVar spread-451 9.15 ( 0.00%) 25.12 (174.49%)
CoeffVar spread-482 26.01 ( 0.00%) 18.24 (-29.86%)
CoeffVar spread-513 26.62 ( 0.00%) 21.60 (-18.85%)
CoeffVar spread-544 26.14 ( 0.00%) 18.03 (-31.02%)
CoeffVar spread-575 24.10 ( 0.00%) 21.88 ( -9.21%)
CoeffVar spread-606 24.89 ( 0.00%) 27.30 ( 9.67%)
CoeffVar spread-637 36.24 ( 0.00%) 28.41 (-21.60%)
CoeffVar spread-640 26.97 ( 0.00%) 62.12 (130.29%)
Max spread-1 0.00 ( 0.00%) 0.00 ( 0.00%)
Max spread-4 4390.00 ( 0.00%) 6802.00 (-54.94%)
Max spread-7 3165.00 ( 0.00%) 4721.00 (-49.16%)
Max spread-12 5377.00 ( 0.00%) 5913.00 ( -9.97%)
Max spread-21 12629.00 ( 0.00%) 12796.00 ( -1.32%)
Max spread-30 8385.00 ( 0.00%) 9572.00 (-14.16%)
Max spread-48 6557.00 ( 0.00%) 6670.00 ( -1.72%)
Max spread-79 4169.00 ( 0.00%) 5230.00 (-25.45%)
Max spread-110 3997.00 ( 0.00%) 3688.00 ( 7.73%)
Max spread-141 3151.00 ( 0.00%) 3416.00 ( -8.41%)
Max spread-172 2798.00 ( 0.00%) 2686.00 ( 4.00%)
Max spread-203 2061.00 ( 0.00%) 2681.00 (-30.08%)
Max spread-234 2049.00 ( 0.00%) 1904.00 ( 7.08%)
Max spread-265 1762.00 ( 0.00%) 1952.00 (-10.78%)
Max spread-296 1963.00 ( 0.00%) 1737.00 ( 11.51%)
Max spread-327 1592.00 ( 0.00%) 1670.00 ( -4.90%)
Max spread-358 1197.00 ( 0.00%) 1366.00 (-14.12%)
Max spread-389 1486.00 ( 0.00%) 1116.00 ( 24.90%)
Max spread-420 1067.00 ( 0.00%) 879.00 ( 17.62%)
Max spread-451 1033.00 ( 0.00%) 1181.00 (-14.33%)
Max spread-482 1060.00 ( 0.00%) 1084.00 ( -2.26%)
Max spread-513 1137.00 ( 0.00%) 885.00 ( 22.16%)
Max spread-544 1032.00 ( 0.00%) 1103.00 ( -6.88%)
Max spread-575 995.00 ( 0.00%) 789.00 ( 20.70%)
Max spread-606 809.00 ( 0.00%) 887.00 ( -9.64%)
Max spread-637 951.00 ( 0.00%) 518.00 ( 45.53%)
Max spread-640 625.00 ( 0.00%) 991.00 (-58.56%) 4.3.0 4.3.0
13672-gc669d3513708-g8e488c2User 69638.44 67856.26
System 403739.55 405630.82
Elapsed 4055.69 4058.90
4.3.0 4.3.0
13672-gc669d3513708-g8e488c2
Minor Faults 2846674710 2831764228
Major Faults 0 2
Swap Ins 0 0
Swap Outs 0 0
Allocation stalls 0 0
DMA allocs 2846385041 2831501075
DMA32 allocs 0 0
Normal allocs 0 0
Movable allocs 0 0
Direct pages scanned 0 0
Kswapd pages scanned 0 0
Kswapd pages reclaimed 0 0
Direct pages reclaimed 0 0Kswapd efficiency 100% 100%
Kswapd velocity 0.000 0.000
Direct efficiency 100% 100%
Direct velocity 0.000 0.000
Percentage direct scans 0% 0%
Zone normal velocity 0.000 0.000 Zone dma32 velocity 0.000 0.000 Zone dma velocity 0.000 0.000 Page writes by reclaim 0.000 0.000 Page writes file 0 0 Page writes anon 0 0 Page reclaim immediate 0 0 Sector Reads 444 956 Sector Writes 14220 14232 Page rescued immediate 0 0 Slabs scanned 0 0 Direct inode steals 0 0 Kswapd inode steals 0 0 Kswapd skipped wait 0 0
THP fault alloc 39 70
THP collapse alloc 3 3
THP splits 40 69
THP fault fallback 209 246
THP collapse fail 0 0 Compaction stalls 0 0 Compaction success 0 0 Compaction failures 0 0 Page migrate success 23141 24465 Page migrate failure 0 0 Compaction pages isolated 0 0 Compaction migrate scanned 0 0 Compaction free scanned 0 0 Compaction cost 24 25 NUMA alloc hit 2846238192 2831325148 NUMA alloc miss 0 0 NUMA interleave hit 0 0 NUMA alloc local 1507528120 1466361674 NUMA base PTE updates 572055 572474 NUMA huge PMD updates 0 0 NUMA page range updates 572055 572474 NUMA hint faults 459654 460212 NUMA hint local faults 229186 224799 NUMA hint local percent 49 48 NUMA pages migrated 23141 24465 AutoNUMA cost 2302% 2305%
-------------- 64k hash pte powernv config start--------------------
Aneesh Kumar K.V (35):
powerpc/mm: move pte headers to book3s directory powerpc/mm: move pte headers to book3s directory (part 2) powerpc/mm: make a separate copy for book3s
powerpc/mm: make a separate copy for book3s (part 2) powerpc/mm: Move hash specific pte width and other defines to book3s powerpc/mm: Delete booke bits from book3s
powerpc/mm: Don't have generic headers introduce functions touching
pte bits
powerpc/mm: Drop pte-common.h from BOOK3S 64
powerpc/mm: Don't use pte_val as lvalue
powerpc/mm: Don't use pmd_val,pud_val and pgd_val as lvalue powerpc/mm: Move hash64 PTE bits from book3s/64/pgtable.h to hash.h powerpc/mm: Move PTE bits from generic functions to hash64 functions. powerpc/booke: Move nohash headers (part 1)
powerpc/booke: Move nohash headers (part 2)
powerpc/booke: Move nohash headers (part 3)
powerpc/booke: Move nohash headers (part 4)
powerpc/booke: Move nohash headers (part 5)
powerpc/mm: Convert 4k hash insert to C
powerpc/mm: Remove the dependency on pte bit position in asm code powerpc/mm: Don't track subpage valid bit in pte_t powerpc/mm: Remove pte_val usage for the second half of pgtable_t powerpc/mm: Increase the width of #define
powerpc/mm: Convert __hash_page_64K to C
powerpc/mm: Convert 4k insert from asm to C
powerpc/mm: Add helper for converting pte bit to hpte bits powerpc/mm: Move WIMG update to helper.
powerpc/mm: Move hugetlb related headers
powerpc/mm: Move THP headers around
powerpc/mm: Add a _PAGE_PTE bit
powerpc/mm: Don't hardcode page table size
powerpc/mm: Don't hardcode the hash pte slot shift powerpc/nohash: Update 64K nohash config to have 32 pte fragement powerpc/nohash: we don't use real_pte_t for nohash powerpc/mm: Use H_READ with H_READ_4
powerpc/mm: Don't open code pgtable_t size
.../include/asm/{pte-hash32.h => book3s/32/hash.h} | 6 +-
.../asm/{pgtable-ppc32.h => book3s/32/pgtable.h} | 288 ++++--
.../{pgtable-ppc64-4k.h => book3s/64/hash-4k.h} | 58 +-
arch/powerpc/include/asm/book3s/64/hash-64k.h | 312 ++++++
arch/powerpc/include/asm/book3s/64/hash.h | 526 ++++++++++
arch/powerpc/include/asm/book3s/64/pgtable.h | 266 ++++++
arch/powerpc/include/asm/book3s/pgtable.h | 29 +
arch/powerpc/include/asm/mmu-hash64.h | 2 +-
.../asm/{pgtable-ppc32.h => nohash/32/pgtable.h} | 25 +-
arch/powerpc/include/asm/{ => nohash/32}/pte-40x.h | 6 +-
arch/powerpc/include/asm/{ => nohash/32}/pte-44x.h | 6 +-
arch/powerpc/include/asm/{ => nohash/32}/pte-8xx.h | 6 +-
.../include/asm/{ => nohash/32}/pte-fsl-booke.h | 6 +-
.../{pgtable-ppc64-4k.h => nohash/64/pgtable-4k.h} | 12 +-
.../64/pgtable-64k.h} | 27 +-
.../asm/{pgtable-ppc64.h => nohash/64/pgtable.h} | 336 +------
arch/powerpc/include/asm/{ => nohash}/pgtable.h | 175 ++--
arch/powerpc/include/asm/{ => nohash}/pte-book3e.h | 6 +-
arch/powerpc/include/asm/page.h | 86 +-
arch/powerpc/include/asm/pgalloc-32.h | 34 +-
arch/powerpc/include/asm/pgalloc-64.h | 27 +-
arch/powerpc/include/asm/pgtable.h | 200 +---
arch/powerpc/include/asm/plpar_wrappers.h | 17 +
arch/powerpc/include/asm/pte-common.h | 5 +
arch/powerpc/include/asm/pte-hash64-4k.h | 17 -
arch/powerpc/include/asm/pte-hash64-64k.h | 102 --
arch/powerpc/include/asm/pte-hash64.h | 54 --
arch/powerpc/kernel/exceptions-64s.S | 18 +-
arch/powerpc/mm/40x_mmu.c | 10 +-
arch/powerpc/mm/Makefile | 9 +-
arch/powerpc/mm/hash64_4k.c | 123 +++
arch/powerpc/mm/hash64_64k.c | 322 +++++++
arch/powerpc/mm/hash_low_64.S | 1003 --------------------
arch/powerpc/mm/hash_native_64.c | 10 +
arch/powerpc/mm/hash_utils_64.c | 105 +-
arch/powerpc/mm/hugepage-hash64.c | 22 +-
arch/powerpc/mm/hugetlbpage-hash64.c | 33 +-
arch/powerpc/mm/hugetlbpage.c | 76 +-
arch/powerpc/mm/init_64.c | 4 -
arch/powerpc/mm/pgtable.c | 4 +
arch/powerpc/mm/pgtable_64.c | 28 +-
arch/powerpc/platforms/pseries/lpar.c | 64 +-
42 files changed, 2276 insertions(+), 2189 deletions(-)
rename arch/powerpc/include/asm/{pte-hash32.h => book3s/32/hash.h} (93%)
copy arch/powerpc/include/asm/{pgtable-ppc32.h => book3s/32/pgtable.h} (62%)
copy arch/powerpc/include/asm/{pgtable-ppc64-4k.h => book3s/64/hash-4k.h} (71%)
create mode 100644 arch/powerpc/include/asm/book3s/64/hash-64k.h
create mode 100644 arch/powerpc/include/asm/book3s/64/hash.h
create mode 100644 arch/powerpc/include/asm/book3s/64/pgtable.h
create mode 100644 arch/powerpc/include/asm/book3s/pgtable.h
rename arch/powerpc/include/asm/{pgtable-ppc32.h => nohash/32/pgtable.h} (96%)
rename arch/powerpc/include/asm/{ => nohash/32}/pte-40x.h (95%)
rename arch/powerpc/include/asm/{ => nohash/32}/pte-44x.h (96%)
rename arch/powerpc/include/asm/{ => nohash/32}/pte-8xx.h (95%)
rename arch/powerpc/include/asm/{ => nohash/32}/pte-fsl-booke.h (88%)
rename arch/powerpc/include/asm/{pgtable-ppc64-4k.h => nohash/64/pgtable-4k.h} (92%)
rename arch/powerpc/include/asm/{pgtable-ppc64-64k.h => nohash/64/pgtable-64k.h} (64%)
rename arch/powerpc/include/asm/{pgtable-ppc64.h => nohash/64/pgtable.h} (51%)
copy arch/powerpc/include/asm/{ => nohash}/pgtable.h (62%)
rename arch/powerpc/include/asm/{ => nohash}/pte-book3e.h (95%)
delete mode 100644 arch/powerpc/include/asm/pte-hash64-4k.h
delete mode 100644 arch/powerpc/include/asm/pte-hash64-64k.h
delete mode 100644 arch/powerpc/include/asm/pte-hash64.h
create mode 100644 arch/powerpc/mm/hash64_4k.c
create mode 100644 arch/powerpc/mm/hash64_64k.c
delete mode 100644 arch/powerpc/mm/hash_low_64.SSignature · 1 line
2.5.0
PATCH v6 1/35 powerpc/mm: move pte headers to book3s directory
---
arch/powerpc/include/asm/{pte-hash32.h => book3s/32/hash.h} | 0
arch/powerpc/include/asm/{pte-hash64.h => book3s/64/hash.h} | 0
arch/powerpc/include/asm/pgtable-ppc32.h | 2 +-
arch/powerpc/include/asm/pgtable-ppc64.h | 2 +-
4 files changed, 2 insertions(+), 2 deletions(-)
rename arch/powerpc/include/asm/{pte-hash32.h => book3s/32/hash.h} (100%)
rename arch/powerpc/include/asm/{pte-hash64.h => book3s/64/hash.h} (100%)arch/powerpc/include/asm/book3s/32/hash.h renamed from arch/powerpc/include/asm/pte-hash32.h +0 -0
arch/powerpc/include/asm/book3s/64/hash.h renamed from arch/powerpc/include/asm/pte-hash64.h +0 -0
arch/powerpc/include/asm/pgtable-ppc32.h +1 -1
@@ -116,7 +116,7 @@ extern int icache_44x_need_flush;#elif defined(CONFIG_8xx)#include <asm/pte-8xx.h> #else /* CONFIG_6xx */ -#include <asm/pte-hash32.h> +#include <asm/book3s/32/hash.h> #endif /* And here we include common definitions */
arch/powerpc/include/asm/pgtable-ppc64.h +1 -1
@@ -98,7 +98,7 @@* Include the PTE bits definitions */ #ifdef CONFIG_PPC_BOOK3S -#include <asm/pte-hash64.h> +#include <asm/book3s/64/hash.h> #else #include <asm/pte-book3e.h> #endif
Signature · 1 line
2.5.0
Michael Ellerman
reviewerRe: [V6,01/35] powerpc/mm: move pte headers to book3s directory
Quoted 2 lines · Aneesh Kumar K.V
Acked-by Scott Wood <[email protected]>Signed-off-by Aneesh Kumar K.V <[email protected]>
Entire series applied to powerpc next, thanks.
https://git.kernel.org/powerpc/c/26b6a3d9bb48f8b4624a6228
cheers
Re: [PATCH V6 01/35] powerpc/mm: move pte headers to book3s directory
In reply to Aneesh Kumar K.VQuoted 10 lines · Aneesh Kumar K.V
Acked-by Scott Wood <[email protected]>Signed-off-by Aneesh Kumar K.V <[email protected]>---
arch/powerpc/include/asm/{pte-hash32.h =>book3s/32/hash.h} | 0arch/powerpc/include/asm/{pte-hash64.h=>book3s/64/hash.h} | 0 arch/powerpc/include/asm/pgtable-ppc32.h | 2 +- arch/powerpc/include/asm/pgtable-ppc64.h | 2 +-4 files changed, 2 insertions(+), 2 deletions(-)rename
arch/powerpc/include/asm/{pte-hash32.h=>book3s/32/hash.h} (100%) renamearch/powerpc/include/asm/{pte-hash64.h=>book3s/64/hash.h} (100%)
Why not pte-hash.h?
Balbir Singh.
PATCH v6 2/35 powerpc/mm: move pte headers to book3s directory (part 2)
In reply to Aneesh Kumar K.VSplitting this so that rename can track changes to file. Before merging we will fold this
---
arch/powerpc/include/asm/book3s/32/hash.h | 6 +++---
.../include/asm/{pte-hash64-4k.h => book3s/64/hash-4k.h} | 0
.../include/asm/{pte-hash64-64k.h => book3s/64/hash-64k.h} | 0
arch/powerpc/include/asm/book3s/64/hash.h | 10 +++++-----
4 files changed, 8 insertions(+), 8 deletions(-)
rename arch/powerpc/include/asm/{pte-hash64-4k.h => book3s/64/hash-4k.h} (100%)
rename arch/powerpc/include/asm/{pte-hash64-64k.h => book3s/64/hash-64k.h} (100%)arch/powerpc/include/asm/book3s/32/hash.h +3 -3
@@ -1,5 +1,5 @@ -#ifndef _ASM_POWERPC_PTE_HASH32_H -#define _ASM_POWERPC_PTE_HASH32_H +#ifndef _ASM_POWERPC_BOOK3S_32_HASH_H +#define _ASM_POWERPC_BOOK3S_32_HASH_H#ifdef __KERNEL__/*
@@ -43,4 +43,4 @@#define PTE_ATOMIC_UPDATES 1#endif /* __KERNEL__ */ -#endif /* _ASM_POWERPC_PTE_HASH32_H */ +#endif /* _ASM_POWERPC_BOOK3S_32_HASH_H */
arch/powerpc/include/asm/book3s/64/hash-4k.h renamed from arch/powerpc/include/asm/pte-hash64-4k.h +0 -0
arch/powerpc/include/asm/book3s/64/hash-64k.h renamed from arch/powerpc/include/asm/pte-hash64-64k.h +0 -0
arch/powerpc/include/asm/book3s/64/hash.h +5 -5
@@ -1,5 +1,5 @@ -#ifndef _ASM_POWERPC_PTE_HASH64_H -#define _ASM_POWERPC_PTE_HASH64_H +#ifndef _ASM_POWERPC_BOOK3S_64_HASH_H +#define _ASM_POWERPC_BOOK3S_64_HASH_H#ifdef __KERNEL__/*
@@ -45,10 +45,10 @@#define PTE_ATOMIC_UPDATES 1#ifdef CONFIG_PPC_64K_PAGES -#include <asm/pte-hash64-64k.h> +#include <asm/book3s/64/hash-64k.h> #else -#include <asm/pte-hash64-4k.h> +#include <asm/book3s/64/hash-4k.h> #endif #endif /* __KERNEL__ */ -#endif /* _ASM_POWERPC_PTE_HASH64_H */ +#endif /* _ASM_POWERPC_BOOK3S_64_HASH_H */
Signature · 1 line
2.5.0
PATCH v6 3/35 powerpc/mm: make a separate copy for book3s
In reply to Aneesh Kumar K.VIn this patch we do:
cp pgtable-ppc32.h book3s/32/pgtable.h
cp pgtable-ppc64.h book3s/64/pgtable.h
This enable us to do further changes to hash specific config. We will change the page table format for 64bit hash in later patches.
--- arch/powerpc/include/asm/book3s/32/pgtable.h | 340 +++++++++++++++ arch/powerpc/include/asm/book3s/64/pgtable.h | 626 +++++++++++++++++++++++++++ arch/powerpc/include/asm/book3s/pgtable.h | 10 + arch/powerpc/include/asm/mmu-hash64.h | 2 +- arch/powerpc/include/asm/pgtable.h | 4 + 5 files changed, 981 insertions(+), 1 deletion(-) create mode 100644 arch/powerpc/include/asm/book3s/32/pgtable.h create mode 100644 arch/powerpc/include/asm/book3s/64/pgtable.h create mode 100644 arch/powerpc/include/asm/book3s/pgtable.h
arch/powerpc/include/asm/book3s/32/pgtable.h new file +340 -0
@@ -0,0 +1,340 @@ +#ifndef _ASM_POWERPC_PGTABLE_PPC32_H +#define _ASM_POWERPC_PGTABLE_PPC32_H + +#include <asm-generic/pgtable-nopmd.h> + +#ifndef __ASSEMBLY__ +#include <linux/sched.h> +#include <linux/threads.h> +#include <asm/io.h> /* For sub-arch specific PPC_PIN_SIZE */ + +extern unsigned long ioremap_bot; + +#ifdef CONFIG_44x +extern int icache_44x_need_flush; +#endif + +#endif /* __ASSEMBLY__ */ + +/* +* The normal case is that PTEs are 32-bits and we have a 1-page +* 1024-entry pgdir pointing to 1-page 1024-entry PTE pages. -- paulus +* +* For any >32-bit physical address platform, we can use the following +* two level page table layout where the pgdir is 8KB and the MS 13 bits +* are an index to the second level table. The combined pgdir/pmd first +* level has 2048 entries and the second level has 512 64-bit PTE entries. +* -Matt +*/ +/* PGDIR_SHIFT determines what a top-level page table entry can map */ +#define PGDIR_SHIFT (PAGE_SHIFT + PTE_SHIFT) +#define PGDIR_SIZE (1UL << PGDIR_SHIFT) +#define PGDIR_MASK (~(PGDIR_SIZE-1)) + +/* +* entries per page directory level: our page-table tree is two-level, so +* we don't really have any PMD directory. +*/ +#ifndef __ASSEMBLY__ +#define PTE_TABLE_SIZE (sizeof(pte_t) << PTE_SHIFT) +#define PGD_TABLE_SIZE (sizeof(pgd_t) << (32 - PGDIR_SHIFT)) +#endif /* __ASSEMBLY__ */ + +#define PTRS_PER_PTE (1 << PTE_SHIFT) +#define PTRS_PER_PMD 1 +#define PTRS_PER_PGD (1 << (32 - PGDIR_SHIFT)) + +#define USER_PTRS_PER_PGD (TASK_SIZE / PGDIR_SIZE) +#define FIRST_USER_ADDRESS 0UL + +#define pte_ERROR (e) \ + pr_err("%s:%d: bad pte %llx.\n", __FILE__, __LINE__, \ + (unsigned long long)pte_val(e)) +#define pgd_ERROR (e) \ + pr_err("%s:%d: bad pgd %08lx.\n", __FILE__, __LINE__, pgd_val(e)) + +/* +* This is the bottom of the PKMAP area with HIGHMEM or an arbitrary +* value (for now) on others, from where we can start layout kernel +* virtual space that goes below PKMAP and FIXMAP +*/ +#ifdef CONFIG_HIGHMEM +#define KVIRT_TOP PKMAP_BASE +#else +#define KVIRT_TOP (0xfe000000UL)/* for now, could be FIXMAP_BASE ? */ +#endif + +/* +* ioremap_bot starts at that address. Early ioremaps move down from there, +* until mem_init() at which point this becomes the top of the vmalloc +* and ioremap space +*/ +#ifdef CONFIG_NOT_COHERENT_CACHE +#define IOREMAP_TOP ((KVIRT_TOP - CONFIG_CONSISTENT_SIZE) & PAGE_MASK) +#else +#define IOREMAP_TOP KVIRT_TOP +#endif + +/* +* Just any arbitrary offset to the start of the vmalloc VM area: the +* current 16MB value just means that there will be a 64MB "hole" after the +* physical memory until the kernel virtual memory starts. That means that +* any out-of-bounds memory accesses will hopefully be caught. +* The vmalloc() routines leaves a hole of 4kB between each vmalloced +* area for the same reason. ;) +* +* We no longer map larger than phys RAM with the BATs so we don't have +* to worry about the VMALLOC_OFFSET causing problems. We do have to worry +* about clashes between our early calls to ioremap() that start growing down +* from ioremap_base being run into the VM area allocations (growing upwards +* from VMALLOC_START). For this reason we have ioremap_bot to check when +* we actually run into our mappings setup in the early boot with the VM +* system. This really does become a problem for machines with good amounts +* of RAM. -- Cort +*/ +#define VMALLOC_OFFSET (0x1000000)/* 16M */ +#ifdef PPC_PIN_SIZE +#define VMALLOC_START (((_ALIGN((long)high_memory, PPC_PIN_SIZE) + VMALLOC_OFFSET) & ~(VMALLOC_OFFSET-1))) +#else +#define VMALLOC_START ((((long)high_memory + VMALLOC_OFFSET) & ~(VMALLOC_OFFSET-1))) +#endif +#define VMALLOC_END ioremap_bot + +/* +* Bits in a linux-style PTE. These match the bits in the +* (hardware-defined) PowerPC PTE as closely as possible. +*/ + +#if defined(CONFIG_40x) +#include <asm/pte-40x.h> +#elif defined(CONFIG_44x) +#include <asm/pte-44x.h> +#elif defined(CONFIG_FSL_BOOKE ) && defined(CONFIG_PTE_64BIT ) +#include <asm/pte-book3e.h> +#elif defined(CONFIG_FSL_BOOKE ) +#include <asm/pte-fsl-booke.h> +#elif defined(CONFIG_8xx) +#include <asm/pte-8xx.h> +#else /* CONFIG_6xx */ +#include <asm/book3s/32/hash.h> +#endif + +/* And here we include common definitions */ +#include <asm/pte-common.h> + +#ifndef __ASSEMBLY__ + +#define pte_clear (mm, addr, ptep) \ + do { pte_update(ptep, ~_PAGE_HASHPTE, 0); } while (0) + +#define pmd_none (pmd) (!pmd_val(pmd)) +#define pmd_bad (pmd) (pmd_val(pmd) & _PMD_BAD) +#define pmd_present (pmd) (pmd_val(pmd) & _PMD_PRESENT_MASK) +#define pmd_clear (pmdp) do { pmd_val(*(pmdp)) = 0; } while (0) + +/* +* When flushing the tlb entry for a page, we also need to flush the hash +* table entry. flush_hash_pages is assembler (for speed) in hashtable.S. +*/ +extern int flush_hash_pages (unsigned context,unsigned long va, +unsigned long pmdval,int count); + +/* Add an HPTE to the hash table */ +extern void add_hash_page (unsigned context,unsigned long va, +unsigned long pmdval); + +/* Flush an entry from the TLB/hash table */ +extern void flush_hash_entry (struct mm_struct *mm,pte_t *ptep, +unsigned long address); + +/* +* PTE updates. This function is called whenever an existing +* valid PTE is updated. This does -not- include set_pte_at() +* which nowadays only sets a new PTE. +* +* Depending on the type of MMU, we may need to use atomic updates +* and the PTE may be either 32 or 64 bit wide. In the later case, +* when using atomic updates, only the low part of the PTE is +* accessed atomically. +* +* In addition, on 44x, we also maintain a global flag indicating +* that an executable user mapping was modified, which is needed +* to properly flush the virtually tagged instruction cache of +* those implementations. +*/ +#ifndef CONFIG_PTE_64BIT +static inline unsigned long pte_update (pte_t *p, +unsigned long clr, +unsigned long set) +{ +#ifdef PTE_ATOMIC_UPDATES +unsigned long old, tmp; + + __asm__ __volatile__("\ +1: lwarx %0,0,%3\n\ +andc %1,%0,%4\n\ +or %1,%1,%5\n" + PPC405_ERR77(0,%3) +" stwcx. %1,0,%3\n\ + bne- 1b" + : "=&r" (old), "=&r" (tmp), "=m" (*p) + : "r" (p), "r" (clr), "r" (set), "m" (*p) + : "cc" ); +#else /* PTE_ATOMIC_UPDATES */ +unsigned long old =pte_val (*p); + *p =__pte ((old & ~clr) | set); +#endif /* !PTE_ATOMIC_UPDATES */ + +#ifdef CONFIG_44x +if ((old & _PAGE_USER) && (old & _PAGE_EXEC)) + icache_44x_need_flush =1 ; +#endif +return old; +} +#else /* CONFIG_PTE_64BIT */ +static inline unsigned long long pte_update (pte_t *p, +unsigned long clr, +unsigned long set) +{ +#ifdef PTE_ATOMIC_UPDATES +unsigned long long old; +unsigned long tmp; + + __asm__ __volatile__("\ +1: lwarx %L0,0,%4\n\ +lwzx %0,0,%3\n\ +andc %1,%L0,%5\n\ +or %1,%1,%6\n" + PPC405_ERR77(0,%3) +" stwcx. %1,0,%4\n\ + bne- 1b" + : "=&r" (old), "=&r" (tmp), "=m" (*p) + : "r" (p), "r" ((unsigned long)(p) + 4), "r" (clr), "r" (set), "m" (*p) + : "cc" ); +#else /* PTE_ATOMIC_UPDATES */ +unsigned long long old =pte_val (*p); + *p =__pte ((old & ~(unsigned long long )clr) | set); +#endif /* !PTE_ATOMIC_UPDATES */ + +#ifdef CONFIG_44x +if ((old & _PAGE_USER) && (old & _PAGE_EXEC)) + icache_44x_need_flush =1 ; +#endif +return old; +} +#endif /* CONFIG_PTE_64BIT */ + +/* +* 2.6 calls this without flushing the TLB entry; this is wrong +* for our hash-based implementation, we fix that up here. +*/ +#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG +static inline int __ptep_test_and_clear_young (unsigned int context,unsigned long addr,pte_t *ptep) +{ +unsigned long old; + old =pte_update (ptep, _PAGE_ACCESSED,0 ); +#if _PAGE_HASHPTE !=0 +if (old & _PAGE_HASHPTE) { +unsigned long ptephys =__pa (ptep) &PAGE_MASK ; +flush_hash_pages (context, addr, ptephys,1 ); + } +#endif +return (old & _PAGE_ACCESSED) !=0 ; +} +#define ptep_test_and_clear_young (__vma, __addr, __ptep) \ + __ptep_test_and_clear_young((__vma)->vm_mm->context.id, __addr, __ptep) + +#define __HAVE_ARCH_PTEP_GET_AND_CLEAR +static inline pte_t ptep_get_and_clear (struct mm_struct *mm,unsigned long addr, +pte_t *ptep) +{ +return __pte (pte_update (ptep, ~_PAGE_HASHPTE,0 )); +} + +#define __HAVE_ARCH_PTEP_SET_WRPROTECT +static inline void ptep_set_wrprotect (struct mm_struct *mm,unsigned long addr, +pte_t *ptep) +{ +pte_update (ptep, (_PAGE_RW | _PAGE_HWWRITE), _PAGE_RO); +} +static inline void huge_ptep_set_wrprotect (struct mm_struct *mm, +unsigned long addr,pte_t *ptep) +{ +ptep_set_wrprotect (mm, addr, ptep); +} + + +static inline void __ptep_set_access_flags (pte_t *ptep,pte_t entry) +{ +unsigned long set =pte_val (entry) & + (_PAGE_DIRTY | _PAGE_ACCESSED | _PAGE_RW | _PAGE_EXEC); +unsigned long clr = ~pte_val (entry) & _PAGE_RO; + +pte_update (ptep, clr, set); +} + +#define __HAVE_ARCH_PTE_SAME +#define pte_same (A ,B ) (((pte_val(A) ^ pte_val(B)) & ~_PAGE_HASHPTE) == 0) + +/* +* Note that on Book E processors, the pmd contains the kernel virtual +* (lowmem) address of the pte page. The physical address is less useful +* because everything runs with translation enabled (even the TLB miss +* handler). On everything else the pmd contains the physical address +* of the pte page. -- paulus +*/ +#ifndef CONFIG_BOOKE +#define pmd_page_vaddr (pmd) \ + ((unsigned long) __va(pmd_val(pmd) & PAGE_MASK)) +#define pmd_page (pmd) \ + pfn_to_page(pmd_val(pmd) >> PAGE_SHIFT) +#else +#define pmd_page_vaddr (pmd) \ + ((unsigned long) (pmd_val(pmd) & PAGE_MASK)) +#define pmd_page (pmd) \ + pfn_to_page((__pa(pmd_val(pmd)) >> PAGE_SHIFT)) +#endif + +/* to find an entry in a kernel page-table-directory */ +#define pgd_offset_k (address) pgd_offset(&init_mm, address) + +/* to find an entry in a page-table-directory */ +#define pgd_index (address) ((address) >> PGDIR_SHIFT) +#define pgd_offset (mm, address) ((mm)->pgd + pgd_index(address)) + +/* Find an entry in the third-level page table.. */ +#define pte_index (address) \ + (((address) >> PAGE_SHIFT) & (PTRS_PER_PTE - 1)) +#define pte_offset_kernel (dir, addr) \ + ((pte_t *) pmd_page_vaddr(*(dir)) + pte_index(addr)) +#define pte_offset_map (dir, addr) \ + ((pte_t *) kmap_atomic(pmd_page(*(dir))) + pte_index(addr)) +#define pte_unmap (pte) kunmap_atomic(pte) + +/* +* Encode and decode a swap entry. +* Note that the bits we use in a PTE for representing a swap entry +* must not include the _PAGE_PRESENT bit or the _PAGE_HASHPTE bit (if used). +* -- paulus +*/ +#define __swp_type (entry) ((entry).val & 0x1f) +#define __swp_offset (entry) ((entry).val >> 5) +#define __swp_entry (type, offset) ((swp_entry_t) { (type) | ((offset) << 5) }) +#define __pte_to_swp_entry (pte) ((swp_entry_t) { pte_val(pte) >> 3 }) +#define __swp_entry_to_pte (x) ((pte_t) { (x).val << 3 }) + +#ifndef CONFIG_PPC_4K_PAGES +void pgtable_cache_init (void ); +#else +/* +* No page table caches to initialise +*/ +#define pgtable_cache_init () do { } while (0) +#endif + +extern int get_pteptr (struct mm_struct *mm,unsigned long addr,pte_t **ptep, +pmd_t **pmdp); + +#endif /* !__ASSEMBLY__ */ + +#endif /* _ASM_POWERPC_PGTABLE_PPC32_H */
arch/powerpc/include/asm/book3s/64/pgtable.h new file +626 -0
626 lines not on this page · raw message
arch/powerpc/include/asm/book3s/pgtable.h new file +10 -0
@@ -0,0 +1,10 @@ +#ifndef _ASM_POWERPC_BOOK3S_PGTABLE_H +#define _ASM_POWERPC_BOOK3S_PGTABLE_H + +#ifdef CONFIG_PPC64 +#include <asm/book3s/64/pgtable.h> +#else +#include <asm/book3s/32/pgtable.h> +#endif + +#endif
arch/powerpc/include/asm/mmu-hash64.h +1 -1
@@ -21,7 +21,7 @@* need for various slices related matters. Note that this isn't the * complete pgtable.h but only a portion of it. */ -#include <asm/pgtable-ppc64.h> +#include <asm/book3s/64/pgtable.h> #include <asm/bug.h> #include <asm/processor.h>
arch/powerpc/include/asm/pgtable.h +4 -0
@@ -13,11 +13,15 @@ struct mm_struct;#endif /* !__ASSEMBLY__ */ +#ifdef CONFIG_PPC_BOOK3S +#include <asm/book3s/pgtable.h> +#else #if defined(CONFIG_PPC64 )# include <asm/pgtable-ppc64.h> #else # include <asm/pgtable-ppc32.h> #endif +#endif /* !CONFIG_PPC_BOOK3S */ /* * We save the slot number & secondary bit in the second half of the
Signature · 1 line
2.5.0
PATCH v6 4/35 powerpc/mm: make a separate copy for book3s (part 2)
In reply to Aneesh Kumar K.VKeep it seperate to make rebasing easier
--- arch/powerpc/include/asm/book3s/32/pgtable.h | 6 +++--- arch/powerpc/include/asm/book3s/64/pgtable.h | 6 +++--- arch/powerpc/include/asm/pgtable-ppc32.h | 2 -- arch/powerpc/include/asm/pgtable-ppc64.h | 4 ---- 4 files changed, 6 insertions(+), 12 deletions(-)
arch/powerpc/include/asm/book3s/32/pgtable.h +3 -3
@@ -1,5 +1,5 @@ -#ifndef _ASM_POWERPC_PGTABLE_PPC32_H -#define _ASM_POWERPC_PGTABLE_PPC32_H +#ifndef _ASM_POWERPC_BOOK3S_32_PGTABLE_H +#define _ASM_POWERPC_BOOK3S_32_PGTABLE_H#include <asm-generic/pgtable-nopmd.h>
@@ -337,4 +337,4 @@ extern int get_pteptr(struct mm_struct *mm, unsigned long addr, pte_t **ptep,#endif /* !__ASSEMBLY__ */ -#endif /* _ASM_POWERPC_PGTABLE_PPC32_H */ +#endif /* _ASM_POWERPC_BOOK3S_32_PGTABLE_H */
arch/powerpc/include/asm/book3s/64/pgtable.h +3 -3
@@ -1,5 +1,5 @@ -#ifndef _ASM_POWERPC_PGTABLE_PPC64_H_ -#define _ASM_POWERPC_PGTABLE_PPC64_H_ +#ifndef _ASM_POWERPC_BOOK3S_64_PGTABLE_H_ +#define _ASM_POWERPC_BOOK3S_64_PGTABLE_H_/* * This file contains the functions and defines necessary to modify and use * the ppc64 hashed page table.
