forked from riscv-boom/riscv-boom
-
Notifications
You must be signed in to change notification settings - Fork 2
membench
David Metz edited this page Apr 20, 2020
·
2 revisions
mem_test to verify all memory is actually addressable and writable mem_sys to measure random access latency
tinymembench v0.4.9 (simple benchmark for memory throughput and latency)
==========================================================================
== Memory bandwidth tests ==
== ==
== Note 1: 1MB = 1000000 bytes ==
== Note 2: Results for 'copy' tests show how many bytes can be ==
== copied per second (adding together read and writen ==
== bytes would have provided twice higher numbers) ==
== Note 3: 2-pass copy means that we are using a small temporary buffer ==
== to first fetch data into it, and only then write it to the ==
== destination (source -> L1 cache, L1 cache -> destination) ==
== Note 4: If sample standard deviation exceeds 0.1%, it is shown in ==
== brackets ==
==========================================================================
C copy backwards : 44.9 MB/s
C copy backwards (32 byte blocks) : 44.1 MB/s
C copy backwards (64 byte blocks) : 44.1 MB/s
C copy : 44.9 MB/s
C copy prefetched (32 bytes step) : 44.0 MB/s
C copy prefetched (64 bytes step) : 44.0 MB/s
C 2-pass copy : 29.5 MB/s
C 2-pass copy prefetched (32 bytes step) : 29.6 MB/s
C 2-pass copy prefetched (64 bytes step) : 29.6 MB/s
C fill : 90.7 MB/s
C fill (shuffle within 16 byte blocks) : 90.7 MB/s
C fill (shuffle within 32 byte blocks) : 90.7 MB/s
C fill (shuffle within 64 byte blocks) : 89.8 MB/s (0.1%)
---
standard memcpy : 41.6 MB/s
standard memset : 90.1 MB/s
==========================================================================
== Memory latency test ==
== ==
== Average time is measured for random memory accesses in the buffers ==
== of different sizes. The larger is the buffer, the more significant ==
== are relative contributions of TLB, L1/L2 cache misses and SDRAM ==
== accesses. For extremely large buffer sizes we are expecting to see ==
== page table walk with several requests to SDRAM for almost every ==
== memory access (though 64MiB is not nearly large enough to experience ==
== this effect to its fullest). ==
== ==
== Note 1: All the numbers are representing extra time, which needs to ==
== be added to L1 cache latency. The cycle timings for L1 cache ==
== latency can be usually found in the processor documentation. ==
== Note 2: Dual random read means that we are simultaneously performing ==
== two independent memory accesses at a time. In the case if ==
== the memory subsystem can't handle multiple outstanding ==
== requests, dual random read has the same timings as two ==
== single reads performed one after another. ==
==========================================================================
block size : single random read / dual random read
1024 : 0.0 ns / 0.0 ns
2048 : 0.0 ns / 0.0 ns
4096 : 0.0 ns / 0.0 ns
8192 : 0.0 ns / 0.0 ns
16384 : 0.0 ns / 0.0 ns
32768 : 375.1 ns / 578.3 ns
65536 : 565.7 ns / 781.4 ns
131072 : 654.6 ns / 855.8 ns
262144 : 697.2 ns / 885.2 ns
524288 : 718.0 ns / 898.4 ns
1048576 : 728.2 ns / 905.0 ns
2097152 : 733.2 ns / 908.5 ns
4194304 : 955.5 ns / 1432.9 ns
8388608 : 1119.5 ns / 1816.1 ns
16777216 : 1306.0 ns / 2202.2 ns
33554432 : 1418.9 ns / 2439.7 ns
67108864 : 1493.8 ns / 2592.4 ns
mem_sys:
csr_time 14
size: 64, time: 22
size: 64, time: 8 adj
