Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in
Toggle navigation
Menu
Open sidebar
tsoc
openmm
Commits
36b9db84
Commit
36b9db84
authored
Jul 13, 2018
by
Peter Eastman
Browse files
Bug fixes and code simplification
parent
1b12ede8
Changes
2
Hide whitespace changes
Inline
Side-by-side
Showing
2 changed files
with
7 additions
and
11 deletions
+7
-11
platforms/cuda/src/CudaKernels.cpp
platforms/cuda/src/CudaKernels.cpp
+4
-5
platforms/cuda/src/kernels/pme.cu
platforms/cuda/src/kernels/pme.cu
+3
-6
No files found.
platforms/cuda/src/CudaKernels.cpp
View file @
36b9db84
...
@@ -1749,7 +1749,6 @@ void CudaCalcNonbondedForceKernel::initialize(const System& system, const Nonbon
...
@@ -1749,7 +1749,6 @@ void CudaCalcNonbondedForceKernel::initialize(const System& system, const Nonbon
pmeDefines["GRID_SIZE_X"] = cu.intToString(dispersionGridSizeX);
pmeDefines["GRID_SIZE_X"] = cu.intToString(dispersionGridSizeX);
pmeDefines["GRID_SIZE_Y"] = cu.intToString(dispersionGridSizeY);
pmeDefines["GRID_SIZE_Y"] = cu.intToString(dispersionGridSizeY);
pmeDefines["GRID_SIZE_Z"] = cu.intToString(dispersionGridSizeZ);
pmeDefines["GRID_SIZE_Z"] = cu.intToString(dispersionGridSizeZ);
pmeDefines
[
"EPSILON_FACTOR"
]
=
"1"
;
pmeDefines["RECIP_EXP_FACTOR"] = cu.doubleToString(M_PI*M_PI/(dispersionAlpha*dispersionAlpha));
pmeDefines["RECIP_EXP_FACTOR"] = cu.doubleToString(M_PI*M_PI/(dispersionAlpha*dispersionAlpha));
pmeDefines["USE_LJPME"] = "1";
pmeDefines["USE_LJPME"] = "1";
double invRCut6 = pow(force.getCutoffDistance(), -6);
double invRCut6 = pow(force.getCutoffDistance(), -6);
...
@@ -1778,7 +1777,7 @@ void CudaCalcNonbondedForceKernel::initialize(const System& system, const Nonbon
...
@@ -1778,7 +1777,7 @@ void CudaCalcNonbondedForceKernel::initialize(const System& system, const Nonbon
roundedZSize = PmeOrder*(int) ceil(dispersionGridSizeZ/(double) PmeOrder);
roundedZSize = PmeOrder*(int) ceil(dispersionGridSizeZ/(double) PmeOrder);
gridElements = max(gridElements, dispersionGridSizeX*dispersionGridSizeY*roundedZSize);
gridElements = max(gridElements, dispersionGridSizeX*dispersionGridSizeY*roundedZSize);
}
}
pmeGrid1
.
initialize
(
cu
,
gridElements
,
2
*
elementSize
,
"
originalP
meGrid"
);
pmeGrid1.initialize(cu, gridElements, 2*elementSize, "
p
meGrid
1
");
pmeGrid2.initialize(cu, gridElements, 2*elementSize, "pmeGrid2");
pmeGrid2.initialize(cu, gridElements, 2*elementSize, "pmeGrid2");
cu.addAutoclearBuffer(pmeGrid2);
cu.addAutoclearBuffer(pmeGrid2);
pmeBsplineModuliX.initialize(cu, gridSizeX, elementSize, "pmeBsplineModuliX");
pmeBsplineModuliX.initialize(cu, gridSizeX, elementSize, "pmeBsplineModuliX");
...
@@ -2196,14 +2195,14 @@ double CudaCalcNonbondedForceKernel::execute(ContextImpl& context, bool includeF
...
@@ -2196,14 +2195,14 @@ double CudaCalcNonbondedForceKernel::execute(ContextImpl& context, bool includeF
sort->sort(pmeAtomGridIndex);
sort->sort(pmeAtomGridIndex);
}
}
cu
.
clearBuffer
(
pmeGrid
1
);
cu.clearBuffer(pmeGrid
2
);
void
*
spreadArgs
[]
=
{
&
cu
.
getPosq
().
getDevicePointer
(),
&
pmeGrid
1
.
getDevicePointer
(),
cu
.