@@ -623,4 +623,4 @@ static inline int pmd_move_must_withdraw(struct spinlock *new_pmd_ptl,return true; }#endif /* __ASSEMBLY__ */ -#endif /* _ASM_POWERPC_PGTABLE_PPC64_H_ */ +#endif /* _ASM_POWERPC_BOOK3S_64_PGTABLE_H_ */
arch/powerpc/include/asm/pgtable-ppc32.h +0 -2
@@ -115,8 +115,6 @@ extern int icache_44x_need_flush;#include <asm/pte-fsl-booke.h> #elif defined(CONFIG_8xx)#include <asm/pte-8xx.h> -#else /* CONFIG_6xx */ -#include <asm/book3s/32/hash.h> #endif /* And here we include common definitions */
arch/powerpc/include/asm/pgtable-ppc64.h +0 -4
@@ -97,11 +97,7 @@/* * Include the PTE bits definitions */ -#ifdef CONFIG_PPC_BOOK3S -#include <asm/book3s/64/hash.h> -#else #include <asm/pte-book3e.h> -#endif #include <asm/pte-common.h> #ifdef CONFIG_PPC_MM_SLICES
Signature · 1 line
2.5.0
Re: [PATCH V6 04/35] powerpc/mm: make a separate copy for book3s (part 2)
Quoted 11 lines · Aneesh Kumar K.V
Keep it seperate to make rebasing easier
Acked-by Scott Wood <[email protected]>Signed-off-by Aneesh Kumar K.V <[email protected]>---
arch/powerpc/include/asm/book3s/32/pgtable.h | 6 +++---arch/powerpc/include/asm/book3s/64/pgtable.h| 6 +++--- arch/powerpc/include/asm/pgtable-ppc32.h | 2 -- arch/powerpc/include/asm/pgtable-ppc64.h | 4 ----4 files changed, 6 insertions(+), 12 deletions(-)
One side effect is that someone could by mistake include asm/book3s/32/pgtable.h and asm/pgtable-ppc32.h and get duplicate definitions and have the compiler complain
Ditto for the 64 bit headers
Balbir
PATCH v6 5/35 powerpc/mm: Move hash specific pte width and other defines to book3s
In reply to Aneesh Kumar K.VThis further make a copy of pte defines to book3s/64/hash*.h. This remove the dependency on pgtable-ppc64-4k.h and pgtable-ppc64-64k.h
--- arch/powerpc/include/asm/book3s/64/hash-4k.h | 86 ++++++++++++++++++++++++++- arch/powerpc/include/asm/book3s/64/hash-64k.h | 46 +++++++++++++- arch/powerpc/include/asm/book3s/64/pgtable.h | 6 +- 3 files changed, 129 insertions(+), 9 deletions(-)
arch/powerpc/include/asm/book3s/64/hash-4k.h +85 -1
@@ -1,4 +1,51 @@ -/* To be include by pgtable-hash64.h only */ +#ifndef _ASM_POWERPC_BOOK3S_64_HASH_4K_H +#define _ASM_POWERPC_BOOK3S_64_HASH_4K_H +/* +* Entries per page directory level. The PTE level must use a 64b record +* for each page table entry. The PMD and PGD level use a 32b record for +* each entry by assuming that each entry is page aligned. +*/ +#define PTE_INDEX_SIZE 9 +#define PMD_INDEX_SIZE 7 +#define PUD_INDEX_SIZE 9 +#define PGD_INDEX_SIZE 9 + +#ifndef __ASSEMBLY__ +#define PTE_TABLE_SIZE (sizeof(pte_t) << PTE_INDEX_SIZE) +#define PMD_TABLE_SIZE (sizeof(pmd_t) << PMD_INDEX_SIZE) +#define PUD_TABLE_SIZE (sizeof(pud_t) << PUD_INDEX_SIZE) +#define PGD_TABLE_SIZE (sizeof(pgd_t) << PGD_INDEX_SIZE) +#endif /* __ASSEMBLY__ */ + +#define PTRS_PER_PTE (1 << PTE_INDEX_SIZE) +#define PTRS_PER_PMD (1 << PMD_INDEX_SIZE) +#define PTRS_PER_PUD (1 << PUD_INDEX_SIZE) +#define PTRS_PER_PGD (1 << PGD_INDEX_SIZE) + +/* PMD_SHIFT determines what a second-level page table entry can map */ +#define PMD_SHIFT (PAGE_SHIFT + PTE_INDEX_SIZE) +#define PMD_SIZE (1UL << PMD_SHIFT) +#define PMD_MASK (~(PMD_SIZE-1)) + +/* With 4k base page size, hugepage PTEs go at the PMD level */ +#define MIN_HUGEPTE_SHIFT PMD_SHIFT + +/* PUD_SHIFT determines what a third-level page table entry can map */ +#define PUD_SHIFT (PMD_SHIFT + PMD_INDEX_SIZE) +#define PUD_SIZE (1UL << PUD_SHIFT) +#define PUD_MASK (~(PUD_SIZE-1)) + +/* PGDIR_SHIFT determines what a fourth-level page table entry can map */ +#define PGDIR_SHIFT (PUD_SHIFT + PUD_INDEX_SIZE) +#define PGDIR_SIZE (1UL << PGDIR_SHIFT) +#define PGDIR_MASK (~(PGDIR_SIZE-1)) + +/* Bits to mask out from a PMD to get to the PTE page */ +#define PMD_MASKED_BITS 0 +/* Bits to mask out from a PUD to get to the PMD page */ +#define PUD_MASKED_BITS 0 +/* Bits to mask out from a PGD to get to the PUD page */ +#define PGD_MASKED_BITS 0/* PTE bits */ #define _PAGE_HASHPTE 0x0400/* software: pte has an associated HPTE */
@@ -15,3 +62,40 @@/* shift to put page number into pte */ #define PTE_RPN_SHIFT (17) +#ifndef __ASSEMBLY__ +/* +* 4-level page tables related bits +*/ + +#define pgd_none (pgd) (!pgd_val(pgd)) +#define pgd_bad (pgd) (pgd_val(pgd) == 0) +#define pgd_present (pgd) (pgd_val(pgd) != 0) +#define pgd_clear (pgdp) (pgd_val(*(pgdp)) = 0) +#define pgd_page_vaddr (pgd) (pgd_val(pgd) & ~PGD_MASKED_BITS) + +static inline pte_t pgd_pte (pgd_t pgd) +{ +return __pte (pgd_val (pgd)); +} + +static inline pgd_t pte_pgd (pte_t pte) +{ +return __pgd (pte_val (pte)); +} +extern struct page *pgd_page (pgd_t pgd); + +#define pud_offset (pgdp, addr) \ + (((pud_t *) pgd_page_vaddr(*(pgdp))) + \ + (((addr) >> PUD_SHIFT) & (PTRS_PER_PUD - 1))) + +#define pud_ERROR (e) \ + pr_err("%s:%d: bad pud %08lx.\n", __FILE__, __LINE__, pud_val(e)) + +/* +* On all 4K setups, remap_4k_pfn() equates to remap_pfn_range() */ +#define remap_4k_pfn (vma, addr, pfn, prot) \ + remap_pfn_range((vma), (addr), (pfn), PAGE_SIZE, (prot)) + +#endif /* !__ASSEMBLY__ */ + +#endif /* _ASM_POWERPC_BOOK3S_64_HASH_4K_H */
arch/powerpc/include/asm/book3s/64/hash-64k.h +43 -3
59 lines not on this page · raw message
arch/powerpc/include/asm/book3s/64/pgtable.h +1 -5
@@ -5,11 +5,7 @@* the ppc64 hashed page table. */ -#ifdef CONFIG_PPC_64K_PAGES -#include <asm/pgtable-ppc64-64k.h> -#else -#include <asm/pgtable-ppc64-4k.h> -#endif +#include <asm/book3s/64/hash.h> #include <asm/barrier.h> #define FIRST_USER_ADDRESS 0UL
Signature · 1 line
2.5.0
Re: [PATCH V6 05/35] powerpc/mm: Move hash specific pte width and other defines to book3s
Quoted 16 lines · Aneesh Kumar K.V
This further make a copy of pte defines to
book3s/64/hash*.h. This remove the dependency onpgtable-ppc64-4k.handpgtable-ppc64-64k.hAcked-by Scott Wood <[email protected]>Signed-off-by Aneesh Kumar K.V <[email protected]>--- arch/powerpc/include/asm/book3s/64/hash-4k.h | 86 ++++++++++++++++++++++++++- arch/powerpc/include/asm/book3s/64/hash-64k.h | 46 +++++++++++++- arch/powerpc/include/asm/book3s/64/pgtable.h | 6 +- 3 files changed, 129 insertions(+), 9 deletions(-)arch/powerpc/include/asm/book3s/64/hash-4k.h +0
-0
arch/powerpc/include/asm/book3s/64/hash-4k.h line 7
Quoted 8 lines
arch/powerpc/include/asm/book3s/64/hash-4k.h +7
-1@@ -1,4 +1,51 @@ -/* To be include by pgtable-hash64.h only */ +#ifndef _ASM_POWERPC_BOOK3S_64_HASH_4K_H +#define _ASM_POWERPC_BOOK3S_64_HASH_4K_H +/* +* Entries per page directory level. The PTE level must use a 64b record +* for each page table entry. The PMD and PGD level use a 32b record for +* each entry by assuming that each entry is page aligned. +*/
More clarity on this please
arch/powerpc/include/asm/book3s/64/hash-4k.h line 12
Quoted 5 lines
arch/powerpc/include/asm/book3s/64/hash-4k.h +5
-0+#define PTE_INDEX_SIZE 9 +#define PMD_INDEX_SIZE 7 +#define PUD_INDEX_SIZE 9 +#define PGD_INDEX_SIZE 9 +
We use 9+7+9+9+(12) bits?
Quoted 27 lines
arch/powerpc/include/asm/book3s/64/hash-4k.h +27
-0+#ifndef __ASSEMBLY__ +#define PTE_TABLE_SIZE (sizeof(pte_t) << PTE_INDEX_SIZE) +#define PMD_TABLE_SIZE (sizeof(pmd_t) << PMD_INDEX_SIZE) +#define PUD_TABLE_SIZE (sizeof(pud_t) << PUD_INDEX_SIZE) +#define PGD_TABLE_SIZE (sizeof(pgd_t) << PGD_INDEX_SIZE) +#endif /* __ASSEMBLY__ */ + +#define PTRS_PER_PTE (1 << PTE_INDEX_SIZE) +#define PTRS_PER_PMD (1 << PMD_INDEX_SIZE) +#define PTRS_PER_PUD (1 << PUD_INDEX_SIZE) +#define PTRS_PER_PGD (1 << PGD_INDEX_SIZE) + +/* PMD_SHIFT determines what a second-level page table entry can map */ +#define PMD_SHIFT (PAGE_SHIFT + PTE_INDEX_SIZE) +#define PMD_SIZE (1UL << PMD_SHIFT) +#define PMD_MASK (~(PMD_SIZE-1)) + +/* With 4k base page size, hugepage PTEs go at the PMD level */ +#define MIN_HUGEPTE_SHIFT PMD_SHIFT + +/* PUD_SHIFT determines what a third-level page table entry can map */ +#define PUD_SHIFT (PMD_SHIFT + PMD_INDEX_SIZE) +#define PUD_SIZE (1UL << PUD_SHIFT) +#define PUD_MASK (~(PUD_SIZE-1)) + +/* PGDIR_SHIFT determines what a fourth-level page table entry can map */ +#define PGDIR_SHIFT (PUD_SHIFT + PUD_INDEX_SIZE)
arch/powerpc/include/asm/book3s/64/hash-4k.h line 49
Quoted 10 lines
arch/powerpc/include/asm/book3s/64/hash-4k.h +9
-0+#define PGDIR_SIZE (1UL << PGDIR_SHIFT) +#define PGDIR_MASK (~(PGDIR_SIZE-1)) + +/* Bits to mask out from a PMD to get to the PTE page */ +#define PMD_MASKED_BITS 0 +/* Bits to mask out from a PUD to get to the PMD page */ +#define PUD_MASKED_BITS 0 +/* Bits to mask out from a PGD to get to the PUD page */ +#define PGD_MASKED_BITS 0
Don't get why these are all 0?
Quoted 51 lines
arch/powerpc/include/asm/book3s/64/hash-4k.h +37
-0/* PTE bits */ #define _PAGE_HASHPTE 0x0400/* software: pte has an associated HPTE */ @@ -15,3 +62,40 @@/* shift to put page number into pte */ #define PTE_RPN_SHIFT (17) +#ifndef __ASSEMBLY__ +/* +* 4-level page tables related bits +*/ + +#define pgd_none (pgd) (!pgd_val(pgd)) +#define pgd_bad (pgd) (pgd_val(pgd) == 0) +#define pgd_present (pgd) (pgd_val(pgd) != 0) +#define pgd_clear (pgdp) (pgd_val(*(pgdp)) = 0) +#define pgd_page_vaddr (pgd) (pgd_val(pgd) & ~PGD_MASKED_BITS) + +static inline pte_t pgd_pte (pgd_t pgd) +{ +return __pte (pgd_val (pgd)); +} + +static inline pgd_t pte_pgd (pte_t pte) +{ +return __pgd (pte_val (pte)); +} +extern struct page *pgd_page (pgd_t pgd); + +#define pud_offset (pgdp, addr) \ + (((pud_t *) pgd_page_vaddr(*(pgdp))) + \ + (((addr) >> PUD_SHIFT) & (PTRS_PER_PUD - 1))) + +#define pud_ERROR (e) \ + pr_err("%s:%d: bad pud %08lx.\n", __FILE__, __LINE__, pud_val(e)) + +/* +* On all 4K setups, remap_4k_pfn() equates to remap_pfn_range() */ +#define remap_4k_pfn (vma, addr, pfn, prot) \ + remap_pfn_range((vma), (addr), (pfn), PAGE_SIZE, (prot)) + +#endif /* !__ASSEMBLY__ */ + +#endif /* _ASM_POWERPC_BOOK3S_64_HASH_4K_H */ arch/powerpc/include/asm/book3s/64/hash-64k.h +0
-1@@ -1,4 +1,35 @@ -/* To be include by pgtable-hash64.h only */
arch/powerpc/include/asm/book3s/64/hash-64k.h line 10
Quoted 10 lines
arch/powerpc/include/asm/book3s/64/hash-64k.h +10
-0+#ifndef _ASM_POWERPC_BOOK3S_64_HASH_64K_H +#define _ASM_POWERPC_BOOK3S_64_HASH_64K_H + +#include <asm-generic/pgtable-nopud.h> + +#define PTE_INDEX_SIZE 8 +#define PMD_INDEX_SIZE 10 +#define PUD_INDEX_SIZE 0 +#define PGD_INDEX_SIZE 12 +
OK.. So there is PGD but no PUD?
arch/powerpc/include/asm/book3s/64/hash-64k.h line 17
Quoted 7 lines
arch/powerpc/include/asm/book3s/64/hash-64k.h +7
-0+#define PTRS_PER_PTE (1 << PTE_INDEX_SIZE) +#define PTRS_PER_PMD (1 << PMD_INDEX_SIZE) +#define PTRS_PER_PGD (1 << PGD_INDEX_SIZE) + +/* With 4k base page size, hugepage PTEs go at the PMD level */ +#define MIN_HUGEPTE_SHIFT PAGE_SHIFT +
Typo, refers to 4k base page size.
Quoted 60 lines
arch/powerpc/include/asm/book3s/64/hash-64k.h +26
-2+/* PMD_SHIFT determines what a second-level page table entry can map */ +#define PMD_SHIFT (PAGE_SHIFT + PTE_INDEX_SIZE) +#define PMD_SIZE (1UL << PMD_SHIFT) +#define PMD_MASK (~(PMD_SIZE-1)) + +/* PGDIR_SHIFT determines what a third-level page table entry can map */ +#define PGDIR_SHIFT (PMD_SHIFT + PMD_INDEX_SIZE) +#define PGDIR_SIZE (1UL << PGDIR_SHIFT) +#define PGDIR_MASK (~(PGDIR_SIZE-1)) + +/* Bits to mask out from a PMD to get to the PTE page */ +/* PMDs point to PTE table fragments which are 4K aligned. */ +#define PMD_MASKED_BITS 0xfff +/* Bits to mask out from a PGD/PUD to get to the PMD page */ +#define PUD_MASKED_BITS 0x1ff/* Additional PTE bits (don't change without checking asm in hash_low.S) */ #define _PAGE_SPECIAL 0x00000400/* software: special page */ @@ -74,8 +105,8 @@ static inline unsigned long __rpte_to_hidx(real_pte_t rpte, unsigned long index)#define __rpte_to_pte (r) ((r).pte)#define __rpte_sub_valid (rpte, index) \ (pte_val(rpte.pte) & (_PAGE_HPTE_SUB0 >> (index))) - -/* Trick: we set __end to va + 64k, which happens works for +/* +* Trick: we set __end to va + 64k, which happens works for * a 16M page as well as we want only one iteration */ #define pte_iterate_hashed_subpages (rpte, psize, vpn, index, shift) \@@ -99,4 +130,13 @@ static inline unsigned long __rpte_to_hidx(real_pte_t rpte, unsigned long index) remap_pfn_range((vma), (addr), (pfn), PAGE_SIZE, \ __pgprot(pgprot_val((prot)) | _PAGE_4K_PFN))) +#define PTE_TABLE_SIZE (sizeof(real_pte_t) << PTE_INDEX_SIZE) +#define PMD_TABLE_SIZE (sizeof(pmd_t) << PMD_INDEX_SIZE) +#define PGD_TABLE_SIZE (sizeof(pgd_t) << PGD_INDEX_SIZE) + +#define pgd_pte (pgd) (pud_pte(((pud_t){ pgd }))) +#define pte_pgd (pte) ((pgd_t)pte_pud(pte)) +#endif /* __ASSEMBLY__ */ + +#endif /* _ASM_POWERPC_BOOK3S_64_HASH_64K_H */ arch/powerpc/include/asm/book3s/64/pgtable.h +1
-5@@ -5,11 +5,7 @@* the ppc64 hashed page table. */ -#ifdef CONFIG_PPC_64K_PAGES -#include <asm/pgtable-ppc64-64k.h> -#else -#include <asm/pgtable-ppc64-4k.h> -#endif +#include <asm/book3s/64/hash.h> #include <asm/barrier.h> #define FIRST_USER_ADDRESS 0UL
PATCH v6 6/35 powerpc/mm: Delete booke bits from book3s
In reply to Aneesh Kumar K.VWe also move __ASSEMBLY__ towards the end of header. This avoid having #ifndef __ASSEMBLY___ all over the header
--- arch/powerpc/include/asm/book3s/32/pgtable.h | 93 +++++++--------------------- arch/powerpc/include/asm/book3s/64/pgtable.h | 86 +++++++------------------ arch/powerpc/include/asm/book3s/pgtable.h | 1 + 3 files changed, 49 insertions(+), 131 deletions(-)
arch/powerpc/include/asm/book3s/32/pgtable.h +24 -69
@@ -3,18 +3,10 @@#include <asm-generic/pgtable-nopmd.h> -#ifndef __ASSEMBLY__ -#include <linux/sched.h> -#include <linux/threads.h> -#include <asm/io.h> /* For sub-arch specific PPC_PIN_SIZE */ - -extern unsigned long ioremap_bot; - -#ifdef CONFIG_44x -extern int icache_44x_need_flush; -#endif +#include <asm/book3s/32/hash.h> -#endif /* __ASSEMBLY__ */ +/* And here we include common definitions */ +#include <asm/pte-common.h> /* * The normal case is that PTEs are 32-bits and we have a 1-page
@@ -31,28 +23,11 @@ extern int icache_44x_need_flush;#define PGDIR_SIZE (1UL << PGDIR_SHIFT)#define PGDIR_MASK (~(PGDIR_SIZE-1)) -/* -* entries per page directory level: our page-table tree is two-level, so -* we don't really have any PMD directory. -*/ -#ifndef __ASSEMBLY__ -#define PTE_TABLE_SIZE (sizeof(pte_t) << PTE_SHIFT) -#define PGD_TABLE_SIZE (sizeof(pgd_t) << (32 - PGDIR_SHIFT)) -#endif /* __ASSEMBLY__ */ -#define PTRS_PER_PTE (1 << PTE_SHIFT)#define PTRS_PER_PMD 1#define PTRS_PER_PGD (1 << (32 - PGDIR_SHIFT))#define USER_PTRS_PER_PGD (TASK_SIZE / PGDIR_SIZE) -#define FIRST_USER_ADDRESS 0UL - -#define pte_ERROR (e) \ - pr_err("%s:%d: bad pte %llx.\n", __FILE__, __LINE__, \ - (unsigned long long)pte_val(e)) -#define pgd_ERROR (e) \ - pr_err("%s:%d: bad pgd %08lx.\n", __FILE__, __LINE__, pgd_val(e)) -/* * This is the bottom of the PKMAP area with HIGHMEM or an arbitrary * value (for now) on others, from where we can start layout kernel
@@ -100,30 +75,30 @@ extern int icache_44x_need_flush;#endif #define VMALLOC_END ioremap_bot +#ifndef __ASSEMBLY__ +#include <linux/sched.h> +#include <linux/threads.h> +#include <asm/io.h> /* For sub-arch specific PPC_PIN_SIZE */ + +extern unsigned long ioremap_bot; + +/* +* entries per page directory level: our page-table tree is two-level, so +* we don't really have any PMD directory. +*/ +#define PTE_TABLE_SIZE (sizeof(pte_t) << PTE_SHIFT) +#define PGD_TABLE_SIZE (sizeof(pgd_t) << (32 - PGDIR_SHIFT)) + +#define pte_ERROR (e) \ + pr_err("%s:%d: bad pte %llx.\n", __FILE__, __LINE__, \ + (unsigned long long)pte_val(e)) +#define pgd_ERROR (e) \ + pr_err("%s:%d: bad pgd %08lx.\n", __FILE__, __LINE__, pgd_val(e))/* * Bits in a linux-style PTE. These match the bits in the * (hardware-defined) PowerPC PTE as closely as possible. */ -#if defined(CONFIG_40x) -#include <asm/pte-40x.h> -#elif defined(CONFIG_44x) -#include <asm/pte-44x.h> -#elif defined(CONFIG_FSL_BOOKE ) && defined(CONFIG_PTE_64BIT ) -#include <asm/pte-book3e.h> -#elif defined(CONFIG_FSL_BOOKE ) -#include <asm/pte-fsl-booke.h> -#elif defined(CONFIG_8xx) -#include <asm/pte-8xx.h> -#else /* CONFIG_6xx */ -#include <asm/book3s/32/hash.h> -#endif - -/* And here we include common definitions */ -#include <asm/pte-common.h> - -#ifndef __ASSEMBLY__ -#define pte_clear (mm, addr, ptep) \ do { pte_update(ptep, ~_PAGE_HASHPTE, 0); } while (0)
@@ -167,7 +142,6 @@ static inline unsigned long pte_update(pte_t *p, unsigned long clr, unsigned long set) { -#ifdef PTE_ATOMIC_UPDATES unsigned long old, tmp; __asm__ __volatile__("\
@@ -180,15 +154,7 @@ static inline unsigned long pte_update(pte_t *p, : "=&r" (old), "=&r" (tmp), "=m" (*p) : "r" (p), "r" (clr), "r" (set), "m" (*p) : "cc" ); -#else /* PTE_ATOMIC_UPDATES */ -unsigned long old =pte_val (*p); - *p =__pte ((old & ~clr) | set); -#endif /* !PTE_ATOMIC_UPDATES */ - -#ifdef CONFIG_44x -if ((old & _PAGE_USER) && (old & _PAGE_EXEC)) - icache_44x_need_flush =1 ; -#endif + return old; }#else /* CONFIG_PTE_64BIT */
@@ -196,7 +162,6 @@ static inline unsigned long long pte_update(pte_t *p, unsigned long clr, unsigned long set) { -#ifdef PTE_ATOMIC_UPDATES unsigned long long old;unsigned long tmp;
@@ -211,15 +176,7 @@ static inline unsigned long long pte_update(pte_t *p, : "=&r" (old), "=&r" (tmp), "=m" (*p) : "r" (p), "r" ((unsigned long)(p) + 4), "r" (clr), "r" (set), "m" (*p) : "cc" ); -#else /* PTE_ATOMIC_UPDATES */ -unsigned long long old =pte_val (*p); - *p =__pte ((old & ~(unsigned long long )clr) | set); -#endif /* !PTE_ATOMIC_UPDATES */ - -#ifdef CONFIG_44x -if ((old & _PAGE_USER) && (old & _PAGE_EXEC)) - icache_44x_need_flush =1 ; -#endif + return old; }#endif /* CONFIG_PTE_64BIT */
@@ -233,12 +190,10 @@ static inline int __ptep_test_and_clear_young(unsigned int context, unsigned lon {unsigned long old; old =pte_update (ptep, _PAGE_ACCESSED,0 ); -#if _PAGE_HASHPTE !=0 if (old & _PAGE_HASHPTE) {unsigned long ptephys =__pa (ptep) &PAGE_MASK ;flush_hash_pages (context, addr, ptephys,1 ); } -#endif return (old & _PAGE_ACCESSED) !=0 ; }#define ptep_test_and_clear_young (__vma, __addr, __ptep) \
arch/powerpc/include/asm/book3s/64/pgtable.h +24 -62
174 lines not on this page · raw message
arch/powerpc/include/asm/book3s/pgtable.h +1 -0
@@ -7,4 +7,5 @@#include <asm/book3s/32/pgtable.h> #endif +#define FIRST_USER_ADDRESS 0UL#endif
Signature · 1 line
2.5.0
PATCH v6 7/35 powerpc/mm: Don't have generic headers introduce functions touching pte bits
In reply to Aneesh Kumar K.VWe are going to drop pte_common.h in the later patch. The idea is to enable hash code not require to define all PTE bits. Having PTE bits defined in pte_common.h made the code unnecessarily complex.
