Task_

to 06 jun 2002

Summary
Task_to 06 jun 2002
Variables
taskidCOMP_1D DIAG E1 E2
prionumPriority value for the factorization
prionum2Priority value for solve steps
cblknumAttached column block
bloknumAttached block
ftgtcntNumber of fan-in targets
ctrbcntTotal number of contributions
btagptrsi non local, pointeur sur la structure (NB reception)
indnum
tasknextchainage des E1 ou E2, si fin = -1 => liberer les btagptr
taskmstrMaster task for E1 or E2 tasks
threadidIndex of the bubble which contains the task
candThread candidate in static version
fcandnumFirst thread candidate
lcandnumLast thread candidate
idGlobal cblknum of the attached column block
SolverCblk_
Variables
strideColumn block stride
colorColor of column block (PICL trace)
procdiagProcessor owner of diagonal block
cblkdiagColumn block owner of diagonal block
SolverBlok_
Variables
coefindIndex in coeftab
SolverMatrix_
Variables
symbmtxSymbolic matrix
cscmtxCompress Sparse Column matrix
coefnbrNumber of coefficients
ftgtnbrNumber of fanintargets
ftgtcntNumber of fanintargets to receive
coefmaxWorking block max size (cblk coeff 1D)
bpftmaxMaximum of block size for btag to receive
cpftmaxMaximum of block size for ftgt to receive
nbftmaxMaximum block number in ftgt
arftmaxMaximum block area in ftgt
clustnumcurrent processor number
clustnbrnumber of processors
procnbrNumber of physical processor used
thrdnbrNumber of local computation threads
bublnbrNumber of local computation threads
btagnbrNumber of Blocktargets
btgsnbrNumber of Blocktargets to send
btgrnbrNumber of Blocktargets to recv
bcofnbr
indnbr
tasknbrNumber of Tasks
ttsktabTask access vector by thread
ttsknbrNumber of tasks by thread
proc2clustproc -> cluster
gridldimDimensions of the virtual processors
gridcdimgrid if dense end block
updovctUpDown vector

Variables

taskid

INT taskid

COMP_1D DIAG E1 E2

prionum

INT prionum

Priority value for the factorization

prionum2

INT prionum2

Priority value for solve steps

cblknum

INT cblknum

Attached column block

bloknum

INT bloknum

Attached block

ftgtcnt

Number of fan-in targets

ctrbcnt

Total number of contributions

btagptr

si non local, pointeur sur la structure (NB reception)

indnum

INT indnum

tasknext

INT tasknext

chainage des E1 ou E2, si fin = -1 => liberer les btagptr

taskmstr

INT taskmstr

Master task for E1 or E2 tasks

threadid

INT threadid

Index of the bubble which contains the task

cand

INT cand

Thread candidate in static version

fcandnum

INT fcandnum

First thread candidate

lcandnum

INT lcandnum

Last thread candidate

id

INT id

Global cblknum of the attached column block

SolverCblk_

Summary
Variables
strideColumn block stride
colorColor of column block (PICL trace)
procdiagProcessor owner of diagonal block
cblkdiagColumn block owner of diagonal block

Variables

stride

INT stride

Column block stride

color

INT color

Color of column block (PICL trace)

procdiag

INT procdiag

Processor owner of diagonal block

cblkdiag

INT cblkdiag

Column block owner of diagonal block

SolverBlok_

Summary
Variables
coefindIndex in coeftab

Variables

coefind

INT coefind

Index in coeftab

SolverMatrix_

Summary
Variables
symbmtxSymbolic matrix
cscmtxCompress Sparse Column matrix
coefnbrNumber of coefficients
ftgtnbrNumber of fanintargets
ftgtcntNumber of fanintargets to receive
coefmaxWorking block max size (cblk coeff 1D)
bpftmaxMaximum of block size for btag to receive
cpftmaxMaximum of block size for ftgt to receive
nbftmaxMaximum block number in ftgt
arftmaxMaximum block area in ftgt
clustnumcurrent processor number
clustnbrnumber of processors
procnbrNumber of physical processor used
thrdnbrNumber of local computation threads
bublnbrNumber of local computation threads
btagnbrNumber of Blocktargets
btgsnbrNumber of Blocktargets to send
btgrnbrNumber of Blocktargets to recv
bcofnbr
indnbr
tasknbrNumber of Tasks
ttsktabTask access vector by thread
ttsknbrNumber of tasks by thread
proc2clustproc -> cluster
gridldimDimensions of the virtual processors
gridcdimgrid if dense end block
updovctUpDown vector

Variables

symbmtx

SymbolMatrix symbmtx

Symbolic matrix

cscmtx

CscMatrix cscmtx

Compress Sparse Column matrix

coefnbr

INT coefnbr

Number of coefficients

ftgtnbr

INT ftgtnbr

Number of fanintargets

ftgtcnt

INT ftgtcnt

Number of fanintargets to receive

coefmax

INT coefmax

Working block max size (cblk coeff 1D)

bpftmax

INT bpftmax

Maximum of block size for btag to receive

cpftmax

INT cpftmax

Maximum of block size for ftgt to receive

nbftmax

INT nbftmax

Maximum block number in ftgt

arftmax

INT arftmax

Maximum block area in ftgt

clustnum

INT clustnum

current processor number

clustnbr

INT clustnbr

number of processors

procnbr

INT procnbr

Number of physical processor used

thrdnbr

INT thrdnbr

Number of local computation threads

bublnbr

INT bublnbr

Number of local computation threads

btagnbr

INT btagnbr

Number of Blocktargets

btgsnbr

INT btgsnbr

Number of Blocktargets to send

btgrnbr

INT btgrnbr

Number of Blocktargets to recv

bcofnbr

INT bcofnbr

indnbr

INT indnbr

tasknbr

INT tasknbr

Number of Tasks

ttsktab

Task access vector by thread

ttsknbr

Number of tasks by thread

proc2clust

proc -> cluster

gridldim

INT gridldim

Dimensions of the virtual processors

gridcdim

INT gridcdim

grid if dense end block

updovct

UpDownVector updovct

UpDown vector

INT taskid
COMP_1D DIAG E1 E2
INT prionum
Priority value for the factorization
INT prionum2
Priority value for solve steps
INT cblknum
Attached column block
INT bloknum
Attached block
INT indnum
INT tasknext
chainage des E1 ou E2, si fin = -1 => liberer les btagptr
INT taskmstr
Master task for E1 or E2 tasks
INT threadid
Index of the bubble which contains the task
INT cand
Thread candidate in static version
INT fcandnum
First thread candidate
INT lcandnum
Last thread candidate
INT id
Global cblknum of the attached column block
INT stride
Column block stride
INT color
Color of column block (PICL trace)
INT procdiag
Processor owner of diagonal block
INT cblkdiag
Column block owner of diagonal block
INT coefind
Index in coeftab
SymbolMatrix symbmtx
Symbolic matrix
CscMatrix cscmtx
Compress Sparse Column matrix
INT coefnbr
Number of coefficients
INT ftgtnbr
Number of fanintargets
INT ftgtcnt
Number of fanintargets to receive
INT coefmax
Working block max size (cblk coeff 1D)
INT bpftmax
Maximum of block size for btag to receive
INT cpftmax
Maximum of block size for ftgt to receive
INT nbftmax
Maximum block number in ftgt
INT arftmax
Maximum block area in ftgt
INT clustnum
current processor number
INT clustnbr
number of processors
INT procnbr
Number of physical processor used
INT thrdnbr
Number of local computation threads
INT bublnbr
Number of local computation threads
INT btagnbr
Number of Blocktargets
INT btgsnbr
Number of Blocktargets to send
INT btgrnbr
Number of Blocktargets to recv
INT bcofnbr
INT indnbr
INT tasknbr
Number of Tasks
INT gridldim
Dimensions of the virtual processors
INT gridcdim
grid if dense end block
UpDownVector updovct
UpDown vector
Close