size: 128, time: 21
size: 128, time: 7 adj
size: 256, time: 21
size: 256, time: 7 adj
size: 512, time: 21
size: 512, time: 7 adj
size: 1024, time: 22
size: 1024, time: 8 adj
size: 2048, time: 23
size: 2048, time: 9 adj
size: 4096, time: 43
size: 4096, time: 29 adj
size: 8192, time: 54
size: 8192, time: 40 adj
size: 16384, time: 60
size: 16384, time: 46 adj
size: 32768, time: 63
size: 32768, time: 49 adj
size: 65536, time: 64
size: 65536, time: 50 adj
size: 131072, time: 65
size: 131072, time: 51 adj
size: 262144, time: 65
size: 262144, time: 51 adj
size: 524288, time: 66
size: 524288, time: 52 adj
size: 1048576, time: 66
size: 1048576, time: 52 adj
size: 2097152, time: 66
size: 2097152, time: 52 adj
size: 4194304, time: 66
size: 4194304, time: 52 adj
#mem_sys.riscv
root@zynq:~# /sdcard/riscv-spec-test/flash-bitstream.sh RocketZynqConfig.bit
root@zynq:~# ./fesvr-zynq mem_sys.riscv
csr_time 1
size: 64, time: 2
size: 64, time: 1 adj
size: 128, time: 2
size: 128, time: 1 adj
size: 256, time: 2
size: 256, time: 1 adj
size: 512, time: 2
size: 512, time: 1 adj
size: 1024, time: 2
size: 1024, time: 1 adj
size: 2048, time: 3
size: 2048, time: 2 adj
size: 4096, time: 17
size: 4096, time: 16 adj
size: 8192, time: 25
size: 8192, time: 24 adj
size: 16384, time: 29
size: 16384, time: 28 adj
size: 32768, time: 31
size: 32768, time: 30 adj
size: 65536, time: 32
size: 65536, time: 31 adj
size: 131072, time: 32
size: 131072, time: 31 adj
size: 262144, time: 33
size: 262144, time: 32 adj
size: 524288, time: 33
size: 524288, time: 32 adj
size: 1048576, time: 33
size: 1048576, time: 32 adj
size: 2097152, time: 33
size: 2097152, time: 32 adj
size: 4194304, time: 33
size: 4194304, time: 32 adj
size: 8388608, time: 33
size: 8388608, time: 32 adj
root@zynq:~# /sdcard/riscv-spec-test/flash-bitstream.sh WithL2TLB_WithL2Cache_With1GbRam_RocketZynqConfig.bit
root@zynq:~# ./fesvr-zynq mem_sys.riscv
csr_time 1
size: 64, time: 3
size: 64, time: 2 adj
size: 128, time: 2
size: 128, time: 1 adj
size: 256, time: 2
size: 256, time: 1 adj
size: 512, time: 2
size: 512, time: 1 adj
size: 1024, time: 2
size: 1024, time: 1 adj
size: 2048, time: 3
size: 2048, time: 2 adj
size: 4096, time: 10
size: 4096, time: 9 adj
size: 8192, time: 14
size: 8192, time: 13 adj
size: 16384, time: 16
size: 16384, time: 15 adj
size: 32768, time: 17
size: 32768, time: 16 adj
size: 65536, time: 19
size: 65536, time: 18 adj
size: 131072, time: 31
size: 131072, time: 30 adj
size: 262144, time: 38
size: 262144, time: 37 adj
size: 524288, time: 42
size: 524288, time: 41 adj
size: 1048576, time: 44
size: 1048576, time: 43 adj
size: 2097152, time: 45
size: 2097152, time: 44 adj
size: 4194304, time: 45
size: 4194304, time: 44 adj
size: 8388608, time: 45
size: 8388608, time: 44 adj
root@zynq:~# /sdcard/riscv-spec-test/flash-bitstream.sh WithL2TLB_WithL2Cache_With1GbRam_SmallBoomZynqConfig.bit
root@zynq:~# ./fesvr-zynq mem_sys.riscv
csr_time 14
size: 64, time: 22
size: 64, time: 8 adj
size: 128, time: 21
size: 128, time: 7 adj
size: 256, time: 21
size: 256, time: 7 adj
size: 512, time: 21
size: 512, time: 7 adj
size: 1024, time: 21
size: 1024, time: 7 adj
size: 2048, time: 22
size: 2048, time: 8 adj
size: 4096, time: 36
size: 4096, time: 22 adj
size: 8192, time: 43
size: 8192, time: 29 adj
size: 16384, time: 47
size: 16384, time: 33 adj
size: 32768, time: 49
size: 32768, time: 35 adj
size: 65536, time: 51
