Mentions légales du service

Skip to content
Snippets Groups Projects
Commit 8dd01784 authored by Mathieu Faverge's avatar Mathieu Faverge
Browse files

Merge branch 'work/improve_ci' into 'master'

remove otf parser and update pipeline

See merge request !42
parents ee511e02 e1f3054e
No related branches found
No related tags found
1 merge request!42remove otf parser and update pipeline
Pipeline #830721 passed
Showing
with 15 additions and 2859 deletions
......@@ -14,7 +14,6 @@ before_script:
include:
- .gitlab/common.yml
- .gitlab/build.yml
- .gitlab/plugins.yml
- .gitlab/sonarqube.yml
- .gitlab/pages.yml
- .gitlab/coverity.yml
......@@ -37,3 +37,15 @@ build_vite_with_vulkan:
variables:
BUILD_OPTIONS: "-DUSE_VULKAN=ON"
NAME: vulkan
build_vite_with_all_plugins:
<<: *build_vite
variables:
BUILD_OPTIONS: "-DUSE_OPENGL=ON -DVITE_ENABLE_VBO=ON -DVITE_PLUGINS_TRACEINFOS=ON -DVITE_PLUGINS_MATRIX_VISUALIZER=ON -DVITE_PLUGINS_CRITICAL_PATH=ON -DVITE_PLUGINS_DISTRIBUTIONS=ON"
NAME: plugins
build_vite_with_otf_tau_parsers:
<<: *build_vite
variables:
BUILD_OPTIONS: "-DUSE_OPENGL=ON -DVITE_ENABLE_VBO=ON -DVITE_ENABLE_MT_PARSERS=ON -DVITE_ENABLE_TAU=ON -DVITE_ENABLE_OTF2=ON"
NAME: all_parsers
......@@ -18,3 +18,6 @@ endif()
option(VITE_ENABLE_WARNING "Enable warning messages" ON)
option(VITE_ENABLE_COVERAGE "Enable flags for coverage test" ON)
# Parser includes
set(OTF2_DIR "$ENV{OTF2_DIR}" CACHE STRING "")
set(TAU_DIR "$ENV{TAU_DIR}" CACHE STRING "")
---
.build_vite_plugin_template: &build_vite_plugin
stage: build
tags: ["docker"]
interruptible: true
artifacts:
name: vite_build_plugin_with_${NAME}
expire_in: 42 minutes
untracked: true
script:
- git submodule update --init --recursive
- mkdir -p build
- cd build
- cmake -DVITE_CI_BRANCH=branch
-C ../.gitlab/ci-test-initial-cache.cmake ..
- cmake .. -DCMAKE_EXPORT_COMPILE_COMMANDS=ON
-DUSE_OPENGL=ON -DVITE_ENABLE_VBO=ON ${BUILD_OPTIONS}
- make | tee ../vite-build-plugin-with-${NAME}.log
- make install | tee -a ../vite-build-plugin-with-${NAME}
only:
- merge_requests
- master@solverstack/vite
build_vite_plugin_with_traceinfos:
<<: *build_vite_plugin
variables:
BUILD_OPTIONS: "-DVITE_PLUGINS_TRACEINFOS=ON"
NAME: traceinfos
build_vite_plugin_with_matrixvisualizer:
<<: *build_vite_plugin
variables:
BUILD_OPTIONS: "-DVITE_PLUGINS_MATRIX_VISUALIZER=ON"
NAME: matrixvisualizer
build_vite_plugin_with_criticalpath:
<<: *build_vite_plugin
variables:
BUILD_OPTIONS: "-DVITE_PLUGINS_CRITICAL_PATH=ON"
NAME: criticalpath
build_vite_plugin_with_distributions:
<<: *build_vite_plugin
variables:
BUILD_OPTIONS: "-DVITE_PLUGINS_DISTRIBUTIONS=ON"
NAME: distributions
......@@ -164,8 +164,6 @@ cmake_dependent_option(VITE_DBG_MEMORY_TRACE
"Enable trace generation of memory usage (requires MEMORY_USAGE)." OFF "VITE_DBG_MEMORY_USAGE" OFF)
### Trace format options
option(VITE_ENABLE_OTF
"Enable the support of OTF file format." OFF)
option(VITE_ENABLE_OTF2
"Enable the support of OTF2 file format." OFF)
option(VITE_ENABLE_TAU
......@@ -243,10 +241,6 @@ if(VITE_ENABLE_VBO)
endif()
endif()
if(VITE_ENABLE_OTF)
find_package(OTF)
endif()
if(VITE_ENABLE_OTF2)
find_package(OTF2)
endif()
......