---
arch/powerpc/include/asm/book3s/pgtable.h | 176 +++++++++++++++++++
.../include/asm/{pgtable.h => pgtable-book3e.h} | 96 +----------
arch/powerpc/include/asm/pgtable.h | 192 +--------------------
3 files changed, 185 insertions(+), 279 deletions(-)
copy arch/powerpc/include/asm/{pgtable.h => pgtable-book3e.h} (70%)arch/powerpc/include/asm/book3s/pgtable.h +176 -0
@@ -8,4 +8,180 @@#endif #define FIRST_USER_ADDRESS 0UL +#ifndef __ASSEMBLY__ + +/* Generic accessors to PTE bits */ +static inline int pte_write (pte_t pte) +{ +return (pte_val (pte) & (_PAGE_RW | _PAGE_RO)) != _PAGE_RO; +} +static inline int pte_dirty (pte_t pte) {return pte_val (pte) & _PAGE_DIRTY; } +static inline int pte_young (pte_t pte) {return pte_val (pte) & _PAGE_ACCESSED; } +static inline int pte_special (pte_t pte) {return pte_val (pte) & _PAGE_SPECIAL; } +static inline int pte_none (pte_t pte) {return (pte_val (pte) & ~_PTE_NONE_MASK) ==0 ; } +static inline pgprot_t pte_pgprot (pte_t pte) {return __pgprot (pte_val (pte) &PAGE_PROT_BITS ); } + +#ifdef CONFIG_NUMA_BALANCING +/* +* These work without NUMA balancing but the kernel does not care. See the +* comment in include/asm-generic/pgtable.h . On powerpc, this will only +* work for user pages and always return true for kernel pages. +*/ +static inline int pte_protnone (pte_t pte) +{ +return (pte_val (pte) & + (_PAGE_PRESENT | _PAGE_USER)) == _PAGE_PRESENT; +} + +static inline int pmd_protnone (pmd_t pmd) +{ +return pte_protnone (pmd_pte (pmd)); +} +#endif /* CONFIG_NUMA_BALANCING */ + +static inline int pte_present (pte_t pte) +{ +return pte_val (pte) & _PAGE_PRESENT; +} + +/* Conversion functions: convert a page and protection to a page entry, +* and a page entry and page directory to the page they refer to. +* +* Even if PTEs can be unsigned long long, a PFN is always an unsigned +* long for now. +*/ +static inline pte_t pfn_pte (unsigned long pfn,pgprot_t pgprot) { +return __pte (((pte_basic_t)(pfn) <<PTE_RPN_SHIFT ) | +pgprot_val (pgprot)); } +static inline unsigned long pte_pfn (pte_t pte) { +return pte_val (pte) >>PTE_RPN_SHIFT ; } + +/* Generic modifiers for PTE bits */ +static inline pte_t pte_wrprotect (pte_t pte) { +pte_val (pte) &= ~(_PAGE_RW | _PAGE_HWWRITE); +pte_val (pte) |= _PAGE_RO;return pte; } +static inline pte_t pte_mkclean (pte_t pte) { +pte_val (pte) &= ~(_PAGE_DIRTY | _PAGE_HWWRITE);return pte; } +static inline pte_t pte_mkold (pte_t pte) { +pte_val (pte) &= ~_PAGE_ACCESSED;return pte; } +static inline pte_t pte_mkwrite (pte_t pte) { +pte_val (pte) &= ~_PAGE_RO; +pte_val (pte) |= _PAGE_RW;return pte; } +static inline pte_t pte_mkdirty (pte_t pte) { +pte_val (pte) |= _PAGE_DIRTY;return pte; } +static inline pte_t pte_mkyoung (pte_t pte) { +pte_val (pte) |= _PAGE_ACCESSED;return pte; } +static inline pte_t pte_mkspecial (pte_t pte) { +pte_val (pte) |= _PAGE_SPECIAL;return pte; } +static inline pte_t pte_mkhuge (pte_t pte) { +return pte; } +static inline pte_t pte_modify (pte_t pte,pgprot_t newprot) +{ +pte_val (pte) = (pte_val (pte) & _PAGE_CHG_MASK) |pgprot_val (newprot); +return pte; +} + + +/* Insert a PTE, top-level function is out of line. It uses an inline +* low level function in the respective pgtable-* files +*/ +extern void set_pte_at (struct mm_struct *mm,unsigned long addr,pte_t *ptep, +pte_t pte); + +/* This low level function performs the actual PTE insertion +* Setting the PTE depends on the MMU type and other factors. It's +* an horrible mess that I'm not going to try to clean up now but +* I'm keeping it in one place rather than spread around +*/ +static inline void __set_pte_at (struct mm_struct *mm,unsigned long addr, +pte_t *ptep,pte_t pte,int percpu) +{ +#if defined(CONFIG_PPC_STD_MMU_32 ) && defined(CONFIG_SMP ) && !defined(CONFIG_PTE_64BIT ) +/* First case is 32-bit Hash MMU in SMP mode with 32-bit PTEs. We use the +* helper pte_update() which does an atomic update. We need to do that +* because a concurrent invalidation can clear _PAGE_HASHPTE. If it's a +* per-CPU PTE such as a kmap_atomic, we do a simple update preserving +* the hash bits instead (ie, same as the non-SMP case) +*/ +if (percpu) + *ptep =__pte ((pte_val (*ptep) & _PAGE_HASHPTE) + | (pte_val (pte) & ~_PAGE_HASHPTE)); +else +pte_update (ptep, ~_PAGE_HASHPTE,pte_val (pte)); + +#elif defined(CONFIG_PPC32 ) && defined(CONFIG_PTE_64BIT ) +/* Second case is 32-bit with 64-bit PTE. In this case, we +* can just store as long as we do the two halves in the right order +* with a barrier in between. This is possible because we take care, +* in the hash code, to pre-invalidate if the PTE was already hashed, +* which synchronizes us with any concurrent invalidation. +* In the percpu case, we also fallback to the simple update preserving +* the hash bits +*/ +if (percpu) { + *ptep =__pte ((pte_val (*ptep) & _PAGE_HASHPTE) + | (pte_val (pte) & ~_PAGE_HASHPTE)); +return ; + } +if (pte_val (*ptep) & _PAGE_HASHPTE) +flush_hash_entry (mm, ptep, addr); + __asm__ __volatile__("\ +stw%U0%X0 %2,%0\n\ +eieio\n\ +stw%U0%X0 %L2,%1" + :"=m" (*ptep),"=m" (*((unsigned char *)ptep+4 )) + :"r" (pte) :"memory" ); + +#elif defined(CONFIG_PPC_STD_MMU_32 ) +/* Third case is 32-bit hash table in UP mode, we need to preserve +* the _PAGE_HASHPTE bit since we may not have invalidated the previous +* translation in the hash yet (done in a subsequent flush_tlb_xxx()) +* and see we need to keep track that this PTE needs invalidating +*/ + *ptep =__pte ((pte_val (*ptep) & _PAGE_HASHPTE) + | (pte_val (pte) & ~_PAGE_HASHPTE)); + +#else +/* Anything else just stores the PTE normally. That covers all 64-bit +* cases, and 32-bit non-hash with 32-bit PTEs. +*/ + *ptep = pte; +#endif +} + + +#define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS +extern int ptep_set_access_flags (struct vm_area_struct *vma,unsigned long address, +pte_t *ptep,pte_t entry,int dirty); + +/* +* Macro to mark a page protection value as "uncacheable". +*/ + +#define _PAGE_CACHE_CTL (_PAGE_COHERENT | _PAGE_GUARDED | _PAGE_NO_CACHE | \ + _PAGE_WRITETHRU) + +#define pgprot_noncached (prot) (__pgprot((pgprot_val(prot) & ~_PAGE_CACHE_CTL) | \ + _PAGE_NO_CACHE | _PAGE_GUARDED)) + +#define pgprot_noncached_wc (prot) (__pgprot((pgprot_val(prot) & ~_PAGE_CACHE_CTL) | \ + _PAGE_NO_CACHE)) + +#define pgprot_cached (prot) (__pgprot((pgprot_val(prot) & ~_PAGE_CACHE_CTL) | \ + _PAGE_COHERENT)) + +#define pgprot_cached_wthru (prot) (__pgprot((pgprot_val(prot) & ~_PAGE_CACHE_CTL) | \ + _PAGE_COHERENT | _PAGE_WRITETHRU)) + +#define pgprot_cached_noncoherent (prot) \ + (__pgprot(pgprot_val(prot) & ~_PAGE_CACHE_CTL)) + +#define pgprot_writecombine pgprot_noncached_wc + +struct file ; +extern pgprot_t phys_mem_access_prot (struct file *file,unsigned long pfn, +unsigned long size,pgprot_t vma_prot); +#define __HAVE_PHYS_MEM_ACCESS_PROT + +#endif /* __ASSEMBLY__ */ #endif
arch/powerpc/include/asm/pgtable-book3e.h renamed from arch/powerpc/include/asm/pgtable.h +8 -88
118 lines not on this page · raw message
arch/powerpc/include/asm/pgtable.h +1 -191
217 lines not on this page · raw message
Signature · 1 line
2.5.0
Re: [PATCH V6 07/35] powerpc/mm: Don't have generic headers introduce functions touching pte bits
Quoted 48 lines · Aneesh Kumar K.V
We are going to drop
pte_common.hin the later patch. The idea is to enable hash code not require to define all PTE bits. Having PTE bits defined inpte_common.hmade the code unnecessarily complex.Acked-by Scott Wood <[email protected]>Signed-off-by Aneesh Kumar K.V <[email protected]>--- arch/powerpc/include/asm/book3s/pgtable.h | 176 +++++++++++++++++++ .../include/asm/{pgtable.h => pgtable-book3e.h} | 96 +---------- arch/powerpc/include/asm/pgtable.h | 192 +-------------------- 3 files changed, 185 insertions(+), 279 deletions(-) copy arch/powerpc/include/asm/{pgtable.h => pgtable-book3e.h} (70%)arch/powerpc/include/asm/book3s/pgtable.h +27
-0@@ -8,4 +8,180 @@#endif #define FIRST_USER_ADDRESS 0UL +#ifndef __ASSEMBLY__ + +/* Generic accessors to PTE bits */ +static inline int pte_write(pte_t pte) +{ + return (pte_val(pte) & (_PAGE_RW | _PAGE_RO)) != _PAGE_RO; +} +static inline int pte_dirty(pte_t pte) { return pte_val(pte) & _PAGE_DIRTY; } +static inline int pte_young(pte_t pte) { return pte_val(pte) & _PAGE_ACCESSED; } +static inline int pte_special(pte_t pte) { return pte_val(pte) & _PAGE_SPECIAL; } +static inline int pte_none(pte_t pte) { return (pte_val(pte) & ~_PTE_NONE_MASK) == 0; } +static inline pgprot_t pte_pgprot(pte_t pte) { return __pgprot(pte_val(pte) & PAGE_PROT_BITS); } + +#ifdef CONFIG_NUMA_BALANCING +/* + * These work without NUMA balancing but the kernel does not care. See the + * comment in include/asm-generic/pgtable.h . On powerpc, this will only + * work for user pages and always return true for kernel pages. + */ +static inline int pte_protnone(pte_t pte) +{ + return (pte_val(pte) & + (_PAGE_PRESENT | _PAGE_USER)) == _PAGE_PRESENT; +} + +static inline int pmd_protnone(pmd_t pmd) +{
arch/powerpc/include/asm/book3s/pgtable.h line 47
Quoted 10 lines
arch/powerpc/include/asm/book3s/pgtable.h +10
-0+ return pte_protnone(pmd_pte(pmd)); +} +#endif /* CONFIG_NUMA_BALANCING */ + +static inline int pte_present (pte_t pte) +{ +return pte_val (pte) & _PAGE_PRESENT; +} + +/* Conversion functions: convert a page and protection to a page entry,
Typo.. comment style, comment starts on the same line for multi-block comments?
Quoted 28 lines
arch/powerpc/include/asm/book3s/pgtable.h +28
-0+* and a page entry and page directory to the page they refer to. +* +* Even if PTEs can be unsigned long long, a PFN is always an unsigned +* long for now. +*/ +static inline pte_t pfn_pte (unsigned long pfn,pgprot_t pgprot) { +return __pte (((pte_basic_t)(pfn) <<PTE_RPN_SHIFT ) | +pgprot_val (pgprot)); } +static inline unsigned long pte_pfn (pte_t pte) { +return pte_val (pte) >>PTE_RPN_SHIFT ; } + +/* Generic modifiers for PTE bits */ +static inline pte_t pte_wrprotect (pte_t pte) { +pte_val (pte) &= ~(_PAGE_RW | _PAGE_HWWRITE); +pte_val (pte) |= _PAGE_RO;return pte; } +static inline pte_t pte_mkclean (pte_t pte) { +pte_val (pte) &= ~(_PAGE_DIRTY | _PAGE_HWWRITE);return pte; } +static inline pte_t pte_mkold (pte_t pte) { +pte_val (pte) &= ~_PAGE_ACCESSED;return pte; } +static inline pte_t pte_mkwrite (pte_t pte) { +pte_val (pte) &= ~_PAGE_RO; +pte_val (pte) |= _PAGE_RW;return pte; } +static inline pte_t pte_mkdirty (pte_t pte) { +pte_val (pte) |= _PAGE_DIRTY;return pte; } +static inline pte_t pte_mkyoung (pte_t pte) { +pte_val (pte) |= _PAGE_ACCESSED;return pte; } +static inline pte_t pte_mkspecial (pte_t pte) { +pte_val (pte) |= _PAGE_SPECIAL;return pte; }
arch/powerpc/include/asm/book3s/pgtable.h line 85
Quoted 10 lines
arch/powerpc/include/asm/book3s/pgtable.h +10
-0+static inline pte_t pte_mkhuge (pte_t pte) { +return pte; } +static inline pte_t pte_modify (pte_t pte,pgprot_t newprot) +{ +pte_val (pte) = (pte_val (pte) & _PAGE_CHG_MASK) |pgprot_val (newprot); +return pte; +} + + +/* Insert a PTE, top-level function is out of line. It uses an inline
Comment style?
Quoted 215 lines
arch/powerpc/include/asm/book3s/pgtable.h +101
-0+* low level function in the respective pgtable-* files +*/ +extern void set_pte_at (struct mm_struct *mm,unsigned long addr,pte_t *ptep, +pte_t pte); + +/* This low level function performs the actual PTE insertion +* Setting the PTE depends on the MMU type and other factors. It's +* an horrible mess that I'm not going to try to clean up now but +* I'm keeping it in one place rather than spread around +*/ +static inline void __set_pte_at (struct mm_struct *mm,unsigned long addr, +pte_t *ptep,pte_t pte,int percpu) +{ +#if defined(CONFIG_PPC_STD_MMU_32 ) && defined(CONFIG_SMP ) && !defined(CONFIG_PTE_64BIT ) +/* First case is 32-bit Hash MMU in SMP mode with 32-bit PTEs. We use the +* helper pte_update() which does an atomic update. We need to do that +* because a concurrent invalidation can clear _PAGE_HASHPTE. If it's a +* per-CPU PTE such as a kmap_atomic, we do a simple update preserving +* the hash bits instead (ie, same as the non-SMP case) +*/ +if (percpu) + *ptep =__pte ((pte_val (*ptep) & _PAGE_HASHPTE) + | (pte_val (pte) & ~_PAGE_HASHPTE)); +else +pte_update (ptep, ~_PAGE_HASHPTE,pte_val (pte)); + +#elif defined(CONFIG_PPC32 ) && defined(CONFIG_PTE_64BIT ) +/* Second case is 32-bit with 64-bit PTE. In this case, we +* can just store as long as we do the two halves in the right order +* with a barrier in between. This is possible because we take care, +* in the hash code, to pre-invalidate if the PTE was already hashed, +* which synchronizes us with any concurrent invalidation. +* In the percpu case, we also fallback to the simple update preserving +* the hash bits +*/ +if (percpu) { + *ptep =__pte ((pte_val (*ptep) & _PAGE_HASHPTE) + | (pte_val (pte) & ~_PAGE_HASHPTE)); +return ; + } +if (pte_val (*ptep) & _PAGE_HASHPTE) +flush_hash_entry (mm, ptep, addr); + __asm__ __volatile__("\ +stw%U0%X0 %2,%0\n\ +eieio\n\ +stw%U0%X0 %L2,%1" + :"=m" (*ptep),"=m" (*((unsigned char *)ptep+4 )) + :"r" (pte) :"memory" ); + +#elif defined(CONFIG_PPC_STD_MMU_32 ) +/* Third case is 32-bit hash table in UP mode, we need to preserve +* the _PAGE_HASHPTE bit since we may not have invalidated the previous +* translation in the hash yet (done in a subsequent flush_tlb_xxx()) +* and see we need to keep track that this PTE needs invalidating +*/ + *ptep =__pte ((pte_val (*ptep) & _PAGE_HASHPTE) + | (pte_val (pte) & ~_PAGE_HASHPTE)); + +#else +/* Anything else just stores the PTE normally. That covers all 64-bit +* cases, and 32-bit non-hash with 32-bit PTEs. +*/ + *ptep = pte; +#endif +} + + +#define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS +extern int ptep_set_access_flags (struct vm_area_struct *vma,unsigned long address, +pte_t *ptep,pte_t entry,int dirty); + +/* +* Macro to mark a page protection value as "uncacheable". +*/ + +#define _PAGE_CACHE_CTL (_PAGE_COHERENT | _PAGE_GUARDED | _PAGE_NO_CACHE | \ + _PAGE_WRITETHRU) + +#define pgprot_noncached (prot) (__pgprot((pgprot_val(prot) & ~_PAGE_CACHE_CTL) | \ + _PAGE_NO_CACHE | _PAGE_GUARDED)) + +#define pgprot_noncached_wc (prot) (__pgprot((pgprot_val(prot) & ~_PAGE_CACHE_CTL) | \ + _PAGE_NO_CACHE)) + +#define pgprot_cached (prot) (__pgprot((pgprot_val(prot) & ~_PAGE_CACHE_CTL) | \ + _PAGE_COHERENT)) + +#define pgprot_cached_wthru (prot) (__pgprot((pgprot_val(prot) & ~_PAGE_CACHE_CTL) | \ + _PAGE_COHERENT | _PAGE_WRITETHRU)) + +#define pgprot_cached_noncoherent (prot) \ + (__pgprot(pgprot_val(prot) & ~_PAGE_CACHE_CTL)) + +#define pgprot_writecombine pgprot_noncached_wc + +struct file ; +extern pgprot_t phys_mem_access_prot (struct file *file,unsigned long pfn, +unsigned long size,pgprot_t vma_prot); +#define __HAVE_PHYS_MEM_ACCESS_PROT + +#endif /* __ASSEMBLY__ */ #endif arch/powerpc/include/asm/pgtable-book3e.h renamed from arch/powerpc/include/asm/pgtable.h +7
-75@@ -1,41 +1,19 @@ -#ifndef _ASM_POWERPC_PGTABLE_H -#define _ASM_POWERPC_PGTABLE_H -#ifdef __KERNEL__ +#ifndef _ASM_POWERPC_PGTABLE_BOOK3E_H +#define _ASM_POWERPC_PGTABLE_BOOK3E_H -#ifndef __ASSEMBLY__ -#include <linux/mmdebug.h> -#include <linux/mmzone.h> -#include <asm/processor.h> /* For TASK_SIZE */ -#include <asm/mmu.h> -#include <asm/page.h> - -struct mm_struct ; - -#endif /* !__ASSEMBLY__ */ - -#ifdef CONFIG_PPC_BOOK3S -#include <asm/book3s/pgtable.h> -#else #if defined(CONFIG_PPC64 ) -# include <asm/pgtable-ppc64.h> +#include <asm/pgtable-ppc64.h> #else -# include <asm/pgtable-ppc32.h> +#include <asm/pgtable-ppc32.h> #endif -#endif /* !CONFIG_PPC_BOOK3S */ - -/* -* We save the slot number & secondary bit in the second half of the -* PTE page. We use the 8 bytes per each pte entry. -*/ -#define PTE_PAGE_HIDX_OFFSET (PTRS_PER_PTE * 8)#ifndef __ASSEMBLY__ -#include <asm/tlbflush.h> -/* Generic accessors to PTE bits */ static inline int pte_write (pte_t pte) -{return (pte_val (pte) & (_PAGE_RW | _PAGE_RO)) != _PAGE_RO; } +{ +return (pte_val (pte) & (_PAGE_RW | _PAGE_RO)) != _PAGE_RO; +}static inline int pte_dirty (pte_t pte) {return pte_val (pte) & _PAGE_DIRTY; }static inline int pte_young (pte_t pte) {return pte_val (pte) & _PAGE_ACCESSED; }static inline int pte_special (pte_t pte) {return pte_val (pte) & _PAGE_SPECIAL; }@@ -77,10 +55,6 @@ static inline pte_t pfn_pte(unsigned long pfn, pgprot_t pgprot) {static inline unsigned long pte_pfn (pte_t pte) {return pte_val (pte) >>PTE_RPN_SHIFT ; } -/* Keep these as a macros to avoid include dependency mess */ -#define pte_page (x) pfn_to_page(pte_pfn(x)) -#define mk_pte (page, pgprot) pfn_pte(page_to_pfn(page), (pgprot)) -/* Generic modifiers for PTE bits */ static inline pte_t pte_wrprotect (pte_t pte) {pte_val (pte) &= ~(_PAGE_RW | _PAGE_HWWRITE);@@ -221,59 +195,5 @@ extern pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn, unsigned long size, pgprot_t vma_prot);#define __HAVE_PHYS_MEM_ACCESS_PROT -/* -* ZERO_PAGE is a global shared page that is always zero: used -* for zero-mapped memory areas etc.. -*/ -extern unsigned long empty_zero_page[]; -#define ZERO_PAGE (vaddr) (virt_to_page(empty_zero_page)) - -extern pgd_t swapper_pg_dir[]; - -void limit_zone_pfn (enum zone_type zone,unsigned long max_pfn); -int dma_pfn_limit_to_zone (u64 pfn_limit); -extern void paging_init (void ); - -/* -* kern_addr_valid is intended to indicate whether an address is a valid -* kernel address. Most 32-bit archs define it as always true (like this) -* but most 64-bit archs actually perform a test. What should we do here? -*/ -#define kern_addr_valid (addr) (1) - -#include <asm-generic/pgtable.h> - - -/* -* This gets called at the end of handling a page fault, when -* the kernel has put a new PTE into the page table for the process. -* We use it to ensure coherency between the i-cache and d-cache -* for the page which has just been mapped in. -* On machines which use an MMU hash table, we use this to put a -* corresponding HPTE into the hash table ahead of time, instead of -* waiting for the inevitable extra hash-table miss exception. -*/ -extern void update_mmu_cache (struct vm_area_struct *,unsigned long ,pte_t *); - -extern int gup_hugepte (pte_t *ptep,unsigned long sz,unsigned long addr, -unsigned long end,int write, -struct page **pages,int *nr); -#ifndef CONFIG_TRANSPARENT_HUGEPAGE -#define pmd_large (pmd) 0 -#define has_transparent_hugepage () 0 -#endif -pte_t *__find_linux_pte_or_hugepte(pgd_t *pgdir, unsigned long ea,
arch/powerpc/include/asm/pgtable-book3e.h extern pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn, · line 275
Quoted 10 lines
arch/powerpc/include/asm/pgtable-book3e.h renamed from arch/powerpc/include/asm/pgtable.h +0
-10- bool *is_thp, unsigned *shift); -static inline pte_t *find_linux_pte_or_hugepte (pgd_t *pgdir,unsigned long ea, -bool *is_thp,unsigned *shift) -{ -if (!arch_irqs_disabled ()) { -pr_info ("%s called with irq enabled\n" , __func__); -dump_stack (); - } -return __find_linux_pte_or_hugepte (pgdir, ea, is_thp, shift); -}
Where's this function gone?
Quoted 1 line
<snip>
Balbir Singh
PATCH v6 8/35 powerpc/mm: Drop pte-common.h from BOOK3S 64
In reply to Aneesh Kumar K.VWe copy only needed PTE bits define from pte-common.h to respective hash related header. This should greatly simply later patches in which we are going to change the pte format for hash config
--- arch/powerpc/include/asm/book3s/64/hash-4k.h | 1 + arch/powerpc/include/asm/book3s/64/hash.h | 2 + arch/powerpc/include/asm/book3s/64/pgtable.h | 104 ++++++++++++++++++++++++++- arch/powerpc/include/asm/book3s/pgtable.h | 16 ++--- 4 files changed, 111 insertions(+), 12 deletions(-)
arch/powerpc/include/asm/book3s/64/hash-4k.h +1 -0
@@ -62,6 +62,7 @@/* shift to put page number into pte */ #define PTE_RPN_SHIFT (17) +#define _PAGE_4K_PFN 0#ifndef __ASSEMBLY__/* * 4-level page tables related bits
arch/powerpc/include/asm/book3s/64/hash.h +2 -0
@@ -20,6 +20,7 @@#define _PAGE_EXEC 0x0004/* No execute on POWER4 and newer (we invert) */ #define _PAGE_GUARDED 0x0008/* We can derive Memory coherence from _PAGE_NO_CACHE */ +#define _PAGE_COHERENT 0x0#define _PAGE_NO_CACHE 0x0020/* I: cache inhibit */ #define _PAGE_WRITETHRU 0x0040/* W: cache write-through */ #define _PAGE_DIRTY 0x0080/* C: page changed */
@@ -30,6 +31,7 @@/* No separate kernel read-only */ #define _PAGE_KERNEL_RW (_PAGE_RW | _PAGE_DIRTY)/* user access blocked by key */ #define _PAGE_KERNEL_RO _PAGE_KERNEL_RW +#define _PAGE_KERNEL_RWX (_PAGE_DIRTY | _PAGE_RW | _PAGE_EXEC)/* Strong Access Ordering */ #define _PAGE_SAO (_PAGE_WRITETHRU | _PAGE_NO_CACHE | _PAGE_COHERENT)
arch/powerpc/include/asm/book3s/64/pgtable.h +101 -3
112 lines not on this page · raw message
arch/powerpc/include/asm/book3s/pgtable.h +7 -9
@@ -11,10 +11,7 @@#ifndef __ASSEMBLY__/* Generic accessors to PTE bits */ -static inline int pte_write (pte_t pte) -{ -return (pte_val (pte) & (_PAGE_RW | _PAGE_RO)) != _PAGE_RO; -} +static inline int pte_write (pte_t pte) {return !!(pte_val (pte) & _PAGE_RW);}static inline int pte_dirty (pte_t pte) {return pte_val (pte) & _PAGE_DIRTY; }static inline int pte_young (pte_t pte) {return pte_val (pte) & _PAGE_ACCESSED; }static inline int pte_special (pte_t pte) {return pte_val (pte) & _PAGE_SPECIAL; }
@@ -57,15 +54,16 @@ static inline unsigned long pte_pfn(pte_t pte) {return pte_val (pte) >>PTE_RPN_SHIFT ; }/* Generic modifiers for PTE bits */ -static inline pte_t pte_wrprotect (pte_t pte) { -pte_val (pte) &= ~(_PAGE_RW | _PAGE_HWWRITE); -pte_val (pte) |= _PAGE_RO;return pte; } +static inline pte_t pte_wrprotect (pte_t pte) +{ +pte_val (pte) &= ~_PAGE_RW; +return pte; +}static inline pte_t pte_mkclean (pte_t pte) { -pte_val (pte) &= ~(_PAGE_DIRTY | _PAGE_HWWRITE);return pte; } +pte_val (pte) &= ~_PAGE_DIRTY;return pte; }static inline pte_t pte_mkold (pte_t pte) {pte_val (pte) &= ~_PAGE_ACCESSED;return pte; }static inline pte_t pte_mkwrite (pte_t pte) { -pte_val (pte) &= ~_PAGE_RO;pte_val (pte) |= _PAGE_RW;return pte; }static inline pte_t pte_mkdirty (pte_t pte) {pte_val (pte) |= _PAGE_DIRTY;return pte; }
Signature · 1 line
2.5.0
PATCH v6 9/35 powerpc/mm: Don't use pte_val as lvalue
In reply to Aneesh Kumar K.VWe also convert few #define to static inline in this patch for better type checking
--- arch/powerpc/include/asm/book3s/pgtable.h | 118 +++++++++++++++++++++--------- arch/powerpc/include/asm/page.h | 10 ++- arch/powerpc/include/asm/pgtable-book3e.h | 68 ++++++++++++----- 3 files changed, 139 insertions(+), 57 deletions(-)
arch/powerpc/include/asm/book3s/pgtable.h +83 -35
@@ -12,9 +12,9 @@/* Generic accessors to PTE bits */ static inline int pte_write (pte_t pte) {return !!(pte_val (pte) & _PAGE_RW);} -static inline int pte_dirty (pte_t pte) {return pte_val (pte) & _PAGE_DIRTY; } -static inline int pte_young (pte_t pte) {return pte_val (pte) & _PAGE_ACCESSED; } -static inline int pte_special (pte_t pte) {return pte_val (pte) & _PAGE_SPECIAL; } +static inline int pte_dirty (pte_t pte) {return !!(pte_val (pte) & _PAGE_DIRTY); } +static inline int pte_young (pte_t pte) {return !!(pte_val (pte) & _PAGE_ACCESSED); } +static inline int pte_special (pte_t pte) {return !!(pte_val (pte) & _PAGE_SPECIAL); }static inline int pte_none (pte_t pte) {return (pte_val (pte) & ~_PTE_NONE_MASK) ==0 ; }static inline pgprot_t pte_pgprot (pte_t pte) {return __pgprot (pte_val (pte) &PAGE_PROT_BITS ); }
@@ -47,36 +47,61 @@ static inline int pte_present(pte_t pte)* Even if PTEs can be unsigned long long, a PFN is always an unsigned * long for now. */ -static inline pte_t pfn_pte (unsigned long pfn,pgprot_t pgprot) { +static inline pte_t pfn_pte (unsigned long pfn,pgprot_t pgprot) +{return __pte (((pte_basic_t)(pfn) <<PTE_RPN_SHIFT ) | -pgprot_val (pgprot)); } -static inline unsigned long pte_pfn (pte_t pte) { -return pte_val (pte) >>PTE_RPN_SHIFT ; } +pgprot_val (pgprot)); +} + +static inline unsigned long pte_pfn (pte_t pte) +{ +return pte_val (pte) >>PTE_RPN_SHIFT ; +}/* Generic modifiers for PTE bits */ static inline pte_t pte_wrprotect (pte_t pte) { -pte_val (pte) &= ~_PAGE_RW; +return __pte (pte_val (pte) & ~_PAGE_RW); +} + +static inline pte_t pte_mkclean (pte_t pte) +{ +return __pte (pte_val (pte) & ~_PAGE_DIRTY); +} + +static inline pte_t pte_mkold (pte_t pte) +{ +return __pte (pte_val (pte) & ~_PAGE_ACCESSED); +} + +static inline pte_t pte_mkwrite (pte_t pte) +{ +return __pte (pte_val (pte) | _PAGE_RW); +} + +static inline pte_t pte_mkdirty (pte_t pte) +{ +return __pte (pte_val (pte) | _PAGE_DIRTY); +} + +static inline pte_t pte_mkyoung (pte_t pte) +{ +return __pte (pte_val (pte) | _PAGE_ACCESSED); +} + +static inline pte_t pte_mkspecial (pte_t pte) +{ +return __pte (pte_val (pte) | _PAGE_SPECIAL); +} + +static inline pte_t pte_mkhuge (pte_t pte) +{return pte; } -static inline pte_t pte_mkclean (pte_t pte) { -pte_val (pte) &= ~_PAGE_DIRTY;return pte; } -static inline pte_t pte_mkold (pte_t pte) { -pte_val (pte) &= ~_PAGE_ACCESSED;return pte; } -static inline pte_t pte_mkwrite (pte_t pte) { -pte_val (pte) |= _PAGE_RW;return pte; } -static inline pte_t pte_mkdirty (pte_t pte) { -pte_val (pte) |= _PAGE_DIRTY;return pte; } -static inline pte_t pte_mkyoung (pte_t pte) { -pte_val (pte) |= _PAGE_ACCESSED;return pte; } -static inline pte_t pte_mkspecial (pte_t pte) { -pte_val (pte) |= _PAGE_SPECIAL;return pte; } -static inline pte_t pte_mkhuge (pte_t pte) { -return pte; } +static inline pte_t pte_modify (pte_t pte,pgprot_t newprot) { -pte_val (pte) = (pte_val (pte) & _PAGE_CHG_MASK) |pgprot_val (newprot); -return pte; +return __pte ((pte_val (pte) & _PAGE_CHG_MASK) |pgprot_val (newprot)); }
@@ -159,22 +184,45 @@ extern int ptep_set_access_flags(struct vm_area_struct *vma, unsigned long addre#define _PAGE_CACHE_CTL (_PAGE_COHERENT | _PAGE_GUARDED | _PAGE_NO_CACHE | \ _PAGE_WRITETHRU) -#define pgprot_noncached (prot) (__pgprot((pgprot_val(prot) & ~_PAGE_CACHE_CTL) | \ - _PAGE_NO_CACHE | _PAGE_GUARDED)) +#define pgprot_noncached pgprot_noncached +static inline pgprot_t pgprot_noncached (pgprot_t prot) +{ +return __pgprot ((pgprot_val (prot) & ~_PAGE_CACHE_CTL) | + _PAGE_NO_CACHE | _PAGE_GUARDED); +} -#define pgprot_noncached_wc (prot) (__pgprot((pgprot_val(prot) & ~_PAGE_CACHE_CTL) | \ - _PAGE_NO_CACHE)) +#define pgprot_noncached_wc pgprot_noncached_wc +static inline pgprot_t pgprot_noncached_wc (pgprot_t prot) +{ +return __pgprot ((pgprot_val (prot) & ~_PAGE_CACHE_CTL) | + _PAGE_NO_CACHE); +} -#define pgprot_cached (prot) (__pgprot((pgprot_val(prot) & ~_PAGE_CACHE_CTL) | \ - _PAGE_COHERENT)) +#define pgprot_cached pgprot_cached +static inline pgprot_t pgprot_cached (pgprot_t prot) +{ +return __pgprot ((pgprot_val (prot) & ~_PAGE_CACHE_CTL) | + _PAGE_COHERENT); +} -#define pgprot_cached_wthru (prot) (__pgprot((pgprot_val(prot) & ~_PAGE_CACHE_CTL) | \ - _PAGE_COHERENT | _PAGE_WRITETHRU)) +#define pgprot_cached_wthru pgprot_cached_wthru +static inline pgprot_t pgprot_cached_wthru (pgprot_t prot) +{ +return __pgprot ((pgprot_val (prot) & ~_PAGE_CACHE_CTL) | + _PAGE_COHERENT | _PAGE_WRITETHRU); +} -#define pgprot_cached_noncoherent (prot) \ - (__pgprot(pgprot_val(prot) & ~_PAGE_CACHE_CTL)) +#define pgprot_cached_noncoherent pgprot_cached_noncoherent +static inline pgprot_t pgprot_cached_noncoherent (pgprot_t prot) +{ +return __pgprot (pgprot_val (prot) & ~_PAGE_CACHE_CTL); +} -#define pgprot_writecombine pgprot_noncached_wc +#define pgprot_writecombine pgprot_writecombine +static inline pgprot_t pgprot_writecombine (pgprot_t prot) +{ +return pgprot_noncached_wc (prot); +}struct file ; extern pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
arch/powerpc/include/asm/page.h +8 -2
@@ -285,8 +285,11 @@ extern long long virt_phys_offset;/* PTE level */ typedef struct {pte_basic_t pte; }pte_t ; -#define pte_val (x) ((x).pte)#define __pte (x) ((pte_t) { (x) }) +static inline pte_basic_t pte_val (pte_t x) +{ +return x.pte; +}/* 64k pages additionally define a bigger "real PTE" type that gathers * the "second half" part of the PTE for pseudo 64k pages
@@ -328,8 +331,11 @@ typedef struct { unsigned long pgprot; } pgprot_t;*/ typedef pte_basic_t pte_t ; -#define pte_val (x) (x)#define __pte (x) (x) +static inline pte_basic_t pte_val (pte_t pte) +{ +return pte; +}#if defined(CONFIG_PPC_64K_PAGES ) && defined(CONFIG_PPC_STD_MMU_64 )typedef struct {pte_t pte;unsigned long hidx; }real_pte_t ;