size: 65536, time: 37 adj
size: 131072, time: 66
size: 131072, time: 52 adj
size: 262144, time: 74
size: 262144, time: 60 adj
size: 524288, time: 79
size: 524288, time: 65 adj
size: 1048576, time: 81
size: 1048576, time: 67 adj
size: 2097152, time: 82
size: 2097152, time: 68 adj
size: 4194304, time: 82
size: 4194304, time: 68 adj
size: 8388608, time: 83
size: 8388608, time: 69 adj
root@zynq:~# /sdcard/riscv-spec-test/flash-bitstream.sh SmallBoomZynqConfig.bit
root@zynq:~# ./fesvr-zynq mem_sys.riscv
csr_time 14
size: 64, time: 22
size: 64, time: 8 adj
size: 128, time: 21
size: 128, time: 7 adj
size: 256, time: 21
size: 256, time: 7 adj
size: 512, time: 21
size: 512, time: 7 adj
size: 1024, time: 22
size: 1024, time: 8 adj
size: 2048, time: 23
size: 2048, time: 9 adj
size: 4096, time: 43
size: 4096, time: 29 adj
size: 8192, time: 54
size: 8192, time: 40 adj
size: 16384, time: 60
size: 16384, time: 46 adj
size: 32768, time: 63
size: 32768, time: 49 adj
size: 65536, time: 64
size: 65536, time: 50 adj
size: 131072, time: 65
size: 131072, time: 51 adj
size: 262144, time: 65
size: 262144, time: 51 adj
size: 524288, time: 66
size: 524288, time: 52 adj
size: 1048576, time: 66
size: 1048576, time: 52 adj
size: 2097152, time: 66
size: 2097152, time: 52 adj
size: 4194304, time: 66
size: 4194304, time: 52 adj
size: 8388608, time: 66
size: 8388608, time: 52 adj
root@zynq:~# ./tinymembench-arm
tinymembench v0.4.9 (simple benchmark for memory throughput and latency)
==========================================================================
== Memory bandwidth tests ==
== ==
== Note 1: 1MB = 1000000 bytes ==
== Note 2: Results for 'copy' tests show how many bytes can be ==
== copied per second (adding together read and writen ==
== bytes would have provided twice higher numbers) ==
== Note 3: 2-pass copy means that we are using a small temporary buffer ==
== to first fetch data into it, and only then write it to the ==
== destination (source -> L1 cache, L1 cache -> destination) ==
== Note 4: If sample standard deviation exceeds 0.1%, it is shown in ==
== brackets ==
==========================================================================
C copy backwards : 330.3 MB/s
C copy backwards (32 byte blocks) : 330.0 MB/s
C copy backwards (64 byte blocks) : 324.3 MB/s (0.4%)
C copy : 333.2 MB/s
C copy prefetched (32 bytes step) : 513.2 MB/s (0.2%)
C copy prefetched (64 bytes step) : 341.0 MB/s
C 2-pass copy : 262.4 MB/s
C 2-pass copy prefetched (32 bytes step) : 450.9 MB/s
C 2-pass copy prefetched (64 bytes step) : 403.0 MB/s
C fill : 2125.9 MB/s
C fill (shuffle within 16 byte blocks) : 2125.7 MB/s
C fill (shuffle within 32 byte blocks) : 2125.9 MB/s
C fill (shuffle within 64 byte blocks) : 685.9 MB/s (0.9%)
---
standard memcpy : 360.2 MB/s
standard memset : 2126.0 MB/s
---
NEON read : 408.0 MB/s
NEON read prefetched (32 bytes step) : 916.4 MB/s
NEON read prefetched (64 bytes step) : 981.9 MB/s
NEON read 2 data streams : 570.3 MB/s
NEON read 2 data streams prefetched (32 bytes step) : 877.8 MB/s
NEON read 2 data streams prefetched (64 bytes step) : 970.2 MB/s
NEON copy : 309.3 MB/s
NEON copy prefetched (32 bytes step) : 503.0 MB/s
NEON copy prefetched (64 bytes step) : 586.4 MB/s
NEON unrolled copy : 283.6 MB/s
NEON unrolled copy prefetched (32 bytes step) : 504.2 MB/s
NEON unrolled copy prefetched (64 bytes step) : 568.2 MB/s
NEON copy backwards : 308.2 MB/s
