harmony-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Weldon Washburn <weldon...@gmail.com>
Subject Re: [modules] classloader/jit interface
Date Tue, 28 Jun 2005 00:27:17 GMT
Tom,

Thanks for being on topic.  This thread has drifted all over the place.

I wasn't thinking clearly on the proposed interface.  Your suggestion
is the correct way to go.  I was originally thinking that the below
interface would be macros that are used to pull the relevant bits out
of the "modifier int".  I now realize that the below looks like a
virtual method call on instances of class/method/field... oops...

I will take another stab at the modifiers API and also pick another
aspect of the JIT/Classloader interface for discussion.  Maybe
class_lookup_field_recursive...

On 27 Jun 2005 09:30:01 -0600, Tom Tromey <tromey@redhat.com> wrote:
> >>>>> "Weldon" == Weldon Washburn <weldonwjw@gmail.com> writes:
> 
> Weldon> Below is a first stab at the API to retrieve the values
> Weldon> contained in internal classloader data structures.
> 
> Weldon> Comment/questions are appreciated.
> 
> Weldon> Class access and property modifiers (table 4.1)
> Weldon> bool Class::is_public();                // ACC_PUBLIC
> Weldon> bool Class::is_final();         // ACC_FINAL
> [ ... ]
> 
> Why not have a single API for retrieving all the flags encoded as a
> word?  We already need this capability to implement things like
> Class.getModifiers.  The flag values are unlikely to change, and we
> won't even have to code them in the source anyway (we can generate
> them from Modifier.class).
> 
> Also, it is worth clarifying whether, for Class, this returns the
> "real" modifiers or the modifiers from the InnerClasses attribute.
> The former are typically needed internally but the latter are needed
> in some situations (Class.newInstance I think).
> 
> Tom
>

Mime
View raw message