arch/powerpc/include/asm/pgtable-book3e.h +48 -20
78 lines not on this page · raw message
Signature · 1 line
2.5.0
PATCH v6 10/35 powerpc/mm: Don't use pmd_val, pud_val and pgd_val as lvalue
In reply to Aneesh Kumar K.VWe convert them static inline function here as we did with pte_val in the previous patch
--- arch/powerpc/include/asm/book3s/32/pgtable.h | 6 ++++- arch/powerpc/include/asm/book3s/64/hash-4k.h | 6 ++++- arch/powerpc/include/asm/book3s/64/pgtable.h | 36 +++++++++++++++++++++------- arch/powerpc/include/asm/page.h | 34 +++++++++++++++++++------- arch/powerpc/include/asm/pgalloc-32.h | 34 +++++++++++++++++++------- arch/powerpc/include/asm/pgalloc-64.h | 17 +++++++++---- arch/powerpc/include/asm/pgtable-ppc32.h | 7 +++++- arch/powerpc/include/asm/pgtable-ppc64-4k.h | 6 ++++- arch/powerpc/include/asm/pgtable-ppc64.h | 36 +++++++++++++++++++++------- arch/powerpc/mm/40x_mmu.c | 10 ++++---- arch/powerpc/mm/pgtable_64.c | 19 +++++++-------- 11 files changed, 154 insertions(+), 57 deletions(-)
arch/powerpc/include/asm/book3s/32/pgtable.h +5 -1
@@ -105,7 +105,11 @@ extern unsigned long ioremap_bot;#define pmd_none (pmd) (!pmd_val(pmd))#define pmd_bad (pmd) (pmd_val(pmd) & _PMD_BAD)#define pmd_present (pmd) (pmd_val(pmd) & _PMD_PRESENT_MASK) -#define pmd_clear (pmdp) do { pmd_val(*(pmdp)) = 0; } while (0) +static inline void pmd_clear (pmd_t *pmdp) +{ + *pmdp =__pmd (0 ); +} +/* * When flushing the tlb entry for a page, we also need to flush the hash
arch/powerpc/include/asm/book3s/64/hash-4k.h +5 -1
@@ -71,9 +71,13 @@#define pgd_none (pgd) (!pgd_val(pgd))#define pgd_bad (pgd) (pgd_val(pgd) == 0)#define pgd_present (pgd) (pgd_val(pgd) != 0) -#define pgd_clear (pgdp) (pgd_val(*(pgdp)) = 0)#define pgd_page_vaddr (pgd) (pgd_val(pgd) & ~PGD_MASKED_BITS) +static inline void pgd_clear (pgd_t *pgdp) +{ + *pgdp =__pgd (0 ); +} +static inline pte_t pgd_pte (pgd_t pgd) {return __pte (pgd_val (pgd));
arch/powerpc/include/asm/book3s/64/pgtable.h +27 -9
@@ -236,21 +236,38 @@#define PMD_BAD_BITS (PTE_TABLE_SIZE-1)#define PUD_BAD_BITS (PMD_TABLE_SIZE-1) -#define pmd_set (pmdp, pmdval) (pmd_val(*(pmdp)) = (pmdval)) +static inline void pmd_set (pmd_t *pmdp,unsigned long val) +{ + *pmdp =__pmd (val); +} + +static inline void pmd_clear (pmd_t *pmdp) +{ + *pmdp =__pmd (0 ); +} + +#define pmd_none (pmd) (!pmd_val(pmd))#define pmd_bad (pmd) (!is_kernel_addr(pmd_val(pmd)) \ || (pmd_val(pmd) & PMD_BAD_BITS))#define pmd_present (pmd) (!pmd_none(pmd)) -#define pmd_clear (pmdp) (pmd_val(*(pmdp)) = 0)#define pmd_page_vaddr (pmd) (pmd_val(pmd) & ~PMD_MASKED_BITS)extern struct page *pmd_page (pmd_t pmd); -#define pud_set (pudp, pudval) (pud_val(*(pudp)) = (pudval)) +static inline void pud_set (pud_t *pudp,unsigned long val) +{ + *pudp =__pud (val); +} + +static inline void pud_clear (pud_t *pudp) +{ + *pudp =__pud (0 ); +} +#define pud_none (pud) (!pud_val(pud))#define pud_bad (pud) (!is_kernel_addr(pud_val(pud)) \ || (pud_val(pud) & PUD_BAD_BITS))#define pud_present (pud) (pud_val(pud) != 0) -#define pud_clear (pudp) (pud_val(*(pudp)) = 0)#define pud_page_vaddr (pud) (pud_val(pud) & ~PUD_MASKED_BITS)extern struct page *pud_page (pud_t pud);
@@ -265,8 +282,11 @@ static inline pud_t pte_pud(pte_t pte)return __pud (pte_val (pte)); }#define pud_write (pud) pte_write(pud_pte(pud)) -#define pgd_set (pgdp, pudp) ({pgd_val(*(pgdp)) = (unsigned long)(pudp);})#define pgd_write (pgd) pte_write(pgd_pte(pgd)) +static inline void pgd_set (pgd_t *pgdp,unsigned long val) +{ + *pgdp =__pgd (val); +}/* * Find an entry in a page-table-directory. We combine the address region
@@ -588,14 +608,12 @@ static inline pmd_t pmd_mkhuge(pmd_t pmd)static inline pmd_t pmd_mknotpresent (pmd_t pmd) { -pmd_val (pmd) &= ~_PAGE_PRESENT; -return pmd; +return __pmd (pmd_val (pmd) & ~_PAGE_PRESENT); }static inline pmd_t pmd_mksplitting (pmd_t pmd) { -pmd_val (pmd) |= _PAGE_SPLITTING; -return pmd; +return __pmd (pmd_val (pmd) | _PAGE_SPLITTING); }#define __HAVE_ARCH_PMD_SAME
arch/powerpc/include/asm/page.h +26 -8
@@ -303,21 +303,30 @@ typedef struct { pte_t pte; } real_pte_t;/* PMD level */ #ifdef CONFIG_PPC64 typedef struct {unsigned long pmd; }pmd_t ; -#define pmd_val (x) ((x).pmd)#define __pmd (x) ((pmd_t) { (x) }) +static inline unsigned long pmd_val (pmd_t x) +{ +return x.pmd; +}/* PUD level exusts only on 4k pages */ #ifndef CONFIG_PPC_64K_PAGES typedef struct {unsigned long pud; }pud_t ; -#define pud_val (x) ((x).pud)#define __pud (x) ((pud_t) { (x) }) +static inline unsigned long pud_val (pud_t x) +{ +return x.pud; +}#endif /* !CONFIG_PPC_64K_PAGES */ #endif /* CONFIG_PPC64 */ /* PGD level */ typedef struct {unsigned long pgd; }pgd_t ; -#define pgd_val (x) ((x).pgd)#define __pgd (x) ((pgd_t) { (x) }) +static inline unsigned long pgd_val (pgd_t x) +{ +return x.pgd; +}/* Page protection bits */ typedef struct {unsigned long pgprot; }pgprot_t ;
@@ -346,22 +355,31 @@ typedef pte_t real_pte_t;#ifdef CONFIG_PPC64 typedef unsigned long pmd_t ; -#define pmd_val (x) (x)#define __pmd (x) (x) +static inline unsigned long pmd_val (pmd_t pmd) +{ +return pmd; +}#ifndef CONFIG_PPC_64K_PAGES typedef unsigned long pud_t ; -#define pud_val (x) (x)#define __pud (x) (x) +static inline unsigned long pud_val (pud_t pud) +{ +return pud; +}#endif /* !CONFIG_PPC_64K_PAGES */ #endif /* CONFIG_PPC64 */ typedef unsigned long pgd_t ; -#define pgd_val (x) (x) -#define pgprot_val (x) (x) +#define __pgd (x) (x) +static inline unsigned long pgd_val (pgd_t pgd) +{ +return pgd; +}typedef unsigned long pgprot_t ; -#define __pgd (x) (x) +#define pgprot_val (x) (x)#define __pgprot (x) (x)#endif
arch/powerpc/include/asm/pgalloc-32.h +26 -8
@@ -21,16 +21,34 @@ extern void pgd_free(struct mm_struct *mm, pgd_t *pgd);/* #define pgd_populate(mm, pmd, pte) BUG() */ #ifndef CONFIG_BOOKE -#define pmd_populate_kernel (mm, pmd, pte) \ - (pmd_val(*(pmd)) = __pa(pte) | _PMD_PRESENT) -#define pmd_populate (mm, pmd, pte) \ - (pmd_val(*(pmd)) = (page_to_pfn(pte) << PAGE_SHIFT) | _PMD_PRESENT) + +static inline void pmd_populate_kernel (struct mm_struct *mm,pmd_t *pmdp, +pte_t *pte) +{ + *pmdp =__pmd (__pa (pte) | _PMD_PRESENT); +} + +static inline void pmd_populate (struct mm_struct *mm,pmd_t *pmdp, +pgtable_t pte_page) +{ + *pmdp =__pmd ((page_to_pfn (pte_page) <<PAGE_SHIFT ) | _PMD_PRESENT); +} +#define pmd_pgtable (pmd) pmd_page(pmd)#else -#define pmd_populate_kernel (mm, pmd, pte) \ - (pmd_val(*(pmd)) = (unsigned long)pte | _PMD_PRESENT) -#define pmd_populate (mm, pmd, pte) \ - (pmd_val(*(pmd)) = (unsigned long)lowmem_page_address(pte) | _PMD_PRESENT) + +static inline void pmd_populate_kernel (struct mm_struct *mm,pmd_t *pmdp, +pte_t *pte) +{ + *pmdp =__pmd ((unsigned long )pte | _PMD_PRESENT); +} + +static inline void pmd_populate (struct mm_struct *mm,pmd_t *pmdp, +pgtable_t pte_page) +{ + *pmdp =__pmd ((unsigned long )lowmem_page_address (pte_page) | _PMD_PRESENT); +} +#define pmd_pgtable (pmd) pmd_page(pmd)#endif
arch/powerpc/include/asm/pgalloc-64.h +13 -4
@@ -53,7 +53,7 @@ static inline void pgd_free(struct mm_struct *mm, pgd_t *pgd)#ifndef CONFIG_PPC_64K_PAGES -#define pgd_populate (MM ,PGD ,PUD ) pgd_set(PGD, PUD) +#define pgd_populate (MM ,PGD ,PUD ) pgd_set(PGD, (unsigned long)PUD)static inline pud_t *pud_alloc_one (struct mm_struct *mm,unsigned long addr) {
@@ -71,9 +71,18 @@ static inline void pud_populate(struct mm_struct *mm, pud_t *pud, pmd_t *pmd)pud_set (pud, (unsigned long )pmd); } -#define pmd_populate (mm, pmd, pte_page) \ - pmd_populate_kernel(mm, pmd, page_address(pte_page)) -#define pmd_populate_kernel (mm, pmd, pte) pmd_set(pmd, (unsigned long)(pte)) +static inline void pmd_populate_kernel (struct mm_struct *mm,pmd_t *pmd, +pte_t *pte) +{ +pmd_set (pmd, (unsigned long )pte); +} + +static inline void pmd_populate (struct mm_struct *mm,pmd_t *pmd, +pgtable_t pte_page) +{ +pmd_set (pmd, (unsigned long )page_address (pte_page)); +} +#define pmd_pgtable (pmd) pmd_page(pmd) static inline pte_t *pte_alloc_one_kernel(struct mm_struct *mm,
arch/powerpc/include/asm/pgtable-ppc32.h +6 -1
@@ -128,7 +128,12 @@ extern int icache_44x_need_flush;#define pmd_none (pmd) (!pmd_val(pmd))#define pmd_bad (pmd) (pmd_val(pmd) & _PMD_BAD)#define pmd_present (pmd) (pmd_val(pmd) & _PMD_PRESENT_MASK) -#define pmd_clear (pmdp) do { pmd_val(*(pmdp)) = 0; } while (0) +static inline void pmd_clear (pmd_t *pmdp) +{ + *pmdp =__pmd (0 ); +} + +/* * When flushing the tlb entry for a page, we also need to flush the hash
arch/powerpc/include/asm/pgtable-ppc64-4k.h +5 -1
@@ -55,11 +55,15 @@#define pgd_none (pgd) (!pgd_val(pgd))#define pgd_bad (pgd) (pgd_val(pgd) == 0)#define pgd_present (pgd) (pgd_val(pgd) != 0) -#define pgd_clear (pgdp) (pgd_val(*(pgdp)) = 0)#define pgd_page_vaddr (pgd) (pgd_val(pgd) & ~PGD_MASKED_BITS)#ifndef __ASSEMBLY__ +static inline void pgd_clear (pgd_t *pgdp) +{ + *pgdp =__pgd (0 ); +} +static inline pte_t pgd_pte (pgd_t pgd) {return __pte (pgd_val (pgd));
arch/powerpc/include/asm/pgtable-ppc64.h +27 -9
@@ -144,21 +144,37 @@#define PMD_BAD_BITS (PTE_TABLE_SIZE-1)#define PUD_BAD_BITS (PMD_TABLE_SIZE-1) -#define pmd_set (pmdp, pmdval) (pmd_val(*(pmdp)) = (pmdval)) +static inline void pmd_set (pmd_t *pmdp,unsigned long val) +{ + *pmdp =__pmd (val); +} + +static inline void pmd_clear (pmd_t *pmdp) +{ + *pmdp =__pmd (0 ); +} +#define pmd_none (pmd) (!pmd_val(pmd))#define pmd_bad (pmd) (!is_kernel_addr(pmd_val(pmd)) \ || (pmd_val(pmd) & PMD_BAD_BITS))#define pmd_present (pmd) (!pmd_none(pmd)) -#define pmd_clear (pmdp) (pmd_val(*(pmdp)) = 0)#define pmd_page_vaddr (pmd) (pmd_val(pmd) & ~PMD_MASKED_BITS)extern struct page *pmd_page (pmd_t pmd); -#define pud_set (pudp, pudval) (pud_val(*(pudp)) = (pudval)) +static inline void pud_set (pud_t *pudp,unsigned long val) +{ + *pudp =__pud (val); +} + +static inline void pud_clear (pud_t *pudp) +{ + *pudp =__pud (0 ); +} +#define pud_none (pud) (!pud_val(pud))#define pud_bad (pud) (!is_kernel_addr(pud_val(pud)) \ || (pud_val(pud) & PUD_BAD_BITS))#define pud_present (pud) (pud_val(pud) != 0) -#define pud_clear (pudp) (pud_val(*(pudp)) = 0)#define pud_page_vaddr (pud) (pud_val(pud) & ~PUD_MASKED_BITS)extern struct page *pud_page (pud_t pud);
@@ -173,9 +189,13 @@ static inline pud_t pte_pud(pte_t pte)return __pud (pte_val (pte)); }#define pud_write (pud) pte_write(pud_pte(pud)) -#define pgd_set (pgdp, pudp) ({pgd_val(*(pgdp)) = (unsigned long)(pudp);})#define pgd_write (pgd) pte_write(pgd_pte(pgd)) +static inline void pgd_set (pgd_t *pgdp,unsigned long val) +{ + *pgdp =__pgd (val); +} +/* * Find an entry in a page-table-directory. We combine the address region * (the high order N bits) and the pgd portion of the address.
@@ -528,14 +548,12 @@ static inline pmd_t pmd_mkhuge(pmd_t pmd)static inline pmd_t pmd_mknotpresent (pmd_t pmd) { -pmd_val (pmd) &= ~_PAGE_PRESENT; -return pmd; +return __pmd (pmd_val (pmd) & ~_PAGE_PRESENT); }static inline pmd_t pmd_mksplitting (pmd_t pmd) { -pmd_val (pmd) |= _PAGE_SPLITTING; -return pmd; +return __pmd (pmd_val (pmd) | _PAGE_SPLITTING); }#define __HAVE_ARCH_PMD_SAME
arch/powerpc/mm/40x_mmu.c +5 -5
@@ -110,10 +110,10 @@ unsigned long __init mmu_mapin_ram(unsigned long top)unsigned long val = p | _PMD_SIZE_16M | _PAGE_EXEC | _PAGE_HWWRITE; pmdp =pmd_offset (pud_offset (pgd_offset_k (v), v), v); -pmd_val (*pmdp++) = val; -pmd_val (*pmdp++) = val; -pmd_val (*pmdp++) = val; -pmd_val (*pmdp++) = val; + *pmdp++ =__pmd (val); + *pmdp++ =__pmd (val); + *pmdp++ =__pmd (val); + *pmdp++ =__pmd (val); v +=LARGE_PAGE_SIZE_16M ; p +=LARGE_PAGE_SIZE_16M ;
@@ -125,7 +125,7 @@ unsigned long __init mmu_mapin_ram(unsigned long top)unsigned long val = p | _PMD_SIZE_4M | _PAGE_EXEC | _PAGE_HWWRITE; pmdp =pmd_offset (pud_offset (pgd_offset_k (v), v), v); -pmd_val (*pmdp) = val; + *pmdp =__pmd (val); v +=LARGE_PAGE_SIZE_4M ; p +=LARGE_PAGE_SIZE_4M ;
arch/powerpc/mm/pgtable_64.c +9 -10
@@ -759,22 +759,20 @@ void hpte_do_hugepage_flush(struct mm_struct *mm, unsigned long addr,static pmd_t pmd_set_protbits (pmd_t pmd,pgprot_t pgprot) { -pmd_val (pmd) |=pgprot_val (pgprot); -return pmd; +return __pmd (pmd_val (pmd) |pgprot_val (pgprot)); }pmd_t pfn_pmd (unsigned long pfn,pgprot_t pgprot) { -pmd_t pmd; +unsigned long pmdv;/* * For a valid pte, we would have _PAGE_PRESENT always * set. We use this to check THP page at pmd level. * leaf pte for huge page, bottom two bits != 00 */ -pmd_val (pmd) = pfn <<PTE_RPN_SHIFT ; -pmd_val (pmd) |= _PAGE_THP_HUGE; - pmd =pmd_set_protbits (pmd, pgprot); -return pmd; + pmdv = pfn <<PTE_RPN_SHIFT ; + pmdv |= _PAGE_THP_HUGE; +return pmd_set_protbits (__pmd (pmdv), pgprot); }pmd_t mk_pmd (struct page *page,pgprot_t pgprot)
@@ -784,10 +782,11 @@ pmd_t mk_pmd(struct page *page, pgprot_t pgprot)pmd_t pmd_modify (pmd_t pmd,pgprot_t newprot) { +unsigned long pmdv; -pmd_val (pmd) &= _HPAGE_CHG_MASK; - pmd =pmd_set_protbits (pmd, newprot); -return pmd; + pmdv =pmd_val (pmd); + pmdv &= _HPAGE_CHG_MASK; +return pmd_set_protbits (__pmd (pmdv), newprot); }/*
Signature · 1 line
2.5.0
PATCH v6 11/35 powerpc/mm: Move hash64 PTE bits from book3s/64/pgtable.h to hash.h
In reply to Aneesh Kumar K.VThis enables us to keep hash64 related bits together, and makes it easy to follow.
--- arch/powerpc/include/asm/book3s/64/hash.h | 448 ++++++++++++++++++++++++++- arch/powerpc/include/asm/book3s/64/pgtable.h | 445 +------------------------- arch/powerpc/include/asm/pgtable.h | 6 - 3 files changed, 448 insertions(+), 451 deletions(-)
arch/powerpc/include/asm/book3s/64/hash.h +445 -3
@@ -2,6 +2,61 @@#define _ASM_POWERPC_BOOK3S_64_HASH_H#ifdef __KERNEL__ +#ifdef CONFIG_PPC_64K_PAGES +#include <asm/book3s/64/hash-64k.h> +#else +#include <asm/book3s/64/hash-4k.h> +#endif + +/* +* Size of EA range mapped by our pagetables. +*/ +#define PGTABLE_EADDR_SIZE (PTE_INDEX_SIZE + PMD_INDEX_SIZE + \ + PUD_INDEX_SIZE + PGD_INDEX_SIZE + PAGE_SHIFT) +#define PGTABLE_RANGE (ASM_CONST(1) << PGTABLE_EADDR_SIZE) + +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +#define PMD_CACHE_INDEX (PMD_INDEX_SIZE + 1) +#else +#define PMD_CACHE_INDEX PMD_INDEX_SIZE +#endif +/* +* Define the address range of the kernel non-linear virtual area +*/ +#define KERN_VIRT_START ASM_CONST(0xD000000000000000) +#define KERN_VIRT_SIZE ASM_CONST(0x0000100000000000) + +/* +* The vmalloc space starts at the beginning of that region, and +* occupies half of it on hash CPUs and a quarter of it on Book3E +* (we keep a quarter for the virtual memmap) +*/ +#define VMALLOC_START KERN_VIRT_START +#define VMALLOC_SIZE (KERN_VIRT_SIZE >> 1) +#define VMALLOC_END (VMALLOC_START + VMALLOC_SIZE) + +/* +* Region IDs +*/ +#define REGION_SHIFT 60UL +#define REGION_MASK (0xfUL << REGION_SHIFT) +#define REGION_ID (ea) (((unsigned long)(ea)) >> REGION_SHIFT) + +#define VMALLOC_REGION_ID (REGION_ID(VMALLOC_START)) +#define KERNEL_REGION_ID (REGION_ID(PAGE_OFFSET)) +#define VMEMMAP_REGION_ID (0xfUL)/* Server only */ +#define USER_REGION_ID (0UL) + +/* +* Defines the address of the vmemap area, in its own region on +* hash table CPUs. +*/ +#define VMEMMAP_BASE (VMEMMAP_REGION_ID << REGION_SHIFT) + +#ifdef CONFIG_PPC_MM_SLICES +#define HAVE_ARCH_UNMAPPED_AREA +#define HAVE_ARCH_UNMAPPED_AREA_TOPDOWN +#endif /* CONFIG_PPC_MM_SLICES */ /* * Common bits between 4K and 64K pages in a linux-style PTE. * These match the bits in the (hardware-defined) PowerPC PTE as closely
@@ -46,11 +101,398 @@/* Hash table based platforms need atomic updates of the linux PTE */ #define PTE_ATOMIC_UPDATES 1 -#ifdef CONFIG_PPC_64K_PAGES -#include <asm/book3s/64/hash-64k.h> +/* +* THP pages can't be special. So use the _PAGE_SPECIAL +*/ +#define _PAGE_SPLITTING _PAGE_SPECIAL + +/* +* We need to differentiate between explicit huge page and THP huge +* page, since THP huge page also need to track real subpage details +*/ +#define _PAGE_THP_HUGE _PAGE_4K_PFN + +/* +* set of bits not changed in pmd_modify. +*/ +#define _HPAGE_CHG_MASK (PTE_RPN_MASK | _PAGE_HPTEFLAGS | \ + _PAGE_DIRTY | _PAGE_ACCESSED | _PAGE_SPLITTING | \ + _PAGE_THP_HUGE) +#define _PTE_NONE_MASK _PAGE_HPTEFLAGS +/* +* The mask convered by the RPN must be a ULL on 32-bit platforms with +* 64-bit PTEs +*/ +#define PTE_RPN_MASK (~((1UL << PTE_RPN_SHIFT) - 1)) +/* +* _PAGE_CHG_MASK masks of bits that are to be preserved across +* pgprot changes +*/ +#define _PAGE_CHG_MASK (PTE_RPN_MASK | _PAGE_HPTEFLAGS | _PAGE_DIRTY | \ + _PAGE_ACCESSED | _PAGE_SPECIAL) +/* +* Mask of bits returned by pte_pgprot() +*/ +#define PAGE_PROT_BITS (_PAGE_GUARDED | _PAGE_COHERENT | _PAGE_NO_CACHE | \ + _PAGE_WRITETHRU | _PAGE_4K_PFN | \ + _PAGE_USER | _PAGE_ACCESSED | \ + _PAGE_RW | _PAGE_DIRTY | _PAGE_EXEC) +/* +* We define 2 sets of base prot bits, one for basic pages (ie, +* cacheable kernel and user pages) and one for non cacheable +* pages. We always set _PAGE_COHERENT when SMP is enabled or +* the processor might need it for DMA coherency. +*/ +#define _PAGE_BASE_NC (_PAGE_PRESENT | _PAGE_ACCESSED | _PAGE_PSIZE) +#define _PAGE_BASE (_PAGE_BASE_NC | _PAGE_COHERENT) + +/* Permission masks used to generate the __P and __S table, +* +* Note:__pgprot is defined in arch/powerpc/include/asm/page.h +* +* Write permissions imply read permissions for now (we could make write-only +* pages on BookE but we don't bother for now). Execute permission control is +* possible on platforms that define _PAGE_EXEC +* +* Note due to the way vm flags are laid out, the bits are XWR +*/ +#define PAGE_NONE __pgprot(_PAGE_BASE) +#define PAGE_SHARED __pgprot(_PAGE_BASE | _PAGE_USER | _PAGE_RW) +#define PAGE_SHARED_X __pgprot(_PAGE_BASE | _PAGE_USER | _PAGE_RW | \ + _PAGE_EXEC) +#define PAGE_COPY __pgprot(_PAGE_BASE | _PAGE_USER ) +#define PAGE_COPY_X __pgprot(_PAGE_BASE | _PAGE_USER | _PAGE_EXEC) +#define PAGE_READONLY __pgprot(_PAGE_BASE | _PAGE_USER ) +#define PAGE_READONLY_X __pgprot(_PAGE_BASE | _PAGE_USER | _PAGE_EXEC) + +#define __P000 PAGE_NONE +#define __P001 PAGE_READONLY +#define __P010 PAGE_COPY +#define __P011 PAGE_COPY +#define __P100 PAGE_READONLY_X +#define __P101 PAGE_READONLY_X +#define __P110 PAGE_COPY_X +#define __P111 PAGE_COPY_X + +#define __S000 PAGE_NONE +#define __S001 PAGE_READONLY +#define __S010 PAGE_SHARED +#define __S011 PAGE_SHARED +#define __S100 PAGE_READONLY_X +#define __S101 PAGE_READONLY_X +#define __S110 PAGE_SHARED_X +#define __S111 PAGE_SHARED_X + +/* Permission masks used for kernel mappings */ +#define PAGE_KERNEL __pgprot(_PAGE_BASE | _PAGE_KERNEL_RW) +#define PAGE_KERNEL_NC __pgprot(_PAGE_BASE_NC | _PAGE_KERNEL_RW | \ + _PAGE_NO_CACHE) +#define PAGE_KERNEL_NCG __pgprot(_PAGE_BASE_NC | _PAGE_KERNEL_RW | \ + _PAGE_NO_CACHE | _PAGE_GUARDED) +#define PAGE_KERNEL_X __pgprot(_PAGE_BASE | _PAGE_KERNEL_RWX) +#define PAGE_KERNEL_RO __pgprot(_PAGE_BASE | _PAGE_KERNEL_RO) +#define PAGE_KERNEL_ROX __pgprot(_PAGE_BASE | _PAGE_KERNEL_ROX) + +/* Protection used for kernel text. We want the debuggers to be able to +* set breakpoints anywhere, so don't write protect the kernel text +* on platforms where such control is possible. +*/ +#if defined(CONFIG_KGDB ) || defined(CONFIG_XMON ) || defined(CONFIG_BDI_SWITCH ) ||\ + defined(CONFIG_KPROBES ) || defined(CONFIG_DYNAMIC_FTRACE ) +#define PAGE_KERNEL_TEXT PAGE_KERNEL_X#else -#include <asm/book3s/64/hash-4k.h> +#define PAGE_KERNEL_TEXT PAGE_KERNEL_ROX#endif +/* Make modules code happy. We don't set RO yet */ +#define PAGE_KERNEL_EXEC PAGE_KERNEL_X +#define PAGE_AGP (PAGE_KERNEL_NC) + +#define PMD_BAD_BITS (PTE_TABLE_SIZE-1) +#define PUD_BAD_BITS (PMD_TABLE_SIZE-1) +/* +* We save the slot number & secondary bit in the second half of the +* PTE page. We use the 8 bytes per each pte entry. +*/ +#define PTE_PAGE_HIDX_OFFSET (PTRS_PER_PTE * 8) + +#ifndef __ASSEMBLY__ +#define pmd_bad (pmd) (!is_kernel_addr(pmd_val(pmd)) \ + || (pmd_val(pmd) & PMD_BAD_BITS)) +#define pmd_page_vaddr (pmd) (pmd_val(pmd) & ~PMD_MASKED_BITS) + +#define pud_bad (pud) (!is_kernel_addr(pud_val(pud)) \ + || (pud_val(pud) & PUD_BAD_BITS)) +#define pud_page_vaddr (pud) (pud_val(pud) & ~PUD_MASKED_BITS) + +#define pgd_index (address) (((address) >> (PGDIR_SHIFT)) & (PTRS_PER_PGD - 1)) +#define pmd_index (address) (((address) >> (PMD_SHIFT)) & (PTRS_PER_PMD - 1)) +#define pte_index (address) (((address) >> (PAGE_SHIFT)) & (PTRS_PER_PTE - 1)) + +extern void hpte_need_flush (struct mm_struct *mm,unsigned long addr, +pte_t *ptep,unsigned long pte,int huge); +extern unsigned long pmd_hugepage_update (struct mm_struct *mm, +unsigned long addr, +pmd_t *pmdp, +unsigned long clr, +unsigned long set); +/* Atomic PTE updates */ +static inline unsigned long pte_update (struct mm_struct *mm, +unsigned long addr, +pte_t *ptep,unsigned long clr, +unsigned long set, +int huge) +{ +unsigned long old, tmp; + + __asm__ __volatile__( +"1: ldarx %0,0,%3 # pte_update\n\ +andi. %1,%0,%6\n\ +bne- 1b \n\ +andc %1,%0,%4 \n\ +or %1,%1,%7\n\ +stdcx. %1,0,%3 \n\ +bne- 1b" + :"=&r" (old),"=&r" (tmp),"=m" (*ptep) + :"r" (ptep),"r" (clr),"m" (*ptep),"i" (_PAGE_BUSY),"r" (set) + :"cc" ); +/* huge pages use the old page table lock */ +if (!huge) +assert_pte_locked (mm, addr); + +if (old & _PAGE_HASHPTE) +hpte_need_flush (mm, addr, ptep, old, huge); + +return old; +} + +static inline int __ptep_test_and_clear_young (struct mm_struct *mm, +unsigned long addr,pte_t *ptep) +{ +unsigned long old; + +if ((pte_val (*ptep) & (_PAGE_ACCESSED | _PAGE_HASHPTE)) ==0 ) +return 0 ; + old =pte_update (mm, addr, ptep, _PAGE_ACCESSED,0 ,0 ); +return (old & _PAGE_ACCESSED) !=0 ; +} +#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG +#define ptep_test_and_clear_young (__vma, __addr, __ptep) \ +({ \ + int __r; \ + __r = __ptep_test_and_clear_young((__vma)->vm_mm, __addr, __ptep); \ + __r; \ +}) + +#define __HAVE_ARCH_PTEP_SET_WRPROTECT +static inline void ptep_set_wrprotect (struct mm_struct *mm,unsigned long addr, +pte_t *ptep) +{ + +if ((pte_val (*ptep) & _PAGE_RW) ==0 ) +return ; + +pte_update (mm, addr, ptep, _PAGE_RW,0 ,0 ); +} + +static inline void huge_ptep_set_wrprotect (struct mm_struct *mm, +unsigned long addr,pte_t *ptep) +{ +if ((pte_val (*ptep) & _PAGE_RW) ==0 ) +return ; + +pte_update (mm, addr, ptep, _PAGE_RW,0 ,1 ); +} + +/* +* We currently remove entries from the hashtable regardless of whether +* the entry was young or dirty. The generic routines only flush if the +* entry was young or dirty which is not good enough. +* +* We should be more intelligent about this but for the moment we override +* these functions and force a tlb flush unconditionally +*/ +#define __HAVE_ARCH_PTEP_CLEAR_YOUNG_FLUSH +#define ptep_clear_flush_young (__vma, __address, __ptep) \ +({ \ + int __young = __ptep_test_and_clear_young((__vma)->vm_mm, __address, \ + __ptep); \ + __young; \ +}) + +#define __HAVE_ARCH_PTEP_GET_AND_CLEAR +static inline pte_t ptep_get_and_clear (struct mm_struct *mm, +unsigned long addr,pte_t *ptep) +{ +unsigned long old =pte_update (mm, addr, ptep, ~0UL ,0 ,0 ); +return __pte (old); +} + +static inline void pte_clear (struct mm_struct *mm,unsigned long addr, +pte_t * ptep) +{ +pte_update (mm, addr, ptep, ~0UL ,0 ,0 ); +} + + +/* Set the dirty and/or accessed bits atomically in a linux PTE, this +* function doesn't need to flush the hash entry +*/ +static inline void __ptep_set_access_flags (pte_t *ptep,pte_t entry) +{ +unsigned long bits =pte_val (entry) & + (_PAGE_DIRTY | _PAGE_ACCESSED | _PAGE_RW | _PAGE_EXEC); + +unsigned long old, tmp; + + __asm__ __volatile__( +"1: ldarx %0,0,%4\n\ +andi. %1,%0,%6\n\ +bne- 1b \n\ +or %0,%3,%0\n\ +stdcx. %0,0,%4\n\ +bne- 1b" + :"=&r" (old),"=&r" (tmp),"=m" (*ptep) + :"r" (bits),"r" (ptep),"m" (*ptep),"i" (_PAGE_BUSY) + :"cc" ); +} + +#define __HAVE_ARCH_PTE_SAME +#define pte_same (A ,B ) (((pte_val(A) ^ pte_val(B)) & ~_PAGE_HPTEFLAGS) == 0) + +static inline char *get_hpte_slot_array (pmd_t *pmdp) +{ +/* +* The hpte hindex is stored in the pgtable whose address is in the +* second half of the PMD +* +* Order this load with the test for pmd_trans_huge in the caller +*/ +smp_rmb (); +return *(char **)(pmdp +PTRS_PER_PMD ); + + +} +/* +* The linux hugepage PMD now include the pmd entries followed by the address +* to the stashed pgtable_t. The stashed pgtable_t contains the hpte bits. +* [ 1 bit secondary | 3 bit hidx | 1 bit valid | 000]. We use one byte per +* each HPTE entry. With 16MB hugepage and 64K HPTE we need 256 entries and +* with 4K HPTE we need 4096 entries. Both will fit in a 4K pgtable_t. +* +* The last three bits are intentionally left to zero. This memory location +* are also used as normal page PTE pointers. So if we have any pointers +* left around while we collapse a hugepage, we need to make sure +* _PAGE_PRESENT bit of that is zero when we look at them +*/ +static inline unsigned int hpte_valid (unsigned char *hpte_slot_array,int index) +{ +return (hpte_slot_array[index] >>3 ) &0x1 ; +} + +static inline unsigned int hpte_hash_index (unsigned char *hpte_slot_array, +int index) +{ +return hpte_slot_array[index] >>4 ; +} + +static inline void mark_hpte_slot_valid (unsigned char *hpte_slot_array, +unsigned int index,unsigned int hidx) +{ + hpte_slot_array[index] = hidx <<4 |0x1 <<3 ; +} + +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +/* +* +* For core kernel code by design pmd_trans_huge is never run on any hugetlbfs +* page. The hugetlbfs page table walking and mangling paths are totally +* separated form the core VM paths and they're differentiated by +* VM_HUGETLB being set on vm_flags well before any pmd_trans_huge could run. +* +* pmd_trans_huge() is defined as false at build time if +* CONFIG_TRANSPARENT_HUGEPAGE=n to optimize away code blocks at build +* time in such case. +* +* For ppc64 we need to differntiate from explicit hugepages from THP, because +* for THP we also track the subpage details at the pmd level. We don't do +* that for explicit huge pages. +* +*/ +static inline int pmd_trans_huge (pmd_t pmd) +{ +/* +* leaf pte for huge page, bottom two bits != 00 +*/ +return (pmd_val (pmd) &0x3 ) && (pmd_val (pmd) & _PAGE_THP_HUGE); +} + +static inline int pmd_trans_splitting (pmd_t pmd) +{ +if (pmd_trans_huge (pmd)) +return pmd_val (pmd) & _PAGE_SPLITTING; +return 0 ; +} + +#endif +static inline int pmd_large (pmd_t pmd) +{ +/* +* leaf pte for huge page, bottom two bits != 00 +*/ +return ((pmd_val (pmd) &0x3 ) !=0x0 ); +} + +static inline pmd_t pmd_mknotpresent (pmd_t pmd) +{ +return __pmd (pmd_val (pmd) & ~_PAGE_PRESENT); +} + +static inline pmd_t pmd_mksplitting (pmd_t pmd) +{ +return __pmd (pmd_val (pmd) | _PAGE_SPLITTING); +} + +#define __HAVE_ARCH_PMD_SAME +static inline int pmd_same (pmd_t pmd_a,pmd_t pmd_b) +{ +return (((pmd_val (pmd_a) ^pmd_val (pmd_b)) & ~_PAGE_HPTEFLAGS) ==0 ); +} + +static inline int __pmdp_test_and_clear_young (struct mm_struct *mm, +unsigned long addr,pmd_t *pmdp) +{ +unsigned long old; + +if ((pmd_val (*pmdp) & (_PAGE_ACCESSED | _PAGE_HASHPTE)) ==0 ) +return 0 ; + old =pmd_hugepage_update (mm, addr, pmdp, _PAGE_ACCESSED,0 ); +return ((old & _PAGE_ACCESSED) !=0 ); +} + +#define __HAVE_ARCH_PMDP_SET_WRPROTECT +static inline void pmdp_set_wrprotect (struct mm_struct *mm,unsigned long addr, +pmd_t *pmdp) +{ + +if ((pmd_val (*pmdp) & _PAGE_RW) ==0 ) +return ; + +pmd_hugepage_update (mm, addr, pmdp, _PAGE_RW,0 ); +} + +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +extern void hpte_do_hugepage_flush (struct mm_struct *mm,unsigned long addr, +pmd_t *pmdp,unsigned long old_pmd); +#else +static inline void hpte_do_hugepage_flush (struct mm_struct *mm, +unsigned long addr,pmd_t *pmdp, +unsigned long old_pmd) +{ +WARN (1 ,"%s called with THP disabled\n" , __func__); +} +#endif /* CONFIG_TRANSPARENT_HUGEPAGE */ + +#endif /* !__ASSEMBLY__ */ #endif /* __KERNEL__ */ #endif /* _ASM_POWERPC_BOOK3S_64_HASH_H */
arch/powerpc/include/asm/book3s/64/pgtable.h +3 -442
536 lines not on this page · raw message
arch/powerpc/include/asm/pgtable.h +0 -6
@@ -18,12 +18,6 @@ struct mm_struct;#include <asm/pgtable-book3e.h> #endif /* !CONFIG_PPC_BOOK3S */ -/* -* We save the slot number & secondary bit in the second half of the -* PTE page. We use the 8 bytes per each pte entry. -*/ -#define PTE_PAGE_HIDX_OFFSET (PTRS_PER_PTE * 8) -#ifndef __ASSEMBLY__#include <asm/tlbflush.h>
Signature · 1 line
2.5.0
PATCH v6 12/35 powerpc/mm: Move PTE bits from generic functions to hash64 functions.