NEON copy backwards prefetched (32 bytes step) : 480.7 MB/s
NEON copy backwards prefetched (64 bytes step) : 556.2 MB/s
NEON 2-pass copy : 249.5 MB/s
NEON 2-pass copy prefetched (32 bytes step) : 461.7 MB/s
NEON 2-pass copy prefetched (64 bytes step) : 427.5 MB/s
NEON unrolled 2-pass copy : 244.1 MB/s
NEON unrolled 2-pass copy prefetched (32 bytes step) : 457.8 MB/s
NEON unrolled 2-pass copy prefetched (64 bytes step) : 518.6 MB/s
NEON fill : 1923.6 MB/s
NEON fill backwards : 1920.3 MB/s
VFP copy : 297.2 MB/s
VFP 2-pass copy : 270.4 MB/s
ARM fill (STRD) : 2125.8 MB/s
ARM fill (STM with 8 registers) : 2126.4 MB/s
ARM fill (STM with 4 registers) : 2126.1 MB/s
ARM copy prefetched (incr pld) : 534.4 MB/s
ARM copy prefetched (wrap pld) : 527.9 MB/s
ARM 2-pass copy prefetched (incr pld) : 528.9 MB/s
ARM 2-pass copy prefetched (wrap pld) : 532.7 MB/s
==========================================================================
== Memory latency test ==
== ==
== Average time is measured for random memory accesses in the buffers ==
== of different sizes. The larger is the buffer, the more significant ==
== are relative contributions of TLB, L1/L2 cache misses and SDRAM ==
== accesses. For extremely large buffer sizes we are expecting to see ==
== page table walk with several requests to SDRAM for almost every ==
== memory access (though 64MiB is not nearly large enough to experience ==
== this effect to its fullest). ==
== ==
== Note 1: All the numbers are representing extra time, which needs to ==
== be added to L1 cache latency. The cycle timings for L1 cache ==
== latency can be usually found in the processor documentation. ==
== Note 2: Dual random read means that we are simultaneously performing ==
== two independent memory accesses at a time. In the case if ==
== the memory subsystem can't handle multiple outstanding ==
== requests, dual random read has the same timings as two ==
== single reads performed one after another. ==
==========================================================================
block size : single random read / dual random read
1024 : 0.0 ns / 0.0 ns
2048 : 0.0 ns / 0.0 ns
4096 : 0.0 ns / 0.0 ns
8192 : 0.0 ns / 0.0 ns
16384 : 0.0 ns / 0.0 ns
32768 : 0.0 ns / 0.2 ns
65536 : 20.6 ns / 31.5 ns
131072 : 31.5 ns / 39.4 ns
262144 : 42.2 ns / 47.7 ns
524288 : 49.0 ns / 52.5 ns
1048576 : 108.6 ns / 147.4 ns
2097152 : 139.9 ns / 178.4 ns
4194304 : 157.2 ns / 192.6 ns
8388608 : 170.0 ns / 206.7 ns
16777216 : 183.5 ns / 227.6 ns
33554432 : 196.9 ns / 251.7 ns
67108864 : 210.9 ns / 278.8 ns
/sdcard/riscv-spec-test/flash-bitstream.sh RocketZynqConfig.bit ./fesvr-zynq pointless_pointers.riscv /sdcard/riscv-spec-test/flash-bitstream.sh WithL2TLB_WithL2Cache_With1GbRam_RocketZynqConfig.bit ./fesvr-zynq pointless_pointers.riscv /sdcard/riscv-spec-test/flash-bitstream.sh WithL2TLB_WithL2Cache_With1GbRam_SmallBoomZynqConfig.bit ./fesvr-zynq pointless_pointers.riscv /sdcard/riscv-spec-test/flash-bitstream.sh SmallBoomZynqConfig.bit ./fesvr-zynq pointless_pointers.riscv
root@zynq:~# /sdcard/riscv-spec-test/flash-bitstream.sh RocketZynqConfig.bit
+ cat RocketZynqConfig.bit
++ cat /sys/class/xdevcfg/xdevcfg/device/prog_done
+ is_done=1
+ [[ 1 != \1 ]]
++ /home/root/fesvr-zynq /home/root/pk /home/root/hello
+ hello=hello
+ [[ hello != \h\e\l\l\o ]]
root@zynq:~# ./fesvr-zynq pointless_pointers.riscv
cleaning memory...
setting up 32 pointers...
chaseing pointers...