#
# Find the OTF libraries and include dir
#
# OTF_INCLUDE_DIR - Directories to include to use OTF
# OTF_LIBRARY - Files to link against to use OTF
# OTF_LIBRARY_DIR - Directories to link to use OTF
# OTF_FOUND - When false, don't try to use OTF
#
# OTF_DIR can be used to make it simpler to find the various include
# directories and compiled libraries when OTF was not installed in the
# usual/well-known directories (e.g. because you made an in tree-source
# compilation or because you installed it in an "unusual" directory).
# Just set OTF_DIR it to your specific installation directory
#
FIND_PROGRAM( OTF_CONFIG_EXE otfconfig
PATHS
/usr/bin
/usr/local/bin
${OTF_DIR}/bin
)
# Use otfconfig to find paths
IF( OTF_CONFIG_EXE )
EXECUTE_PROCESS( COMMAND "${OTF_CONFIG_EXE}" "--includes" OUTPUT_VARIABLE OTF_INCLUDE_DIR )
EXECUTE_PROCESS( COMMAND "${OTF_CONFIG_EXE}" "--libs" OUTPUT_VARIABLE OTF_LIBS )
ENDIF( OTF_CONFIG_EXE )
IF( OTF_LIBS ) # OTF_LIBS is "-Lpath -lotf {-lz}"
STRING( REGEX MATCHALL "([^\ ]+\ |[^\ ]+$)" ARG_LIST "${OTF_LIBS}" )
# MESSAGE( "list = ${ARG_LIST}" )
FOREACH( listVar ${ARG_LIST} )
IF( listVar MATCHES "-L.*" ) # Gets the lib path
STRING( REGEX REPLACE "-L" "" listVar "${listVar}" )
STRING( REGEX REPLACE " " "" listVar "${listVar}" )
# MESSAGE( "libpath = ${listVar}" )
LIST( APPEND OTF_LIBRARY_DIR ${listVar} )
ENDIF()
IF( listVar MATCHES "-lo.*" ) # only gets the otf lib. If only -l.*, we also gets -lz and it does not work
STRING( REGEX REPLACE "-l" "" listVar "${listVar}" )
STRING( REGEX REPLACE " " "" listVar "${listVar}" )
# MESSAGE( "libs = ${listVar}" )
LIST( APPEND OTF_LIBRARY ${listVar} )
ENDIF()
ENDFOREACH( listVar )
ENDIF( OTF_LIBS )
IF( OTF_INCLUDE_DIR )
# Remove the -I because cmake handles it
STRING( REGEX REPLACE "-I" "" OTF_INCLUDE_DIR "${OTF_INCLUDE_DIR}" )
IF( OTF_LIBRARY )
SET( OTF_FOUND "YES" )
MARK_AS_ADVANCED( OTF_DIR )
MARK_AS_ADVANCED( OTF_INCLUDE_DIR )
MARK_AS_ADVANCED( OTF_LIBRARY_DIR )
MARK_AS_ADVANCED( OTF_LIBRARY )
ENDIF( OTF_LIBRARY )
ENDIF( OTF_INCLUDE_DIR )
IF( NOT OTF_FOUND )
MESSAGE("OTF installation was not found. Please provide OTF_DIR:")
MESSAGE(" - through the GUI when working with ccmake, ")
MESSAGE(" - as a command line argument when working with cmake e.g. ")
MESSAGE(" cmake .. -DOTF_DIR:PATH=/usr/local/otf ")
MESSAGE("Note: the following message is triggered by cmake on the first ")
MESSAGE(" undefined necessary PATH variable (e.g. OTF_INCLUDE_DIR).")
MESSAGE(" Providing OTF_DIR (as above described) is probably the")
MESSAGE(" simplest solution unless you have a really customized/odd")
MESSAGE(" OTF installation...")
SET(OTF_DIR "" CACHE PATH "Root of OTF install tree." )
ENDIF( NOT OTF_FOUND )
Andreas Knuepfer <andreas.knuepfer AT tu-dresden.de>
Ronny Brendel <ronny.brendel AT tu-dresden.de>
Matthias Jurenz <matthias.jurenz AT tu-dresden.de>
Johannes Spazier <johannes.spazier AT tu-dresden.de>
Michael Heyde <michael.heyde AT tu-dresden.de>
Michael Kluge <michael.kluge AT tu-dresden.de>
Holger Mickler <holger.mickler AT tu-dresden.de>
Holger Brunst <holger.brunst AT tu-dresden.de>
Hartmut Mix <hartmut.mix AT tu-dresden.de>
Copyright (c) 2005-2010, ZIH, Technische Universitaet Dresden,
Federal Republic of Germany
All rights reserved.
Redistribution and use in source and binary forms, with or without modification,
are permitted provided that the following conditions are met:
- Redistributions of source code must retain the above copyright notice,
this list of conditions and the following disclaimer.
- Redistributions in binary form must reproduce the above copyright notice,
this list of conditions and the following disclaimer in the documentation and/or
other materials provided with the distribution.
