Re: Need help with change

2013-05-22 Thread David Holmes
On 22/05/2013 9:47 PM, Coleen Phillimore wrote: On 5/22/2013 7:34 AM, David Holmes wrote: I don't understand. AFAICS the JVMTI spec just says you have to be able to get to these things from the Class object. So if the fields are in Class then we just need to access them - no need for an upcall.

Re: Need help with change

2013-05-22 Thread Staffan Larsen
On 22 maj 2013, at 13:47, Coleen Phillimore wrote: > On 5/22/2013 7:34 AM, David Holmes wrote: >> I don't understand. AFAICS the JVMTI spec just says you have to be able to >> get to these things from the Class object. So if the fields are in Class >> then we just need to access them - no nee

Re: Need help with change

2013-05-22 Thread Coleen Phillimore
On 5/22/2013 7:34 AM, David Holmes wrote: I don't understand. AFAICS the JVMTI spec just says you have to be able to get to these things from the Class object. So if the fields are in Class then we just need to access them - no need for an upcall. We're moving the signers field out of Class a

Re: Need help with change

2013-05-22 Thread David Holmes
On 22/05/2013 9:18 PM, Coleen Phillimore wrote: On 5/22/2013 3:04 AM, Staffan Larsen wrote: On 22 maj 2013, at 06:35, David Holmes wrote: On 21/05/2013 11:56 PM, Coleen Phillimore wrote: On 05/21/2013 09:29 AM, Staffan Larsen wrote: When doing heap iteration with JVMTI, the spec requires ca

Re: Need help with change

2013-05-22 Thread Staffan Larsen
On 22 maj 2013, at 13:18, Coleen Phillimore wrote: > On 5/22/2013 3:04 AM, Staffan Larsen wrote: >> On 22 maj 2013, at 06:35, David Holmes wrote: >> >>> On 21/05/2013 11:56 PM, Coleen Phillimore wrote: On 05/21/2013 09:29 AM, Staffan Larsen wrote: > When doing heap iteration with JVM

Re: Need help with change

2013-05-22 Thread Coleen Phillimore
On 5/22/2013 3:04 AM, Staffan Larsen wrote: On 22 maj 2013, at 06:35, David Holmes wrote: On 21/05/2013 11:56 PM, Coleen Phillimore wrote: On 05/21/2013 09:29 AM, Staffan Larsen wrote: When doing heap iteration with JVMTI, the spec requires callbacks from the VM to the agent identifying the

Re: Need help with change

2013-05-22 Thread Staffan Larsen
On 22 maj 2013, at 06:35, David Holmes wrote: > On 21/05/2013 11:56 PM, Coleen Phillimore wrote: >> >> On 05/21/2013 09:29 AM, Staffan Larsen wrote: >>> When doing heap iteration with JVMTI, the spec requires callbacks from >>> the VM to the agent identifying the signers and protection domain >

Re: Need help with change

2013-05-21 Thread David Holmes
On 21/05/2013 11:56 PM, Coleen Phillimore wrote: On 05/21/2013 09:29 AM, Staffan Larsen wrote: When doing heap iteration with JVMTI, the spec requires callbacks from the VM to the agent identifying the signers and protection domain references. This is what tagMap does, see jvmtiTagMap.cpp:2464.

Re: Need help with change

2013-05-21 Thread Coleen Phillimore
On 05/21/2013 09:29 AM, Staffan Larsen wrote: When doing heap iteration with JVMTI, the spec requires callbacks from the VM to the agent identifying the signers and protection domain references. This is what tagMap does, see jvmtiTagMap.cpp:2464. As long as it it still possible for JVMTI to f

Re: Need help with change

2013-05-21 Thread Staffan Larsen
When doing heap iteration with JVMTI, the spec requires callbacks from the VM to the agent identifying the signers and protection domain references. This is what tagMap does, see jvmtiTagMap.cpp:2464. As long as it it still possible for JVMTI to find these references (with ik->protection_domai

Need help with change

2013-05-21 Thread Coleen Phillimore
I found during code review comment editing for my change that removes signers and protection domain from the InstanceKlass, that JVMTI code seems to have some sort of call back and knowledge of these fields in instanceKlass. Reference from a class to its sig