RFR: 8231356: Fix broken ResourceObj::operator new in debug builds [v2]
lkorinth at openjdk.java.net
Wed Aug 11 11:43:24 UTC 2021
On Wed, 11 Aug 2021 11:36:43 GMT, Leo Korinth <lkorinth at openjdk.org> wrote:
>> ResourceObj::operator new calls ResourceObj::operator new (non array version). In debug builds, each resource object (on C_HEAP) will be initialized with set_allocation_type() (which is correct). What is not correct is that the constructor (and thus) set_allocation_type() is called on the array itself (which is not a ResourceObj). This initialization will be partially overwritten by the header that keeps track of the array size. When the array destructor later is called, it will also chain call the non-array destructor. In debug builds the verification of _allocation_t will fail as it has been overwritten by the code that keeps track of the array size.
>> The following assert will fail:
>> assert(~(_allocation_t | allocation_mask) == (uintptr_t)this, "lost resource object");
>> The reason that it has not been detected is that no one uses ResourceObj::operator new on resource objects with C_HEAP storage.
> Leo Korinth has updated the pull request incrementally with one additional commit since the last revision:
> Remove all other new  operators in RsourceObj and the delete  operator as well
I removed all other operator new  and delete  in ResourceObj. I think I can move the "type" out of the ResourceObj and into a header before the ResourceObj. I think this would improve and simplify the design (and I think I could also make it work for arrays), but let us start with removing the current problem and close this PR.
More information about the hotspot-runtime-dev