In reply to Aneesh Kumar K.Vfunctions which operate on pte bits are moved to hash*.h and other generic functions are moved to pgtable.h
--- arch/powerpc/include/asm/book3s/32/pgtable.h | 183 ++++++++++++++++++++++++ arch/powerpc/include/asm/book3s/64/hash.h | 151 ++++++++++++++++++++ arch/powerpc/include/asm/book3s/64/pgtable.h | 6 + arch/powerpc/include/asm/book3s/pgtable.h | 204 --------------------------- 4 files changed, 340 insertions(+), 204 deletions(-)
arch/powerpc/include/asm/book3s/32/pgtable.h +183 -0
@@ -294,6 +294,189 @@ void pgtable_cache_init(void);extern int get_pteptr (struct mm_struct *mm,unsigned long addr,pte_t **ptep,pmd_t **pmdp); +/* Generic accessors to PTE bits */ +static inline int pte_write (pte_t pte) {return !!(pte_val (pte) & _PAGE_RW);} +static inline int pte_dirty (pte_t pte) {return !!(pte_val (pte) & _PAGE_DIRTY); } +static inline int pte_young (pte_t pte) {return !!(pte_val (pte) & _PAGE_ACCESSED); } +static inline int pte_special (pte_t pte) {return !!(pte_val (pte) & _PAGE_SPECIAL); } +static inline int pte_none (pte_t pte) {return (pte_val (pte) & ~_PTE_NONE_MASK) ==0 ; } +static inline pgprot_t pte_pgprot (pte_t pte) {return __pgprot (pte_val (pte) &PAGE_PROT_BITS ); } + +static inline int pte_present (pte_t pte) +{ +return pte_val (pte) & _PAGE_PRESENT; +} + +/* Conversion functions: convert a page and protection to a page entry, +* and a page entry and page directory to the page they refer to. +* +* Even if PTEs can be unsigned long long, a PFN is always an unsigned +* long for now. +*/ +static inline pte_t pfn_pte (unsigned long pfn,pgprot_t pgprot) +{ +return __pte (((pte_basic_t)(pfn) <<PTE_RPN_SHIFT ) | +pgprot_val (pgprot)); +} + +static inline unsigned long pte_pfn (pte_t pte) +{ +return pte_val (pte) >>PTE_RPN_SHIFT ; +} + +/* Generic modifiers for PTE bits */ +static inline pte_t pte_wrprotect (pte_t pte) +{ +return __pte (pte_val (pte) & ~_PAGE_RW); +} + +static inline pte_t pte_mkclean (pte_t pte) +{ +return __pte (pte_val (pte) & ~_PAGE_DIRTY); +} + +static inline pte_t pte_mkold (pte_t pte) +{ +return __pte (pte_val (pte) & ~_PAGE_ACCESSED); +} + +static inline pte_t pte_mkwrite (pte_t pte) +{ +return __pte (pte_val (pte) | _PAGE_RW); +} + +static inline pte_t pte_mkdirty (pte_t pte) +{ +return __pte (pte_val (pte) | _PAGE_DIRTY); +} + +static inline pte_t pte_mkyoung (pte_t pte) +{ +return __pte (pte_val (pte) | _PAGE_ACCESSED); +} + +static inline pte_t pte_mkspecial (pte_t pte) +{ +return __pte (pte_val (pte) | _PAGE_SPECIAL); +} + +static inline pte_t pte_mkhuge (pte_t pte) +{ +return pte; +} + +static inline pte_t pte_modify (pte_t pte,pgprot_t newprot) +{ +return __pte ((pte_val (pte) & _PAGE_CHG_MASK) |pgprot_val (newprot)); +} + + + +/* This low level function performs the actual PTE insertion +* Setting the PTE depends on the MMU type and other factors. It's +* an horrible mess that I'm not going to try to clean up now but +* I'm keeping it in one place rather than spread around +*/ +static inline void __set_pte_at (struct mm_struct *mm,unsigned long addr, +pte_t *ptep,pte_t pte,int percpu) +{ +#if defined(CONFIG_PPC_STD_MMU_32 ) && defined(CONFIG_SMP ) && !defined(CONFIG_PTE_64BIT ) +/* First case is 32-bit Hash MMU in SMP mode with 32-bit PTEs. We use the +* helper pte_update() which does an atomic update. We need to do that +* because a concurrent invalidation can clear _PAGE_HASHPTE. If it's a +* per-CPU PTE such as a kmap_atomic, we do a simple update preserving +* the hash bits instead (ie, same as the non-SMP case) +*/ +if (percpu) + *ptep =__pte ((pte_val (*ptep) & _PAGE_HASHPTE) + | (pte_val (pte) & ~_PAGE_HASHPTE)); +else +pte_update (ptep, ~_PAGE_HASHPTE,pte_val (pte)); + +#elif defined(CONFIG_PPC32 ) && defined(CONFIG_PTE_64BIT ) +/* Second case is 32-bit with 64-bit PTE. In this case, we +* can just store as long as we do the two halves in the right order +* with a barrier in between. This is possible because we take care, +* in the hash code, to pre-invalidate if the PTE was already hashed, +* which synchronizes us with any concurrent invalidation. +* In the percpu case, we also fallback to the simple update preserving +* the hash bits +*/ +if (percpu) { + *ptep =__pte ((pte_val (*ptep) & _PAGE_HASHPTE) + | (pte_val (pte) & ~_PAGE_HASHPTE)); +return ; + } +if (pte_val (*ptep) & _PAGE_HASHPTE) +flush_hash_entry (mm, ptep, addr); + __asm__ __volatile__("\ +stw%U0%X0 %2,%0\n\ +eieio\n\ +stw%U0%X0 %L2,%1" + :"=m" (*ptep),"=m" (*((unsigned char *)ptep+4 )) + :"r" (pte) :"memory" ); + +#elif defined(CONFIG_PPC_STD_MMU_32 ) +/* Third case is 32-bit hash table in UP mode, we need to preserve +* the _PAGE_HASHPTE bit since we may not have invalidated the previous +* translation in the hash yet (done in a subsequent flush_tlb_xxx()) +* and see we need to keep track that this PTE needs invalidating +*/ + *ptep =__pte ((pte_val (*ptep) & _PAGE_HASHPTE) + | (pte_val (pte) & ~_PAGE_HASHPTE)); + +#else +#error "Not supported " +#endif +} + +/* +* Macro to mark a page protection value as "uncacheable". +*/ + +#define _PAGE_CACHE_CTL (_PAGE_COHERENT | _PAGE_GUARDED | _PAGE_NO_CACHE | \ + _PAGE_WRITETHRU) + +#define pgprot_noncached pgprot_noncached +static inline pgprot_t pgprot_noncached (pgprot_t prot) +{ +return __pgprot ((pgprot_val (prot) & ~_PAGE_CACHE_CTL) | + _PAGE_NO_CACHE | _PAGE_GUARDED); +} + +#define pgprot_noncached_wc pgprot_noncached_wc +static inline pgprot_t pgprot_noncached_wc (pgprot_t prot) +{ +return __pgprot ((pgprot_val (prot) & ~_PAGE_CACHE_CTL) | + _PAGE_NO_CACHE); +} + +#define pgprot_cached pgprot_cached +static inline pgprot_t pgprot_cached (pgprot_t prot) +{ +return __pgprot ((pgprot_val (prot) & ~_PAGE_CACHE_CTL) | + _PAGE_COHERENT); +} + +#define pgprot_cached_wthru pgprot_cached_wthru +static inline pgprot_t pgprot_cached_wthru (pgprot_t prot) +{ +return __pgprot ((pgprot_val (prot) & ~_PAGE_CACHE_CTL) | + _PAGE_COHERENT | _PAGE_WRITETHRU); +} + +#define pgprot_cached_noncoherent pgprot_cached_noncoherent +static inline pgprot_t pgprot_cached_noncoherent (pgprot_t prot) +{ +return __pgprot (pgprot_val (prot) & ~_PAGE_CACHE_CTL); +} + +#define pgprot_writecombine pgprot_writecombine +static inline pgprot_t pgprot_writecombine (pgprot_t prot) +{ +return pgprot_noncached_wc (prot); +} +#endif /* !__ASSEMBLY__ */ #endif /* _ASM_POWERPC_BOOK3S_32_PGTABLE_H */
arch/powerpc/include/asm/book3s/64/hash.h +151 -0
157 lines not on this page · raw message
arch/powerpc/include/asm/book3s/64/pgtable.h +6 -0
@@ -201,6 +201,12 @@ static inline pte_t *pmdp_ptep(pmd_t *pmd)#define pmd_mkdirty (pmd) pte_pmd(pte_mkdirty(pmd_pte(pmd)))#define pmd_mkyoung (pmd) pte_pmd(pte_mkyoung(pmd_pte(pmd)))#define pmd_mkwrite (pmd) pte_pmd(pte_mkwrite(pmd_pte(pmd))) +#ifdef CONFIG_NUMA_BALANCING +static inline int pmd_protnone (pmd_t pmd) +{ +return pte_protnone (pmd_pte (pmd)); +} +#endif /* CONFIG_NUMA_BALANCING */ #define __HAVE_ARCH_PMD_WRITE#define pmd_write (pmd) pte_write(pmd_pte(pmd))
arch/powerpc/include/asm/book3s/pgtable.h +0 -204
221 lines not on this page · raw message
Signature · 1 line
2.5.0
PATCH v6 13/35 powerpc/booke: Move nohash headers (part 1)
In reply to Aneesh Kumar K.VMove the booke related headers below booke/32 or booke/64
We are splitting this change into multiple patch to make the rebasing easier. The following patches can be folded into this if needed. They are kept separate for easier review.
---
arch/powerpc/include/asm/{pgtable-book3e.h => nohash/pgtable.h} | 0
arch/powerpc/include/asm/pgtable.h | 2 +-
2 files changed, 1 insertion(+), 1 deletion(-)
rename arch/powerpc/include/asm/{pgtable-book3e.h => nohash/pgtable.h} (100%)arch/powerpc/include/asm/nohash/pgtable.h renamed from arch/powerpc/include/asm/pgtable-book3e.h +0 -0
arch/powerpc/include/asm/pgtable.h +1 -1
@@ -15,7 +15,7 @@ struct mm_struct;#ifdef CONFIG_PPC_BOOK3S #include <asm/book3s/pgtable.h> #else -#include <asm/pgtable-book3e.h> +#include <asm/nohash/pgtable.h> #endif /* !CONFIG_PPC_BOOK3S */ #ifndef __ASSEMBLY__
Signature · 1 line
2.5.0
PATCH v6 14/35 powerpc/booke: Move nohash headers (part 2)
In reply to Aneesh Kumar K.V---
arch/powerpc/include/asm/{pgtable-ppc32.h => nohash/32/pgtable.h} | 0
arch/powerpc/include/asm/{pgtable-ppc64.h => nohash/64/pgtable.h} | 2 +-
arch/powerpc/include/asm/nohash/pgtable.h | 8 ++++----
3 files changed, 5 insertions(+), 5 deletions(-)
rename arch/powerpc/include/asm/{pgtable-ppc32.h => nohash/32/pgtable.h} (100%)
rename arch/powerpc/include/asm/{pgtable-ppc64.h => nohash/64/pgtable.h} (99%)arch/powerpc/include/asm/nohash/32/pgtable.h renamed from arch/powerpc/include/asm/pgtable-ppc32.h +0 -0
arch/powerpc/include/asm/nohash/64/pgtable.h renamed from arch/powerpc/include/asm/pgtable-ppc64.h +1 -1
@@ -18,7 +18,7 @@* Size of EA range mapped by our pagetables. */ #define PGTABLE_EADDR_SIZE (PTE_INDEX_SIZE + PMD_INDEX_SIZE + \ - PUD_INDEX_SIZE + PGD_INDEX_SIZE + PAGE_SHIFT) + PUD_INDEX_SIZE + PGD_INDEX_SIZE + PAGE_SHIFT)#define PGTABLE_RANGE (ASM_CONST(1) << PGTABLE_EADDR_SIZE)#ifdef CONFIG_TRANSPARENT_HUGEPAGE
arch/powerpc/include/asm/nohash/pgtable.h +4 -4
@@ -1,10 +1,10 @@ -#ifndef _ASM_POWERPC_PGTABLE_BOOK3E_H -#define _ASM_POWERPC_PGTABLE_BOOK3E_H +#ifndef _ASM_POWERPC_NOHASH_PGTABLE_H +#define _ASM_POWERPC_NOHASH_PGTABLE_H#if defined(CONFIG_PPC64 ) -#include <asm/pgtable-ppc64.h> +#include <asm/nohash/64/pgtable.h> #else -#include <asm/pgtable-ppc32.h> +#include <asm/nohash/32/pgtable.h> #endif #ifndef __ASSEMBLY__
Signature · 1 line
2.5.0
PATCH v6 15/35 powerpc/booke: Move nohash headers (part 3)
In reply to Aneesh Kumar K.V---
.../include/asm/{pgtable-ppc64-4k.h => nohash/64/pgtable-4k.h} | 0
.../asm/{pgtable-ppc64-64k.h => nohash/64/pgtable-64k.h} | 0
arch/powerpc/include/asm/nohash/64/pgtable.h | 10 +++++-----
3 files changed, 5 insertions(+), 5 deletions(-)
rename arch/powerpc/include/asm/{pgtable-ppc64-4k.h => nohash/64/pgtable-4k.h} (100%)
rename arch/powerpc/include/asm/{pgtable-ppc64-64k.h => nohash/64/pgtable-64k.h} (100%)arch/powerpc/include/asm/nohash/64/pgtable-4k.h renamed from arch/powerpc/include/asm/pgtable-ppc64-4k.h +0 -0
arch/powerpc/include/asm/nohash/64/pgtable-64k.h renamed from arch/powerpc/include/asm/pgtable-ppc64-64k.h +0 -0
arch/powerpc/include/asm/nohash/64/pgtable.h +5 -5
@@ -1,14 +1,14 @@ -#ifndef _ASM_POWERPC_PGTABLE_PPC64_H_ -#define _ASM_POWERPC_PGTABLE_PPC64_H_ +#ifndef _ASM_POWERPC_NOHASH_64_PGTABLE_H +#define _ASM_POWERPC_NOHASH_64_PGTABLE_H/* * This file contains the functions and defines necessary to modify and use * the ppc64 hashed page table. */ #ifdef CONFIG_PPC_64K_PAGES -#include <asm/pgtable-ppc64-64k.h> +#include <asm/nohash/64/pgtable-64k.h> #else -#include <asm/pgtable-ppc64-4k.h> +#include <asm/nohash/64/pgtable-4k.h> #endif #include <asm/barrier.h>
@@ -637,4 +637,4 @@ static inline int pmd_move_must_withdraw(struct spinlock *new_pmd_ptl,return true; }#endif /* __ASSEMBLY__ */ -#endif /* _ASM_POWERPC_PGTABLE_PPC64_H_ */ +#endif /* _ASM_POWERPC_NOHASH_64_PGTABLE_H */
Signature · 1 line
2.5.0
PATCH v6 16/35 powerpc/booke: Move nohash headers (part 4)
In reply to Aneesh Kumar K.V---
arch/powerpc/include/asm/nohash/32/pgtable.h | 16 ++++++++--------
arch/powerpc/include/asm/{ => nohash/32}/pte-40x.h | 0
arch/powerpc/include/asm/{ => nohash/32}/pte-44x.h | 0
arch/powerpc/include/asm/{ => nohash/32}/pte-8xx.h | 0
arch/powerpc/include/asm/{ => nohash/32}/pte-fsl-booke.h | 0
arch/powerpc/include/asm/nohash/64/pgtable.h | 2 +-
arch/powerpc/include/asm/{ => nohash}/pte-book3e.h | 0
7 files changed, 9 insertions(+), 9 deletions(-)
rename arch/powerpc/include/asm/{ => nohash/32}/pte-40x.h (100%)
rename arch/powerpc/include/asm/{ => nohash/32}/pte-44x.h (100%)
rename arch/powerpc/include/asm/{ => nohash/32}/pte-8xx.h (100%)
rename arch/powerpc/include/asm/{ => nohash/32}/pte-fsl-booke.h (100%)
rename arch/powerpc/include/asm/{ => nohash}/pte-book3e.h (100%)arch/powerpc/include/asm/nohash/32/pgtable.h +8 -8
@@ -1,5 +1,5 @@ -#ifndef _ASM_POWERPC_PGTABLE_PPC32_H -#define _ASM_POWERPC_PGTABLE_PPC32_H +#ifndef _ASM_POWERPC_NOHASH_32_PGTABLE_H +#define _ASM_POWERPC_NOHASH_32_PGTABLE_H#include <asm-generic/pgtable-nopmd.h>
@@ -106,15 +106,15 @@ extern int icache_44x_need_flush;*/ #if defined(CONFIG_40x) -#include <asm/pte-40x.h> +#include <asm/nohash/32/pte-40x.h> #elif defined(CONFIG_44x) -#include <asm/pte-44x.h> +#include <asm/nohash/32/pte-44x.h> #elif defined(CONFIG_FSL_BOOKE ) && defined(CONFIG_PTE_64BIT ) -#include <asm/pte-book3e.h> +#include <asm/nohash/pte-book3e.h> #elif defined(CONFIG_FSL_BOOKE ) -#include <asm/pte-fsl-booke.h> +#include <asm/nohash/32/pte-fsl-booke.h> #elif defined(CONFIG_8xx) -#include <asm/pte-8xx.h> +#include <asm/nohash/32/pte-8xx.h> #endif /* And here we include common definitions */
@@ -340,4 +340,4 @@ extern int get_pteptr(struct mm_struct *mm, unsigned long addr, pte_t **ptep,#endif /* !__ASSEMBLY__ */ -#endif /* _ASM_POWERPC_PGTABLE_PPC32_H */ +#endif /* __ASM_POWERPC_NOHASH_32_PGTABLE_H */
arch/powerpc/include/asm/nohash/32/pte-40x.h renamed from arch/powerpc/include/asm/pte-40x.h +0 -0
arch/powerpc/include/asm/nohash/32/pte-44x.h renamed from arch/powerpc/include/asm/pte-44x.h +0 -0
arch/powerpc/include/asm/nohash/32/pte-8xx.h renamed from arch/powerpc/include/asm/pte-8xx.h +0 -0
arch/powerpc/include/asm/nohash/32/pte-fsl-booke.h renamed from arch/powerpc/include/asm/pte-fsl-booke.h +0 -0
arch/powerpc/include/asm/nohash/64/pgtable.h +1 -1
@@ -97,7 +97,7 @@/* * Include the PTE bits definitions */ -#include <asm/pte-book3e.h> +#include <asm/nohash/pte-book3e.h> #include <asm/pte-common.h> #ifdef CONFIG_PPC_MM_SLICES
arch/powerpc/include/asm/nohash/pte-book3e.h renamed from arch/powerpc/include/asm/pte-book3e.h +0 -0
Signature · 1 line
2.5.0
PATCH v6 17/35 powerpc/booke: Move nohash headers (part 5)
In reply to Aneesh Kumar K.V--- arch/powerpc/include/asm/nohash/32/pte-40x.h | 6 +++--- arch/powerpc/include/asm/nohash/32/pte-44x.h | 6 +++--- arch/powerpc/include/asm/nohash/32/pte-8xx.h | 6 +++--- arch/powerpc/include/asm/nohash/32/pte-fsl-booke.h | 6 +++--- arch/powerpc/include/asm/nohash/64/pgtable-4k.h | 6 +++--- arch/powerpc/include/asm/nohash/64/pgtable-64k.h | 6 +++--- arch/powerpc/include/asm/nohash/pte-book3e.h | 6 +++--- 7 files changed, 21 insertions(+), 21 deletions(-)
arch/powerpc/include/asm/nohash/32/pte-40x.h +3 -3
@@ -1,5 +1,5 @@ -#ifndef _ASM_POWERPC_PTE_40x_H -#define _ASM_POWERPC_PTE_40x_H +#ifndef _ASM_POWERPC_NOHASH_32_PTE_40x_H +#define _ASM_POWERPC_NOHASH_32_PTE_40x_H#ifdef __KERNEL__/*
@@ -61,4 +61,4 @@#define PTE_ATOMIC_UPDATES 1#endif /* __KERNEL__ */ -#endif /* _ASM_POWERPC_PTE_40x_H */ +#endif /* _ASM_POWERPC_NOHASH_32_PTE_40x_H */
arch/powerpc/include/asm/nohash/32/pte-44x.h +3 -3
@@ -1,5 +1,5 @@ -#ifndef _ASM_POWERPC_PTE_44x_H -#define _ASM_POWERPC_PTE_44x_H +#ifndef _ASM_POWERPC_NOHASH_32_PTE_44x_H +#define _ASM_POWERPC_NOHASH_32_PTE_44x_H#ifdef __KERNEL__/*
@@ -94,4 +94,4 @@#endif /* __KERNEL__ */ -#endif /* _ASM_POWERPC_PTE_44x_H */ +#endif /* _ASM_POWERPC_NOHASH_32_PTE_44x_H */
arch/powerpc/include/asm/nohash/32/pte-8xx.h +3 -3
@@ -1,5 +1,5 @@ -#ifndef _ASM_POWERPC_PTE_8xx_H -#define _ASM_POWERPC_PTE_8xx_H +#ifndef _ASM_POWERPC_NOHASH_32_PTE_8xx_H +#define _ASM_POWERPC_NOHASH_32_PTE_8xx_H#ifdef __KERNEL__/*
@@ -62,4 +62,4 @@ _PAGE_HWWRITE | _PAGE_EXEC)#endif /* __KERNEL__ */ -#endif /* _ASM_POWERPC_PTE_8xx_H */ +#endif /* _ASM_POWERPC_NOHASH_32_PTE_8xx_H */
arch/powerpc/include/asm/nohash/32/pte-fsl-booke.h +3 -3
@@ -1,5 +1,5 @@ -#ifndef _ASM_POWERPC_PTE_FSL_BOOKE_H -#define _ASM_POWERPC_PTE_FSL_BOOKE_H +#ifndef _ASM_POWERPC_NOHASH_32_PTE_FSL_BOOKE_H +#define _ASM_POWERPC_NOHASH_32_PTE_FSL_BOOKE_H#ifdef __KERNEL__/* PTE bit definitions for Freescale BookE SW loaded TLB MMU based
@@ -37,4 +37,4 @@#define PTE_WIMGE_SHIFT (6)#endif /* __KERNEL__ */ -#endif /* _ASM_POWERPC_PTE_FSL_BOOKE_H */ +#endif /* _ASM_POWERPC_NOHASH_32_PTE_FSL_BOOKE_H */
arch/powerpc/include/asm/nohash/64/pgtable-4k.h +3 -3
@@ -1,5 +1,5 @@ -#ifndef _ASM_POWERPC_PGTABLE_PPC64_4K_H -#define _ASM_POWERPC_PGTABLE_PPC64_4K_H +#ifndef _ASM_POWERPC_NOHASH_64_PGTABLE_4K_H +#define _ASM_POWERPC_NOHASH_64_PGTABLE_4K_H/* * Entries per page directory level. The PTE level must use a 64b record * for each page table entry. The PMD and PGD level use a 32b record for
@@ -89,4 +89,4 @@ extern struct page *pgd_page(pgd_t pgd);#define remap_4k_pfn (vma, addr, pfn, prot) \ remap_pfn_range((vma), (addr), (pfn), PAGE_SIZE, (prot)) -#endif /* _ASM_POWERPC_PGTABLE_PPC64_4K_H */ +#endif /* _ _ASM_POWERPC_NOHASH_64_PGTABLE_4K_H */
arch/powerpc/include/asm/nohash/64/pgtable-64k.h +3 -3
@@ -1,5 +1,5 @@ -#ifndef _ASM_POWERPC_PGTABLE_PPC64_64K_H -#define _ASM_POWERPC_PGTABLE_PPC64_64K_H +#ifndef _ASM_POWERPC_NOHASH_64_PGTABLE_64K_H +#define _ASM_POWERPC_NOHASH_64_PGTABLE_64K_H#include <asm-generic/pgtable-nopud.h>
@@ -41,4 +41,4 @@#define pgd_pte (pgd) (pud_pte(((pud_t){ pgd })))#define pte_pgd (pte) ((pgd_t)pte_pud(pte)) -#endif /* _ASM_POWERPC_PGTABLE_PPC64_64K_H */ +#endif /* _ASM_POWERPC_NOHASH_64_PGTABLE_64K_H */
arch/powerpc/include/asm/nohash/pte-book3e.h +3 -3
@@ -1,5 +1,5 @@ -#ifndef _ASM_POWERPC_PTE_BOOK3E_H -#define _ASM_POWERPC_PTE_BOOK3E_H +#ifndef _ASM_POWERPC_NOHASH_PTE_BOOK3E_H +#define _ASM_POWERPC_NOHASH_PTE_BOOK3E_H#ifdef __KERNEL__/* PTE bit definitions for processors compliant to the Book3E
@@ -84,4 +84,4 @@#endif #endif /* __KERNEL__ */ -#endif /* _ASM_POWERPC_PTE_FSL_BOOKE_H */ +#endif /* _ASM_POWERPC_NOHASH_PTE_BOOK3E_H */
Signature · 1 line
2.5.0
PATCH v6 18/35 powerpc/mm: Convert 4k hash insert to C
In reply to Aneesh Kumar K.V--- arch/powerpc/mm/Makefile | 3 + arch/powerpc/mm/hash64_64k.c | 202 +++++++++++++++++++++ arch/powerpc/mm/hash_low_64.S | 380 ---------------------------------------- arch/powerpc/mm/hash_utils_64.c | 4 +- 4 files changed, 208 insertions(+), 381 deletions(-) create mode 100644 arch/powerpc/mm/hash64_64k.c
arch/powerpc/mm/Makefile +3 -0
@@ -18,6 +18,9 @@ obj-$(CONFIG_PPC_STD_MMU_32) += ppc_mmu_32.o obj-$(CONFIG_PPC_STD_MMU) += hash_low_$(CONFIG_WORD_SIZE).o \ tlb_hash$(CONFIG_WORD_SIZE).o \ mmu_context_hash$(CONFIG_WORD_SIZE).o +ifeq ($(CONFIG_PPC_STD_MMU_64),y) +obj-$(CONFIG_PPC_64K_PAGES) += hash64_64k.o +endif obj-$(CONFIG_PPC_ICSWX) += icswx.o obj-$(CONFIG_PPC_ICSWX_PID) += icswx_pid.o obj-$(CONFIG_40x) += 40x_mmu.o
arch/powerpc/mm/hash64_64k.c new file +202 -0
202 lines not on this page · raw message
arch/powerpc/mm/hash_low_64.S +0 -380
391 lines not on this page · raw message
arch/powerpc/mm/hash_utils_64.c +3 -1
@@ -653,7 +653,7 @@ static void __init htab_finish_init(void)patch_branch (ht64_call_hpte_updatepp,ppc_function_entry (ppc_md.hpte_updatepp),BRANCH_SET_LINK ); -#endif /* CONFIG_PPC_64K_PAGES */ +#else /* !CONFIG_PPC_64K_PAGES */ patch_branch(htab_call_hpte_insert1, ppc_function_entry(ppc_md.hpte_insert),
@@ -667,6 +667,8 @@ static void __init htab_finish_init(void)patch_branch (htab_call_hpte_updatepp,ppc_function_entry (ppc_md.hpte_updatepp),BRANCH_SET_LINK ); +#endif + }static void __init htab_initialize (void )
Signature · 1 line
2.5.0
PATCH v6 19/35 powerpc/mm: Remove the dependency on pte bit position in asm code
In reply to Aneesh Kumar K.VWe should not expect pte bit position in asm code. Simply by moving part of that to C
--- arch/powerpc/kernel/exceptions-64s.S | 18 ++++-------------- arch/powerpc/mm/hash_utils_64.c | 29 +++++++++++++++++++++++++++++ 2 files changed, 33 insertions(+), 14 deletions(-)
arch/powerpc/kernel/exceptions-64s.S +4 -14
@@ -1556,29 +1556,19 @@ do_hash_page: lwz r0,TI_PREEMPT(r11) /* If we're in an "NMI" */ andis. r0,r0,NMI_MASK@h /* (i.e. an irq when soft-disabled) */ bne 77f /* then don't call hash_page now */ - /* - * We need to set the _PAGE_USER bit if MSR_PR is set or if we are - * accessing a userspace segment (even from the kernel). We assume - * kernel addresses always have the high bit set. - */ - rlwinm r4,r4,32-25+9,31-9,31-9 /* DSISR_STORE -> _PAGE_RW */ - rotldi r0,r3,15 /* Move high bit into MSR_PR posn */ - orc r0,r12,r0 /* MSR_PR | ~high_bit */ - rlwimi r4,r0,32-13,30,30 /* becomes _PAGE_USER access bit */ - ori r4,r4,1 /* add _PAGE_PRESENT */ - rlwimi r4,r5,22+2,31-2,31-2 /* Set _PAGE_EXEC if trap is 0x400 */ /* * r3 contains the faulting address - * r4 contains the required access permissions + * r4 msr * r5 contains the trap number * r6 contains dsisr * * at return r3 = 0 for success, 1 for page fault, negative for error */ + mr r4,r12 ld r6,_DSISR(r1) - bl hash_page /* build HPTE if possible */ - cmpdi r3,0 /* see if hash_page succeeded */ + bl __hash_page /* build HPTE if possible */ + cmpdi r3,0 /* see if __hash_page succeeded */ /* Success */ beq fast_exc_return_irq /* Return from exception on success */
arch/powerpc/mm/hash_utils_64.c +29 -0
@@ -1205,6 +1205,35 @@ int hash_page(unsigned long ea, unsigned long access, unsigned long trap, }EXPORT_SYMBOL_GPL (hash_page); +int __hash_page (unsigned long ea,unsigned long msr,unsigned long trap, +unsigned long dsisr) +{ +unsigned long access = _PAGE_PRESENT; +unsigned long flags =0 ; +struct mm_struct *mm = current->mm; + +if (REGION_ID (ea) ==VMALLOC_REGION_ID ) + mm = &init_mm; + +if (dsisr &DSISR_NOHPTE ) + flags |=HPTE_NOHPTE_UPDATE ; + +if (dsisr &DSISR_ISSTORE ) + access |= _PAGE_RW; +/* +* We need to set the _PAGE_USER bit if MSR_PR is set or if we are +* accessing a userspace segment (even from the kernel). We assume +* kernel addresses always have the high bit set. +*/ +if ((msr &MSR_PR ) || (REGION_ID (ea) ==USER_REGION_ID )) + access |= _PAGE_USER; + +if (trap ==0x400 ) + access |= _PAGE_EXEC; + +return hash_page_mm (mm, ea, access, trap, flags); +} +void hash_preload (struct mm_struct *mm,unsigned long ea,unsigned long access,unsigned long trap) {
Signature · 1 line
2.5.0
PATCH v6 20/35 powerpc/mm: Don't track subpage valid bit in pte_t
In reply to Aneesh Kumar K.VThis free up 11 bits in pte_t. In the later patch we also change the pte_t format so that we can start supporting migration pte at pmd level. We now track 4k subpage valid bit as below
If we have _PAGE_COMBO set, we override the _PAGE_F_GIX_SHIFT and _PAGE_F_SECOND. Together we have 4 bits, each of them used to indicate whether any of the 4 4k subpage in that group is valid. ie,
[ group 1 bit ] [ group 2 bit ] ..... [ group 4 ] [ subpage 1 - 4] [ subpage 5- 8] ..... [ subpage 13 - 16]
We still track each 4k subpage slot number and secondary hash information in the second half of pgtable_t. Removing the subpage tracking have some significant overhead on aim9 and ebizzy benchmark and to support THP with 4K subpage, we do need a pgtable_t of 4096 bytes.