- Neither the name of ZIH, TU Dresden nor the names of its contributors
may be used to endorse or promote products derived from this software without
specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
1.7catfish
- added second version of each record that contains a Key-Value
list to store individual data
- two new definition records which can be used to assign
attributes to processes
- two new snapshot record types: BeginCollopSnapshot,
BeginFileOpSnapshot
- API changes in EndFileOP record and SendSnapshot record because
of a flawed design
- added MPI-parallel version of otfmerge (otfmerge-mpi)
- otfprofile has a lite mode usefuly for highly parallel traces
- added option to otf[de]compress for specifying an output directory
- try to open files with 'O_NOATIME' for reading for increased speed
- try to handle ESTALE errors on NFS gracefully
- fixed OTF long format for several records
- fixed error handling in OTF reader
- fixed bug in otfshrink: a process number can be between
0 and 4294967296 now
1.6.5stingray
- enforce building position independent code (PIC) for the static
version of the OTF library
(so it's possible to add these objects to 3rd-party shared libraries)
- fixed and enhanced calculation of trace file size in otfinfo
- patched libtool:
- to avoid a bug in detection of the PGI 10 C++ compiler
- to detect IBM cross compilers on BlueGene/P
1.6.4stingray
- enhanced reading of process group definition records
1.6.3stingray
- fixed errors while reading the long OTF format
- updated python interface
1.6.2stingray
- fixed build errors on SUN Solaris with older GNU compilers
- fixed skewed arguments for OTF_WStream_writeEndFileOperation
- fixed progress calculation of otfprofile
- improved performance of otfprofile - especially handling of huge
tracefiles
1.6.1stingray
- added tool 'otfinfo' which can be used to get basic information of
a trace
- fix invalid pointer dereferences in OTF_stripFilename() and
OTF_File_open_zlevel()
- fixed doxygen documentation: add OTF_Handler.h to package
1.6stingray
- added tool 'otfshrink' which can be used to reduce a trace to
specified processes
- introduced an error variable (otf_errno, otf_strerr) that indicates
whether an error occurred or not
- fixed build dependencies of otfprofile
- added file operation types OTF_FILEOP_DUP, OTF_FILEOP_SYNC,
OTF_FILEOP_LOCK, and OTF_FILEOP_UNLOCK
- changed datatype for sent/received data volume of collective
operation begin/end records from 'uint32_t' to 'uint64_t'
- replaced usage of PATH_MAX by VT_PATH_MAX to avoid compile errors
1.5.2dolphin
- adapted python bindings to the current API (related to new record
types)
- if possible, otfprofile creates a PS-file
- fixed VPATH issue, when using python bindings
- fixed infinite recursion in python wrappers
pyOTF_COUNTER_VARTYPE_ISSIGNED and pyOTF_COUNTER_VARTYPE_ISUNSIGNED
1.5.1dolphin
- corrected package preparation for libtool
- build shared OTF library by default
- added version information to shared OTF library
1.5
- added file operation begin/end records
- added file operation types OTF_FILEOP_UNLINK, OTF_FILEOP_RENAME,
and OTF_FILEOP_OTHER
- use part of fileop for I/O flags (e.g. async, coll.)
- added collective operation begin/end records
- added remote memory access (RMA) operation records
- added marker records
1.4.1
- bugfix: removed included system headers inside 'extern "C" {}'
- bugfix: added configure check for functions 'fseeko()' and 'ftello()'
(these functions are not available on NEC SX platforms)
- bugfix in otfprofile: fixed infinite loop while drawing the
statistics for "top 50 of counters"
1.4
- added tool 'otfprofile' which can be used to generate CSV/Latex-
summaries about functions and P2P-messages of a trace
- build/install tool 'otfdump' by default
- install documentation
1.3.10
- added configure option '--[enable|disable]-binaries' to control
building of binaries (useful for multi-lib installations)
1.3.9
- fixed a portability bug: use own implementation of function
'strdup()'
- install a symbolic link 'otfdecompress' which is a synonym for
'otfcompress -d'
1.3.8
- generate OTF_inttypes.h during configuring which provides
the integer types like int64_t, int8_t,...
- shared OTF library will be linked with zlib, thus additionally
linking with zlib is not necessary
- bugfix in otf2vtf: correctly handle function leave records with id 0
1.3.7
- added example for using OTF_MasterControl to the doxygen documentation
- general improvement of the documentation
- improved autoconf's m4 scripts for cross buildings
1.3.6
- OTF_{Keywords.h,Parse.h,Platform.h} will no longer be installed
- added a compile time assert making sure the correct sizes of the
integer types ([u]int*_T)
- OTF is now Windows Visual Studio compatible
1.3.5
- added doxygen html documentation (docu/doxygen/html/index.html)
- fixed issue, where newlines were forbidden in def, stats and snaps
- added missing const to OTF_WStream_writeDefProcessGroup()
1.3.4
- bug fix in otfmerge: missing handling of DefCollectiveOperation
- bugfix in the documentation: overview.eps was broken
1.3.3
- fixed configure bug when searching for zlib on mac
- fixed bug in otfmerge, where the parameter "-n 0" caused an error
- added macros for specifing the data type of the counter
(OTF_COUNTER_VARTYPE_*)
- fixed otfconfig help text
- fixed bugs when using make dist
- python bindings disabled by default
- fixed configure bug, when using python bindings
- updated python bindings to support user defined counters
1.3.2
- added small functions to convert various variable types to common
counter type uint64_t. By this means, other types can be transported
as uint64_t and re-casted during reading.
