RFR: 8240110: Improve NULL

[David Holmes]

On 20/04/2020 3:08 am, Leo Korinth wrote:
> 
> 
> On 17/04/2020 08:45, David Holmes wrote:
>> Hi Leo,
>>
>> I've taken a look at the complete set of changes and the only thing I 
>> strongly object to is the change to:
>>
>> src/hotspot/share/jvmci/jvmciExceptions.hpp
>>
>> as it just doesn't make any sense to me. If you were to write the 
>> expanded macros out in full you would write the code as it currently 
>> is defined in the macro. If a method returns a pointer you would write:
>>
>> return NULL;
>>
>> If a method returns a jint/jbyte/jlong etc you would write
>>
>> return 0;
>>
>> Cheers,
>> David
> 
> I totally agree with you and the change needed was smaller than I 
> rememberd, I should have fixed this directly. Returning NULL for an 
> object that is not of pointer type seems just wrong, but returning a 
> JVMCIObjectArray(NULL) seems okay.
> 
> What about this?
> 
> #define JVMCI_CHECK_0             JVMCI_CHECK_(NULL_WORD)

Why does this integer returning macro need to return NULL_WORD? As I 
said above if writing this directly I would write "return 0;" in the 
code. Is this just an attempt under the covers to (eventually) avoid 
misusing JVMCI_CHECK_0 where JVMCI_CHECK_NULL should have been used?

> #define JVMCI_CHECK_NULL          JVMCI_CHECK_(NULL)
> 
> and 3 changes of JVMCI_CHECK_NULL to JVMCI_CHECK_(JVMCIObjectArray(NULL))

I'm not sure specifically where those changes are.

In general I think I prefer to see an explicit constructor passed NULL 
rather than some implicit conversion, but that is a stylistic matter and 
the implicit conversion form seems to be used quite extensively.

Thanks,
David

> (this change is so that we get one conversion less)
> 
> Thanks,
> Leo
> 
>