--- arch/powerpc/include/asm/book3s/64/hash-4k.h | 10 +------- arch/powerpc/include/asm/book3s/64/hash-64k.h | 35 ++++++--------------------- arch/powerpc/include/asm/book3s/64/hash.h | 10 ++++---- arch/powerpc/mm/hash64_64k.c | 34 ++++++++++++++++++++++++-- arch/powerpc/mm/hash_low_64.S | 6 +---- arch/powerpc/mm/hugetlbpage-hash64.c | 5 +--- arch/powerpc/mm/pgtable_64.c | 2 +- 7 files changed, 48 insertions(+), 54 deletions(-)
arch/powerpc/include/asm/book3s/64/hash-4k.h +1 -9
@@ -47,17 +47,9 @@/* Bits to mask out from a PGD to get to the PUD page */ #define PGD_MASKED_BITS 0 -/* PTE bits */ -#define _PAGE_HASHPTE 0x0400/* software: pte has an associated HPTE */ -#define _PAGE_SECONDARY 0x8000/* software: HPTE is in secondary group */ -#define _PAGE_GROUP_IX 0x7000/* software: HPTE index within group */ -#define _PAGE_F_SECOND _PAGE_SECONDARY -#define _PAGE_F_GIX _PAGE_GROUP_IX -#define _PAGE_SPECIAL 0x10000/* software: special page */ -/* PTE flags to conserve for HPTE identification */ #define _PAGE_HPTEFLAGS (_PAGE_BUSY | _PAGE_HASHPTE | \ - _PAGE_SECONDARY | _PAGE_GROUP_IX) + _PAGE_F_SECOND | _PAGE_F_GIX)/* shift to put page number into pte */ #define PTE_RPN_SHIFT (17)
arch/powerpc/include/asm/book3s/64/hash-64k.h +7 -28
@@ -31,33 +31,13 @@/* Bits to mask out from a PGD/PUD to get to the PMD page */ #define PUD_MASKED_BITS 0x1ff -/* Additional PTE bits (don't change without checking asm in hash_low.S) */ -#define _PAGE_SPECIAL 0x00000400/* software: special page */ -#define _PAGE_HPTE_SUB 0x0ffff000/* combo only: sub pages HPTE bits */ -#define _PAGE_HPTE_SUB0 0x08000000/* combo only: first sub page */ -#define _PAGE_COMBO 0x10000000/* this is a combo 4k page */ -#define _PAGE_4K_PFN 0x20000000/* PFN is for a single 4k page */ - -/* For 64K page, we don't have a separate _PAGE_HASHPTE bit. Instead, -* we set that to be the whole sub-bits mask. The C code will only -* test this, so a multi-bit mask will work. For combo pages, this -* is equivalent as effectively, the old _PAGE_HASHPTE was an OR of -* all the sub bits. For real 64k pages, we now have the assembly set -* _PAGE_HPTE_SUB0 in addition to setting the HIDX bits which overlap -* that mask. This is fine as long as the HIDX bits are never set on -* a PTE that isn't hashed, which is the case today. -* -* A little nit is for the huge page C code, which does the hashing -* in C, we need to provide which bit to use. -*/ -#define _PAGE_HASHPTE _PAGE_HPTE_SUB - -/* Note the full page bits must be in the same location as for normal -* 4k pages as the same assembly will be used to insert 64K pages -* whether the kernel has CONFIG_PPC_64K_PAGES or not +#define _PAGE_COMBO 0x00020000/* this is a combo 4k page */ +#define _PAGE_4K_PFN 0x00040000/* PFN is for a single 4k page */ +/* +* Used to track subpage group valid if _PAGE_COMBO is set +* This overloads _PAGE_F_GIX and _PAGE_F_SECOND */ -#define _PAGE_F_SECOND 0x00008000/* full page: hidx bits */ -#define _PAGE_F_GIX 0x00007000/* full page: hidx bits */ +#define _PAGE_COMBO_VALID (_PAGE_F_GIX | _PAGE_F_SECOND)/* PTE flags to conserve for HPTE identification */ #define _PAGE_HPTEFLAGS (_PAGE_BUSY | _PAGE_HASHPTE | _PAGE_COMBO)
@@ -103,8 +83,7 @@ static inline unsigned long __rpte_to_hidx(real_pte_t rpte, unsigned long index) }#define __rpte_to_pte (r) ((r).pte) -#define __rpte_sub_valid (rpte, index) \ - (pte_val(rpte.pte) & (_PAGE_HPTE_SUB0 >> (index))) +extern bool __rpte_sub_valid (real_pte_t rpte,unsigned long index);/* * Trick: we set __end to va + 64k, which happens works for * a 16M page as well as we want only one iteration
arch/powerpc/include/asm/book3s/64/hash.h +5 -5
@@ -81,7 +81,12 @@#define _PAGE_DIRTY 0x0080/* C: page changed */ #define _PAGE_ACCESSED 0x0100/* R: page referenced */ #define _PAGE_RW 0x0200/* software: user write access allowed */ +#define _PAGE_HASHPTE 0x0400/* software: pte has an associated HPTE */ #define _PAGE_BUSY 0x0800/* software: PTE & hash are busy */ +#define _PAGE_F_GIX 0x7000/* full page: hidx bits */ +#define _PAGE_F_GIX_SHIFT 12 +#define _PAGE_F_SECOND 0x8000/* Whether to use secondary hash or not */ +#define _PAGE_SPECIAL 0x10000/* software: special page */ /* No separate kernel read-only */ #define _PAGE_KERNEL_RW (_PAGE_RW | _PAGE_DIRTY)/* user access blocked by key */
@@ -210,11 +215,6 @@#define PMD_BAD_BITS (PTE_TABLE_SIZE-1)#define PUD_BAD_BITS (PMD_TABLE_SIZE-1) -/* -* We save the slot number & secondary bit in the second half of the -* PTE page. We use the 8 bytes per each pte entry. -*/ -#define PTE_PAGE_HIDX_OFFSET (PTRS_PER_PTE * 8)#ifndef __ASSEMBLY__#define pmd_bad (pmd) (!is_kernel_addr(pmd_val(pmd)) \
arch/powerpc/mm/hash64_64k.c +32 -2
@@ -15,6 +15,35 @@#include <linux/mm.h> #include <asm/machdep.h> #include <asm/mmu.h> +/* +* index from 0 - 15 +*/ +bool __rpte_sub_valid (real_pte_t rpte,unsigned long index) +{ +unsigned long g_idx; +unsigned long ptev =pte_val (rpte.pte); + + g_idx = (ptev & _PAGE_COMBO_VALID) >> _PAGE_F_GIX_SHIFT; + index = index >>2 ; +if (g_idx & (0x1 << index)) +return true; +else +return false; +} +/* +* index from 0 - 15 +*/ +static unsigned long mark_subptegroup_valid (unsigned long ptev,unsigned long index) +{ +unsigned long g_idx; + +if (!(ptev & _PAGE_COMBO)) +return ptev; + index = index >>2 ; + g_idx =0x1 << index; + +return ptev | (g_idx << _PAGE_F_GIX_SHIFT); +} int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid, pte_t *ptep, unsigned long trap, unsigned long flags,
@@ -102,7 +131,7 @@ int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid,*/ if (!(old_pte & _PAGE_COMBO)) {flush_hash_page (vpn, rpte,MMU_PAGE_64K , ssize, flags); - old_pte &= ~_PAGE_HPTE_SUB; + old_pte &= ~_PAGE_HASHPTE | _PAGE_F_GIX | _PAGE_F_SECOND; goto htab_insert_hpte; }/*
@@ -192,7 +221,8 @@ repeat:/* __real_pte use pte_val() any idea why ? FIXME!! */ rpte.hidx &= ~(0xfUL << (subpg_index <<2 )); *hidxp = rpte.hidx | (slot << (subpg_index <<2 )); - new_pte |= (_PAGE_HPTE_SUB0 >> subpg_index); + new_pte =mark_subptegroup_valid (new_pte, subpg_index); + new_pte |= _PAGE_HASHPTE;/* * check __real_pte for details on matching smp_rmb() */
arch/powerpc/mm/hash_low_64.S +1 -5
@@ -285,7 +285,7 @@ htab_modify_pte: /* Secondary group ? if yes, get a inverted hash value */ mr r5,r28 - andi. r0,r31,_PAGE_SECONDARY + andi. r0,r31,_PAGE_F_SECOND beq 1f not r5,r5 1:
@@ -473,11 +473,7 @@ ht64_insert_pte: lis r0,_PAGE_HPTEFLAGS@h ori r0,r0,_PAGE_HPTEFLAGS@l andc r30,r30,r0 -#ifdef CONFIG_PPC_64K_PAGES - oris r30,r30,_PAGE_HPTE_SUB0@h -#else ori r30,r30,_PAGE_HASHPTE -#endif /* Phyical address in r5 */ rldicl r5,r31,64-PTE_RPN_SHIFT,PTE_RPN_SHIFT sldi r5,r5,PAGE_SHIFT
arch/powerpc/mm/hugetlbpage-hash64.c +1 -4
@@ -91,11 +91,8 @@ int __hash_page_huge(unsigned long ea, unsigned long access, unsigned long vsid, pa =pte_pfn (__pte (old_pte)) <<PAGE_SHIFT ;/* clear HPTE slot informations in new PTE */ -#ifdef CONFIG_PPC_64K_PAGES - new_pte = (new_pte & ~_PAGE_HPTEFLAGS) | _PAGE_HPTE_SUB0; -#else new_pte = (new_pte & ~_PAGE_HPTEFLAGS) | _PAGE_HASHPTE; -#endif +/* Add in WIMG bits */ rflags |= (new_pte & (_PAGE_WRITETHRU | _PAGE_NO_CACHE | _PAGE_COHERENT | _PAGE_GUARDED));
arch/powerpc/mm/pgtable_64.c +1 -1
@@ -625,7 +625,7 @@ void pmdp_splitting_flush(struct vm_area_struct *vma,"1: ldarx %0,0,%3\n\ andi. %1,%0,%6\n\ bne- 1b \n\ -ori %1,%0,%4 \n\ +oris %1,%0,%4@h \n\ stdcx. %1,0,%3 \n\ bne- 1b" : "=&r" (old), "=&r" (tmp), "=m" (*pmdp)
Signature · 1 line
2.5.0
Re: [PATCH V6 20/35] powerpc/mm: Don't track subpage valid bit in pte_t
Quoted 12 lines · Aneesh Kumar K.V
This free up 11 bits in
pte_t. In the later patch we also change thepte_tformat so that we can start supporting migration pte at pmd level. We now track 4k subpage valid bit as belowIf we have
_PAGE_COMBOset, we override the_PAGE_F_GIX_SHIFTand_PAGE_F_SECOND. Together we have 4 bits, each of them used to indicate whether any of the 4 4k subpage in that group is valid. ie,[ group 1 bit ] [ group 2 bit ] ..... [ group 4 ] [ subpage 1 - 4] [ subpage 5- 8] ..... [ subpage 13 - 16]
Quoted 6 lines
We still track each 4k subpage slot number and secondary hash information in the second half of
pgtable_t. Removing the subpage tracking have some significant overhead on aim9 and ebizzy benchmark and to support THP with 4K subpage, we do need apgtable_tof 4096 bytes.Signed-off-by Aneesh Kumar K.V <[email protected]>
We removed 16 bits which used to track individual sub pages and added back 4 bits to track newly created sub page groups. So we save 12 bits there. How it is 11 bits ? Have we added back one more bit into pte_t some where for sub page purpose which I am missing ?
Quoted 129 lines
--- arch/powerpc/include/asm/book3s/64/hash-4k.h | 10 +------- arch/powerpc/include/asm/book3s/64/hash-64k.h | 35 ++++++--------------------- arch/powerpc/include/asm/book3s/64/hash.h | 10 ++++---- arch/powerpc/mm/hash64_64k.c | 34 ++++++++++++++++++++++++-- arch/powerpc/mm/hash_low_64.S | 6 +---- arch/powerpc/mm/hugetlbpage-hash64.c | 5 +--- arch/powerpc/mm/pgtable_64.c | 2 +- 7 files changed, 48 insertions(+), 54 deletions(-)arch/powerpc/include/asm/book3s/64/hash-4k.h +1
-9@@ -47,17 +47,9 @@/* Bits to mask out from a PGD to get to the PUD page */ #define PGD_MASKED_BITS 0 -/* PTE bits */ -#define _PAGE_HASHPTE 0x0400/* software: pte has an associated HPTE */ -#define _PAGE_SECONDARY 0x8000/* software: HPTE is in secondary group */ -#define _PAGE_GROUP_IX 0x7000/* software: HPTE index within group */ -#define _PAGE_F_SECOND _PAGE_SECONDARY -#define _PAGE_F_GIX _PAGE_GROUP_IX -#define _PAGE_SPECIAL 0x10000/* software: special page */ -/* PTE flags to conserve for HPTE identification */ #define _PAGE_HPTEFLAGS (_PAGE_BUSY | _PAGE_HASHPTE | \ - _PAGE_SECONDARY | _PAGE_GROUP_IX) + _PAGE_F_SECOND | _PAGE_F_GIX)/* shift to put page number into pte */ #define PTE_RPN_SHIFT (17)arch/powerpc/include/asm/book3s/64/hash-64k.h +7
-28@@ -31,33 +31,13 @@/* Bits to mask out from a PGD/PUD to get to the PMD page */ #define PUD_MASKED_BITS 0x1ff -/* Additional PTE bits (don't change without checking asm in hash_low.S) */ -#define _PAGE_SPECIAL 0x00000400/* software: special page */ -#define _PAGE_HPTE_SUB 0x0ffff000/* combo only: sub pages HPTE bits */ -#define _PAGE_HPTE_SUB0 0x08000000/* combo only: first sub page */ -#define _PAGE_COMBO 0x10000000/* this is a combo 4k page */ -#define _PAGE_4K_PFN 0x20000000/* PFN is for a single 4k page */ - -/* For 64K page, we don't have a separate _PAGE_HASHPTE bit. Instead, -* we set that to be the whole sub-bits mask. The C code will only -* test this, so a multi-bit mask will work. For combo pages, this -* is equivalent as effectively, the old _PAGE_HASHPTE was an OR of -* all the sub bits. For real 64k pages, we now have the assembly set -* _PAGE_HPTE_SUB0 in addition to setting the HIDX bits which overlap -* that mask. This is fine as long as the HIDX bits are never set on -* a PTE that isn't hashed, which is the case today. -* -* A little nit is for the huge page C code, which does the hashing -* in C, we need to provide which bit to use. -*/ -#define _PAGE_HASHPTE _PAGE_HPTE_SUB - -/* Note the full page bits must be in the same location as for normal -* 4k pages as the same assembly will be used to insert 64K pages -* whether the kernel has CONFIG_PPC_64K_PAGES or not +#define _PAGE_COMBO 0x00020000/* this is a combo 4k page */ +#define _PAGE_4K_PFN 0x00040000/* PFN is for a single 4k page */ +/* +* Used to track subpage group valid if _PAGE_COMBO is set +* This overloads _PAGE_F_GIX and _PAGE_F_SECOND */ -#define _PAGE_F_SECOND 0x00008000/* full page: hidx bits */ -#define _PAGE_F_GIX 0x00007000/* full page: hidx bits */ +#define _PAGE_COMBO_VALID (_PAGE_F_GIX | _PAGE_F_SECOND)/* PTE flags to conserve for HPTE identification */ #define _PAGE_HPTEFLAGS (_PAGE_BUSY | _PAGE_HASHPTE | _PAGE_COMBO)@@ -103,8 +83,7 @@ static inline unsigned long __rpte_to_hidx(real_pte_t rpte, unsigned long index) }#define __rpte_to_pte (r) ((r).pte) -#define __rpte_sub_valid (rpte, index) \ - (pte_val(rpte.pte) & (_PAGE_HPTE_SUB0 >> (index))) +extern bool __rpte_sub_valid (real_pte_t rpte,unsigned long index);/* * Trick: we set __end to va + 64k, which happens works for * a 16M page as well as we want only one iteration arch/powerpc/include/asm/book3s/64/hash.h +5
-5@@ -81,7 +81,12 @@#define _PAGE_DIRTY 0x0080/* C: page changed */ #define _PAGE_ACCESSED 0x0100/* R: page referenced */ #define _PAGE_RW 0x0200/* software: user write access allowed */ +#define _PAGE_HASHPTE 0x0400/* software: pte has an associated HPTE */ #define _PAGE_BUSY 0x0800/* software: PTE & hash are busy */ +#define _PAGE_F_GIX 0x7000/* full page: hidx bits */ +#define _PAGE_F_GIX_SHIFT 12 +#define _PAGE_F_SECOND 0x8000/* Whether to use secondary hash or not */ +#define _PAGE_SPECIAL 0x10000/* software: special page */ /* No separate kernel read-only */ #define _PAGE_KERNEL_RW (_PAGE_RW | _PAGE_DIRTY)/* user access blocked by key */ @@ -210,11 +215,6 @@#define PMD_BAD_BITS (PTE_TABLE_SIZE-1)#define PUD_BAD_BITS (PMD_TABLE_SIZE-1) -/* -* We save the slot number & secondary bit in the second half of the -* PTE page. We use the 8 bytes per each pte entry. -*/ -#define PTE_PAGE_HIDX_OFFSET (PTRS_PER_PTE * 8)#ifndef __ASSEMBLY__#define pmd_bad (pmd) (!is_kernel_addr(pmd_val(pmd)) \arch/powerpc/mm/hash64_64k.c +5
-0@@ -15,6 +15,35 @@#include <linux/mm.h> #include <asm/machdep.h> #include <asm/mmu.h> +/* +* index from 0 - 15 +*/ +bool __rpte_sub_valid (real_pte_t rpte,unsigned long index) +{
arch/powerpc/mm/hash64_64k.c line 32
Quoted 10 lines
arch/powerpc/mm/hash64_64k.c +10
-0+unsigned long g_idx; +unsigned long ptev =pte_val (rpte.pte); + + g_idx = (ptev & _PAGE_COMBO_VALID) >> _PAGE_F_GIX_SHIFT; + index = index >>2 ; +if (g_idx & (0x1 << index)) +return true; +else +return false; +}
This function checks for validity of the sub page group not individual sub page index. Because we dont track individual sub page index validity any more, wondering if that is even possible.
Quoted 18 lines
arch/powerpc/mm/hash64_64k.c +14
-0+/* +* index from 0 - 15 +*/ +static unsigned long mark_subptegroup_valid (unsigned long ptev,unsigned long index) +{ +unsigned long g_idx; + +if (!(ptev & _PAGE_COMBO)) +return ptev; + index = index >>2 ; + g_idx =0x1 << index; + +return ptev | (g_idx << _PAGE_F_GIX_SHIFT); +} int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid, pte_t *ptep, unsigned long trap, unsigned long flags,
arch/powerpc/mm/hash64_64k.c int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid, · line 134
Quoted 5 lines
arch/powerpc/mm/hash64_64k.c +1
-1@@ -102,7 +131,7 @@ int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid,*/ if (!(old_pte & _PAGE_COMBO)) {flush_hash_page (vpn, rpte,MMU_PAGE_64K , ssize, flags); - old_pte &= ~_PAGE_HPTE_SUB; + old_pte &= ~_PAGE_HASHPTE | _PAGE_F_GIX | _PAGE_F_SECOND;
Or use _PAGE_COMBO_VALID directly instead ?
Quoted 60 lines
arch/powerpc/mm/hash64_64k.c +2
-1goto htab_insert_hpte; }/* @@ -192,7 +221,8 @@ repeat:/* __real_pte use pte_val() any idea why ? FIXME!! */ rpte.hidx &= ~(0xfUL << (subpg_index <<2 )); *hidxp = rpte.hidx | (slot << (subpg_index <<2 )); - new_pte |= (_PAGE_HPTE_SUB0 >> subpg_index); + new_pte =mark_subptegroup_valid (new_pte, subpg_index); + new_pte |= _PAGE_HASHPTE;/* * check __real_pte for details on matching smp_rmb() */ arch/powerpc/mm/hash_low_64.S +1
-5@@ -285,7 +285,7 @@ htab_modify_pte: /* Secondary group ? if yes, get a inverted hash value */ mr r5,r28 - andi. r0,r31,_PAGE_SECONDARY + andi. r0,r31,_PAGE_F_SECOND beq 1f not r5,r5 1:@@ -473,11 +473,7 @@ ht64_insert_pte: lis r0,_PAGE_HPTEFLAGS@h ori r0,r0,_PAGE_HPTEFLAGS@l andc r30,r30,r0 -#ifdef CONFIG_PPC_64K_PAGES - oris r30,r30,_PAGE_HPTE_SUB0@h -#else ori r30,r30,_PAGE_HASHPTE -#endif /* Phyical address in r5 */ rldicl r5,r31,64-PTE_RPN_SHIFT,PTE_RPN_SHIFT sldi r5,r5,PAGE_SHIFTarch/powerpc/mm/hugetlbpage-hash64.c +1
-4@@ -91,11 +91,8 @@ int __hash_page_huge(unsigned long ea, unsigned long access, unsigned long vsid, pa =pte_pfn (__pte (old_pte)) <<PAGE_SHIFT ;/* clear HPTE slot informations in new PTE */ -#ifdef CONFIG_PPC_64K_PAGES - new_pte = (new_pte & ~_PAGE_HPTEFLAGS) | _PAGE_HPTE_SUB0; -#else new_pte = (new_pte & ~_PAGE_HPTEFLAGS) | _PAGE_HASHPTE; -#endif +/* Add in WIMG bits */ rflags |= (new_pte & (_PAGE_WRITETHRU | _PAGE_NO_CACHE | _PAGE_COHERENT | _PAGE_GUARDED));arch/powerpc/mm/pgtable_64.c +0
-0
arch/powerpc/mm/pgtable_64.c void pmdp_splitting_flush(struct vm_area_struct *vma, · line 628
Quoted 5 lines
arch/powerpc/mm/pgtable_64.c +1
-1@@ -625,7 +625,7 @@ void pmdp_splitting_flush(struct vm_area_struct *vma,"1: ldarx %0,0,%3\n\ andi. %1,%0,%6\n\ bne- 1b \n\ -ori %1,%0,%4 \n\ +oris %1,%0,%4@h \n\
Why is this change required ?
Re: [PATCH V6 20/35] powerpc/mm: Don't track subpage valid bit in pte_t
Quoted 18 lines · Anshuman Khandual
Quoted 17 lines · Aneesh Kumar K.V
This free up 11 bits in
pte_t. In the later patch we also change thepte_tformat so that we can start supporting migration pte at pmd level. We now track 4k subpage valid bit as belowIf we have
_PAGE_COMBOset, we override the_PAGE_F_GIX_SHIFTand_PAGE_F_SECOND. Together we have 4 bits, each of them used to indicate whether any of the 4 4k subpage in that group is valid. ie,[ group 1 bit ] [ group 2 bit ] ..... [ group 4 ] [ subpage 1 - 4] [ subpage 5- 8] ..... [ subpage 13 - 16]We still track each 4k subpage slot number and secondary hash information in the second half of
pgtable_t. Removing the subpage tracking have some significant overhead on aim9 and ebizzy benchmark and to support THP with 4K subpage, we do need apgtable_tof 4096 bytes.
Quoted 6 lines
Quoted 1 line
Signed-off-by Aneesh Kumar K.V <[email protected]>We removed 16 bits which used to track individual sub pages and added back 4 bits to track newly created sub page groups. So we save 12 bits there. How it is 11 bits ? Have we added back one more bit into
pte_tsome where for sub page purpose which I am missing ?
We added back 5 bits, (I guess you are missing _PAGE_HASHPTE )
Quoted 6 lines
Quoted 6 lines
--- arch/powerpc/include/asm/book3s/64/hash-4k.h | 10 +------- arch/powerpc/include/asm/book3s/64/hash-64k.h | 35 ++++++--------------------- arch/powerpc/include/asm/book3s/64/hash.h | 10 ++++---- arch/powerpc/mm/hash64_64k.c | 34 ++++++++++++++++++++++++--
.....
......
Quoted 19 lines
arch/powerpc/mm/hash64_64k.c +0
-0Quoted 18 lines
index
9ffeae2cbb57..f1b86ba63430100644arch/powerpc/mm/hash64_64k.c +11
-0@@ -15,6 +15,35 @@#include <linux/mm.h> #include <asm/machdep.h> #include <asm/mmu.h> +/* +* index from 0 - 15 +*/ +bool __rpte_sub_valid (real_pte_t rpte,unsigned long index) +{ +unsigned long g_idx; +unsigned long ptev =pte_val (rpte.pte); + + g_idx = (ptev & _PAGE_COMBO_VALID) >> _PAGE_F_GIX_SHIFT; + index = index >>2 ; +if (g_idx & (0x1 << index))
Quoted 9 lines
arch/powerpc/mm/hash64_64k.c line 32
Quoted 4 lines
arch/powerpc/mm/hash64_64k.c +4
-0+return true; + else + return false; +}This function checks for validity of the sub page group not individual sub page index. Because we dont track individual sub page index validity any more, wondering if that is even possible.
