AI/rocm-systems
2
0
Fork 0
Koodi Ongelmat Pull-pyynnöt Actions Paketit Projektit Julkaisut Wiki Toiminta

Apply .clangformat to all repo source files

Selaa lähdekoodia 
Change-Id: I7e79c6058f0303f9a98911e3b7dd2e8596079344
This commit is contained in:
Maneesh Gupta
2018-03-12 11:29:03 +05:30
vanhempi 18e70b1e6b
commit 1ba06f63c4
293 muutettua tiedostoa jossa 43980 lisäystä ja 45830 poistoa
Download Patch File Download Diff File Expand all files Collapse all files
samples/0_Intro/hcc_dialects/vadd_amp_arrayview.cpp
+19 -18
Näytä tiedosto
@@ -26,8 +26,8 @@ THE SOFTWARE.
// will automatically copy data to and from the host, without the user needing
// to manually perform such copies. This is an excellent mode for developers
// new to GPU programming and matches the memory models provided by recent systems where
// CPU and GPU share the same memory pool. Advanced programmers may prefer
// more explicit control over the data movement - shown in the other vadd_hc_array and
// CPU and GPU share the same memory pool. Advanced programmers may prefer
// more explicit control over the data movement - shown in the other vadd_hc_array and
// vadd_hc_am examples.
// This example shows the similarity between C++AMP and and HC for simple cases where
// implicit data transfer is used - really the only difference is the namespace.
@@ -35,8 +35,7 @@ THE SOFTWARE.
#include <amp.h>
int main(int argc, char *argv[])
{
int main(int argc, char* argv[]) {
int sizeElements = 1000000;
bool pass = true;
@@ -46,28 +45,30 @@ int main(int argc, char *argv[])
concurrency::array_view<float> C(sizeElements);
// Initialize host data
for (int i=0; i<sizeElements; i++) {
A[i] = 1.618f * i;
for (int i = 0; i < sizeElements; i++) {
A[i] = 1.618f * i;
B[i] = 3.142f * i;
}
C.discard_data(); // tell runtime not to copy CPU host data.
C.discard_data(); // tell runtime not to copy CPU host data.
// Launch kernel onto default accelerator
// The HCC runtime will ensure that A and B are available on the accelerator before launching the kernel.
concurrency::parallel_for_each(concurrency::extent<1> (sizeElements),
[=] (concurrency::index<1> idx) restrict(amp) {
int i = idx[0];
C[i] = A[i] + B[i];
});
// The HCC runtime will ensure that A and B are available on the accelerator before launching
// the kernel.
concurrency::parallel_for_each(concurrency::extent<1>(sizeElements),
[=](concurrency::index<1> idx) restrict(amp) {
int i = idx[0];
C[i] = A[i] + B[i];
});
for (int i=0; i<sizeElements; i++) {
float ref= 1.618f * i + 3.142f * i;
// Because C is an array_view, the HCC runtime will copy C back to host at first access here:
for (int i = 0; i < sizeElements; i++) {
float ref = 1.618f * i + 3.142f * i;
// Because C is an array_view, the HCC runtime will copy C back to host at first access
// here:
if (C[i] != ref) {
printf ("error:%d computed=%6.2f, reference=%6.2f\n", i, C[i], ref);
printf("error:%d computed=%6.2f, reference=%6.2f\n", i, C[i], ref);
pass = false;
}
};
if (pass) printf ("PASSED!\n");
if (pass) printf("PASSED!\n");
}