...done - 0x81000168 - 0x81000168
size: 64, cycles: 186, loops: 32, cycles_per_loop: 5
cleaning memory...
setting up 64 pointers...
chaseing pointers...
...done - 0x81000110 - 0x81000110
size: 128, cycles: 203, loops: 64, cycles_per_loop: 3
cleaning memory...
setting up 128 pointers...
chaseing pointers...
...done - 0x81000030 - 0x81000030
size: 256, cycles: 395, loops: 128, cycles_per_loop: 3
cleaning memory...
setting up 256 pointers...
chaseing pointers...
...done - 0x81000960 - 0x81000960
size: 512, cycles: 784, loops: 256, cycles_per_loop: 3
cleaning memory...
setting up 512 pointers...
chaseing pointers...
...done - 0x810017f0 - 0x810017f0
size: 1024, cycles: 1622, loops: 512, cycles_per_loop: 3
cleaning memory...
setting up 1024 pointers...
chaseing pointers...
...done - 0x810020e8 - 0x810020e8
size: 2048, cycles: 5176, loops: 1024, cycles_per_loop: 5
cleaning memory...
setting up 2048 pointers...
chaseing pointers...
...done - 0x810045f0 - 0x810045f0
size: 4096, cycles: 44267, loops: 2048, cycles_per_loop: 21
cleaning memory...
setting up 4096 pointers...
chaseing pointers...
...done - 0x81006af8 - 0x81006af8
size: 8192, cycles: 114585, loops: 4096, cycles_per_loop: 27
cleaning memory...
setting up 8192 pointers...
chaseing pointers...
...done - 0x81015838 - 0x81015838
size: 16384, cycles: 255885, loops: 8192, cycles_per_loop: 31
cleaning memory...
setting up 16384 pointers...
chaseing pointers...
...done - 0x8100a360 - 0x8100a360
size: 32768, cycles: 532958, loops: 16384, cycles_per_loop: 32
cleaning memory...
setting up 32768 pointers...
chaseing pointers...
...done - 0x8104c968 - 0x8104c968
size: 65536, cycles: 1095575, loops: 32768, cycles_per_loop: 33
cleaning memory...
setting up 65536 pointers...
chaseing pointers...
...done - 0x8104a7a0 - 0x8104a7a0
size: 131072, cycles: 2213510, loops: 65536, cycles_per_loop: 33
cleaning memory...
setting up 131072 pointers...
chaseing pointers...
...done - 0x811eb308 - 0x811eb308
size: 262144, cycles: 4451866, loops: 131072, cycles_per_loop: 33
cleaning memory...
setting up 262144 pointers...
chaseing pointers...
...done - 0x811adf10 - 0x811adf10
size: 524288, cycles: 8924008, loops: 262144, cycles_per_loop: 34
cleaning memory...
setting up 524288 pointers...
chaseing pointers...
...done - 0x81181ab8 - 0x81181ab8
size: 1048576, cycles: 17867260, loops: 524288, cycles_per_loop: 34
cleaning memory...
setting up 1048576 pointers...
chaseing pointers...
...done - 0x81fa2870 - 0x81fa2870
size: 2097152, cycles: 35749656, loops: 1048576, cycles_per_loop: 34
cleaning memory...
setting up 2097152 pointers...
chaseing pointers...
...done - 0x81d4beb0 - 0x81d4beb0
size: 4194304, cycles: 71516485, loops: 2097152, cycles_per_loop: 34
cleaning memory...
setting up 4194304 pointers...
chaseing pointers...
...done - 0x81d31938 - 0x81d31938
size: 8388608, cycles: 143054469, loops: 4194304, cycles_per_loop: 34
(null) = 0
(null) = 0
(null) = 0
(null) = 0
(null) = 0
(null) = 0
root@zynq:~# /sdcard/riscv-spec-test/flash-bitstream.sh WithL2TLB_WithL2Cache_With1GbRam_RocketZynqConfig.bit
+ cat WithL2TLB_WithL2Cache_With1GbRam_RocketZynqConfig.bit
++ cat /sys/class/xdevcfg/xdevcfg/device/prog_done
+ is_done=1
+ [[ 1 != \1 ]]
++ /home/root/fesvr-zynq /home/root/pk /home/root/hello
+ hello=hello
+ [[ hello != \h\e\l\l\o ]]
root@zynq:~# ./fesvr-zynq pointless_pointers.riscv
cleaning memory...
setting up 32 pointers...
chaseing pointers...