That is correct. We are not tracking individual subpage validity after this patch. That is explained in the commit message.
Quoted 15 lines
Quoted 15 lines
arch/powerpc/mm/hash64_64k.c +14
-0+/* +* index from 0 - 15 +*/ +static unsigned long mark_subptegroup_valid (unsigned long ptev,unsigned long index) +{ +unsigned long g_idx; + +if (!(ptev & _PAGE_COMBO)) +return ptev; + index = index >>2 ; + g_idx =0x1 << index; + +return ptev | (g_idx << _PAGE_F_GIX_SHIFT); +}
Quoted 10 lines
arch/powerpc/mm/hash64_64k.c int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid, · line 134
Quoted 8 lines
arch/powerpc/mm/hash64_64k.c +1
-1int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid, pte_t *ptep, unsigned long trap, unsigned long flags,@@ -102,7 +131,7 @@ int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid,*/ if (!(old_pte & _PAGE_COMBO)) {flush_hash_page (vpn, rpte,MMU_PAGE_64K , ssize, flags); - old_pte &= ~_PAGE_HPTE_SUB; + old_pte &= ~_PAGE_HASHPTE | _PAGE_F_GIX | _PAGE_F_SECOND;Or use
_PAGE_COMBO_VALIDdirectly instead ?
Nope we are not clearning a combo related detail there. We are invalidating an old 64K mapping there. (yes, the values are matching but we should not use the _PAGE_COMBO_VALID there).
Quoted 60 lines
Quoted 60 lines
arch/powerpc/mm/hash64_64k.c +2
-1goto htab_insert_hpte; }/* @@ -192,7 +221,8 @@ repeat:/* __real_pte use pte_val() any idea why ? FIXME!! */ rpte.hidx &= ~(0xfUL << (subpg_index <<2 )); *hidxp = rpte.hidx | (slot << (subpg_index <<2 )); - new_pte |= (_PAGE_HPTE_SUB0 >> subpg_index); + new_pte =mark_subptegroup_valid (new_pte, subpg_index); + new_pte |= _PAGE_HASHPTE;/* * check __real_pte for details on matching smp_rmb() */ arch/powerpc/mm/hash_low_64.S +1
-5@@ -285,7 +285,7 @@ htab_modify_pte: /* Secondary group ? if yes, get a inverted hash value */ mr r5,r28 - andi. r0,r31,_PAGE_SECONDARY + andi. r0,r31,_PAGE_F_SECOND beq 1f not r5,r5 1:@@ -473,11 +473,7 @@ ht64_insert_pte: lis r0,_PAGE_HPTEFLAGS@h ori r0,r0,_PAGE_HPTEFLAGS@l andc r30,r30,r0 -#ifdef CONFIG_PPC_64K_PAGES - oris r30,r30,_PAGE_HPTE_SUB0@h -#else ori r30,r30,_PAGE_HASHPTE -#endif /* Phyical address in r5 */ rldicl r5,r31,64-PTE_RPN_SHIFT,PTE_RPN_SHIFT sldi r5,r5,PAGE_SHIFTarch/powerpc/mm/hugetlbpage-hash64.c +1
-4@@ -91,11 +91,8 @@ int __hash_page_huge(unsigned long ea, unsigned long access, unsigned long vsid, pa =pte_pfn (__pte (old_pte)) <<PAGE_SHIFT ;/* clear HPTE slot informations in new PTE */ -#ifdef CONFIG_PPC_64K_PAGES - new_pte = (new_pte & ~_PAGE_HPTEFLAGS) | _PAGE_HPTE_SUB0; -#else new_pte = (new_pte & ~_PAGE_HPTEFLAGS) | _PAGE_HASHPTE; -#endif +/* Add in WIMG bits */ rflags |= (new_pte & (_PAGE_WRITETHRU | _PAGE_NO_CACHE | _PAGE_COHERENT | _PAGE_GUARDED));arch/powerpc/mm/pgtable_64.c +0
-0
Quoted 8 lines
arch/powerpc/mm/pgtable_64.c void pmdp_splitting_flush(struct vm_area_struct *vma, · line 628
Quoted 6 lines
arch/powerpc/mm/pgtable_64.c +1
-1@@ -625,7 +625,7 @@ void pmdp_splitting_flush(struct vm_area_struct *vma,"1: ldarx %0,0,%3\n\ andi. %1,%0,%6\n\ bne- 1b \n\ -ori %1,%0,%4 \n\ +oris %1,%0,%4@h \n\ Why is this change required ?
Because _PAGE_SPLITTING is beyond 16 bits now ?
-aneesh
Re: [PATCH V6 20/35] powerpc/mm: Don't track subpage valid bit in pte_t
In reply to Aneesh Kumar K.VQuoted 9 lines · Aneesh Kumar K.V
This free up 11 bits in
pte_t. In the later patch we also change thepte_tformat so that we can start supporting migration pte at pmd level. We now track 4k subpage valid bit as belowIf we have
_PAGE_COMBOset, we override the_PAGE_F_GIX_SHIFTand_PAGE_F_SECOND. Together we have 4 bits, each of them used to indicate whether any of the 4 4k subpage in that group is valid. ie,
Quoted 7 lines
[ group 1 bit ] [ group 2 bit ] ..... [ group 4 ] [ subpage 1 - 4] [ subpage 5- 8] ..... [ subpage 13 - 16]We still track each 4k subpage slot number and secondary hash information in the second half of
pgtable_t. Removing the subpage tracking have some significant overhead on aim9 and ebizzy benchmark and to support THP with 4K subpage, we do need apgtable_tof 4096 bytes.
I know this has already been applied, but this hunk looks wrong:
arch/powerpc/mm/hash64_64k.c int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid, · line 134
Quoted 6 lines
arch/powerpc/mm/hash64_64k.c +1
-1@@ -102,7 +131,7 @@ int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid,*/ if (!(old_pte & _PAGE_COMBO)) {flush_hash_page (vpn, rpte,MMU_PAGE_64K , ssize, flags); - old_pte &= ~_PAGE_HPTE_SUB; + old_pte &= ~_PAGE_HASHPTE | _PAGE_F_GIX | _PAGE_F_SECOND;
Shouldn't this be:
+ old_pte &= ~(_PAGE_HASHPTE | _PAGE_F_GIX | _PAGE_F_SECOND);
instead?
Paul.
Re: [PATCH V6 20/35] powerpc/mm: Don't track subpage valid bit in pte_t
Quoted 31 lines · Paul Mackerras
Quoted 9 lines · Aneesh Kumar K.V
This free up 11 bits in
pte_t. In the later patch we also change thepte_tformat so that we can start supporting migration pte at pmd level. We now track 4k subpage valid bit as belowIf we have
_PAGE_COMBOset, we override the_PAGE_F_GIX_SHIFTand_PAGE_F_SECOND. Together we have 4 bits, each of them used to indicate whether any of the 4 4k subpage in that group is valid. ie,Quoted 7 lines
[ group 1 bit ] [ group 2 bit ] ..... [ group 4 ] [ subpage 1 - 4] [ subpage 5- 8] ..... [ subpage 13 - 16]We still track each 4k subpage slot number and secondary hash information in the second half of
pgtable_t. Removing the subpage tracking have some significant overhead on aim9 and ebizzy benchmark and to support THP with 4K subpage, we do need apgtable_tof 4096 bytes.I know this has already been applied, but this hunk looks wrong:
arch/powerpc/mm/hash64_64k.c int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid, · line 134
Quoted 6 lines
arch/powerpc/mm/hash64_64k.c +1
-1@@ -102,7 +131,7 @@ int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid,*/ if (!(old_pte & _PAGE_COMBO)) {flush_hash_page (vpn, rpte,MMU_PAGE_64K , ssize, flags); - old_pte &= ~_PAGE_HPTE_SUB; + old_pte &= ~_PAGE_HASHPTE | _PAGE_F_GIX | _PAGE_F_SECOND;Shouldn't this be:
(quoted diff) +1
-0+ old_pte &= ~(_PAGE_HASHPTE | _PAGE_F_GIX | _PAGE_F_SECOND);
Quoted 1 line
instead?
Thanks for checking this closely. Yes it should be what you suggested. I will do a patch for this.
-aneesh
PATCH v6 21/35 powerpc/mm: Remove pte_val usage for the second half of pgtable_t
In reply to Aneesh Kumar K.V--- arch/powerpc/include/asm/book3s/64/hash-64k.h | 4 +++- arch/powerpc/mm/hash64_64k.c | 1 - 2 files changed, 3 insertions(+), 2 deletions(-)
arch/powerpc/include/asm/book3s/64/hash-64k.h +3 -1
@@ -61,6 +61,7 @@static inline real_pte_t __real_pte (pte_t pte,pte_t *ptep) {real_pte_t rpte; +unsigned long *hidxp; rpte.pte = pte; rpte.hidx =0 ;
@@ -70,7 +71,8 @@ static inline real_pte_t __real_pte(pte_t pte, pte_t *ptep)* check. The store side ordering is done in __hash_page_4K */ smp_rmb (); - rpte.hidx =pte_val (*((ptep) +PTRS_PER_PTE )); + hidxp = (unsigned long *)(ptep +PTRS_PER_PTE ); + rpte.hidx = *hidxp; }return rpte; }
arch/powerpc/mm/hash64_64k.c +0 -1
@@ -218,7 +218,6 @@ repeat:* nobody is undating hidx. */ hidxp = (unsigned long *)(ptep +PTRS_PER_PTE ); -/* __real_pte use pte_val() any idea why ? FIXME!! */ rpte.hidx &= ~(0xfUL << (subpg_index <<2 )); *hidxp = rpte.hidx | (slot << (subpg_index <<2 )); new_pte =mark_subptegroup_valid (new_pte, subpg_index);
Signature · 1 line
2.5.0
PATCH v6 22/35 powerpc/mm: Increase the width of #define
In reply to Aneesh Kumar K.VNo real change, only style changes
--- arch/powerpc/include/asm/book3s/64/hash.h | 26 +++++++++++++------------- 1 file changed, 13 insertions(+), 13 deletions(-)
arch/powerpc/include/asm/book3s/64/hash.h +13 -13
@@ -69,23 +69,23 @@* We could create separate kernel read-only if we used the 3 PP bits * combinations that newer processors provide but we currently don't. */ -#define _PAGE_PRESENT 0x0001/* software: pte contains a translation */ -#define _PAGE_USER 0x0002/* matches one of the PP bits */ +#define _PAGE_PRESENT 0x00001/* software: pte contains a translation */ +#define _PAGE_USER 0x00002/* matches one of the PP bits */ #define _PAGE_BIT_SWAP_TYPE 2 -#define _PAGE_EXEC 0x0004/* No execute on POWER4 and newer (we invert) */ -#define _PAGE_GUARDED 0x0008 +#define _PAGE_EXEC 0x00004/* No execute on POWER4 and newer (we invert) */ +#define _PAGE_GUARDED 0x00008/* We can derive Memory coherence from _PAGE_NO_CACHE */ #define _PAGE_COHERENT 0x0 -#define _PAGE_NO_CACHE 0x0020/* I: cache inhibit */ -#define _PAGE_WRITETHRU 0x0040/* W: cache write-through */ -#define _PAGE_DIRTY 0x0080/* C: page changed */ -#define _PAGE_ACCESSED 0x0100/* R: page referenced */ -#define _PAGE_RW 0x0200/* software: user write access allowed */ -#define _PAGE_HASHPTE 0x0400/* software: pte has an associated HPTE */ -#define _PAGE_BUSY 0x0800/* software: PTE & hash are busy */ -#define _PAGE_F_GIX 0x7000/* full page: hidx bits */ +#define _PAGE_NO_CACHE 0x00020/* I: cache inhibit */ +#define _PAGE_WRITETHRU 0x00040/* W: cache write-through */ +#define _PAGE_DIRTY 0x00080/* C: page changed */ +#define _PAGE_ACCESSED 0x00100/* R: page referenced */ +#define _PAGE_RW 0x00200/* software: user write access allowed */ +#define _PAGE_HASHPTE 0x00400/* software: pte has an associated HPTE */ +#define _PAGE_BUSY 0x00800/* software: PTE & hash are busy */ +#define _PAGE_F_GIX 0x07000/* full page: hidx bits */ #define _PAGE_F_GIX_SHIFT 12 -#define _PAGE_F_SECOND 0x8000/* Whether to use secondary hash or not */ +#define _PAGE_F_SECOND 0x08000/* Whether to use secondary hash or not */ #define _PAGE_SPECIAL 0x10000/* software: special page */ /* No separate kernel read-only */
Signature · 1 line
2.5.0
PATCH v6 23/35 powerpc/mm: Convert __hash_page_64K to C
In reply to Aneesh Kumar K.VConvert from asm to C
--- arch/powerpc/include/asm/book3s/64/hash-64k.h | 3 +- arch/powerpc/mm/hash64_64k.c | 130 ++++++++++++ arch/powerpc/mm/hash_low_64.S | 290 +------------------------- arch/powerpc/mm/hash_utils_64.c | 19 +- 4 files changed, 134 insertions(+), 308 deletions(-)
arch/powerpc/include/asm/book3s/64/hash-64k.h +2 -1
@@ -40,7 +40,8 @@#define _PAGE_COMBO_VALID (_PAGE_F_GIX | _PAGE_F_SECOND)/* PTE flags to conserve for HPTE identification */ -#define _PAGE_HPTEFLAGS (_PAGE_BUSY | _PAGE_HASHPTE | _PAGE_COMBO) +#define _PAGE_HPTEFLAGS (_PAGE_BUSY | _PAGE_F_SECOND | \ + _PAGE_F_GIX | _PAGE_HASHPTE | _PAGE_COMBO)/* Shift to put page number into pte. *
arch/powerpc/mm/hash64_64k.c +130 -0
133 lines not on this page · raw message
arch/powerpc/mm/hash_low_64.S +1 -289
293 lines not on this page · raw message
arch/powerpc/mm/hash_utils_64.c +1 -18
@@ -633,28 +633,11 @@ extern u32 htab_call_hpte_insert1[];extern u32 htab_call_hpte_insert2[];extern u32 htab_call_hpte_remove[];extern u32 htab_call_hpte_updatepp[]; -extern u32 ht64_call_hpte_insert1[]; -extern u32 ht64_call_hpte_insert2[]; -extern u32 ht64_call_hpte_remove[]; -extern u32 ht64_call_hpte_updatepp[];static void __init htab_finish_init (void ) { -#ifdef CONFIG_PPC_64K_PAGES -patch_branch (ht64_call_hpte_insert1, -ppc_function_entry (ppc_md.hpte_insert), -BRANCH_SET_LINK ); -patch_branch (ht64_call_hpte_insert2, -ppc_function_entry (ppc_md.hpte_insert), -BRANCH_SET_LINK ); -patch_branch (ht64_call_hpte_remove, -ppc_function_entry (ppc_md.hpte_remove), -BRANCH_SET_LINK ); -patch_branch (ht64_call_hpte_updatepp, -ppc_function_entry (ppc_md.hpte_updatepp), -BRANCH_SET_LINK ); -#else /* !CONFIG_PPC_64K_PAGES */ +#ifdef CONFIG_PPC_4K_PAGES patch_branch (htab_call_hpte_insert1,ppc_function_entry (ppc_md.hpte_insert),BRANCH_SET_LINK );
Signature · 1 line
2.5.0
PATCH v6 24/35 powerpc/mm: Convert 4k insert from asm to C
In reply to Aneesh Kumar K.VThis is similar to 64K insert. May be we want to consolidate
--- arch/powerpc/mm/Makefile | 6 +- arch/powerpc/mm/hash64_4k.c | 139 +++++++++++++++++ arch/powerpc/mm/hash_low_64.S | 331 ---------------------------------------- arch/powerpc/mm/hash_utils_64.c | 26 ---- 4 files changed, 142 insertions(+), 360 deletions(-) create mode 100644 arch/powerpc/mm/hash64_4k.c delete mode 100644 arch/powerpc/mm/hash_low_64.S
arch/powerpc/mm/Makefile +3 -3
@@ -14,11 +14,11 @@ obj-$(CONFIG_PPC_MMU_NOHASH) += mmu_context_nohash.o tlb_nohash.o \ obj-$(CONFIG_PPC_BOOK3E) += tlb_low_$(CONFIG_WORD_SIZE)e.o hash64-$(CONFIG_PPC_NATIVE) := hash_native_64.o obj-$(CONFIG_PPC_STD_MMU_64) += hash_utils_64.o slb_low.o slb.o $(hash64-y) -obj-$(CONFIG_PPC_STD_MMU_32) += ppc_mmu_32.o -obj-$(CONFIG_PPC_STD_MMU) += hash_low_$(CONFIG_WORD_SIZE).o \ - tlb_hash$(CONFIG_WORD_SIZE).o \ +obj-$(CONFIG_PPC_STD_MMU_32) += ppc_mmu_32.o hash_low_32.o +obj-$(CONFIG_PPC_STD_MMU) += tlb_hash$(CONFIG_WORD_SIZE).o \ mmu_context_hash$(CONFIG_WORD_SIZE).o ifeq ($(CONFIG_PPC_STD_MMU_64),y) +obj-$(CONFIG_PPC_4K_PAGES) += hash64_4k.o obj-$(CONFIG_PPC_64K_PAGES) += hash64_64k.o endif obj-$(CONFIG_PPC_ICSWX) += icswx.o
arch/powerpc/mm/hash64_4k.c new file +139 -0
139 lines not on this page · raw message
arch/powerpc/mm/hash_low_64.S deleted +0 -331
331 lines not on this page · raw message
arch/powerpc/mm/hash_utils_64.c +0 -26
@@ -629,31 +629,6 @@ int remove_section_mapping(unsigned long start, unsigned long end) }#endif /* CONFIG_MEMORY_HOTPLUG */ -extern u32 htab_call_hpte_insert1[]; -extern u32 htab_call_hpte_insert2[]; -extern u32 htab_call_hpte_remove[]; -extern u32 htab_call_hpte_updatepp[]; - -static void __init htab_finish_init (void ) -{ - -#ifdef CONFIG_PPC_4K_PAGES -patch_branch (htab_call_hpte_insert1, -ppc_function_entry (ppc_md.hpte_insert), -BRANCH_SET_LINK ); -patch_branch (htab_call_hpte_insert2, -ppc_function_entry (ppc_md.hpte_insert), -BRANCH_SET_LINK ); -patch_branch (htab_call_hpte_remove, -ppc_function_entry (ppc_md.hpte_remove), -BRANCH_SET_LINK ); -patch_branch (htab_call_hpte_updatepp, -ppc_function_entry (ppc_md.hpte_updatepp), -BRANCH_SET_LINK ); -#endif - -} -static void __init htab_initialize (void ) {unsigned long table;
@@ -800,7 +775,6 @@ static void __init htab_initialize(void) mmu_linear_psize, mmu_kernel_ssize)); } -htab_finish_init ();DBG (" <- htab_initialize()\n" ); }
Signature · 1 line
2.5.0
PATCH v6 25/35 powerpc/mm: Add helper for converting pte bit to hpte bits
In reply to Aneesh Kumar K.VInstead of open coding it in multiple code paths, export the helper and add more documentation. Also make sure we don't make assumption regarding pte bit position
--- arch/powerpc/include/asm/book3s/64/hash.h | 1 + arch/powerpc/mm/hash64_4k.c | 13 +----------- arch/powerpc/mm/hash64_64k.c | 35 +++---------------------------- arch/powerpc/mm/hash_utils_64.c | 22 ++++++++++++------- arch/powerpc/mm/hugepage-hash64.c | 13 +----------- arch/powerpc/mm/hugetlbpage-hash64.c | 4 +--- 6 files changed, 21 insertions(+), 67 deletions(-)
arch/powerpc/include/asm/book3s/64/hash.h +1 -0
@@ -236,6 +236,7 @@ extern unsigned long pmd_hugepage_update(struct mm_struct *mm,pmd_t *pmdp, unsigned long clr, unsigned long set); +extern unsigned long htab_convert_pte_flags (unsigned long pteflags);/* Atomic PTE updates */ static inline unsigned long pte_update(struct mm_struct *mm, unsigned long addr,
arch/powerpc/mm/hash64_4k.c +1 -12
@@ -53,18 +53,7 @@ int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid,* PP bits. _PAGE_USER is already PP bit 0x2, so we only * need to add in 0x1 if it's a read-only user page */ - rflags = new_pte & _PAGE_USER; -if ((new_pte & _PAGE_USER) && !((new_pte & _PAGE_RW) && - (new_pte & _PAGE_DIRTY))) - rflags |=0x1 ; -/* -* _PAGE_EXEC -> HW_NO_EXEC since it's inverted -*/ - rflags |= ((new_pte & _PAGE_EXEC) ?0 :HPTE_R_N ); -/* -* Always add C and Memory coherence bit -*/ - rflags |=HPTE_R_C |HPTE_R_M ; + rflags =htab_convert_pte_flags (new_pte);/* * Add in WIMG bits */
arch/powerpc/mm/hash64_64k.c +3 -32
@@ -85,22 +85,7 @@ int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid,* Handle the subpage protection bits */ subpg_pte = new_pte & ~subpg_prot; -/* -* PP bits. _PAGE_USER is already PP bit 0x2, so we only -* need to add in 0x1 if it's a read-only user page -*/ - rflags = subpg_pte & _PAGE_USER; -if ((subpg_pte & _PAGE_USER) && !((subpg_pte & _PAGE_RW) && - (subpg_pte & _PAGE_DIRTY))) - rflags |=0x1 ; -/* -* _PAGE_EXEC -> HW_NO_EXEC since it's inverted -*/ - rflags |= ((subpg_pte & _PAGE_EXEC) ?0 :HPTE_R_N ); -/* -* Always add C and Memory coherence bit -*/ - rflags |=HPTE_R_C |HPTE_R_M ; + rflags =htab_convert_pte_flags (subpg_pte);/* * Add in WIMG bits */
@@ -271,22 +256,8 @@ int __hash_page_64K(unsigned long ea, unsigned long access, new_pte |= _PAGE_DIRTY; }while (old_pte !=__cmpxchg_u64 ((unsigned long *)ptep, old_pte, new_pte)); -/* -* PP bits. _PAGE_USER is already PP bit 0x2, so we only -* need to add in 0x1 if it's a read-only user page -*/ - rflags = new_pte & _PAGE_USER; -if ((new_pte & _PAGE_USER) && !((new_pte & _PAGE_RW) && - (new_pte & _PAGE_DIRTY))) - rflags |=0x1 ; -/* -* _PAGE_EXEC -> HW_NO_EXEC since it's inverted -*/ - rflags |= ((new_pte & _PAGE_EXEC) ?0 :HPTE_R_N ); -/* -* Always add C and Memory coherence bit -*/ - rflags |=HPTE_R_C |HPTE_R_M ; + + rflags =htab_convert_pte_flags (new_pte);/* * Add in WIMG bits */
arch/powerpc/mm/hash_utils_64.c +14 -8
@@ -159,20 +159,26 @@ static struct mmu_psize_def mmu_psize_defaults_gp[] = { }, }; -static unsigned long htab_convert_pte_flags (unsigned long pteflags) +unsigned long htab_convert_pte_flags (unsigned long pteflags) { -unsigned long rflags = pteflags &0x1fa ; +unsigned long rflags =0 ;/* _PAGE_EXEC -> NOEXEC */ if ((pteflags & _PAGE_EXEC) ==0 ) rflags |=HPTE_R_N ; - -/* PP bits. PAGE_USER is already PP bit 0x2, so we only -* need to add in 0x1 if it's a read-only user page +/* +* PP bits: +* Linux use slb key 0 for kernel and 1 for user. +* kernel areas are mapped by PP bits 00 +* and and there is no kernel RO (_PAGE_KERNEL_RO). +* User area mapped by 0x2 and read only use by +* 0x3. */ -if ((pteflags & _PAGE_USER) && !((pteflags & _PAGE_RW) && - (pteflags & _PAGE_DIRTY))) - rflags |=1 ; +if (pteflags & _PAGE_USER) { + rflags |=0x2 ; +if (!((pteflags & _PAGE_RW) && (pteflags & _PAGE_DIRTY))) + rflags |=0x1 ; + }/* * Always add "C" bit for perf. Memory coherence is always enabled */
arch/powerpc/mm/hugepage-hash64.c +1 -12
@@ -54,18 +54,7 @@ int __hash_page_thp(unsigned long ea, unsigned long access, unsigned long vsid, new_pmd |= _PAGE_DIRTY; }while (old_pmd !=__cmpxchg_u64 ((unsigned long *)pmdp, old_pmd, new_pmd)); -/* -* PP bits. _PAGE_USER is already PP bit 0x2, so we only -* need to add in 0x1 if it's a read-only user page -*/ - rflags = new_pmd & _PAGE_USER; -if ((new_pmd & _PAGE_USER) && !((new_pmd & _PAGE_RW) && - (new_pmd & _PAGE_DIRTY))) - rflags |=0x1 ; -/* -* _PAGE_EXEC -> HW_NO_EXEC since it's inverted -*/ - rflags |= ((new_pmd & _PAGE_EXEC) ?0 :HPTE_R_N ); + rflags =htab_convert_pte_flags (new_pmd);#if 0 if (!cpu_has_feature (CPU_FTR_COHERENT_ICACHE )) {
arch/powerpc/mm/hugetlbpage-hash64.c +1 -3
@@ -59,10 +59,8 @@ int __hash_page_huge(unsigned long ea, unsigned long access, unsigned long vsid, new_pte |= _PAGE_DIRTY; }while (old_pte !=__cmpxchg_u64 ((unsigned long *)ptep, old_pte, new_pte)); + rflags =htab_convert_pte_flags (new_pte); - rflags =0x2 | (!(new_pte & _PAGE_RW)); -/* _PAGE_EXEC -> HW_NO_EXEC since it's inverted */ - rflags |= ((new_pte & _PAGE_EXEC) ?0 :HPTE_R_N ); sz = ((1UL ) << shift);if (!cpu_has_feature (CPU_FTR_COHERENT_ICACHE ))/* No CPU has hugepages but lacks no execute, so we
Signature · 1 line
2.5.0
PATCH v6 26/35 powerpc/mm: Move WIMG update to helper.
In reply to Aneesh Kumar K.VOnly difference here is, we apply the WIMG mapping early, so rflags passed to updatepp will also be changed.
--- arch/powerpc/mm/hash64_4k.c | 5 ----- arch/powerpc/mm/hash64_64k.c | 10 ---------- arch/powerpc/mm/hash_utils_64.c | 13 ++++++++++++- arch/powerpc/mm/hugepage-hash64.c | 7 ------- arch/powerpc/mm/hugetlbpage-hash64.c | 8 -------- 5 files changed, 12 insertions(+), 31 deletions(-)
arch/powerpc/mm/hash64_4k.c +0 -5
@@ -54,11 +54,6 @@ int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid,* need to add in 0x1 if it's a read-only user page */ rflags =htab_convert_pte_flags (new_pte); -/* -* Add in WIMG bits -*/ - rflags |= (new_pte & (_PAGE_WRITETHRU | _PAGE_NO_CACHE | - _PAGE_COHERENT | _PAGE_GUARDED));if (!cpu_has_feature (CPU_FTR_NOEXECUTE ) && !cpu_has_feature (CPU_FTR_COHERENT_ICACHE ))
arch/powerpc/mm/hash64_64k.c +0 -10
@@ -86,11 +86,6 @@ int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid,*/ subpg_pte = new_pte & ~subpg_prot; rflags =htab_convert_pte_flags (subpg_pte); -/* -* Add in WIMG bits -*/ - rflags |= (subpg_pte & (_PAGE_WRITETHRU | _PAGE_NO_CACHE | - _PAGE_COHERENT | _PAGE_GUARDED));if (!cpu_has_feature (CPU_FTR_NOEXECUTE ) && !cpu_has_feature (CPU_FTR_COHERENT_ICACHE )) {
@@ -258,11 +253,6 @@ int __hash_page_64K(unsigned long ea, unsigned long access, old_pte, new_pte)); rflags =htab_convert_pte_flags (new_pte); -/* -* Add in WIMG bits -*/ - rflags |= (new_pte & (_PAGE_WRITETHRU | _PAGE_NO_CACHE | - _PAGE_COHERENT | _PAGE_GUARDED));if (!cpu_has_feature (CPU_FTR_NOEXECUTE ) && !cpu_has_feature (CPU_FTR_COHERENT_ICACHE ))
arch/powerpc/mm/hash_utils_64.c +12 -1
@@ -182,7 +182,18 @@ unsigned long htab_convert_pte_flags(unsigned long pteflags)/* * Always add "C" bit for perf. Memory coherence is always enabled */ -return rflags |HPTE_R_C |HPTE_R_M ; + rflags |=HPTE_R_C |HPTE_R_M ; +/* +* Add in WIG bits +*/ +if (pteflags & _PAGE_WRITETHRU) + rflags |=HPTE_R_W ; +if (pteflags & _PAGE_NO_CACHE) + rflags |=HPTE_R_I ; +if (pteflags & _PAGE_GUARDED) + rflags |=HPTE_R_G ; + +return rflags; } int htab_bolt_mapping(unsigned long vstart, unsigned long vend,
arch/powerpc/mm/hugepage-hash64.c +0 -7
@@ -120,13 +120,6 @@ int __hash_page_thp(unsigned long ea, unsigned long access, unsigned long vsid, pa =pmd_pfn (__pmd (old_pmd)) <<PAGE_SHIFT ; new_pmd |= _PAGE_HASHPTE; -/* Add in WIMG bits */ - rflags |= (new_pmd & (_PAGE_WRITETHRU | _PAGE_NO_CACHE | - _PAGE_GUARDED)); -/* -* enable the memory coherence always -*/ - rflags |=HPTE_R_M ; repeat: hpte_group = ((hash & htab_hash_mask) *HPTES_PER_GROUP ) & ~0x7UL ;
arch/powerpc/mm/hugetlbpage-hash64.c +0 -8
@@ -91,14 +91,6 @@ int __hash_page_huge(unsigned long ea, unsigned long access, unsigned long vsid,/* clear HPTE slot informations in new PTE */ new_pte = (new_pte & ~_PAGE_HPTEFLAGS) | _PAGE_HASHPTE; -/* Add in WIMG bits */ - rflags |= (new_pte & (_PAGE_WRITETHRU | _PAGE_NO_CACHE | - _PAGE_COHERENT | _PAGE_GUARDED)); -/* -* enable the memory coherence always -*/ - rflags |=HPTE_R_M ; - slot =hpte_insert_repeating (hash, vpn, pa, rflags,0 , mmu_psize, ssize);
Signature · 1 line
2.5.0
PATCH v6 27/35 powerpc/mm: Move hugetlb related headers
In reply to Aneesh Kumar K.VW.r.t hugetlb, we support two format for pmd. With book3s_64 and 64K linux page size, we can have pte at the pmd level. Hence we don't need to support hugepd there. For everything else hugepd is supported and pmd_huge is (0).
