Nick has been so kind as to provide a current non-debug kernel. I´ve performed various benchmarks, both with and without wapbl. I´m attaching the bytebench results as well as the lmbench results in raw and summary form (they are with wapbl enabled). I missed to provide the Linux summary of lmbench in my previous mail, so it comes along with this one. Depening on the test, either are we faster or Linux is. Anyhow, it´s a huge improvment over the HEAD kernel. The compile test passed in 5 min. 40 sec. ( 5m17sec. user, 18.61s. kernel) this time. I still wonder why Linux is so much faster here, especially in terms of user time. Turning off wapbl has also a positive effect, but I´m not having numbers here just now. 2015-03-11 10:59 GMT+01:00 Nick Hudson <skrll%netbsd.org@localhost>: > On 03/11/15 09:53, Stephan wrote: >> >> Sure, I can do that. Can you tell me wether DIAGNOSTIC and DEBUG are >> disabled in these builds: >> >> >> http://nyftp.netbsd.org/pub/NetBSD-daily/HEAD/201503110500Z/evbarm-earmv6hf/binary/gzimg/ > > > Both are currently enabled in HEAD > > > http://nxr.netbsd.org/xref/src/sys/arch/evbarm/conf/RPI#135 > > > >> >> If not, I suspect that I need to build my own kernel. > > > Yeah, you will. > > Nick
# # # # # # # ##### ###### # # #### # #
# # ## # # # # # # # ## # # # # #
# # # # # # ## ##### ##### # # # # ######
# # # # # # ## # # # # # # # # #
# # # ## # # # # # # # ## # # # #
#### # # # # # ##### ###### # # #### # #
4 1 Based on the Byte Magazine Unix Benchmark
44 11
v v 4 4 1
v v 44444 1 v4.1 revisions mostly by David C. Niemi,
v 4 o 111 Reston, VA, USA <niemi%tux.org@localhost>
Dhrystone 2 using register variables 1 2 3 4 5 6 7 8 9 10
Double-Precision Whetstone 1 2 3 4 5 6 7 8 9 10
System Call Overhead 1 2 3 4 5 6 7 8 9 10
Pipe Throughput 1 2 3 4 5 6 7 8 9 10
Pipe-based Context Switching 1 2 3 4 5 6 7 8 9 10
Process Creation 1 2 3
Execl Throughput 1 2 3
Filesystem Throughput 1024 bufsize 2000 maxblocks 1 2 3
Filesystem Throughput 256 bufsize 500 maxblocks 1 2 3
Filesystem Throughput 4096 bufsize 8000 maxblocks 1 2 3
Shell Scripts (1 concurrent) 1 2 3
Shell Scripts (8 concurrent) 1 2 3
Shell Scripts (16 concurrent) 1 2 3
Arithmetic Test (type = short) 1 2 3
Arithmetic Test (type = int) 1 2 3
Arithmetic Test (type = long) 1 2 3
Arithmetic Test (type = float) 1 2 3
Arithmetic Test (type = double) 1 2 3
Arithoh 1 2 3
C Compiler Throughput 1 2 3
Dc: sqrt(2) to 99 decimal places 1 2 3
Recursion Test--Tower of Hanoi 1 2 3
==============================================================
BYTE UNIX Benchmarks (Version 4.1.0)
System -- rpi
Start Benchmark Run: Wed Feb 4 21:39:05 JST 2015
1 interactive users.
9:39PM up 1:43, 1 user, load averages: 0.03, 0.21, 0.41
-r-xr-xr-x 1 root wheel 176971 Feb 2 02:21 /bin/sh
/bin/sh: ELF 32-bit LSB executable, ARM, EABI5 version 1 (SYSV), dynamically linked, interpreter /libexec/ld.elf_so, for NetBSD 7.99.4, compiled for: earmv6hf, not stripped
/dev/ld0a 7325933 2893971 4065666 41% /
Dhrystone 2 using register variables 1319049.0 lps (10.0 secs, 10 samples)
Double-Precision Whetstone 429.9 MWIPS (10.0 secs, 10 samples)
System Call Overhead 163590.2 lps (10.0 secs, 10 samples)
Pipe Throughput 156763.9 lps (10.0 secs, 10 samples)
Pipe-based Context Switching 38326.4 lps (10.0 secs, 10 samples)
Process Creation 353.3 lps (30.1 secs, 3 samples)
Execl Throughput 133.0 lps (29.6 secs, 3 samples)
File Read 1024 bufsize 2000 maxblocks 62669.0 KBps (30.0 secs, 3 samples)
File Write 1024 bufsize 2000 maxblocks 120884.0 KBps (30.0 secs, 3 samples)
File Copy 1024 bufsize 2000 maxblocks 42877.0 KBps (30.0 secs, 3 samples)
File Read 256 bufsize 500 maxblocks 31267.0 KBps (30.0 secs, 3 samples)
File Write 256 bufsize 500 maxblocks 38639.0 KBps (30.0 secs, 3 samples)
File Copy 256 bufsize 500 maxblocks 16154.0 KBps (30.0 secs, 3 samples)
File Read 4096 bufsize 8000 maxblocks 86576.0 KBps (30.0 secs, 3 samples)
File Write 4096 bufsize 8000 maxblocks 286567.0 KBps (30.0 secs, 3 samples)
File Copy 4096 bufsize 8000 maxblocks 40585.0 KBps (30.0 secs, 3 samples)
Shell Scripts (1 concurrent) 382.3 lpm (60.1 secs, 3 samples)
Shell Scripts (8 concurrent) 51.9 lpm (60.1 secs, 3 samples)
Shell Scripts (16 concurrent) 25.9 lpm (60.1 secs, 3 samples)
Arithmetic Test (type = short) 1.0 lps (0.0 secs, 3 samples)
Arithmetic Test (type = int) 1.0 lps (0.0 secs, 3 samples)
Arithmetic Test (type = long) 1.0 lps (0.0 secs, 3 samples)
Arithmetic Test (type = float) 1.0 lps (0.0 secs, 3 samples)
Arithmetic Test (type = double) 1.0 lps (0.0 secs, 3 samples)
Arithoh 1.0 lps (0.0 secs, 3 samples)
C Compiler Throughput 73.9 lpm (60.1 secs, 3 samples)
Dc: sqrt(2) to 99 decimal places 5645.4 lpm (30.1 secs, 3 samples)
Recursion Test--Tower of Hanoi 16263.0 lps (20.0 secs, 3 samples)
INDEX VALUES
TEST BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 1319049.0 113.0
Double-Precision Whetstone 55.0 429.9 78.2
Execl Throughput 43.0 133.0 30.9
File Copy 1024 bufsize 2000 maxblocks 3960.0 42877.0 108.3
File Copy 256 bufsize 500 maxblocks 1655.0 16154.0 97.6
File Copy 4096 bufsize 8000 maxblocks 5800.0 40585.0 70.0
Pipe Throughput 12440.0 156763.9 126.0
Pipe-based Context Switching 4000.0 38326.4 95.8
Process Creation 126.0 353.3 28.0
Shell Scripts (8 concurrent) 6.0 51.9 86.5
System Call Overhead 15000.0 163590.2 109.1
=========
FINAL SCORE 77.9
Attachment:
rpi-new
Description: Binary data
Attachment:
summary-netbsd.out
Description: Binary data
Attachment:
summary-linux.out
Description: Binary data