RtlULongLongMult - NtDoc

#ifndef _NTINTSAFE_H_INCLUDED_
#if WINAPI_FAMILY_PARTITION(WINAPI_PARTITION_APP | WINAPI_PARTITION_SYSTEM | WINAPI_PARTITION_GAMES)

//
// ULONGLONG multiplication
//
_Must_inspect_result_
__inline
NTSTATUS
RtlULongLongMult(
    _In_ ULONGLONG ullMultiplicand,
    _In_ ULONGLONG ullMultiplier,
    _Out_ _Deref_out_range_(==, ullMultiplicand * ullMultiplier) ULONGLONG* pullResult)
{
    NTSTATUS status;
#if defined(_USE_INTRINSIC_MULTIPLY128)
    ULONGLONG ullResultHigh;
    ULONGLONG ullResultLow;
    
    ullResultLow = UnsignedMultiply128(ullMultiplicand, ullMultiplier, &ullResultHigh);
    if (ullResultHigh == 0)
    {
        _Analysis_assume_(ullMultiplicand * ullMultiplier == ullResultLow);
        *pullResult = ullResultLow;
        status = STATUS_SUCCESS;
    }
    else
    {
        *pullResult = ULONGLONG_ERROR;
        status = STATUS_INTEGER_OVERFLOW;
    }
#else
    // 64x64 into 128 is like 32.32 x 32.32.
    //
    // a.b * c.d = a*(c.d) + .b*(c.d) = a*c + a*.d + .b*c + .b*.d
    // back in non-decimal notation where A=a*2^32 and C=c*2^32:  
    // A*C + A*d + b*C + b*d
    // So there are four components to add together.
    //   result = (a*c*2^64) + (a*d*2^32) + (b*c*2^32) + (b*d)
    //
    // a * c must be 0 or there would be bits in the high 64-bits
    // a * d must be less than 2^32 or there would be bits in the high 64-bits
    // b * c must be less than 2^32 or there would be bits in the high 64-bits
    // then there must be no overflow of the resulting values summed up.
    
    ULONG dw_a;
    ULONG dw_b;
    ULONG dw_c;
    ULONG dw_d;
    ULONGLONG ad = 0;
    ULONGLONG bc = 0;
    ULONGLONG bd = 0;
    ULONGLONG ullResult = 0;

    status = STATUS_INTEGER_OVERFLOW;
    
    dw_a = (ULONG)(ullMultiplicand >> 32);
    dw_c = (ULONG)(ullMultiplier >> 32);

    // common case -- if high dwords are both zero, no chance for overflow
    if ((dw_a == 0) && (dw_c == 0))
    {
        dw_b = (DWORD)ullMultiplicand;
        dw_d = (DWORD)ullMultiplier;

        *pullResult = (((ULONGLONG)dw_b) * (ULONGLONG)dw_d);
        status = STATUS_SUCCESS;
    }
    else
    {
        // a * c must be 0 or there would be bits set in the high 64-bits
        if ((dw_a == 0) ||
            (dw_c == 0))
        {
            dw_d = (DWORD)ullMultiplier;

            // a * d must be less than 2^32 or there would be bits set in the high 64-bits
            ad = (((ULONGLONG)dw_a) * (ULONGLONG)dw_d);
            if ((ad & 0xffffffff00000000) == 0)
            {
                dw_b = (DWORD)ullMultiplicand;

                // b * c must be less than 2^32 or there would be bits set in the high 64-bits
                bc = (((ULONGLONG)dw_b) * (ULONGLONG)dw_c);
                if ((bc & 0xffffffff00000000) == 0)
                {
                    // now sum them all up checking for overflow.
                    // shifting is safe because we already checked for overflow above
                    if (NT_SUCCESS(RtlULongLongAdd(bc << 32, ad << 32, &ullResult)))                        
                    {
                        // b * d
                        bd = (((ULONGLONG)dw_b) * (ULONGLONG)dw_d);
                    
                        if (NT_SUCCESS(RtlULongLongAdd(ullResult, bd, &ullResult)))
                        {
                            *pullResult = ullResult;
                            status = STATUS_SUCCESS;
                        }
                    }
                }
            }
        }
    }

    if (!NT_SUCCESS(status))
    {
        *pullResult = ULONGLONG_ERROR;
    }
#pragma warning(suppress:26071)
#endif // _USE_INTRINSIC_MULTIPLY128
    return status;
}

#endif
#endif
View code on GitHub