-
Bug
-
Resolution: Fixed
-
P4
-
6
-
None
-
1.6
-
generic
-
generic
category : java
subcategory : other
release : mustang
type : bug
synopsis : New Verifier flaw received
customer name : Dr Wes Munsil
customer mail : ###@###.###
sdn id :
language : en
company : other
hardware : generic
os : generic
bug id : 0
date created : Mon Oct 31 22:28:02 MST 2005
date evaluated : Wed Nov 09 10:41:50 MST 2005
description :
A DESCRIPTION OF THE FLAW :
This is a specification issue.
The term "offset item of the stack map frame" is used on page 137.
It is nowhere defined. In fact, the stack map frame has no offset item--rather,
the applicable byte code offset is the result of a calculation (or "calulation"
[sic])--see page 133.
Consequently, the definition on page 137, namely
"We say that an instruction in the byte code has a corresponding stack map frame
if the offset in the offset item of the stack map frame is the same as the
offset of the instruction in the byte codes."
is vacuous, and it is not possible to say that any instruction in the byte code
has a corresponding stack map frame. This renders much of the new material on
typechecking rules moot.
subcategory : other
release : mustang
type : bug
synopsis : New Verifier flaw received
customer name : Dr Wes Munsil
customer mail : ###@###.###
sdn id :
language : en
company : other
hardware : generic
os : generic
bug id : 0
date created : Mon Oct 31 22:28:02 MST 2005
date evaluated : Wed Nov 09 10:41:50 MST 2005
description :
A DESCRIPTION OF THE FLAW :
This is a specification issue.
The term "offset item of the stack map frame" is used on page 137.
It is nowhere defined. In fact, the stack map frame has no offset item--rather,
the applicable byte code offset is the result of a calculation (or "calulation"
[sic])--see page 133.
Consequently, the definition on page 137, namely
"We say that an instruction in the byte code has a corresponding stack map frame
if the offset in the offset item of the stack map frame is the same as the
offset of the instruction in the byte codes."
is vacuous, and it is not possible to say that any instruction in the byte code
has a corresponding stack map frame. This renders much of the new material on
typechecking rules moot.