This requires type information as additional counter property flags!
1.3.1
- if an fopen() call fails an error is now printed out every time
- OTF_READ_ERROR was introduced as an error return value for all
OTF_Reader_readXXX and OTF_RStream_readXXX functions. So (uint64_t)-1 is
now a reserved value, thus cannot be used as record limit.
The record limit always has to be smaller than or equal to
OTF_READ_MAXRECORDS.
- fixed issue in otfmerge, when using many streams
1.3
- integrated libtool for shared and static linkage to otf
- removed the possibility to include zlib symbols into otf
- fixed warning which appeared when not using zlib
- added python bindings to the otf library
1.2.25
- added doxygen html documentation (docu/doxygen/html/index.html)
1.2.24
- minor change in otf[de|un]compress
- strictly avoid writing of unsorted time stamps. Now, there will be an
error message and the corresponding buffer/stream will be disabled for
further writing. The write routines provide an invlaid return code but
currently there is no way to "repair" the buffer/stream after an
unsorted time stamp has been issued.
1.2.23
- removed libtool from autoconf to make configure much faster,
libtool was not used by automake anyhow
- removed some autoconf-generated files from CVS,
you might need to re-run 'sh bootstrap' yourself
1.2.22
- fixed inttypes/stdint bug
1.2.21
- fixed "--with-zlib-symbols"-bug
1.2.20
- added own inttypes definitions in case there is no on the platform
(necessary on NEC SX6)
1.2.19
- finished otfdump
- removed debug output in OTF_Reader.c
- re-introduced DefVersion record as read only
1.2.18
- added fwrite check for writting less bytes than expected
- re-write of otfdump. Now, it uses OTF read lib instead of plain
scanf. It produces nicer output and numbers records in read order.
1.2.17
- bugfixed parser (wrong treatment of unknown records)
- added FileGroupOperationSummary record
- changed FileOperation-, OpenFileSnapshot- and FileOperationSummary-
record NOTE: these records are still experimental
- included file operation records into all tools
NOTE: some parameters were added/changed
- the byte-parameter in the FileOperation record does now contain the
new position in the file file after the seek operation
- bugfixed otfmerge
1.2.16
- fixed a problem with comments in otfmerge
- 2 new records introduced (NOTE: these are experimental):
- OpenFileSnapshot
- FileOperationSummary
1.2.15
- 3 new records introduced (NOTE: these are experimental):
- DefFile
- DefFileGroup
- FileOperation
1.2.14
- do not linke with '-lz' if '--with-zlibsymbols' was specified
- added zlib include line to otflib/Makefile.am
1.2.13
- removed intel compiler warnings
- changed OTF_FILETYPE_*-macros
- fixed issues with OTF_getFilename()
1.2.12
- removed intel compiler warnings in otfmerge
- removed debug output
- fixed 64bit issue
1.2.11
- changed OTF_RETURN*( 0=success, !0 = error )
- added these macros to all internal functions and tools for better
consistency
- fixed various memoryleaks in otf and otfmerge
- added otfconfig to tools. otfconfig shows installationparameters
important for developers
- updated documentation
1.2.10
- bugfix: otfmerge does no longer accept traces with local streams
1.2.9
- changeable zlevel
- changeable zbuffersize
1.2.8
- allow suffix '.dylib' for zlib library file (from Mac OS X)
- removed configure warning
1.2.7
- added progress functions to OTF_RStream
- added a progress counter to otfmerge
1.2.6
- support shared libraries
1.2.5
- bugfix: correctly handle process groups with more than 1000 entries
1.2.4
- bugfix: zlib compression bug, wrong sanity check fixed
1.2.3
- bugfix: provided copy handlers returned wrong value
1.2.2
- important bugfix: definitionstream 0 was ignored since version 1.2
1.2.1
- added progress functions using read bytes instead of timestamps
1.2
- introduce transparent zlib compression
1.1.7
- really did the bugfix for 1.1.6, was missing for some reasons
1.1.6
- bugfix: correctly handle process groups with more than 1000 entries
1.1.5
- have UnknownRecord report handle incomplete records or additional
bytes at the end of a file.
1.1.4
- fixed a bug in OTF_Reader which might have caused the very first
time stamp of a trace to be not properly sorted
- introduced '--snapshots' and '--statistics' switches to do only
snapshots or statistics. for statistics a selective mode is allowed
which regards only some streams. By this means statistics can be
created in parallel by calling otfaux multiple times.
