JDK-8220416 : Comparison of klass pointers is not optimized any more
  • Type: Bug
  • Component: hotspot
  • Sub-Component: compiler
  • Affected Version: 8,9,10,11,12,13,14
  • Priority: P4
  • Status: Resolved
  • Resolution: Fixed
  • Submitted: 2019-03-11
  • Updated: 2019-10-23
  • Resolved: 2019-09-12
The Version table provides details related to the release that this issue/RFE will be addressed.

Unresolved : Release in which this issue/RFE will be addressed.
Resolved: Release in which this issue/RFE has been resolved.
Fixed : Release in which this issue/RFE has been fixed. The release containing this fix may be available for download as an Early Access Release or a General Availability Release.

To download the current JDK release, click here.
JDK 14
14 b15Fixed
Related Reports
Relates :  
JDK-6964458 - Reimplement class meta-data storage to use native memory
Relates :  
JDK-8232873 - Add missing test for 8220416
Description
I noticed this:

  if (p0 && p1) {
    Node* in1 = in(1)->uncast();
    Node* in2 = in(2)->uncast();
    AllocateNode* alloc1 = AllocateNode::Ideal_allocation(in1, NULL);
    AllocateNode* alloc2 = AllocateNode::Ideal_allocation(in2, NULL);
    if (MemNode::detect_ptr_independence(in1, alloc1, in2, alloc2, NULL)) {
      return TypeInt::CC_GT;  // different pointers
    }
    ciKlass* klass0 = p0->klass();
    bool    xklass0 = p0->klass_is_exact();
    ciKlass* klass1 = p1->klass();
    bool    xklass1 = p1->klass_is_exact();
    int kps = (p0->isa_klassptr()?1:0) + (p1->isa_klassptr()?1:0);
    if (klass0 && klass1 &&
        kps != 1 &&             // both or neither are klass pointers
        klass0->is_loaded() && !klass0->is_interface() && // do not trust interfaces
        klass1->is_loaded() && !klass1->is_interface() &&
        (!klass0->is_obj_array_klass() ||
         !klass0->as_obj_array_klass()->base_element_klass()->is_interface()) &&
        (!klass1->is_obj_array_klass() ||
         !klass1->as_obj_array_klass()->base_element_klass()->is_interface())) {

in CmpPNode::sub() doesn't trigger for klass pointers comparisons since JDK-6964458 because TypeKlassPtr doesn't subclass TypeOopPtr anymore.
Comments
URL: https://hg.openjdk.java.net/jdk/jdk/rev/46e11f978852 User: thartmann Date: 2019-09-12 10:27:40 +0000
12-09-2019
http://cr.openjdk.java.net/~chagedorn/8220416/webrev.00/
11-09-2019
ILW = C2 optimization is not triggered causing suboptimal code (old regression), comparison of klass pointers, no workaround = LHH = P4
12-03-2019