getPeriodicBoxSizePointer
(),
void* spreadArgs[] = {&cu.getPosq().getDevicePointer(), &pmeGrid
2
.getDevicePointer(), cu.getPeriodicBoxSizePointer(),
cu.getInvPeriodicBoxSizePointer(), cu.getPeriodicBoxVecXPointer(), cu.getPeriodicBoxVecYPointer(), cu.getPeriodicBoxVecZPointer(),
cu.getInvPeriodicBoxSizePointer(), cu.getPeriodicBoxVecXPointer(), cu.getPeriodicBoxVecYPointer(), cu.getPeriodicBoxVecZPointer(),
recipBoxVectorPointer[0], recipBoxVectorPointer[1], recipBoxVectorPointer[2], &pmeAtomGridIndex.getDevicePointer(),
recipBoxVectorPointer[0], recipBoxVectorPointer[1], recipBoxVectorPointer[2], &pmeAtomGridIndex.getDevicePointer(),
&sigmaEpsilon.getDevicePointer()};
&sigmaEpsilon.getDevicePointer()};
cu.executeKernel(pmeDispersionSpreadChargeKernel, spreadArgs, cu.getNumAtoms(), 128);
cu.executeKernel(pmeDispersionSpreadChargeKernel, spreadArgs, cu.getNumAtoms(), 128);
void
*
finishSpreadArgs
[]
=
{
&
pmeGrid1
.
getDevicePointer
()};
void* finishSpreadArgs[] = {
&pmeGrid2.getDevicePointer(),
&pmeGrid1.getDevicePointer()};
cu.executeKernel(pmeDispersionFinishSpreadChargeKernel, finishSpreadArgs, dispersionGridSizeX*dispersionGridSizeY*dispersionGridSizeZ, 256);
cu.executeKernel(pmeDispersionFinishSpreadChargeKernel, finishSpreadArgs, dispersionGridSizeX*dispersionGridSizeY*dispersionGridSizeZ, 256);
if (useCudaFFT) {
if (useCudaFFT) {
...
...
platforms/cuda/src/kernels/pme.cu
View file @
36b9db84
...
@@ -52,7 +52,7 @@ extern "C" __global__ void gridSpreadCharge(const real4* __restrict__ posq, real
...
@@ -52,7 +52,7 @@ extern "C" __global__ void gridSpreadCharge(const real4* __restrict__ posq, real
real4
pos
=
posq
[
atom
];
real4
pos
=
posq
[
atom
];
#ifdef USE_LJPME
#ifdef USE_LJPME
const
float2
sigEps
=
sigmaEpsilon
[
atom
];
const
float2
sigEps
=
sigmaEpsilon
[
atom
];
const
real
charge
=
8
*
sigEps
.
x
*
sigEps
.
x
*
sigEps
.
x
*
sigEps
.
y
*
EPSILON_FACTOR
;
const
real
charge
=
8
*
sigEps
.
x
*
sigEps
.
x
*
sigEps
.
x
*
sigEps
.
y
;
#else
#else
const
real
charge
=
(
CHARGE
)
*
EPSILON_FACTOR
;
const
real
charge
=
(
CHARGE
)
*
EPSILON_FACTOR
;
#endif
#endif
...
@@ -140,11 +140,8 @@ extern "C" __global__ void finishSpreadCharge(
...
@@ -140,11 +140,8 @@ extern "C" __global__ void finishSpreadCharge(
const
unsigned
int
gridSize
=
GRID_SIZE_X
*
GRID_SIZE_Y
*
GRID_SIZE_Z
;
const
unsigned
int
gridSize
=
GRID_SIZE_X
*
GRID_SIZE_Y
*
GRID_SIZE_Z
;
real
scale
=
1
/
(
real
)
0x100000000
;
real
scale
=
1
/
(
real
)
0x100000000
;
for
(
int
index
=
blockIdx
.
x
*
blockDim
.
x
+
threadIdx
.
x
;
index
<
gridSize
;
index
+=
blockDim
.
x
*
gridDim
.
x
)
{
for
(
int
index
=
blockIdx
.
x
*
blockDim
.
x
+
threadIdx
.
x
;
index
<
gridSize
;
index
+=
blockDim
.
x
*
gridDim
.
x
)
{
int
xindex
=
index
/
(
GRID_SIZE_Y
*
GRID_SIZE_Z
);
int
zindex
=
index
%
GRID_SIZE_Z
;
int
remainder
=
index
-
xindex
*
GRID_SIZE_Y
*
GRID_SIZE_Z
;
int
loadIndex
=
zindexTable
[
zindex
]
+
blockSize
*
(
int
)
(
index
/
GRID_SIZE_Z
);
int
yindex
=
remainder
/
GRID_SIZE_Z
;
int
zindex
=
remainder
-
yindex
*
GRID_SIZE_Z
;
int
loadIndex
=
zindexTable
[
zindex
]
+
(
xindex
*
GRID_SIZE_Y
+
yindex
)
*
blockSize
;
#if defined(USE_DOUBLE_PRECISION) || defined(USE_DETERMINISTIC_FORCES)
#if defined(USE_DOUBLE_PRECISION) || defined(USE_DETERMINISTIC_FORCES)
grid2
[
index
]
=
scale
*
grid1
[
loadIndex
];
grid2
[
index
]
=
scale
*
grid1
[
loadIndex
];
#else
#else
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment