RFR (XS) 8022283: Assertion failed: assert(is_loaded() && field->holder()->is_loaded() && klass()->is_subclass_of (field->holder())) failed: invalid access

[Christian Thalinger]

On Aug 20, 2013, at 9:28 AM, Vladimir Ivanov <vladimir.x.ivanov at oracle.com> wrote:

> Rickard,

Looks good.

> 
> As a cleanup, can you split failing assert in 3 different asserts?
> 
> ciConstant ciInstance::field_value(ciField* field) {
>  assert(is_loaded() &&
>         field->holder()->is_loaded() &&
>         klass()->is_subclass_of(field->holder()),
>         "invalid access");

Right.  I remember we talked about this.  Rickard, can you do this change?

-- Chris

> 
> 
> Otherwise, looks good (not a Reviewer).
> 
> Best regards,
> Vladimir Ivanov
> 
> On 8/20/13 8:04 PM, Rickard Bäckman wrote:
>> http://cr.openjdk.java.net/~rbackman/8022283/
>> http://bugs.sun.com/view_bug.do?bug_id=8022283
>> 
>> Add a check that verifies that the constant is "is_loaded".
>> 
>> Comment from ciObject.hpp
>> 
>>   // Note: some ciObjects refer to oops which have yet to be created.
>>   // We refer to these as "unloaded".  Specifically, there are
>>   // unloaded instances of java.lang.Class,
>>   // java.lang.invoke.MethodHandle, and java.lang.invoke.MethodType.
>>   // By convention the ciNullObject is considered loaded, and
>>   // primitive types are considered loaded.
>> 
>> In this case we had a java.lang.invoke.MethodType.
>> 
>> <ciField name=java/lang/invoke/MethodHandle.type signature=Ljava/lang/invoke/MethodType; offset=12 type=java/lang/invoke/MethodType is_constant=true>$1 = void
>> 
>> Thanks
>> /R
>>