...done - 0x81000168 - 0x81000168
size: 64, cycles: 147, loops: 32, cycles_per_loop: 4
cleaning memory...
setting up 64 pointers...
chaseing pointers...
...done - 0x81000110 - 0x81000110
size: 128, cycles: 203, loops: 64, cycles_per_loop: 3
cleaning memory...
setting up 128 pointers...
chaseing pointers...
...done - 0x81000030 - 0x81000030
size: 256, cycles: 395, loops: 128, cycles_per_loop: 3
cleaning memory...
setting up 256 pointers...
chaseing pointers...
...done - 0x81000960 - 0x81000960
size: 512, cycles: 779, loops: 256, cycles_per_loop: 3
cleaning memory...
setting up 512 pointers...
chaseing pointers...
...done - 0x810017f0 - 0x810017f0
size: 1024, cycles: 1613, loops: 512, cycles_per_loop: 3
cleaning memory...
setting up 1024 pointers...
chaseing pointers...
...done - 0x810020e8 - 0x810020e8
size: 2048, cycles: 5047, loops: 1024, cycles_per_loop: 4
cleaning memory...
setting up 2048 pointers...
chaseing pointers...
...done - 0x810045f0 - 0x810045f0
size: 4096, cycles: 26772, loops: 2048, cycles_per_loop: 13
cleaning memory...
setting up 4096 pointers...
chaseing pointers...
...done - 0x81006af8 - 0x81006af8
size: 8192, cycles: 65694, loops: 4096, cycles_per_loop: 16
cleaning memory...
setting up 8192 pointers...
chaseing pointers...
...done - 0x81015838 - 0x81015838
size: 16384, cycles: 143047, loops: 8192, cycles_per_loop: 17
cleaning memory...
setting up 16384 pointers...
chaseing pointers...
...done - 0x8100a360 - 0x8100a360
size: 32768, cycles: 299066, loops: 16384, cycles_per_loop: 18
cleaning memory...
setting up 32768 pointers...
chaseing pointers...
...done - 0x8104c968 - 0x8104c968
size: 65536, cycles: 673032, loops: 32768, cycles_per_loop: 20
cleaning memory...
setting up 65536 pointers...
chaseing pointers...
...done - 0x8104a7a0 - 0x8104a7a0
size: 131072, cycles: 2197578, loops: 65536, cycles_per_loop: 33
cleaning memory...
setting up 131072 pointers...
chaseing pointers...
...done - 0x811eb308 - 0x811eb308
size: 262144, cycles: 5360410, loops: 131072, cycles_per_loop: 40
cleaning memory...
setting up 262144 pointers...
chaseing pointers...
...done - 0x811adf10 - 0x811adf10
size: 524288, cycles: 11740561, loops: 262144, cycles_per_loop: 44
cleaning memory...
setting up 524288 pointers...
chaseing pointers...
...done - 0x81181ab8 - 0x81181ab8
size: 1048576, cycles: 24738103, loops: 524288, cycles_per_loop: 47
cleaning memory...
setting up 1048576 pointers...
chaseing pointers...
...done - 0x81fa2870 - 0x81fa2870
size: 2097152, cycles: 50904095, loops: 1048576, cycles_per_loop: 48
cleaning memory...
setting up 2097152 pointers...
chaseing pointers...
...done - 0x81d4beb0 - 0x81d4beb0
size: 4194304, cycles: 103314272, loops: 2097152, cycles_per_loop: 49
cleaning memory...
setting up 4194304 pointers...
chaseing pointers...
...done - 0x81d31938 - 0x81d31938
size: 8388608, cycles: 208166844, loops: 4194304, cycles_per_loop: 49
(null) = 0
(null) = 0
(null) = 0
(null) = 0
(null) = 0
(null) = 0
root@zynq:~# /sdcard/riscv-spec-test/flash-bitstream.sh WithL2TLB_WithL2Cache_With1GbRam_SmallBoomZynqConfig.bit
+ cat WithL2TLB_WithL2Cache_With1GbRam_SmallBoomZynqConfig.bit
++ cat /sys/class/xdevcfg/xdevcfg/device/prog_done
+ is_done=1
+ [[ 1 != \1 ]]
++ /home/root/fesvr-zynq /home/root/pk /home/root/hello
+ hello=hello
+ [[ hello != \h\e\l\l\o ]]
root@zynq:~# ./fesvr-zynq pointless_pointers.riscv
cleaning memory...
setting up 32 pointers...
chaseing pointers...