--- arch/powerpc/include/asm/book3s/64/hash-4k.h | 31 ++++++++++++ arch/powerpc/include/asm/book3s/64/hash-64k.h | 51 +++++++++++++++++++ arch/powerpc/include/asm/nohash/pgtable.h | 25 ++++++++++ arch/powerpc/include/asm/page.h | 42 ++-------------- arch/powerpc/mm/hugetlbpage-hash64.c | 18 +++++++ arch/powerpc/mm/hugetlbpage.c | 72 --------------------------- 6 files changed, 129 insertions(+), 110 deletions(-)
arch/powerpc/include/asm/book3s/64/hash-4k.h +31 -0
@@ -93,6 +93,37 @@ extern struct page *pgd_page(pgd_t pgd);#define remap_4k_pfn (vma, addr, pfn, prot) \ remap_pfn_range((vma), (addr), (pfn), PAGE_SIZE, (prot)) +#ifdef CONFIG_HUGETLB_PAGE +/* +* For 4k page size, we support explicit hugepage via hugepd +*/ +static inline int pmd_huge (pmd_t pmd) +{ +return 0 ; +} + +static inline int pud_huge (pud_t pud) +{ +return 0 ; +} + +static inline int pgd_huge (pgd_t pgd) +{ +return 0 ; +} +#define pgd_huge pgd_huge + +static inline int hugepd_ok (hugepd_t hpd) +{ +/* +* hugepd pointer, bottom two bits == 00 and next 4 bits +* indicate size of table +*/ +return (((hpd.pd &0x3 ) ==0x0 ) && ((hpd.pd &HUGEPD_SHIFT_MASK ) !=0 )); +} +#define is_hugepd (hpd) (hugepd_ok(hpd)) +#endif +#endif /* !__ASSEMBLY__ */ #endif /* _ASM_POWERPC_BOOK3S_64_HASH_4K_H */
arch/powerpc/include/asm/book3s/64/hash-64k.h +51 -0
57 lines not on this page · raw message
arch/powerpc/include/asm/nohash/pgtable.h +25 -0
@@ -223,5 +223,30 @@ extern pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn, unsigned long size, pgprot_t vma_prot);#define __HAVE_PHYS_MEM_ACCESS_PROT +#ifdef CONFIG_HUGETLB_PAGE +static inline int hugepd_ok (hugepd_t hpd) +{ +return (hpd.pd >0 ); +} + +static inline int pmd_huge (pmd_t pmd) +{ +return 0 ; +} + +static inline int pud_huge (pud_t pud) +{ +return 0 ; +} + +static inline int pgd_huge (pgd_t pgd) +{ +return 0 ; +} +#define pgd_huge pgd_huge + +#define is_hugepd (hpd) (hugepd_ok(hpd)) +#endif +#endif /* __ASSEMBLY__ */ #endif
arch/powerpc/include/asm/page.h +4 -38
49 lines not on this page · raw message
arch/powerpc/mm/hugetlbpage-hash64.c +18 -0
@@ -114,3 +114,21 @@ int __hash_page_huge(unsigned long ea, unsigned long access, unsigned long vsid, *ptep =__pte (new_pte & ~_PAGE_BUSY);return 0 ; } + +#if defined(CONFIG_PPC_64K_PAGES ) && defined(CONFIG_DEBUG_VM ) +/* +* This enables us to catch the wrong page directory format +* Moved here so that we can use WARN() in the call. +*/ +int hugepd_ok (hugepd_t hpd) +{ +bool is_hugepd; + +/* +* We should not find this format in page directory, warn otherwise. +*/ + is_hugepd = (((hpd.pd &0x3 ) ==0x0 ) && ((hpd.pd &HUGEPD_SHIFT_MASK ) !=0 )); +WARN (is_hugepd,"Found wrong page directory format\n" ); +return 0 ; +} +#endif
arch/powerpc/mm/hugetlbpage.c +0 -72
78 lines not on this page · raw message
Signature · 1 line
2.5.0
PATCH v6 28/35 powerpc/mm: Move THP headers around
In reply to Aneesh Kumar K.VWe support THP only with book3s_64 and 64K page size. Move THP details to hash64-64k.h to clarify the same.
--- arch/powerpc/include/asm/book3s/64/hash-64k.h | 126 +++++++++++++ arch/powerpc/include/asm/book3s/64/hash.h | 223 +++++------------------ arch/powerpc/include/asm/nohash/64/pgtable.h | 253 +------------------------- arch/powerpc/mm/hash_native_64.c | 10 + arch/powerpc/mm/pgtable_64.c | 2 +- arch/powerpc/platforms/pseries/lpar.c | 10 + 6 files changed, 201 insertions(+), 423 deletions(-)
arch/powerpc/include/asm/book3s/64/hash-64k.h +126 -0
@@ -170,6 +170,132 @@ static inline int hugepd_ok(hugepd_t hpd)#endif /* CONFIG_HUGETLB_PAGE */ +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +extern unsigned long pmd_hugepage_update (struct mm_struct *mm, +unsigned long addr, +pmd_t *pmdp, +unsigned long clr, +unsigned long set); +static inline char *get_hpte_slot_array (pmd_t *pmdp) +{ +/* +* The hpte hindex is stored in the pgtable whose address is in the +* second half of the PMD +* +* Order this load with the test for pmd_trans_huge in the caller +*/ +smp_rmb (); +return *(char **)(pmdp +PTRS_PER_PMD ); + + +} +/* +* The linux hugepage PMD now include the pmd entries followed by the address +* to the stashed pgtable_t. The stashed pgtable_t contains the hpte bits. +* [ 1 bit secondary | 3 bit hidx | 1 bit valid | 000]. We use one byte per +* each HPTE entry. With 16MB hugepage and 64K HPTE we need 256 entries and +* with 4K HPTE we need 4096 entries. Both will fit in a 4K pgtable_t. +* +* The last three bits are intentionally left to zero. This memory location +* are also used as normal page PTE pointers. So if we have any pointers +* left around while we collapse a hugepage, we need to make sure +* _PAGE_PRESENT bit of that is zero when we look at them +*/ +static inline unsigned int hpte_valid (unsigned char *hpte_slot_array,int index) +{ +return (hpte_slot_array[index] >>3 ) &0x1 ; +} + +static inline unsigned int hpte_hash_index (unsigned char *hpte_slot_array, +int index) +{ +return hpte_slot_array[index] >>4 ; +} + +static inline void mark_hpte_slot_valid (unsigned char *hpte_slot_array, +unsigned int index,unsigned int hidx) +{ + hpte_slot_array[index] = hidx <<4 |0x1 <<3 ; +} + +/* +* +* For core kernel code by design pmd_trans_huge is never run on any hugetlbfs +* page. The hugetlbfs page table walking and mangling paths are totally +* separated form the core VM paths and they're differentiated by +* VM_HUGETLB being set on vm_flags well before any pmd_trans_huge could run. +* +* pmd_trans_huge() is defined as false at build time if +* CONFIG_TRANSPARENT_HUGEPAGE=n to optimize away code blocks at build +* time in such case. +* +* For ppc64 we need to differntiate from explicit hugepages from THP, because +* for THP we also track the subpage details at the pmd level. We don't do +* that for explicit huge pages. +* +*/ +static inline int pmd_trans_huge (pmd_t pmd) +{ +/* +* leaf pte for huge page, bottom two bits != 00 +*/ +return (pmd_val (pmd) &0x3 ) && (pmd_val (pmd) & _PAGE_THP_HUGE); +} + +static inline int pmd_trans_splitting (pmd_t pmd) +{ +if (pmd_trans_huge (pmd)) +return pmd_val (pmd) & _PAGE_SPLITTING; +return 0 ; +} + +static inline int pmd_large (pmd_t pmd) +{ +/* +* leaf pte for huge page, bottom two bits != 00 +*/ +return ((pmd_val (pmd) &0x3 ) !=0x0 ); +} + +static inline pmd_t pmd_mknotpresent (pmd_t pmd) +{ +return __pmd (pmd_val (pmd) & ~_PAGE_PRESENT); +} + +static inline pmd_t pmd_mksplitting (pmd_t pmd) +{ +return __pmd (pmd_val (pmd) | _PAGE_SPLITTING); +} + +#define __HAVE_ARCH_PMD_SAME +static inline int pmd_same (pmd_t pmd_a,pmd_t pmd_b) +{ +return (((pmd_val (pmd_a) ^pmd_val (pmd_b)) & ~_PAGE_HPTEFLAGS) ==0 ); +} + +static inline int __pmdp_test_and_clear_young (struct mm_struct *mm, +unsigned long addr,pmd_t *pmdp) +{ +unsigned long old; + +if ((pmd_val (*pmdp) & (_PAGE_ACCESSED | _PAGE_HASHPTE)) ==0 ) +return 0 ; + old =pmd_hugepage_update (mm, addr, pmdp, _PAGE_ACCESSED,0 ); +return ((old & _PAGE_ACCESSED) !=0 ); +} + +#define __HAVE_ARCH_PMDP_SET_WRPROTECT +static inline void pmdp_set_wrprotect (struct mm_struct *mm,unsigned long addr, +pmd_t *pmdp) +{ + +if ((pmd_val (*pmdp) & _PAGE_RW) ==0 ) +return ; + +pmd_hugepage_update (mm, addr, pmdp, _PAGE_RW,0 ); +} + +#endif /* CONFIG_TRANSPARENT_HUGEPAGE */ #endif /* __ASSEMBLY__ */ #endif /* _ASM_POWERPC_BOOK3S_64_HASH_64K_H */
arch/powerpc/include/asm/book3s/64/hash.h +49 -174
253 lines not on this page · raw message
arch/powerpc/include/asm/nohash/64/pgtable.h +5 -248
263 lines not on this page · raw message
arch/powerpc/mm/hash_native_64.c +10 -0
@@ -429,6 +429,7 @@ static void native_hpte_invalidate(unsigned long slot, unsigned long vpn,local_irq_restore (flags); } +#ifdef CONFIG_TRANSPARENT_HUGEPAGE static void native_hugepage_invalidate(unsigned long vsid, unsigned long addr, unsigned char *hpte_slot_array,
@@ -482,6 +483,15 @@ static void native_hugepage_invalidate(unsigned long vsid, }local_irq_restore (flags); } +#else +static void native_hugepage_invalidate (unsigned long vsid, +unsigned long addr, +unsigned char *hpte_slot_array, +int psize,int ssize,int local) +{ +WARN (1 ,"%s called without THP support\n" , __func__); +} +#endif static inline int __hpte_actual_psize (unsigned int lp,int psize) {
arch/powerpc/mm/pgtable_64.c +1 -1
@@ -359,7 +359,7 @@ struct page *pud_page(pud_t pud)struct page *pmd_page (pmd_t pmd) {if (pmd_trans_huge (pmd) ||pmd_huge (pmd)) -return pfn_to_page (pmd_pfn (pmd)); +return pte_page (pmd_pte (pmd));return virt_to_page (pmd_page_vaddr (pmd)); }
arch/powerpc/platforms/pseries/lpar.c +10 -0
@@ -396,6 +396,7 @@ static void pSeries_lpar_hpte_invalidate(unsigned long slot, unsigned long vpn,BUG_ON (lpar_rc !=H_SUCCESS ); } +#ifdef CONFIG_TRANSPARENT_HUGEPAGE /* * Limit iterations holding pSeries_lpar_tlbie_lock to 3. We also need * to make sure that we avoid bouncing the hypervisor tlbie lock.
@@ -494,6 +495,15 @@ static void pSeries_lpar_hugepage_invalidate(unsigned long vsid,__pSeries_lpar_hugepage_invalidate (slot_array, vpn_array, index, psize, ssize); } +#else +static void pSeries_lpar_hugepage_invalidate (unsigned long vsid, +unsigned long addr, +unsigned char *hpte_slot_array, +int psize,int ssize,int local) +{ +WARN (1 ,"%s called without THP support\n" , __func__); +} +#endif static void pSeries_lpar_hpte_removebolted (unsigned long ea,int psize,int ssize)
Signature · 1 line
2.5.0
PATCH v6 29/35 powerpc/mm: Add a _PAGE_PTE bit
In reply to Aneesh Kumar K.VFor a pte entry we will have _PAGE_PTE set. Our pte page address have a minimum alignment requirement of HUGEPD_SHIFT_MASK + 1. We use the lower 7 bits to indicate hugepd. ie.
For pmd and pgd we can find:
1) _PAGE_PTE set pte -> indicate PTE
2) bits [2..6] non zero -> indicate hugepd.
They also encode the size. We skip bit 1 (_PAGE_PRESENT).
3) othewise pointer to next table.
--- arch/powerpc/include/asm/book3s/64/hash-4k.h | 9 ++++++--- arch/powerpc/include/asm/book3s/64/hash-64k.h | 23 +++++++++-------------- arch/powerpc/include/asm/book3s/64/hash.h | 13 +++++++------ arch/powerpc/include/asm/book3s/64/pgtable.h | 3 +-- arch/powerpc/include/asm/pte-common.h | 5 +++++ arch/powerpc/mm/hugetlbpage.c | 4 ++-- arch/powerpc/mm/pgtable.c | 4 ++++ arch/powerpc/mm/pgtable_64.c | 7 +------ 8 files changed, 35 insertions(+), 33 deletions(-)
arch/powerpc/include/asm/book3s/64/hash-4k.h +6 -3
@@ -116,10 +116,13 @@ static inline int pgd_huge(pgd_t pgd)static inline int hugepd_ok (hugepd_t hpd) {/* -* hugepd pointer, bottom two bits == 00 and next 4 bits -* indicate size of table +* if it is not a pte and have hugepd shift mask +* set, then it is a hugepd directory pointer */ -return (((hpd.pd &0x3 ) ==0x0 ) && ((hpd.pd &HUGEPD_SHIFT_MASK ) !=0 )); +if (!(hpd.pd & _PAGE_PTE) && + ((hpd.pd &HUGEPD_SHIFT_MASK ) !=0 )) +return true; +return false; }#define is_hugepd (hpd) (hugepd_ok(hpd))#endif
arch/powerpc/include/asm/book3s/64/hash-64k.h +9 -14
@@ -130,25 +130,25 @@ extern bool __rpte_sub_valid(real_pte_t rpte, unsigned long index);static inline int pmd_huge (pmd_t pmd) {/* -* leaf pte for huge page, bottom two bits != 00 +* leaf pte for huge page */ -return ((pmd_val (pmd) &0x3 ) !=0x0 ); +return !!(pmd_val (pmd) & _PAGE_PTE); }static inline int pud_huge (pud_t pud) {/* -* leaf pte for huge page, bottom two bits != 00 +* leaf pte for huge page */ -return ((pud_val (pud) &0x3 ) !=0x0 ); +return !!(pud_val (pud) & _PAGE_PTE); }static inline int pgd_huge (pgd_t pgd) {/* -* leaf pte for huge page, bottom two bits != 00 +* leaf pte for huge page */ -return ((pgd_val (pgd) &0x3 ) !=0x0 ); +return !!(pgd_val (pgd) & _PAGE_PTE); }#define pgd_huge pgd_huge
@@ -236,10 +236,8 @@ static inline void mark_hpte_slot_valid(unsigned char *hpte_slot_array,*/ static inline int pmd_trans_huge (pmd_t pmd) { -/* -* leaf pte for huge page, bottom two bits != 00 -*/ -return (pmd_val (pmd) &0x3 ) && (pmd_val (pmd) & _PAGE_THP_HUGE); +return !!((pmd_val (pmd) & (_PAGE_PTE | _PAGE_THP_HUGE)) == + (_PAGE_PTE | _PAGE_THP_HUGE)); }static inline int pmd_trans_splitting (pmd_t pmd)
@@ -251,10 +249,7 @@ static inline int pmd_trans_splitting(pmd_t pmd)static inline int pmd_large (pmd_t pmd) { -/* -* leaf pte for huge page, bottom two bits != 00 -*/ -return ((pmd_val (pmd) &0x3 ) !=0x0 ); +return !!(pmd_val (pmd) & _PAGE_PTE); }static inline pmd_t pmd_mknotpresent (pmd_t pmd)
arch/powerpc/include/asm/book3s/64/hash.h +7 -6
@@ -14,11 +14,12 @@* We could create separate kernel read-only if we used the 3 PP bits * combinations that newer processors provide but we currently don't. */ -#define _PAGE_PRESENT 0x00001/* software: pte contains a translation */ -#define _PAGE_USER 0x00002/* matches one of the PP bits */ +#define _PAGE_PTE 0x00001 +#define _PAGE_PRESENT 0x00002/* software: pte contains a translation */ #define _PAGE_BIT_SWAP_TYPE 2 -#define _PAGE_EXEC 0x00004/* No execute on POWER4 and newer (we invert) */ -#define _PAGE_GUARDED 0x00008 +#define _PAGE_USER 0x00004/* matches one of the PP bits */ +#define _PAGE_EXEC 0x00008/* No execute on POWER4 and newer (we invert) */ +#define _PAGE_GUARDED 0x00010/* We can derive Memory coherence from _PAGE_NO_CACHE */ #define _PAGE_COHERENT 0x0#define _PAGE_NO_CACHE 0x00020/* I: cache inhibit */
@@ -49,7 +50,7 @@*/ #define _HPAGE_CHG_MASK (PTE_RPN_MASK | _PAGE_HPTEFLAGS | \ _PAGE_DIRTY | _PAGE_ACCESSED | _PAGE_SPLITTING | \ - _PAGE_THP_HUGE) + _PAGE_THP_HUGE | _PAGE_PTE)#ifdef CONFIG_PPC_64K_PAGES #include <asm/book3s/64/hash-64k.h>
@@ -135,7 +136,7 @@* pgprot changes */ #define _PAGE_CHG_MASK (PTE_RPN_MASK | _PAGE_HPTEFLAGS | _PAGE_DIRTY | \ - _PAGE_ACCESSED | _PAGE_SPECIAL) + _PAGE_ACCESSED | _PAGE_SPECIAL | _PAGE_PTE)/* * Mask of bits returned by pte_pgprot() */
arch/powerpc/include/asm/book3s/64/pgtable.h +1 -2
@@ -213,8 +213,7 @@ static inline int pmd_protnone(pmd_t pmd)static inline pmd_t pmd_mkhuge (pmd_t pmd) { -/* Do nothing, mk_pmd() does this part. */ -return pmd; +return __pmd (pmd_val (pmd) | (_PAGE_PTE | _PAGE_THP_HUGE)); }#define __HAVE_ARCH_PMDP_SET_ACCESS_FLAGS
arch/powerpc/include/asm/pte-common.h +5 -0
@@ -40,6 +40,11 @@#else #define _PAGE_RW 0#endif + +#ifndef _PAGE_PTE +#define _PAGE_PTE 0 +#endif +#ifndef _PMD_PRESENT_MASK#define _PMD_PRESENT_MASK _PMD_PRESENT#endif
arch/powerpc/mm/hugetlbpage.c +2 -2
@@ -894,8 +894,8 @@ void flush_dcache_icache_hugepage(struct page *page)* We have 4 cases for pgds and pmds: * (1) invalid (all zeroes) * (2) pointer to next table, as normal; bottom 6 bits == 0 -* (3) leaf pte for huge page, bottom two bits != 00 -* (4) hugepd pointer, bottom two bits == 00, next 4 bits indicate size of table +* (3) leaf pte for huge page _PAGE_PTE set +* (4) hugepd pointer, _PAGE_PTE = 0 and bits [2..6] indicate size of table * * So long as we atomically load page table pointers we are safe against teardown, * we can follow the address down to the the page and take a ref on it.
arch/powerpc/mm/pgtable.c +4 -0
@@ -179,6 +179,10 @@ void set_pte_at(struct mm_struct *mm, unsigned long addr, pte_t *ptep,*/ VM_WARN_ON ((pte_val (*ptep) & (_PAGE_PRESENT | _PAGE_USER)) == (_PAGE_PRESENT | _PAGE_USER)); +/* +* Add the pte bit when tryint set a pte +*/ + pte =__pte (pte_val (pte) | _PAGE_PTE);/* Note: mm->context.id might not yet have been assigned as * this context might not have been activated yet when this
arch/powerpc/mm/pgtable_64.c +1 -6
@@ -765,13 +765,8 @@ static pmd_t pmd_set_protbits(pmd_t pmd, pgprot_t pgprot)pmd_t pfn_pmd (unsigned long pfn,pgprot_t pgprot) {unsigned long pmdv; -/* -* For a valid pte, we would have _PAGE_PRESENT always -* set. We use this to check THP page at pmd level. -* leaf pte for huge page, bottom two bits != 00 -*/ + pmdv = pfn <<PTE_RPN_SHIFT ; - pmdv |= _PAGE_THP_HUGE;return pmd_set_protbits (__pmd (pmdv), pgprot); }
Signature · 1 line
2.5.0
PATCH v6 30/35 powerpc/mm: Don't hardcode page table size
In reply to Aneesh Kumar K.Vpte and pmd table size are dependent on config items. Don't hard code the same. This make sure we use the right value when masking pmd entries and also while checking pmd_bad
--- arch/powerpc/include/asm/book3s/64/hash-64k.h | 30 ++++++++++++++++++------ arch/powerpc/include/asm/nohash/64/pgtable-64k.h | 21 +++++++++++++---- arch/powerpc/include/asm/pgalloc-64.h | 10 -------- arch/powerpc/mm/init_64.c | 4 ---- 4 files changed, 40 insertions(+), 25 deletions(-)
arch/powerpc/include/asm/book3s/64/hash-64k.h +23 -7
@@ -25,12 +25,6 @@#define PGDIR_SIZE (1UL << PGDIR_SHIFT)#define PGDIR_MASK (~(PGDIR_SIZE-1)) -/* Bits to mask out from a PMD to get to the PTE page */ -/* PMDs point to PTE table fragments which are 4K aligned. */ -#define PMD_MASKED_BITS 0xfff -/* Bits to mask out from a PGD/PUD to get to the PMD page */ -#define PUD_MASKED_BITS 0x1ff -#define _PAGE_COMBO 0x00020000/* this is a combo 4k page */ #define _PAGE_4K_PFN 0x00040000/* PFN is for a single 4k page */ /*
@@ -49,6 +43,24 @@* of addressable physical space, or 46 bits for the special 4k PFNs. */ #define PTE_RPN_SHIFT (30) +/* +* we support 16 fragments per PTE page of 64K size. +*/ +#define PTE_FRAG_NR 16 +/* +* We use a 2K PTE page fragment and another 2K for storing +* real_pte_t hash index +*/ +#define PTE_FRAG_SIZE_SHIFT 12 +#define PTE_FRAG_SIZE (1UL << PTE_FRAG_SIZE_SHIFT) + +/* +* Bits to mask out from a PMD to get to the PTE page +* PMDs point to PTE table fragments which are PTE_FRAG_SIZE aligned. +*/ +#define PMD_MASKED_BITS (PTE_FRAG_SIZE - 1) +/* Bits to mask out from a PGD/PUD to get to the PMD page */ +#define PUD_MASKED_BITS 0x1ff#ifndef __ASSEMBLY__
@@ -112,8 +124,12 @@ extern bool __rpte_sub_valid(real_pte_t rpte, unsigned long index); remap_pfn_range((vma), (addr), (pfn), PAGE_SIZE, \ __pgprot(pgprot_val((prot)) | _PAGE_4K_PFN))) -#define PTE_TABLE_SIZE (sizeof(real_pte_t) << PTE_INDEX_SIZE) +#define PTE_TABLE_SIZE PTE_FRAG_SIZE +#ifdef CONFIG_TRANSPARENT_HUGEPAGE +#define PMD_TABLE_SIZE ((sizeof(pmd_t) << PMD_INDEX_SIZE) + (sizeof(unsigned long) << PMD_INDEX_SIZE)) +#else #define PMD_TABLE_SIZE (sizeof(pmd_t) << PMD_INDEX_SIZE) +#endif #define PGD_TABLE_SIZE (sizeof(pgd_t) << PGD_INDEX_SIZE)#define pgd_pte (pgd) (pud_pte(((pud_t){ pgd })))
arch/powerpc/include/asm/nohash/64/pgtable-64k.h +17 -4
@@ -9,8 +9,19 @@#define PUD_INDEX_SIZE 0#define PGD_INDEX_SIZE 12 +/* +* we support 16 fragments per PTE page of 64K size. +*/ +#define PTE_FRAG_NR 16 +/* +* We use a 2K PTE page fragment and another 2K for storing +* real_pte_t hash index +*/ +#define PTE_FRAG_SIZE_SHIFT 12 +#define PTE_FRAG_SIZE (1UL << PTE_FRAG_SIZE_SHIFT) +#ifndef __ASSEMBLY__ -#define PTE_TABLE_SIZE (sizeof(real_pte_t) << PTE_INDEX_SIZE) +#define PTE_TABLE_SIZE PTE_FRAG_SIZE#define PMD_TABLE_SIZE (sizeof(pmd_t) << PMD_INDEX_SIZE)#define PGD_TABLE_SIZE (sizeof(pgd_t) << PGD_INDEX_SIZE)#endif /* __ASSEMBLY__ */
@@ -32,9 +43,11 @@#define PGDIR_SIZE (1UL << PGDIR_SHIFT)#define PGDIR_MASK (~(PGDIR_SIZE-1)) -/* Bits to mask out from a PMD to get to the PTE page */ -/* PMDs point to PTE table fragments which are 4K aligned. */ -#define PMD_MASKED_BITS 0xfff +/* +* Bits to mask out from a PMD to get to the PTE page +* PMDs point to PTE table fragments which are PTE_FRAG_SIZE aligned. +*/ +#define PMD_MASKED_BITS (PTE_FRAG_SIZE - 1)/* Bits to mask out from a PGD/PUD to get to the PMD page */ #define PUD_MASKED_BITS 0x1ff
arch/powerpc/include/asm/pgalloc-64.h +0 -10
@@ -163,16 +163,6 @@ static inline void __pte_free_tlb(struct mmu_gather *tlb, pgtable_t table, }#else /* if CONFIG_PPC_64K_PAGES */ -/* -* we support 16 fragments per PTE page. -*/ -#define PTE_FRAG_NR 16 -/* -* We use a 2K PTE page fragment and another 2K for storing -* real_pte_t hash index -*/ -#define PTE_FRAG_SIZE_SHIFT 12 -#define PTE_FRAG_SIZE (2 * PTRS_PER_PTE * sizeof(pte_t))extern pte_t *page_table_alloc (struct mm_struct *,unsigned long ,int );extern void page_table_free (struct mm_struct *,unsigned long *,int );
arch/powerpc/mm/init_64.c +0 -4
@@ -87,11 +87,7 @@ static void pgd_ctor(void *addr)static void pmd_ctor (void *addr) { -#ifdef CONFIG_TRANSPARENT_HUGEPAGE -memset (addr,0 ,PMD_TABLE_SIZE *2 ); -#else memset (addr,0 ,PMD_TABLE_SIZE ); -#endif }struct kmem_cache *pgtable_cache[MAX_PGTABLE_INDEX_SIZE ];
Signature · 1 line
2.5.0
PATCH v6 31/35 powerpc/mm: Don't hardcode the hash pte slot shift
In reply to Aneesh Kumar K.VUse the #define instead of open-coding the same
--- arch/powerpc/include/asm/book3s/64/hash-64k.h | 2 +- arch/powerpc/include/asm/book3s/64/pgtable.h | 2 +- arch/powerpc/include/asm/nohash/64/pgtable.h | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-)
arch/powerpc/include/asm/book3s/64/hash-64k.h +1 -1
@@ -94,7 +94,7 @@ static inline unsigned long __rpte_to_hidx(real_pte_t rpte, unsigned long index) {if ((pte_val (rpte.pte) & _PAGE_COMBO))return (rpte.hidx >> (index<<2 )) &0xf ; -return (pte_val (rpte.pte) >>12 ) &0xf ; +return (pte_val (rpte.pte) >> _PAGE_F_GIX_SHIFT) &0xf ; }#define __rpte_to_pte (r) ((r).pte)
arch/powerpc/include/asm/book3s/64/pgtable.h +1 -1
@@ -50,7 +50,7 @@#define __real_pte (e,p) (e)#define __rpte_to_pte (r) (__pte(r))#endif -#define __rpte_to_hidx (r,index) (pte_val(__rpte_to_pte(r)) >> 12) +#define __rpte_to_hidx (r,index) (pte_val(__rpte_to_pte(r)) >>_PAGE_F_GIX_SHIFT)#define pte_iterate_hashed_subpages (rpte, psize, va, index, shift) \ do { \
arch/powerpc/include/asm/nohash/64/pgtable.h +1 -1
@@ -121,7 +121,7 @@#define __real_pte (e,p) (e)#define __rpte_to_pte (r) (__pte(r))#endif -#define __rpte_to_hidx (r,index) (pte_val(__rpte_to_pte(r)) >> 12) +#define __rpte_to_hidx (r,index) (pte_val(__rpte_to_pte(r)) >> _PAGE_F_GIX_SHIFT)#define pte_iterate_hashed_subpages (rpte, psize, va, index, shift) \ do { \
Signature · 1 line
2.5.0
PATCH v6 32/35 powerpc/nohash: Update 64K nohash config to have 32 pte fragement
In reply to Aneesh Kumar K.VThey don't need to track 4k subpage slot details and hence don't need second half of pgtable_t.
--- arch/powerpc/include/asm/nohash/64/pgtable-64k.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
arch/powerpc/include/asm/nohash/64/pgtable-64k.h +3 -3
@@ -10,14 +10,14 @@#define PGD_INDEX_SIZE 12/* -* we support 16 fragments per PTE page of 64K size. +* we support 32 fragments per PTE page of 64K size */ -#define PTE_FRAG_NR 16 +#define PTE_FRAG_NR 32/* * We use a 2K PTE page fragment and another 2K for storing * real_pte_t hash index */ -#define PTE_FRAG_SIZE_SHIFT 12 +#define PTE_FRAG_SIZE_SHIFT 11#define PTE_FRAG_SIZE (1UL << PTE_FRAG_SIZE_SHIFT)#ifndef __ASSEMBLY__
Signature · 1 line
2.5.0
PATCH v6 33/35 powerpc/nohash: we don't use real_pte_t for nohash
In reply to Aneesh Kumar K.VRemove the related functions and #defines
--- arch/powerpc/include/asm/nohash/64/pgtable.h | 33 ---------------------------- 1 file changed, 33 deletions(-)
arch/powerpc/include/asm/nohash/64/pgtable.h +0 -33
@@ -106,39 +106,6 @@#endif /* CONFIG_PPC_MM_SLICES */ #ifndef __ASSEMBLY__ - -/* -* This is the default implementation of various PTE accessors, it's -* used in all cases except Book3S with 64K pages where we have a -* concept of sub-pages -*/ -#ifndef __real_pte - -#ifdef CONFIG_STRICT_MM_TYPECHECKS -#define __real_pte (e,p) ((real_pte_t){(e)}) -#define __rpte_to_pte (r) ((r).pte) -#else -#define __real_pte (e,p) (e) -#define __rpte_to_pte (r) (__pte(r)) -#endif -#define __rpte_to_hidx (r,index) (pte_val(__rpte_to_pte(r)) >> _PAGE_F_GIX_SHIFT) - -#define pte_iterate_hashed_subpages (rpte, psize, va, index, shift) \ - do { \ - index = 0; \ - shift = mmu_psize_defs[psize].shift; \ - -#define pte_iterate_hashed_end () } while(0) - -/* -* We expect this to be called only for user addresses or kernel virtual -* addresses other than the linear mapping. -*/ -#define pte_pagesize_index (mm, addr, pte) MMU_PAGE_4K - -#endif /* __real_pte */ - -/* pte_clear moved to later in this file */ #define PMD_BAD_BITS (PTE_TABLE_SIZE-1)
Signature · 1 line
2.5.0
PATCH v6 34/35 powerpc/mm: Use H_READ with H_READ_4
In reply to Aneesh Kumar K.VThis will bulk read 4 hash pte slot entries and should reduce the loop
--- arch/powerpc/include/asm/plpar_wrappers.h | 17 ++++++++++ arch/powerpc/platforms/pseries/lpar.c | 54 +++++++++++++++---------------- 2 files changed, 44 insertions(+), 27 deletions(-)
arch/powerpc/include/asm/plpar_wrappers.h +17 -0
@@ -202,6 +202,23 @@ static inline long plpar_pte_read_raw(unsigned long flags, unsigned long ptex, }/* +* ptes must be 8*sizeof(unsigned long) +*/ +static inline long plpar_pte_read_4 (unsigned long flags,unsigned long ptex, +unsigned long *ptes) + +{ +long rc; +unsigned long retbuf[PLPAR_HCALL9_BUFSIZE ]; + + rc =plpar_hcall9 (H_READ , retbuf, flags |H_READ_4 , ptex); + +memcpy (ptes, retbuf,8 *sizeof (unsigned long )); + +return rc; +} + +/* * plpar_pte_read_4_raw can be called in real mode. * ptes must be 8*sizeof(unsigned long) */
arch/powerpc/platforms/pseries/lpar.c +27 -27
75 lines not on this page · raw message
Signature · 1 line
2.5.0
PATCH v6 35/35 powerpc/mm: Don't open code pgtable_t size
In reply to Aneesh Kumar K.VThe slot information of base page size hash pte is stored in the pgtable_t w.r.t transparent hugepage. We need to make sure we don't index beyond pgtable_t size.
--- arch/powerpc/mm/hugepage-hash64.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/powerpc/mm/hugepage-hash64.c +1 -1
@@ -71,7 +71,7 @@ int __hash_page_thp(unsigned long ea, unsigned long access, unsigned long vsid,*/ shift = mmu_psize_defs[psize].shift; index = (ea & ~HPAGE_PMD_MASK ) >> shift; -BUG_ON (index >=4096 ); +BUG_ON (index >=PTE_FRAG_SIZE ); vpn =hpt_vpn (ea, vsid, ssize); hpte_slot_array =get_hpte_slot_array (pmdp);
Signature · 1 line
2.5.0