fdurastante
/
psblas3
mirror of https://github.com/sfilippone/psblas3.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

psblas-testpre:

Browse Source 
psb/base/modules/X_base_vect_mod.p90

Added doxygen refs.
psblas3-type-indexed
Salvatore Filippone 13 years ago
parent e2ae7d75c3
commit 9cafbc190c
2 changed files with 907 additions and 847 deletions
Show Stats Download Patch File Download Diff File

151
base/modules/psb_indx_map_mod.f90
Unescape Escape View File

@ -34,70 +34,95 @@
! package: psb_indx_map_mod ! package: psb_indx_map_mod
! Defines the PSB_INDX_MAP class. ! Defines the PSB_INDX_MAP class.
! !
! In particular, besides the communicator, it contains the data relevant
! to the following queries:
! 1. How many global rows/columns?
! 2. How many local rows/columns?
! 3. Convert between local and global indices
! 4. Add to local indices.
! 5. Find (one of) the owner(s) of a given index
! 6. Query the indx state.
! 7. Does the dynamic class support extensions of the rows? I.e., can
! it have overlap? for instance, the BLOCK cannot, it would run afoul
! of the glob_to_loc translation.
!
! Checking for the existence of overlap is very expensive, thus
! it is done at build time (for extended-halo cases it can be inferred from
! the construction process).
!
! 1. Each global index I is owned by at least one process;
!
! 2. On each process, indices from 1 to N_ROW (desc%indxmap%get_lr())
! are locally owned; the value of N_ROW can be determined upon allocation
! based on the index distribution (see also the interface to CDALL).
!
! 3. If a global index is owned by more than one process, we have an OVERLAP
! in which case the sum of all the N_ROW values is greater than the total
! size of the index space;
!
! 4. During the buildup of the descriptor, according to the user specified
! stencil, we also take notice of indices that are not owned by the current
! process, but whose value is needed to proceed with the computation; these
! form the HALO of the current process. Halo indices are assigned local indices
! from N_ROW+1 to N_COL (inclusive).
!
! 5. The upper bound N_COL moves during the descriptor build process (see CDINS).
!
!
! This is the base version of the class; as such, it only contains
! methods for getting/setting the common attributes, whereas
! the index translation methods are only implemented at the derived
! class level.
! Note that the INIT method is defined in the derived methods, and
! is specialized for them; a better solution would have to have
! a constructor for each specific class, with the name of the class,
! but this is not yet working on many compilers, most notably GNU.
!
! Note: the CLONE method was implemented as a workaround for a problem
! with SOURCE= allocation on GNU. Might be removed later on.
! !
! !
module psb_indx_map_mod module psb_indx_map_mod
use psb_const_mod use psb_const_mod
use psb_desc_const_mod use psb_desc_const_mod
!
!> \namespace psb_base_mod \class psb_indx_map
!! \brief Object to handle the mapping between global and local indices.
!!
!!
!! In particular, besides the communicator, it contains the data relevant
!! to the following queries:
!!
!! - How many global rows/columns?
!!
!! - How many local rows/columns?
!!
!! - Convert between local and global indices
!!
!! - Add to local indices.
!!
!! - Find (one of) the owner(s) of a given index
!!
!! - Query the indx state.
!!
!! - Does the dynamic class support extensions of the rows? I.e., can
!! it have overlap? for instance, the BLOCK cannot, it would run afoul
!! of the glob_to_loc translation.
!!
!! Checking for the existence of overlap is very expensive, thus
!! it is done at build time (for extended-halo cases it can be inferred from
!! the construction process).
!!
!! The object can be in the NULL, BUILD or ASSEMBLED state.
!!
!! Rules/constraints:
!!
!! 1. Each global index I is owned by at least one process;
!!
!! 2. On each process, indices from 1 to N_ROW (desc%indxmap%get_lr())
!! are locally owned; the value of N_ROW can be determined upon allocation
!! based on the index distribution (see also the interface to CDALL).
!!
!! 3. If a global index is owned by more than one process, we have an OVERLAP
!! in which case the sum of all the N_ROW values is greater than the total