1.1.3
- fixed a minor bug in otfaux
1.1.2
- inverted return value of call-back handlers:
'0' is non-error, '!= 0' is regarded as an error, now!
(this makes OTF conform with the VTF3 scheme.)
1.1.1
- OTF_Reader now considers the return values of the handlers
- added OTF_VERBOSE, OTF_DEBUG macro for error treatment
- introduced 'UnknownRecord' handler which allows to catch
unknown record types
1.0
- initial version
(See specific OTF related information at the end of this file.)
Copyright (C) 1994, 1995, 1996, 1999, 2000, 2001, 2002 Free Software
Foundation, Inc.
This file is free documentation; the Free Software Foundation gives
unlimited permission to copy, distribute and modify it.
Basic Installation
==================
These are generic installation instructions. OTF-specific options can be
found at the end of this document.
The `configure' shell script attempts to guess correct values for
various system-dependent variables used during compilation. It uses
those values to create a `Makefile' in each directory of the package.
It may also create one or more `.h' files containing system-dependent
definitions. Finally, it creates a shell script `config.status' that
you can run in the future to recreate the current configuration, and a
file `config.log' containing compiler output (useful mainly for
debugging `configure').
It can also use an optional file (typically called `config.cache'
and enabled with `--cache-file=config.cache' or simply `-C') that saves
the results of its tests to speed up reconfiguring. (Caching is
disabled by default to prevent problems with accidental use of stale
cache files.)
If you need to do unusual things to compile the package, please try
to figure out how `configure' could check whether to do them, and mail
diffs or instructions to the address given in the `README' so they can
be considered for the next release. If you are using the cache, and at
some point `config.cache' contains results you don't want to keep, you
may remove or edit it.
The file `configure.ac' (or `configure.in') is used to create
`configure' by a program called `autoconf'. You only need
`configure.ac' if you want to change it or regenerate `configure' using
a newer version of `autoconf'.
The simplest way to compile this package is:
1. `cd' to the directory containing the package's source code and type
`./configure' to configure the package for your system. If you're
using `csh' on an old version of System V, you might need to type
`sh ./configure' instead to prevent `csh' from trying to execute
`configure' itself.
Running `configure' takes awhile. While running, it prints some
messages telling which features it is checking for.
2. Type `make' to compile the package.
3. Optionally, type `make check' to run any self-tests that come with
the package.
4. Type `make install' to install the programs and any data files and
documentation.
5. You can remove the program binaries and object files from the
source code directory by typing `make clean'. To also remove the
files that `configure' created (so you can compile the package for
a different kind of computer), type `make distclean'. There is
also a `make maintainer-clean' target, but that is intended mainly
for the package's developers. If you use it, you may have to get
all sorts of other programs in order to regenerate files that came
with the distribution.
Compilers and Options
=====================
Some systems require unusual options for compilation or linking that
the `configure' script does not know about. Run `./configure --help'
for details on some of the pertinent environment variables.
You can give `configure' initial values for configuration parameters
by setting variables in the command line or in the environment. Here
is an example:
./configure CC=c89 CFLAGS=-O2 LIBS=-lposix
*Note Defining Variables::, for more details.
Compiling For Multiple Architectures
====================================
You can compile the package for more than one kind of computer at the
same time, by placing the object files for each architecture in their
own directory. To do this, you must use a version of `make' that
supports the `VPATH' variable, such as GNU `make'. `cd' to the
directory where you want the object files and executables to go and run
the `configure' script. `configure' automatically checks for the
source code in the directory that `configure' is in and in `..'.
If you have to use a `make' that does not support the `VPATH'
variable, you have to compile the package for one architecture at a
time in the source code directory. After you have installed the
package for one architecture, use `make distclean' before reconfiguring
for another architecture.
Installation Names
==================
By default, `make install' will install the package's files in
`/usr/local/bin', `/usr/local/man', etc. You can specify an
installation prefix other than `/usr/local' by giving `configure' the
option `--prefix=PATH'.
You can specify separate installation prefixes for
architecture-specific files and architecture-independent files. If you
give `configure' the option `--exec-prefix=PATH', the package will use
PATH as the prefix for installing programs and libraries.
Documentation and other data files will still use the regular prefix.
In addition, if you use an unusual directory layout you can give
options like `--bindir=PATH' to specify different values for particular
kinds of files. Run `configure --help' for a list of the directories
you can set and what kinds of files go in them.
If the package supports it, you can cause programs to be installed
with an extra prefix or suffix on their names by giving `configure' the
option `--program-prefix=PREFIX' or `--program-suffix=SUFFIX'.
Optional Features
=================
Some packages pay attention to `--enable-FEATURE' options to
`configure', where FEATURE indicates an optional part of the package.
They may also pay attention to `--with-PACKAGE' options, where PACKAGE
is something like `gnu-as' or `x' (for the X Window System). The
`README' should mention any `--enable-' and `--with-' options that the
package recognizes.
For packages that use the X Window System, `configure' can usually
find the X include and library files automatically, but if it doesn't,
you can use the `configure' options `--x-includes=DIR' and
`--x-libraries=DIR' to specify their locations.
Specifying the System Type
==========================
There may be some features `configure' cannot figure out
automatically, but needs to determine by the type of machine the package
will run on. Usually, assuming the package is built to be run on the
_same_ architectures, `configure' can figure that out, but if it prints
a message saying it cannot guess the machine type, give it the
`--build=TYPE' option. TYPE can either be a short name for the system
type, such as `sun4', or a canonical name which has the form:
CPU-COMPANY-SYSTEM
where SYSTEM can have one of these forms:
OS KERNEL-OS
See the file `config.sub' for the possible values of each field. If
`config.sub' isn't included in this package, then this package doesn't
need to know the machine type.
If you are _building_ compiler tools for cross-compiling, you should
use the `--target=TYPE' option to select the type of system they will
produce code for.
If you want to _use_ a cross compiler, that generates code for a
platform different from the build platform, you should specify the
"host" platform (i.e., that on which the generated programs will
eventually be run) with `--host=TYPE'.
Sharing Defaults
================
If you want to set default values for `configure' scripts to share,
you can create a site shell script called `config.site' that gives
default values for variables like `CC', `cache_file', and `prefix'.
`configure' looks for `PREFIX/share/config.site' if it exists, then
`PREFIX/etc/config.site' if it exists. Or, you can set the
`CONFIG_SITE' environment variable to the location of the site script.
A warning: not all `configure' scripts look for a site script.
Defining Variables
==================
Variables not defined in a site shell script can be set in the
environment passed to `configure'. However, some packages may run
configure again during the build, and the customized values of these
variables may be lost. In order to avoid this problem, you should set
them in the `configure' command line, using `VAR=value'. For example:
./configure CC=/usr/local2/bin/gcc
will cause the specified gcc to be used as the C compiler (unless it is
overridden in the site shell script).
`configure' Invocation
======================
`configure' recognizes the following options to control how it
operates.
`--help'
`-h'
Print a summary of the options to `configure', and exit.
`--version'
`-V'
Print the version of Autoconf used to generate the `configure'
script, and exit.
`--cache-file=FILE'
Enable the cache: use and save the results of the tests in FILE,
traditionally `config.cache'. FILE defaults to `/dev/null' to
disable caching.
`--config-cache'
`-C'
Alias for `--cache-file=config.cache'.
`--quiet'
`--silent'
`-q'
Do not print messages saying which checks are being made. To
suppress all normal output, redirect it to `/dev/null' (any error
messages will still be shown).
`--srcdir=DIR'
Look for the package's source code in directory DIR. Usually
`configure' can determine that directory automatically.
`configure' also accepts some other, not widely useful, options. Run
`configure --help' for more details.
OTF specific `configure' Examples
=================================
The '--with-vtf3' option allows to specify VTF3 include path and library path.
By default '../vtf3' is searched for headers & library unless '--without-vtf3'
is specified. Additional options '--with-vtf3-lib-dir' and
'--with-vtf3-include-dir' allow to give headers location and library location
separately.
Further options '--with-debug' and '--with-verbose' make the OTF library and
tools perform stricter checks and produce more verbose output (warnings/errors)
on stderr respectively.
On some platforms an error like "dereferencing type-punned pointer will break
strict-aliasing rules" may occur. If that is the case add "-fno-strict-aliasing"
(thats the parameter for gcc. It may be a bit different for other compilers)
to the CFLAGS and CXXFLAGS.
The '--with-zlib' option (default) enables tranparent compression/decompression
by ZLib, '--without-zlib' disables this feature. If '--with-zlibsymbols' is
specified in addition, all symbols from libz.a are directly included into
libotf.a. This makes it unnecessary to append '-lz' to another programs linker
list. However, it might cause collisions if '-lz' is present with the other
program already.
Note that '--with-zlibsymbols' is not compatible with shared libraries. The option
'--disable-shared' is necessary when ZLib symbols are included into OTF library.
Installing 32-bit and 64-bit Libraries Concurrently
===================================================
Some platforms support 32-bit and 64-bit executables and libraries. Please chose
one flavor as default and set appropriate options via './configure'. Then do
'make' and 'make install'. This will have everything built and installed with
default flavor: library, header files and tools.
In order to have alternative library flavor created, re-run './configure' with
alternative options and do 'make' but NOT 'make install'. Rather, copy the
library manually to the desired location. Make sure not to overwrite the default
library. Header files and tools are not affected by this issue because always
only the default version will be necessary.
