JIT should optimize away the cast that occurs after Array.Clone #45302
Description
The typical way to clone an array is to use Array.Clone, which returns an object that needs to be cast back to the original array type.
string[] originalArray = GetArray();
string[] clonedArray = (string[])originalArray.Clone();After the call to Clone, the JIT emits a call to the ChkCastAny helper to ensure that the cast is legitimate.
public static object[] M(object[] o) {
return (Object[])o.Clone();
}M(System.Object[])
L0000: sub rsp, 0x28
L0004: mov eax, [rcx]
L0006: call 0x00007ffb1bc0e0c0 ; Array.Clone()
L000b: mov rdx, rax
L000e: mov rcx, 0x7ffabc1af762 ; type handle for object[]
L0018: call 0x00007ffb1bc16750 ; ChkCastAny
L001d: nop
L001e: add rsp, 0x28
L0022: retThere is an opportunity for improvement here. If the target of the Array.Clone call is known by the JIT to be a local of type T, and if the return value of Clone is immediately cast to T or type U (where reinterpret_cast<U>(T) is legal), then the JIT should no-op the cast. In the sample above, that means labels L000b - L001d would disappear.
Some examples:
string[] stringArray;
string[] clone1 = (string[])stringArray.Clone(); // JIT should elide check
object[] clone2 = (object[])stringArray.Clone(); // JIT should elide check
object[] objectArray;
string[] clone3 = (string[])objectArray.Clone(); // check should take place, could succeed or fail at runtime
object[] clone4 = (object[])objectArray.Clone(); // JIT should elide check
int[] intArray;
int[] clone5 = (int[])intArray.Clone(); // JIT should elide check
uint[] clone6 = (uint[])intArray.Clone(); // JIT should elide check
object[] clone7 = (object[])intArray.Clone(); // check should take place and will fail at runtime