!! size of the index space;
!!
!! 4. During the buildup of the descriptor, according to the user specified
!! stencil, we also take notice of indices that are not owned by the current
!! process, but whose value is needed to proceed with the computation; these
!! form the HALO of the current process. Halo indices are assigned local indices
!! from N_ROW+1 to N_COL (inclusive).
!!
!! 5. The upper bound N_COL moves during the descriptor build process (see CDINS).
!!
!!
!! This is the base version of the class; as such, it only contains
!! methods for getting/setting the common attributes, whereas
!! the index translation methods are only implemented at the derived
!! class level.
!! Note that the INIT method is defined in the derived methods, and
!! is specialized for them; a better solution would have to have
!! a constructor for each specific class, with the name of the class,
!! but this is not yet working on many compilers, most notably GNU.
!!
!! Note: the CLONE method was implemented as a workaround for a problem
!! with SOURCE= allocation on GNU. Might be removed later on.
!!
type :: psb_indx_map type :: psb_indx_map
!> State of the map
integer(psb_ipk_) :: state = psb_desc_null_ integer(psb_ipk_) :: state = psb_desc_null_
!> Communication context
integer(psb_mpik_) :: ictxt = -1 integer(psb_mpik_) :: ictxt = -1
!> MPI communicator
integer(psb_mpik_) :: mpic = -1 integer(psb_mpik_) :: mpic = -1
!> Number of global rows
integer(psb_ipk_) :: global_rows = -1 integer(psb_ipk_) :: global_rows = -1
!> Number of global columns
integer(psb_ipk_) :: global_cols = -1 integer(psb_ipk_) :: global_cols = -1
!> Number of local rows
integer(psb_ipk_) :: local_rows = -1 integer(psb_ipk_) :: local_rows = -1
!> Number of local columns
integer(psb_ipk_) :: local_cols = -1 integer(psb_ipk_) :: local_cols = -1
!> A pointer to the user-defined parts subroutine
procedure(psb_parts), nopass, pointer :: parts => null() procedure(psb_parts), nopass, pointer :: parts => null()
!> The global vector assigning indices to processes, temp copy
integer(psb_ipk_), allocatable :: tempvg(:) integer(psb_ipk_), allocatable :: tempvg(:)
!> Reserved for future use.
integer(psb_ipk_), allocatable :: oracle(:,:) integer(psb_ipk_), allocatable :: oracle(:,:)
contains contains
@ -171,6 +196,24 @@ module psb_indx_map_mod
& base_g2lv2_ins, base_init_vl, base_is_null,& & base_g2lv2_ins, base_init_vl, base_is_null,&
& base_row_extendable, base_clone & base_row_extendable, base_clone
!> Function: psb_indx_map_fnd_owner
!! \memberof psb_indx_map
!! \brief Find the process owning indices
!!
!! Given a list of indices IDX, return the processes owning
!! them. The base class provides the default implementation,
!! which is simply aggregating the requests and converting on
!! each proces who then builds its part of the solution.
!! This implies that in general this routine is a
!! synchronization point; for some derived classes it is
!! possible to answer locally, but this should not be relied
!! upon.
!!
!! \param idx(:) The set of indices (local to each process)
!! \param iprc(:) The processes owning them
!! \param info return code.
!!
interface interface
subroutine psb_indx_map_fnd_owner(idx,iprc,idxmap,info) subroutine psb_indx_map_fnd_owner(idx,iprc,idxmap,info)
import :: psb_indx_map, psb_ipk_ import :: psb_indx_map, psb_ipk_
@ -185,6 +228,9 @@ module psb_indx_map_mod
contains contains
!>
!! \memberof psb_indx_map
!! \brief Print a descriptive name
function base_get_fmt() result(res) function base_get_fmt() result(res)
implicit none implicit none
character(len=5) :: res character(len=5) :: res
@ -319,6 +365,9 @@ contains
end subroutine base_set_mpic end subroutine base_set_mpic
!>
!! \memberof psb_indx_map
!! \brief Is the class capable of having overlapped rows?
function base_row_extendable() result(val) function base_row_extendable() result(val)
implicit none implicit none
logical :: val logical :: val
@ -387,9 +436,9 @@ contains
end function base_sizeof end function base_sizeof
! !!!!!!!!!!!!!!!! !>
! !! \memberof psb_indx_map
! !!!!!!!!!!!!!!!! !! \brief Local to global, scalar, in place
subroutine base_l2gs1(idx,idxmap,info,mask,owned) subroutine base_l2gs1(idx,idxmap,info,mask,owned)
use psb_error_mod use psb_error_mod
implicit none implicit none