...done - 0x81000168 - 0x81000168
size: 64, cycles: 283, loops: 32, cycles_per_loop: 8
cleaning memory...
setting up 64 pointers...
chaseing pointers...
...done - 0x81000110 - 0x81000110
size: 128, cycles: 385, loops: 64, cycles_per_loop: 6
cleaning memory...
setting up 128 pointers...
chaseing pointers...
...done - 0x81000030 - 0x81000030
size: 256, cycles: 582, loops: 128, cycles_per_loop: 4
cleaning memory...
setting up 256 pointers...
chaseing pointers...
...done - 0x81000960 - 0x81000960
size: 512, cycles: 1164, loops: 256, cycles_per_loop: 4
cleaning memory...
setting up 512 pointers...
chaseing pointers...
...done - 0x810017f0 - 0x810017f0
size: 1024, cycles: 2214, loops: 512, cycles_per_loop: 4
cleaning memory...
setting up 1024 pointers...
chaseing pointers...
...done - 0x810020e8 - 0x810020e8
size: 2048, cycles: 5171, loops: 1024, cycles_per_loop: 5
cleaning memory...
setting up 2048 pointers...
chaseing pointers...
...done - 0x810045f0 - 0x810045f0
size: 4096, cycles: 28639, loops: 2048, cycles_per_loop: 13
cleaning memory...
setting up 4096 pointers...
chaseing pointers...
...done - 0x81006af8 - 0x81006af8
size: 8192, cycles: 73906, loops: 4096, cycles_per_loop: 18
cleaning memory...
setting up 8192 pointers...
chaseing pointers...
...done - 0x81015838 - 0x81015838
size: 16384, cycles: 162783, loops: 8192, cycles_per_loop: 19
cleaning memory...
setting up 16384 pointers...
chaseing pointers...
...done - 0x8100a360 - 0x8100a360
size: 32768, cycles: 339731, loops: 16384, cycles_per_loop: 20
cleaning memory...
setting up 32768 pointers...
chaseing pointers...
...done - 0x8104c968 - 0x8104c968
size: 65536, cycles: 766438, loops: 32768, cycles_per_loop: 23
cleaning memory...
setting up 65536 pointers...
chaseing pointers...
...done - 0x8104a7a0 - 0x8104a7a0
size: 131072, cycles: 2455013, loops: 65536, cycles_per_loop: 37
cleaning memory...
setting up 131072 pointers...
chaseing pointers...
...done - 0x811eb308 - 0x811eb308
size: 262144, cycles: 5923739, loops: 131072, cycles_per_loop: 45
cleaning memory...
setting up 262144 pointers...
chaseing pointers...
...done - 0x811adf10 - 0x811adf10
size: 524288, cycles: 12803436, loops: 262144, cycles_per_loop: 48
cleaning memory...
setting up 524288 pointers...
chaseing pointers...
...done - 0x81181ab8 - 0x81181ab8
size: 1048576, cycles: 26532261, loops: 524288, cycles_per_loop: 50
cleaning memory...
setting up 1048576 pointers...
chaseing pointers...
...done - 0x81fa2870 - 0x81fa2870
size: 2097152, cycles: 53985677, loops: 1048576, cycles_per_loop: 51
cleaning memory...
setting up 2097152 pointers...
chaseing pointers...
...done - 0x81d4beb0 - 0x81d4beb0
size: 4194304, cycles: 108891798, loops: 2097152, cycles_per_loop: 51
cleaning memory...
setting up 4194304 pointers...
chaseing pointers...
