aboutsummaryrefslogtreecommitdiff
path: root/src
diff options
context:
space:
mode:
authorThomas White <taw@physics.org>2010-02-18 12:19:36 +0100
committerThomas White <taw@physics.org>2010-02-18 12:19:36 +0100
commit0d471a2f7722981b0d766d040a1c69424406f6b6 (patch)
treef947a97e0bda074930663a49d2e016694f6aedfa /src
parent1cbd38bd817f4b73403e348fa715c6c5356be98a (diff)
Fixes for previous commit
Diffstat (limited to 'src')
-rw-r--r--src/diffraction-gpu.c53
1 files changed, 49 insertions, 4 deletions
diff --git a/src/diffraction-gpu.c b/src/diffraction-gpu.c
index 89655601..2408538f 100644
--- a/src/diffraction-gpu.c
+++ b/src/diffraction-gpu.c
@@ -29,7 +29,7 @@
#define BANDWIDTH (1.0 / 100.0)
-static const char *clstrerr(cl_int err)
+static const char *clError(cl_int err)
{
switch ( err ) {
case CL_SUCCESS : return "no error";
@@ -142,6 +142,7 @@ void get_diffraction_gpu(struct image *image, int na, int nb, int nc,
double a, b, c, d;
float kc;
const size_t dims[2] = {1024, 1024};
+ cl_event event_d;
cl_mem sfacs;
size_t sfac_size;
@@ -255,20 +256,64 @@ void get_diffraction_gpu(struct image *image, int na, int nb, int nc,
orientation[2] = image->orientation.y;
orientation[3] = image->orientation.z;
- clSetKernelArg(kern, 0, sizeof(cl_mem), &diff);
+ err = clSetKernelArg(kern, 0, sizeof(cl_mem), &diff);
+ if ( err != CL_SUCCESS ) {
+ ERROR("Couldn't set arg 0: %s\n", clError(err));
+ return;
+ }
clSetKernelArg(kern, 1, sizeof(cl_mem), &tt);
+ if ( err != CL_SUCCESS ) {
+ ERROR("Couldn't set arg 1: %s\n", clError(err));
+ return;
+ }
clSetKernelArg(kern, 2, sizeof(cl_float), &kc);
+ if ( err != CL_SUCCESS ) {
+ ERROR("Couldn't set arg 2: %s\n", clError(err));
+ return;
+ }
clSetKernelArg(kern, 3, sizeof(cl_int), &image->width);
+ if ( err != CL_SUCCESS ) {
+ ERROR("Couldn't set arg 3: %s\n", clError(err));
+ return;
+ }
clSetKernelArg(kern, 4, sizeof(cl_float), &image->det.panels[0].cx);
+ if ( err != CL_SUCCESS ) {
+ ERROR("Couldn't set arg 4: %s\n", clError(err));
+ return;
+ }
clSetKernelArg(kern, 5, sizeof(cl_float), &image->det.panels[0].cy);
+ if ( err != CL_SUCCESS ) {
+ ERROR("Couldn't set arg 5: %s\n", clError(err));
+ return;
+ }
clSetKernelArg(kern, 6, sizeof(cl_float), &image->det.panels[0].res);
+ if ( err != CL_SUCCESS ) {
+ ERROR("Couldn't set arg 6: %s\n", clError(err));
+ return;
+ }
clSetKernelArg(kern, 7, sizeof(cl_float), &image->det.panels[0].clen);
+ if ( err != CL_SUCCESS ) {
+ ERROR("Couldn't set arg 7: %s\n", clError(err));
+ return;
+ }
clSetKernelArg(kern, 8, sizeof(cl_float16), &cell);
+ if ( err != CL_SUCCESS ) {
+ ERROR("Couldn't set arg 8: %s\n", clError(err));
+ return;
+ }
clSetKernelArg(kern, 9, sizeof(cl_mem), &sfacs);
+ if ( err != CL_SUCCESS ) {
+ ERROR("Couldn't set arg 9: %s\n", clError(err));
+ return;
+ }
clSetKernelArg(kern, 10, sizeof(cl_float4), &orientation);
+ if ( err != CL_SUCCESS ) {
+ ERROR("Couldn't set arg 10: %s\n", clError(err));
+ return;
+ }
err = clEnqueueNDRangeKernel(cq, kern, 2, NULL, dims, NULL,
- 1, &event_q, &event_d);
+ 0, NULL, &event_d);
if ( err != CL_SUCCESS ) {
ERROR("Couldn't enqueue diffraction kernel\n");
return;
@@ -277,7 +322,7 @@ void get_diffraction_gpu(struct image *image, int na, int nb, int nc,
diff_ptr = clEnqueueMapBuffer(cq, diff, CL_TRUE, CL_MAP_READ, 0,
diff_size, 1, &event_d, NULL, &err);
if ( err != CL_SUCCESS ) {
- ERROR("Couldn't map diffraction buffer: %s\n", clstrerr(err));
+ ERROR("Couldn't map diffraction buffer: %s\n", clError(err));
return;
}
tt_ptr = clEnqueueMapBuffer(cq, tt, CL_TRUE, CL_MAP_READ, 0,