Hello,
I am running easy random access benchmarking in intel SGX, I am interested in the performance when the data passed into enclave is exceed enclave size. But I always meet Aborted (core dumped) when the input size exceeds enclave size, is there any configuration or parameter to avoid this issue?
Here is my benchmark code:
// Enclave/test/test.cpp
void ecall_ret_memory_access_2(int *arr, size_t sz)
{
uint32_t val;
for(int i = 0; i < (int)sz; i++){
arr = arr + sgx_read_rand((unsigned char *) &val, 4);
}
}
// Enclave/test/test.edl
enclave {
trusted {
public void ecall_ret_memory_access_2([in, out,count=sz] int *arr, size_t sz);
};
};
