Home | History | Annotate | only in /external/netperf
Up to higher level directory
NameDateSize
acinclude.m421-Oct-201613.7K
aclocal.m421-Oct-201634.6K
Android.mk21-Oct-20161.5K
AUTHORS21-Oct-20168.6K
autogen.sh21-Oct-201695
ChangeLog21-Oct-201627
config.guess21-Oct-201642.5K
config.h21-Oct-201611.2K
config.h.in21-Oct-201610.5K
config.sub21-Oct-201631K
configure21-Oct-2016233.7K
configure.ac21-Oct-201615.3K
COPYING21-Oct-20162K
depcomp21-Oct-201613.5K
doc/21-Oct-2016
inet_ntop.c21-Oct-201630
INSTALL21-Oct-20169K
install-sh21-Oct-201613.3K
m4/21-Oct-2016
Makefile.am21-Oct-2016120
Makefile.in21-Oct-201622.8K
missing21-Oct-201610K
mkinstalldirs21-Oct-20161.9K
MODULE_LICENSE_HP21-Oct-20162K
netperf.spec.in21-Oct-20161.8K
NEWS21-Oct-20160
NOTICE21-Oct-20162K
README21-Oct-20162.1K
README.aix21-Oct-20161.6K
README.hpux21-Oct-20161.7K
README.osx21-Oct-2016258
README.ovms21-Oct-20162.3K
README.solaris21-Oct-20161.2K
README.version21-Oct-201690
README.vmware21-Oct-2016720
README.windows21-Oct-20164.2K
Release_Notes21-Oct-201647.1K
src/21-Oct-2016

README

      1 
      2 BE SURE TO READ THE MANUAL. EVEN THOUGH IT MAY BE OUTDATED.
      3 
      4 This is a brief readme file for the netperf TCP/UDP/sockets/etc
      5 performance benchmark. This is here mostly as a boot-strap. The real
      6 information is in the manual, which can be found in netperf.ps and
      7 online from http://www.netperf.org/netperf/NetperfPage.html. The
      8 sources, and a limited number of binaries, can be found from
      9 ftp://ftp.cup.hp.com/dist/networking/benchmarks/netperf/ .
     10 
     11 BE SURE TO READ THE MANUAL. EVEN THOUGH IT MAY BE OUTDATED.
     12 
     13 There is a COPYRIGHT file included.  It is called COPYING because that
     14 is what autosomethingorother wanted. It is based on what the HP Legal
     15 Eagles gave me. I am not sure if the legalese is clear, but the intent
     16 is to say "Here is a benchmark. Use it in good health. Pass it along,
     17 port it, enhance it. You didn't pay for this tool, so don't expect it
     18 to be perfect ;-)" The rest of it is there to keep the layers happy...  
     19 
     20 While the copyright is pretty much in spirit an "open source" one, it
     21 is not in letter - I never took the time to try to get it approved.
     22 If you feel strongly about the license of the code you use and want
     23 something under the GPL, consider netperf4:
     24 
     25 http://www.netperf.org/svn/netperf4/trunk
     26 
     27 Feel free to report netperf results in public forums, but please be
     28 excruciatingly complete in your description of the test envorinment.
     29 The old netperf database at:
     30 
     31   http://www.netperf.org/netperf/NetperfPage.html
     32 
     33 is no more - or rather the utilities for accessing it no longer run.
     34 The data is still present in the tree, albeit _VERY_ old now.
     35 
     36 There is an Internet mailing list devoted to netperf. It is called
     37 netperf-talk and it is hosted on netperf.org. Subscription requests
     38 should go to netperf-talk-request (a] netperf.org.
     39 
     40 Please DO NOT SEND subscription requests to netperf-talk!
     41 
     42 If you run into severe difficulties, or are just feeling chatty,
     43 please feel free to drop some email to me - Rick Jones
     44 <rick.jones2 (a] hp.com>. Be sure to include a meaningful subject lines.
     45 
     46 happy benchmarking, 
     47 rick jones
     48 
     49 BE SURE TO READ THE MANUAL. EVEN THOUGH IT MAY BE OUTDATED.
     50