47
docs/doxypsb
Unescape Escape View File

@ -1,14 +1,14 @@
# Doxyfile 1.7.5 # Doxyfile 1.7.5
# This file describes the settings to be used by the documentation system # This file describes the settings to be used by the documentation system
# doxygen (www.doxygen.org) for a project # doxygen (www.doxygen.org) for a project.
# #
# All text after a hash (#) is considered a comment and will be ignored # All text after a hash (#) is considered a comment and will be ignored.
# The format is: # The format is:
# TAG = value [value, ...] # TAG = value [value, ...]
# For lists items can also be appended using: # For lists items can also be appended using:
# TAG += value [value, ...] # TAG += value [value, ...]
# Values that contain spaces should be placed between quotes (" ") # Values that contain spaces should be placed between quotes (" ").
#--------------------------------------------------------------------------- #---------------------------------------------------------------------------
# Project related configuration options # Project related configuration options
@ -535,7 +535,8 @@ SHOW_DIRECTORIES = YES
SHOW_FILES = YES SHOW_FILES = YES
# Set the SHOW_NAMESPACES tag to NO to disable the generation of the # Set the SHOW_NAMESPACES tag to NO to disable the generation of the
# Namespaces page. This will remove the Namespaces entry from the Quick Index # Namespaces page.
# This will remove the Namespaces entry from the Quick Index
# and from the Folder Tree View (if specified). The default is YES. # and from the Folder Tree View (if specified). The default is YES.
SHOW_NAMESPACES = YES SHOW_NAMESPACES = YES
@ -628,7 +629,10 @@ WARN_LOGFILE =
# directories like "/usr/src/myproject". Separate the files or directories # directories like "/usr/src/myproject". Separate the files or directories
# with spaces. # with spaces.
INPUT = ../base ../prec ../krylov ../util INPUT = ../base \
../prec \
../krylov \
../util
# This tag can be used to specify the character encoding of the source files # This tag can be used to specify the character encoding of the source files
# that doxygen parses. Internally doxygen uses the UTF-8 encoding, which is # that doxygen parses. Internally doxygen uses the UTF-8 encoding, which is
@ -716,14 +720,17 @@ IMAGE_PATH =
# by executing (via popen()) the command <filter> <input-file>, where <filter> # by executing (via popen()) the command <filter> <input-file>, where <filter>
# is the value of the INPUT_FILTER tag, and <input-file> is the name of an # is the value of the INPUT_FILTER tag, and <input-file> is the name of an
# input file. Doxygen will then use the output that the filter program writes # input file. Doxygen will then use the output that the filter program writes
# to standard output. If FILTER_PATTERNS is specified, this tag will be # to standard output.
# If FILTER_PATTERNS is specified, this tag will be
# ignored. # ignored.
INPUT_FILTER = INPUT_FILTER =
# The FILTER_PATTERNS tag can be used to specify filters on a per file pattern # The FILTER_PATTERNS tag can be used to specify filters on a per file pattern
# basis. Doxygen will compare the file name with each pattern and apply the # basis.
# filter if there is a match. The filters are a list of the form: # Doxygen will compare the file name with each pattern and apply the
# filter if there is a match.
# The filters are a list of the form:
# pattern=filter (like *.cpp=my_cpp_filter). See INPUT_FILTER for further # pattern=filter (like *.cpp=my_cpp_filter). See INPUT_FILTER for further
# info on how filters are used. If FILTER_PATTERNS is empty or if # info on how filters are used. If FILTER_PATTERNS is empty or if
# non of the patterns match the file name, INPUT_FILTER is applied. # non of the patterns match the file name, INPUT_FILTER is applied.
@ -781,7 +788,8 @@ REFERENCES_RELATION = YES
# If the REFERENCES_LINK_SOURCE tag is set to YES (the default) # If the REFERENCES_LINK_SOURCE tag is set to YES (the default)
# and SOURCE_BROWSER tag is set to YES, then the hyperlinks from # and SOURCE_BROWSER tag is set to YES, then the hyperlinks from
# functions in REFERENCES_RELATION and REFERENCED_BY_RELATION lists will # functions in REFERENCES_RELATION and REFERENCED_BY_RELATION lists will
# link to the source code. Otherwise they will link to the documentation. # link to the source code.
# Otherwise they will link to the documentation.
REFERENCES_LINK_SOURCE = YES REFERENCES_LINK_SOURCE = YES
@ -846,7 +854,7 @@ HTML_FILE_EXTENSION = .html
# The HTML_HEADER tag can be used to specify a personal HTML header for # The HTML_HEADER tag can be used to specify a personal HTML header for
# each generated HTML page. If it is left blank doxygen will generate a # each generated HTML page. If it is left blank doxygen will generate a
# standard header. Note that when using a custom header you are responsible # standard header. Note that when using a custom header you are responsible
# for the proper inclusion of any scripts and style sheets that doxygen # for the proper inclusion of any scripts and style sheets that doxygen
# needs, which is dependent on the configuration options used. # needs, which is dependent on the configuration options used.
# It is adviced to generate a default header using "doxygen -w html # It is adviced to generate a default header using "doxygen -w html
# header.html footer.html stylesheet.css YourConfigFile" and then modify # header.html footer.html stylesheet.css YourConfigFile" and then modify
@ -862,10 +870,9 @@ HTML_HEADER =
HTML_FOOTER = HTML_FOOTER =
# If the HTML_TIMESTAMP tag is set to YES then the generated HTML # If the HTML_TIMESTAMP tag is set to YES then the generated HTML documentation will contain the timesstamp.
# documentation will contain the timesstamp.
HTML_TIMESTAMP = YES HTML_TIMESTAMP = NO
# The HTML_STYLESHEET tag can be used to specify a user-defined cascading # The HTML_STYLESHEET tag can be used to specify a user-defined cascading
# style sheet that is used by each HTML page. It can be used to # style sheet that is used by each HTML page. It can be used to
@ -1065,7 +1072,7 @@ QHP_SECT_FILTER_ATTRS =
QHG_LOCATION = QHG_LOCATION =
# If the GENERATE_ECLIPSEHELP tag is set to YES, additional index files # If the GENERATE_ECLIPSEHELP tag is set to YES, additional index files
# will be generated, which together with the HTML files, form an Eclipse help # will be generated, which together with the HTML files, form an Eclipse help
# plugin. To install this plugin and make it available under the help contents # plugin. To install this plugin and make it available under the help contents
# menu in Eclipse, the contents of the directory containing the HTML and XML # menu in Eclipse, the contents of the directory containing the HTML and XML
# files needs to be copied into the plugins directory of eclipse. The name of # files needs to be copied into the plugins directory of eclipse. The name of
@ -1416,8 +1423,10 @@ GENERATE_PERLMOD = NO
PERLMOD_LATEX = NO PERLMOD_LATEX = NO
# If the PERLMOD_PRETTY tag is set to YES the Perl module output will be # If the PERLMOD_PRETTY tag is set to YES the Perl module output will be
# nicely formatted so it can be parsed by a human reader. This is useful # nicely formatted so it can be parsed by a human reader.
# if you want to understand what is going on. On the other hand, if this # This is useful
# if you want to understand what is going on.
# On the other hand, if this
# tag is set to NO the size of the Perl module output will be much smaller # tag is set to NO the size of the Perl module output will be much smaller
# and Perl will parse it just the same. # and Perl will parse it just the same.
@ -1504,9 +1513,11 @@ SKIP_FUNCTION_MACROS = YES
# Optionally an initial location of the external documentation # Optionally an initial location of the external documentation
# can be added for each tagfile. The format of a tag file without # can be added for each tagfile. The format of a tag file without
# this location is as follows: # this location is as follows:
# TAGFILES = file1 file2 ... #
# TAGFILES = file1 file2 ...
# Adding location for the tag files is done as follows: # Adding location for the tag files is done as follows:
# TAGFILES = file1=loc1 "file2 = loc2" ... #
# TAGFILES = file1=loc1 "file2 = loc2" ...
# where "loc1" and "loc2" can be relative or absolute paths or # where "loc1" and "loc2" can be relative or absolute paths or
# URLs. If a location is present for each tag, the installdox tool # URLs. If a location is present for each tag, the installdox tool
# does not have to be run to correct the links. # does not have to be run to correct the links.

Write Preview
Loading…
Cancel
Save