Example configure Command Line
==============================
example on Linux:
%> ./configure CFLAGS="-O3 -g -Wall -Werror" CXXFLAGS="-O3 -g -Wall -Werror"
--with-vtf3=$HOME/prog/vtf/ --with-zlib --with-zlibsymbols --disable-shared
--prefix=$HOME
example on SGI Irix:
%> ./configure CC=cc CFLAGS="-O3 -n32 -g -ansiW" --with-vtf3=$HOME/prog/vtf/
example on NEC SX6i front end with cross compiler
%> ./configure AR="sxar" RANLIB="sxar st" CC="sxcc -sx5" CXX="sxc++ -sx5"
--host=sx6-nec-superux14.1 --build=i686-pc-linux-gnu
--without-zlib --prefix=/home_sx6i/knuepfe/
Copyright (c) 2005-2010, ZIH, Technische Universitaet Dresden,
Federal Republic of Germany
All rights reserved.
Redistribution and use in source and binary forms, with or without modification,
are permitted provided that the following conditions are met:
- Redistributions of source code must retain the above copyright notice,
this list of conditions and the following disclaimer.
- Redistributions in binary form must reproduce the above copyright notice,
this list of conditions and the following disclaimer in the documentation and/or
other materials provided with the distribution.
- Neither the name of ZIH, TU Dresden nor the names of its contributors
may be used to endorse or promote products derived from this software without
specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
ACLOCAL_AMFLAGS = -I config/m4
if AMHAVEPYTHONSWIG
OTFLIB_PY = otflib_py
else
OTFLIB_PY =
endif
SUBDIRS = \
otflib \
$(OTFLIB_PY) \
tools \
docu
EXTRA_DIST= \
otf_vc08.sln \
config/otf_get_version.sh \
tests/hello/Makefile \
tests/hello/hello_otf.c \
tests/otf_python/groups.py \
tests/otf_python/test_read.py \
tests/otf_python/test_read_qt.py \
tests/otf_python/test_write.py \
VERSION
This diff is collapsed.
Open Trace Format (OTF)
Detailed program analysis of massively parallel programs requires the recording of event based performance data during run-time and its visualisation with appropriate software tools like the Vampir framework.
The specification of a convenient and powerful trace file format has to fulfill a large number of requirements. It must allow an efficient collection of the event based performance data on a parallel program environment. On the other hand it has to provide a very fast and convenient access to the large amount of performance data and the corresponding event definitions by the analysis software tools.
To improve scalability for very large and massively parallel traces the Open Trace Format (OTF) has been developed at ZIH as a successor format to the Vampir Trace Format (VTF3).
The Open Trace Format makes use of a portable ASCII encoding. It distributes single traces to multiple so called streams with one or more files each. Merging of records from multiple files is done transparently by the OTF library. The number of possible streams is not limited by the number of available file handles.
The provided read/write library should be used as a convenient interface for third party software. The library supports efficient parallel and distributed access to trace data and offers selective reading access regarding arbitrary time intervals, process selection and record types.
Optional auxiliary information can assist this selective access.
The software package contains additional tools for trace date conversion or preparation.
More information can be found in the provided documentation.
OTF version is available as open source software under the BSD license, see 'COPYING'. The software comes as a source code package. It can be installed by using autoconf procedures, see 'INSTALL'.
- re-design OTF's error/success return values for handlers and functions
- ideas for future development
- add memory allocation/deallocation record types, how?
- have more semantics
- have less semantics
- add general purpose record types like (type/owner,key,value) with no semantics,
-- can be used freely, ignored if unknown,
-- key/value are either uint64_t, array of uint64_t or text
- add record to write wall clock time or epoch or year:month:day:hour:minute:second
with this, VT can record time at start and end of trace (request by Robert Henschel)
- create a (graphical) tool for changing definition records:
- grouping functions
- ...
- solve the problem that when only reading statistic records one cannot
determine the begin and end of a process !!!!!
You might introduce a new record for this. ( maybe like begin/endprocess
for events? )
You could also write a statistics comment record when the first event appears,
(in otfaux)
- add new statistics record type that gives the caller-callee relation
- only write this at the end of a trace run
- use asynchronous write (aio_write) and maybe even read for performance
# This is the VERSION file for OTF, describing the precise version of OTF in
# this distribution. The various components of the version number below are
# combined to form a single version number string.
# major, minor, and sub are generally combined in the form
# <major>.<minor>.<sub>. If sub is zero, then it is omitted.
major=1
minor=7
sub=0
# string is used for alpha, beta, or release tags. If it is non-empty, it will
# be appended to the version number.
#
# history of release tags:
# 0.2.* octopussy
# 0.3.*-1.1.* starfish
# 1.2.* pufferfish
# 1.3.* jellyfish
# 1.4.* shark
# 1.5.* dolphin
# 1.6.* stingray
# 1.7.* catfish
#
string=catfish
# library is used for the library versioning. These three numbers will be added
# to the name of the library (e.g. libotf.so.1.2.3).
# The meaning of these numbers is as follows (from left to right):
#
# current The number of the current API exported by the library.
# That must be incremented, if the API has changed.
# revision The implementation number of the most recent API
# exported by the library. A value of '0' means that this
# is the first implementation of the API.
# If any of the source for the library has changed,
# revision must be incremented.
# age The number of previous additional APIs supported by the
# library. If age is '2', the library can be linked into
# executables built with a release of the library that
# exported the current API number (current), or any of
# the preceding two APIs.
# If the new API is backward-compatible with the preceding
# release, age must be incremented. Otherwise, reset age
# to '0'.
library=1:0:0
m4_include(config/m4/acinclude.debug.m4)
m4_include(config/m4/acinclude.math.m4)
m4_include(config/m4/acinclude.mpi.m4)
m4_include(config/m4/acinclude.omp.m4)
m4_include(config/m4/acinclude.pydev.m4)
m4_include(config/m4/acinclude.swig.m4)
m4_include(config/m4/acinclude.swig_python.m4)
m4_include(config/m4/acinclude.vtf3.m4)
m4_include(config/m4/acinclude.verbose.m4)
m4_include(config/m4/acinclude.zlib.m4)
m4_include(config/m4/acarchive/ax_mpi.m4)
m4_include(config/m4/acarchive/ax_openmp.m4)
This diff is collapsed.
/* config.h.in. Generated from configure.in by autoheader. */
/* Define to 1 if you have the `asprintf' function. */
#undef HAVE_ASPRINTF
/* Define to 1 if you have the declaration of `O_NOATIME', and to 0 if you
don't. */
#undef HAVE_DECL_O_NOATIME
/* Define to 1 if you have the <dlfcn.h> header file. */
#undef HAVE_DLFCN_H
/* Define to 1 if you have the <fcntl.h> header file. */
#undef HAVE_FCNTL_H
/* Define to 1 if fseeko (and presumably ftello) exists and is declared. */
#undef HAVE_FSEEKO
/* Define to 1 if you have the `gettimeofday' function. */
#undef HAVE_GETTIMEOFDAY
/* Define to 1 if you have the <inttypes.h> header file. */
#undef HAVE_INTTYPES_H
/* Define to 1 if you have the <memory.h> header file. */
#undef HAVE_MEMORY_H
/* Define to 1 if you have the `snprintf' function. */
#undef HAVE_SNPRINTF
/* Define to 1 if you have the <stddef.h> header file. */
#undef HAVE_STDDEF_H
/* Define to 1 if you have the <stdint.h> header file. */
#undef HAVE_STDINT_H
/* Define to 1 if you have the <stdlib.h> header file. */
#undef HAVE_STDLIB_H
/* Define to 1 if you have the `strdup' function. */
#undef HAVE_STRDUP
/* Define to 1 if you have the <strings.h> header file. */
#undef HAVE_STRINGS_H
/* Define to 1 if you have the <string.h> header file. */
#undef HAVE_STRING_H
/* Define to 1 if you have the <sys/param.h> header file. */
#undef HAVE_SYS_PARAM_H
/* Define to 1 if you have the <sys/stat.h> header file. */
#undef HAVE_SYS_STAT_H
/* Define to 1 if you have the <sys/time.h> header file. */
#undef HAVE_SYS_TIME_H
/* Define to 1 if you have the <sys/types.h> header file. */
#undef HAVE_SYS_TYPES_H
/* Define to 1 if you have the `__va_copy' function. */
#undef HAVE_UNDERSCORE_VA_COPY
/* Define to 1 if you have the <unistd.h> header file. */
#undef HAVE_UNISTD_H
/* Define to 1 if you have the `vasprintf' function. */
#undef HAVE_VASPRINTF
/* Define to 1 if you have the `va_copy' function. */
#undef HAVE_VA_COPY
/* Define to 1 if you have the `vsnprintf' function. */
#undef HAVE_VSNPRINTF
/* */
#undef HAVE_VTF3
/* */
#undef HAVE_ZLIB
/* Define to the sub-directory in which libtool stores uninstalled libraries.
*/
#undef LT_OBJDIR
/* */
#undef OTF_DEBUG
/* */
#undef OTF_VERBOSE
/* Name of package */
#undef PACKAGE
/* Define to the address where bug reports for this package should be sent. */
#undef PACKAGE_BUGREPORT
/* Define to the full name of this package. */
#undef PACKAGE_NAME
/* Define to the full name and version of this package. */
#undef PACKAGE_STRING
/* Define to the one symbol short name of this package. */
#undef PACKAGE_TARNAME
/* Define to the home page for this package. */
#undef PACKAGE_URL
/* Define to the version of this package. */
#undef PACKAGE_VERSION
/* The size of `long', as computed by sizeof. */
#undef SIZEOF_LONG
/* Define to 1 if you have the ANSI C header files. */
#undef STDC_HEADERS
/* Version number of package */
#undef VERSION
/* Define to 1 to make fseeko visible on some hosts (e.g. glibc 2.2). */
#undef _LARGEFILE_SOURCE
/* Define to `unsigned int' if <sys/types.h> does not define. */
#undef size_t
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment