mirror of https://gitlab.com/QEF/q-e.git
Setting the chunk size as a constant
This commit is contained in:
parent
fa21b8d52a
commit
9a94d4d047
|
@ -36,7 +36,8 @@ SUBROUTINE cegterg( h_psi, s_psi, uspp, g_psi, &
|
|||
! maximum dimension of the reduced basis set :
|
||||
! (the basis set is refreshed when its dimension would exceed nvecx)
|
||||
! umber of spin polarizations
|
||||
INTEGER :: numblock, blocksize
|
||||
INTEGER, PARAMETER :: blocksize = 256
|
||||
INTEGER :: numblock
|
||||
! chunking parameters
|
||||
COMPLEX(DP), INTENT(INOUT) :: evc(npwx,npol,nvec)
|
||||
! evc contains the refined estimates of the eigenvectors
|
||||
|
@ -121,7 +122,6 @@ SUBROUTINE cegterg( h_psi, s_psi, uspp, g_psi, &
|
|||
END IF
|
||||
!
|
||||
! setting chunck size
|
||||
blocksize = 256
|
||||
numblock = (npw+blocksize-1)/blocksize
|
||||
!
|
||||
ALLOCATE( psi( npwx, npol, nvecx ), STAT=ierr )
|
||||
|
@ -597,7 +597,8 @@ SUBROUTINE pcegterg(h_psi, s_psi, uspp, g_psi, &
|
|||
! maximum dimension of the reduced basis set
|
||||
! (the basis set is refreshed when its dimension would exceed nvecx)
|
||||
! number of spin polarizations
|
||||
INTEGER :: numblock, blocksize
|
||||
INTEGER, PARAMETER :: blocksize = 256
|
||||
INTEGER :: numblock
|
||||
! chunking parameters
|
||||
COMPLEX(DP), INTENT(INOUT) :: evc(npwx,npol,nvec)
|
||||
! evc contains the refined estimates of the eigenvectors
|
||||
|
@ -687,7 +688,6 @@ SUBROUTINE pcegterg(h_psi, s_psi, uspp, g_psi, &
|
|||
END IF
|
||||
!
|
||||
! setting chunck size
|
||||
blocksize = 256
|
||||
numblock = (npw+blocksize-1)/blocksize
|
||||
!
|
||||
ALLOCATE( psi( npwx, npol, nvecx ), STAT=ierr )
|
||||
|
|
|
@ -37,13 +37,13 @@ subroutine g_psi (lda, n, m, npol, psi, e)
|
|||
integer :: k, i
|
||||
! counter on psi functions
|
||||
! counter on G vectors
|
||||
integer :: iblock, numblock, blocksize
|
||||
integer, parameter :: blocksize = 256
|
||||
integer :: iblock, numblock
|
||||
! chunking parameters
|
||||
!
|
||||
call start_clock ('g_psi')
|
||||
!
|
||||
! setting chunck size
|
||||
blocksize = 256
|
||||
numblock = (n+blocksize-1)/blocksize
|
||||
!
|
||||
#ifdef TEST_NEW_PRECONDITIONING
|
||||
|
|
|
@ -50,7 +50,8 @@ subroutine init_us_2 (npw_, igk_, q_, vkb_)
|
|||
integer :: iq
|
||||
|
||||
! cache blocking parameters
|
||||
INTEGER :: iblock, numblock, blocksize, realblocksize
|
||||
INTEGER, PARAMETER :: blocksize = 256
|
||||
INTEGER :: iblock, numblock, realblocksize
|
||||
!
|
||||
if (lmaxkb.lt.0) return
|
||||
call start_clock ('init_us_2')
|
||||
|
@ -58,7 +59,6 @@ subroutine init_us_2 (npw_, igk_, q_, vkb_)
|
|||
! write(*,'(3i4,i5,3f10.5)') size(tab,1), size(tab,2), size(tab,3), size(vq), q_
|
||||
|
||||
! setting cache blocking size
|
||||
blocksize = 256
|
||||
numblock = (npw_+blocksize-1)/blocksize
|
||||
|
||||
if (spline_ps) then
|
||||
|
|
|
@ -34,10 +34,10 @@ SUBROUTINE usnldiag (npw, h_diag, s_diag)
|
|||
INTEGER :: ikb, jkb, ih, jh, na, nt, ig, ipol
|
||||
COMPLEX(DP) :: ps1(2), ps2(2), ar
|
||||
! cache blocking parameters
|
||||
INTEGER :: iblock, numblock, blocksize
|
||||
INTEGER, PARAMETER :: blocksize = 256
|
||||
INTEGER :: iblock, numblock
|
||||
!
|
||||
! setting cache blocking size
|
||||
blocksize = 256
|
||||
numblock = (npw+blocksize-1)/blocksize
|
||||
!
|
||||
!$omp parallel do private(ikb, jkb, ps1, ps2, ar)
|
||||
|
|
Loading…
Reference in New Issue