Changed the iterator to iterate over the same size tensors in the same order independent of memory layout.
This commit is contained in:
@@ -39,7 +39,9 @@ void tensoriter_test_scalar_get(void)
|
||||
uint32_t contained = 0;
|
||||
tensoriter_scalar iter = tensoriter_scalar_create(t);
|
||||
do {
|
||||
uint32_t value = (uint32_t) tensoriter_scalar_get(iter);
|
||||
bool success;
|
||||
uint32_t value = (uint32_t) tensoriter_scalar_get(iter, &success);
|
||||
tensor_assert(success, "mute");
|
||||
tensor_assert(((1 << (value - 1)) & contained) == 0, "mute");
|
||||
contained |= 1 << (value - 1);
|
||||
} while (tensoriter_scalar_next(iter));
|
||||
|
||||
Reference in New Issue
Block a user