README.aix

      1 Pseudo-random things about netperf on AIX:
      2 
      3 While it _should_ not be necessary in the release bits, the rc bits
      4 for netperf 2.4.0 required:
      5 
      6     configure _may_ need:
      7 
      8     CFLAGS="-qcpluscmt -lperfstat"
      9 
     10     until such time as Rick Jones can figure-out or be told how to
     11     automagically add those using the configure script (hint hint :)
     12 
     13 The release bits should be OK without the above.  Depending on the
     14 name used to invoke the compiler, the -qmumble option may be implicit.
     15 
     16 AIX include files have a VERY unfortuneate set of #define's in them
     17 for phrases network oriented programs are QUITE likely to have in
     18 their source - "rem_addr" and "rem_size" A "cousin" of the Netperf
     19 Contributing Editor reports this interferes with --enable-dlpi
     20 compilation and that it was also a problem for MySQL compiltion.
     21 While we await IBM's APAR with bated breath, netperf has been kludged
     22 to workaround this bug in IBM's include files.  It has been reported
     23 that a "PMR" 34940,212,848 has been submitted to IBM in relation to
     24 this bug in their header files.
     25 
     26 Coutesy of Jan Krueger of TUI Infotec, Germany we have a pointer to
     27 a workaround for problems with compilation on AIX 6.1:
     28 
     29 https://www-304.ibm.com/support/docview.wss?uid=isg1IV01736
     30 
     31 when you get errors like:
     32 config.status: creating src/netperf_version.h
     33         gcc -DHAVE_CONFIG_H -I. -I..       -MT netlib.o -MD -MP -MF .deps/netlib.Tpo -c -o netlib.o netlib.c
     34 In file included from /usr/include/sys/corral.h:25,
     35                  from /usr/include/libperfstat.h:28,
     36                  from netlib.h:679,
     37                  from netlib.c:166:
     38 /usr/include/netinet/in6_var.h:65: error: array type has incomplete element type
     39

README.hpux

      1 A note about CPU utilization...
      2 
      3 For HP-UX 11.0 <= system < 11.23 the configure script will select the
      4 "pstat" CPU utilization mechanism.  This mechanism is the familiar
      5 HP-UX idle counter mechanism (for all incense and porpoises) and
      6 requires calibration.  See src/netcpu_pstat.c for all the details.
      7 
      8 For HP-UX 11.23 >= system, the configure script will select the
      9 "pstatnew" CPU utilization mechanism.  11.23 adds cycle counts for
     10 user, kernel and interrupt modes to the idle cycle counter.  As such,
     11 it _should_ be possible to simply take the sum of the four and the
     12 fractions and know how much time was spent in each mode.
     13 HOWEVER... there is a bug in the accounting for interrupt cycles,
     14 where interrupt cycles go missing.  SOOO, since there is an accurate
     15 way to know what the total number of cycles should have been over the
     16 interval, and we know (ass-u-me) that the idle cycle counter is good
     17 (since the pstat mechanism has tested that one OK), we will take the
     18 ratio of idle to total cycles to compute CPU util.  
     19 
     20 We will still calculate fractions for user, kernel and interrupt, and
     21 report them in debug (-d) output, but with a warning for interrupt
     22 time.  See src/netcpu_pstatnew.c for all the details.
     23 
     24 Up through HP-UX 11.23 (aka 11iV2) if you enable burst mode, and
     25 happen to send sub-MSS requests and/or responses you _cannot_ assume
     26 that the packet per second rate on the wire will match the transaction
     27 per second rate reported by netperf, even if you set TCP_NODELAY with
     28 the test-specific -D option.  The HP-UX 11.X TCP stack likely will be
     29 generating some immediate 'standalone' ACKnowledgements which may not
     30 be generated by other stacks.  This has been reported to the HP-UX TCP
     31 folks, and an announcement will be made when that issue is resolved.
     32

README.osx

      1 If you are reading this, it suggests you are using a version of
      2 netperf in which the issue of file/directory name case insensitivity
      3 in OSX has been worked-around by renaming the NetPerf and NetServer
      4 directories to NetPerfDir and NetServerDir respectively.
      5

README.ovms

      1 February 11, 2003
      2 
      3 At the time of the initial port, I was not aware of a make facility
      4 for OpenVMS.  So, I would just compile and link the various files by
      5 hand:
      6 
      7    $ cc netperf.c
      8    $ cc netlib.c
      9    $ cc netsh.c
     10    $ cc nettest_bsd.c
     11    $ cc netserver.c
     12    $ link/exe=netperf netperf.obj,netsh.obj,netlib.obj,nettest_bsd.obj
     13    $ link/exe=netserver netserver.obj,netsh.obj,netlib.obj,nettest_bsd.obj
     14 
     15 Installation for OpenVMS has a few differences from installation under
     16 say Unix.  There is no inetd for VMS - however, there is the concept
     17 of an adding an auxilliary service that seems quite similar.
     18 
     19 To configure netperf for operation as an auxilliary service, you will
     20 need to edit/use the netserver_run.com file and alter the "path" to
     21 netserver accordingly.  The version that ships is setup for where Rick
     22 Jones did his initial porting work and most likely is not apropriate
     23 for you :)
     24 
     25    $  define sys$output sys$sysroot:[netperf]hello_service.log 
     26    $  define sys$error  sys$sysroot:[netperf]hello_service.log 
     27    $  run sys$sysroot:[netperf]netserver.exe 
     28 
     29 Then it will be necessary to "define" netperf (netserver) as an
     30 auxilliary service. This will need to be customized as apropriate for
     31 your system
     32 
     33    $ tcpip set service netserver - 
     34    _$ /port=12865 - 
     35    _$ /protocol=tcp - 
     36    _$ /user=system - 
     37    _$ /limit=48 -
     38    _$ /process_name=netserver - 
     39    _$ /file=sys$sysroot:[netperf]netserver_run.com 
     40 
     41 And then it is necessary to enable the service:
     42 
     43 $ tcpip enable service netserver
     44 
     45 If you want to disable the service, you can issue the command
     46 
     47 $ tcpip set noservice netserver
     48 
     49 By default, OpenVMS is case-insensitive with commandlines, and will
     50 downshift everything to lower case.  This does not interact well with
     51 netperf's use of command-line options like "-H" and "-h" with rather
     52 different meanings.
     53 
     54 To workaround that, the following defines are believed to be
     55 sufficient.
     56 
     57 $ define  DECC$ARGV_PARSE_STYLE       ENABLE
     58 $ define  DECC$EFS_CASE_PRESERVE      ENABLE
     59 $ define  DECC$POSIX_SEEK_STREAM_FILE ENABLE
     60 $ define  DECC$EFS_CHARSET            ENABLE
     61 $ set process /parse_style=extended
     62 
     63 I do not know if this will be something one can also do for the
     64 netserver - presumeably one could put these things in the
     65 netserver_run.com file (guessing). At present though I've not tried
     66 that, and I'm also not sure that netserver has any upper-case options.
     67

README.solaris

      1 Until the release bits the following was true:
      2 
      3     Until such time as Rick Jones can figure-out or be told how to make it
      4     automagic in the configure script, prior to configure on solaris, you
      5     may need:
      6 
      7     CFLAGS="-lsocket -lnsl -lkstat"
      8 
      9     and if you are trying to compile the SCTP tests:
     10 
     11     CFLAGS="-lxnet -lsocket -lnsl -lkstat -D_XOPEN_SOURCE=500 -D__EXTENSIONS__"
     12 
     13 as the release bits have a "smarter" configure script, and the SCTP
     14 tests use libsctp, the above no longer applies.  It should all just be
     15 automagic (although for SCTP you still must --enable-sctp at configure
     16 time)
     17 
     18 Beware CPU util figures on anything before Solaris 10 that does not
     19 say 100%, and still be a triffle cautious with Solaris 10 CPU util
     20 reports.  The CPU time accounting mechanisms either do not track time
     21 spent servicing interrupts, or do so in parallel with time spent in
     22 user/kernel/idle which means that some idle time isn't _really_ idle
     23 time.
     24 
     25 And beyond that, it is still not clear if the CPU utilization reported
     26 on systems with hardware threading support (eg UltraSPARC-T1) is
     27 really accurate even ignoring the issue with interrupt time.  It is
     28 likely that to be truely accurate, it is necessary to know how much
     29 "real work" any one strand performed.
     30

README.version

      1 URL: ftp://ftp.netperf.org/netperf/netperf-2.6.0.tar.gz
      2 Version: 2.4.4
      3 BugComponent: 1351
      4

README.vmware

      1 Compiling for VMware is somewhat like compiling for Windows - there is
      2 a separate, standalone makefile one uses.  In this case, it is
      3 src/Makefile.uw.  So, to build the bits, cd to src/ and make -f
      4 Makefile.uw.
      5 
      6 At present, the makefile is setup to use a number of the "none" files
      7 - in particular netcpu_none.c.  When/if we enable the "omni" tests
      8 we'll perhaps see the addition of a number of other "none" files as
      9 well.
     10 
     11 Also, seems the way things are "run" under VMware is enough different
     12 that the scripts, should you chose to use them, will need to be
     13 modified.  The initial set of patches make some arbitrary changes that
     14 need to be re-worked with some "To run this under VMware uncomment
     15 this line" or somesuch.
     16

