lattice/latticegpu.jl
1
0
Fork 0
mirror of https://igit.ific.uv.es/alramos/latticegpu.jl.git synced 2025-06-29 21:39:27 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

Wrap threadIdx, etc in an inline function

Browse source
This commit is contained in:
Alberto Ramos 2022-01-05 20:04:55 +01:00
parent 65a619e554
commit 584b92edb0
7 changed files with 21 additions and 18 deletions
Download patch file Download diff file Expand all files Collapse all files

10
src/YM/YMflow.jl
View file

@ -89,7 +89,7 @@ end
function krnl_add_zth!(frc, frc2::AbstractArray{TA}, U::AbstractArray{TG}, lp::SpaceParm{N,M,B,D}) where {TA,TG,N,M,B,D}
b, r = CUDA.threadIdx().x, CUDA.blockIdx().x
b, r = assign_thx()
it = point_time((b, r), lp)
SFBC = ((B == BC_SF_AFWB) || (B == BC_SF_ORBI) )
@ -261,7 +261,7 @@ Eoft_plaq(U, gp::GaugeParm{T,G,NN}, lp::SpaceParm{N,M,B,D}, ymws::YMworkspace) w
function krnl_plaq_pln!(plx, U::AbstractArray{T}, Ubnd, ztw, ipl, lp::SpaceParm{N,M,B,D}) where {T,N,M,B,D}
b, r = CUDA.threadIdx().x, CUDA.blockIdx().x
b, r = assign_thx()
id1, id2 = lp.plidx[ipl]
SFBC = ((B == BC_SF_AFWB) || (B == BC_SF_ORBI)) && (id1 == lp.iL[end])
@ -387,7 +387,7 @@ Eoft_clover(U, gp::GaugeParm, lp::SpaceParm{N,M,B,D}, ymws::YMworkspace{T}) wher
function krnl_add_et!(rm, op, frc1, U, lp::SpaceParm{4,M,B,D}) where {M,B,D}
b, r = CUDA.threadIdx().x, CUDA.blockIdx().x
b, r = assign_thx()
X1 = (frc1[b,1,r]+frc1[b,2,r]+frc1[b,3,r]+frc1[b,4,r])
@ -399,7 +399,7 @@ end
function krnl_add_qd!(rm, op, frc1, frc2, U, lp::SpaceParm{4,M,B,D}) where {M,B,D}
b, r = CUDA.threadIdx().x, CUDA.blockIdx().x
b, r = assign_thx()
I = point_coord((b,r), lp)
rm[I] += op(dot( (frc1[b,1,r]+frc1[b,2,r]+frc1[b,3,r]+frc1[b,4,r]),
@ -409,7 +409,7 @@ end
function krnl_field_tensor!(frc1::AbstractArray{TA}, frc2, U::AbstractArray{T}, Ubnd, ipl1, ipl2, ztw1, ztw2, lp::SpaceParm{4,M,B,D}) where {TA,T,M,B,D}
b, r = CUDA.threadIdx().x, CUDA.blockIdx().x
b, r = assign_thx()
it = point_time((b,r), lp)
SFBC = ((B == BC_SF_AFWB) || (B == BC_SF_ORBI) )