...done - 0x81d31938 - 0x81d31938
size: 8388608, cycles: 218680150, loops: 4194304, cycles_per_loop: 52
(null) = 0
(null) = 0
(null) = 0
(null) = 0
(null) = 0
(null) = 0
root@zynq:~# /sdcard/riscv-spec-test/flash-bitstream.sh SmallBoomZynqConfig.bit
+ cat SmallBoomZynqConfig.bit
++ cat /sys/class/xdevcfg/xdevcfg/device/prog_done
+ is_done=1
+ [[ 1 != \1 ]]
++ /home/root/fesvr-zynq /home/root/pk /home/root/hello
+ hello=hello
+ [[ hello != \h\e\l\l\o ]]
root@zynq:~# ./fesvr-zynq pointless_pointers.riscv
cleaning memory...
setting up 32 pointers...
chaseing pointers...
...done - 0x81000168 - 0x81000168
size: 64, cycles: 333, loops: 32, cycles_per_loop: 10
cleaning memory...
setting up 64 pointers...
chaseing pointers...
...done - 0x81000110 - 0x81000110
size: 128, cycles: 385, loops: 64, cycles_per_loop: 6
cleaning memory...
setting up 128 pointers...
chaseing pointers...
...done - 0x81000030 - 0x81000030
size: 256, cycles: 576, loops: 128, cycles_per_loop: 4
cleaning memory...
setting up 256 pointers...
chaseing pointers...
...done - 0x81000960 - 0x81000960
size: 512, cycles: 1108, loops: 256, cycles_per_loop: 4
cleaning memory...
setting up 512 pointers...
chaseing pointers...
...done - 0x810017f0 - 0x810017f0
size: 1024, cycles: 2355, loops: 512, cycles_per_loop: 4
cleaning memory...
setting up 1024 pointers...
chaseing pointers...
...done - 0x810020e8 - 0x810020e8
size: 2048, cycles: 6559, loops: 1024, cycles_per_loop: 6
cleaning memory...
setting up 2048 pointers...
chaseing pointers...
...done - 0x810045f0 - 0x810045f0
size: 4096, cycles: 43901, loops: 2048, cycles_per_loop: 21
cleaning memory...
setting up 4096 pointers...
chaseing pointers...
...done - 0x81006af8 - 0x81006af8
size: 8192, cycles: 116326, loops: 4096, cycles_per_loop: 28
cleaning memory...
setting up 8192 pointers...
chaseing pointers...
...done - 0x81015838 - 0x81015838
size: 16384, cycles: 260622, loops: 8192, cycles_per_loop: 31
cleaning memory...
setting up 16384 pointers...
chaseing pointers...
...done - 0x8100a360 - 0x8100a360
size: 32768, cycles: 549865, loops: 16384, cycles_per_loop: 33
cleaning memory...
setting up 32768 pointers...
chaseing pointers...
...done - 0x8104c968 - 0x8104c968
size: 65536, cycles: 1130473, loops: 32768, cycles_per_loop: 34
cleaning memory...
setting up 65536 pointers...
chaseing pointers...
...done - 0x8104a7a0 - 0x8104a7a0
size: 131072, cycles: 2286140, loops: 65536, cycles_per_loop: 34
cleaning memory...
setting up 131072 pointers...
chaseing pointers...
...done - 0x811eb308 - 0x811eb308
size: 262144, cycles: 4603646, loops: 131072, cycles_per_loop: 35
cleaning memory...
setting up 262144 pointers...
chaseing pointers...
...done - 0x811adf10 - 0x811adf10
size: 524288, cycles: 9232769, loops: 262144, cycles_per_loop: 35
cleaning memory...
setting up 524288 pointers...
chaseing pointers...
...done - 0x81181ab8 - 0x81181ab8
size: 1048576, cycles: 18493368, loops: 524288, cycles_per_loop: 35
cleaning memory...
setting up 1048576 pointers...
chaseing pointers...
...done - 0x81fa2870 - 0x81fa2870
size: 2097152, cycles: 37009522, loops: 1048576, cycles_per_loop: 35
cleaning memory...
setting up 2097152 pointers...
chaseing pointers...
...done - 0x81d4beb0 - 0x81d4beb0
size: 4194304, cycles: 74053102, loops: 2097152, cycles_per_loop: 35
cleaning memory...
setting up 4194304 pointers...
chaseing pointers...
...done - 0x81d31938 - 0x81d31938
size: 8388608, cycles: 148132652, loops: 4194304, cycles_per_loop: 35
(null) = 0
(null) = 0
(null) = 0
(null) = 0
(null) = 0
(null) = 0