README.windows

      1 It has been reported that versions of netperf have configured and
      2 compiled under Cygwin.  
      3 
      4 It is also known that netperf has compiled using the Windows DDK.
      5 Here is a skeleton of the instructions to do so:
      6 
      7 Steps are:
      8 
      9 A) Install the Windows driver developer kit (if not already done).
     10 
     11 B) open a Cmd windows (i.e., a DOS box) for the target environment
     12    (target OS version; free vs checked build; x86, AMD64, or IA64).
     13    This is picked from the "Start\Developer Kits" path.
     14 
     15 C) enter the src\NetPerfDir directory
     16 
     17 D) Edit sources to enable any desired optional features (eg
     18    -DWANT_HISTOGRAM) or to remove features which your version of
     19    Windows might not support (eg -DHAVE_STRUCT_SOCKADDR_STORAGE)
     20 
     21 E) while still in the src\NetperfDir directory type "build /cD".
     22 
     23 F) Repeat steps C through E in src\NetServerDir
     24 
     25 G) the target files will be in a directory like:
     26    NetPerfDir\objchk_wnet_IA64\IA64, NetServerDir\objchk_wnet_IA64\IA64
     27    NetPerfDir\objfre_wnet_x86\i386, or NetPerfDir\objfre_wnet_AMD64\amd64
     28 
     29 NOTE: If any components of the path (ie the full names of the files,
     30 including parent directories) contain spaces (eg "My Documents"),
     31 build will charge off into the weeds.
     32 
     33 If you do not want the corresponding features, edit the sources files
     34 and remove the -Ds for WANT_DEMO, WANT_HISTOGRAM and/or WANT_INTERVALS
     35 
     36 And if that weren't enough, it is also known that netperf has been
     37 compiled using MS Visual Studio 2003.  Here are the instructions from
     38 the person who made that work (See Authors):
     39 
     40 1.  Under the PROJECT tab, <project name> PROPERTIES, LINKER folder,
     41 Select COMMAND LINE and add WS2_32.lib in the whitespace labeled
     42 Additional Options:
     43  
     44  
     45 2.  Under the PROJECT tab, <project name> PROPERTIES, C/C++ foleder,
     46 Select Preprocessor, On the right, add DO_IPV6; at the end of the
     47 Preprocessor Definitions whitespace.
     48  
     49  
     50 He goes on to say:
     51  
     52 NOTE: WHEN COMPLING NETSERVER, it works, but I got issued the
     53 foillowing warnigns in my build:
     54  
     55 
     56 ------ Rebuild All started: Project: netserver, Configuration: Debug Win32 ------
     57 
     58 Deleting intermediate files and output files for project 'netserver', configuration 'Debug|Win32'.
     59 
     60 Compiling...
     61 
     62 nettest_bsd.c
     63 
     64 g:\Program Files\netperf\netperf-2.4.1rc1\src\nettest_bsd.c(846) : warning C4267: 'function' : conversion from 'size_t' to 'int', possible loss of data
     65 
     66 g:\Program Files\netperf\netperf-2.4.1rc1\src\nettest_bsd.c(1303) : warning C4267: 'function' : conversion from 'size_t' to 'int', possible loss of data
     67 
     68 g:\Program Files\netperf\netperf-2.4.1rc1\src\nettest_bsd.c(2020) : warning C4267: 'function' : conversion from 'size_t' to 'int', possible loss of data
     69 
     70 g:\Program Files\netperf\netperf-2.4.1rc1\src\nettest_bsd.c(5080) : warning C4267: 'function' : conversion from 'size_t' to 'int', possible loss of data
     71 
     72 g:\Program Files\netperf\netperf-2.4.1rc1\src\nettest_bsd.c(5715) : warning C4267: 'function' : conversion from 'size_t' to 'int', possible loss of data
     73 
     74 g:\Program Files\netperf\netperf-2.4.1rc1\src\nettest_bsd.c(6591) : warning C4267: 'function' : conversion from 'size_t' to 'int', possible loss of data
     75 
     76 g:\Program Files\netperf\netperf-2.4.1rc1\src\nettest_bsd.c(8013) : warning C4267: 'function' : conversion from 'size_t' to 'int', possible loss of data
     77 
     78 g:\Program Files\netperf\netperf-2.4.1rc1\src\nettest_bsd.c(11123) : warning C4267: 'function' : conversion from 'size_t' to 'int', possible loss of data
     79 
     80 netsh.c
     81 
     82 netserver.c
     83 
     84 g:\Program Files\netperf\netperf-2.4.1rc1\src\netserver.c(457) : warning C4267: 'function' : conversion from 'size_t' to 'int', possible loss of data
     85 
     86 netlib.c
     87 
     88 g:\Program Files\netperf\netperf-2.4.1rc1\src\netlib.c(2470) : warning C4267: 'function' : conversion from 'size_t' to 'int', possible loss of data
     89 
     90 g:\Program Files\netperf\netperf-2.4.1rc1\src\netlib.c(2480) : warning C4267: 'function' : conversion from 'size_t' to 'int', possible loss of data
     91 
     92 netcpu_ntperf.c
     93 
     94 inet_ntop.c
     95 
     96 Generating Code...
     97 
     98 Linking...
     99 
    100 Build log was saved at "file://g:\Documents and Settings\Administrator\My Documents\Visual Studio Projects\netserver\Debug\BuildLog.htm"
    101 
    102 netserver - 0 error(s), 11 warning(s)
    103 
    104  
    105 
    106 ---------------------- Done ----------------------
    107 
    108 Rebuild All: 1 succeeded, 0 failed, 0 skipped
    109 
    110