hs16-b09 has been promoted...

[Erik Trimble]

Build 09 of Hotspot 16 has just been promoted.

Note that there is a known runtime issue with the SA code; a PIT waiver 
has been issued for this, and a promised fix is in the works for the 
next build.

Testing was done against the 6u18 JDK codebase.


Given this is the first integration using the new external-facing HSX 
repositories, here is the relevant tip changeset of HSX16:

http://hg.openjdk.java.net/hsx/hsx16/master/rev/5979352804a0


------------


Component : VM
Status    : 1 major failures, 0 minor failures
Date      : 09/08/2009 at 15:30
Tested By : VM SQE
Cost(total man-days): 1

Workspace : /java/jdk/ws/hotspot/ws/16/master
Bundles   : 
/net/jprt-web.sfbay/jprt/archive/2009/09/2009-09-04-072951.et151817.hs16b09/bundles/
Platforms :
Solaris Sparc 11(32), -client
Solaris Sparc 11(32), -server
Solaris Sparc 10(32), -client
Solaris Sparc 10(32), -server
Solaris x86 11(32), -client
Solaris x86 11(32), -server
Solaris x86 10(32), -client
Solaris x86 10(32), -server
WinXP Prof(32), -client
WinXP Prof(32), -server
WinXP Home(32), -client
WinXP Home(32), -server
Win Server 2003(32), -client
Win Server 2003(32), -server
Windows Vista 32 bit, -client
Windows Vista 32 bit, -server
Windows Vista 64 bit, -server
RH AS4.0 (32), -client
RH AS4.0 (32), -server
Solaris AMD64(32jdk), -server
Solaris AMD64(32jdk), -client
Solaris AMD64(64jdk), -d64/-server
Sol Sparc 10(64OS)(32jdk), -server
Sol Sparc 10(64OS)(32jdk), -client
Sol Sparc 10(64OS)(64jdk), -d64/-server
win server2003 AMD(64OS)(32jdk), -server
win server2003 AMD(64OS)(32jdk), -client
win server2003 AMD(64OS)(64jdk), -d64/-server
RH AS4.0 AMD(64OS)(32jdk), -server
RH AS4.0 AMD(64OS)(32jdk), -client
RH AS4.0 AMD(64OS)(64jdk), -d64/-server
Others

Tests     : /net/sqenfs-1.sfbay/export1/comp/vm/testbase/
Browsers  : NA
Patches   : NA
Logs      : 
http://sqeweb.sfbay/nfs/tools/gtee/results/HSX/PIT/VM/16/b09/jdk6u18b02/product/
Number of Tests Executed : 271439 product tests, 0 unit tests, 0 tck tests

Bug verification status:
======================================
  Tested, Pass:
6795465: Crash in assembler_sparc.cpp with client compiler on solaris-sparc
6845368: large objects cause a crash or unexpected exception
6862945: conversion of jmethodID to methodOop in JVMTI is too expensive
6875329: fix for 6795465 broke exception handler cloning
6875866: Intrinsic for String.indexOf() is broken on x86 with SSE4.2

  Tested, Pass (partial fixes):

  Tested, Fail:

  Untested bug fixes:
     Bugs/rfes with no unit tests:
6798898: CMS: bugs related to class unloading
6819085: G1: use larger and/or user settable region size
6827605: new String intrinsics may prevent EA scalar replacement
6830542: Performance: new JVM_DefineClass with conditional verify
6841313: G1: dirty cards of survivor regions in parallel
6861660: OopMapBlock count/size confusion
6862863: C2 compiler fails in elide_copy()
6871111: G1: remove the concurrent overhead tracker
6875393: JNI itable index cache is broken
6876584: parameters order is incorrect for enc_String_Equals() in x86_32.ad

     Other reasons:
6803688: Integrate latest JAX-WS (2.1.6) in to JDK 6u14
6879058: Change default JPRT release target for 6Update series VMs
6879061: Bump the HS16 build number to 09

  New bugs filed:
     Bugs in PIT build:
6879572: SA fails with "RuntimeException: field "_is_marked_dependent" 
not found in type instanceKlass"

     Bugs in earlier promoted build:

Number of PIT requested: 2
Integration target J2SE build number: 1.6.0_18-b02

Issues and Notes:
   This is hs16-b09 PIT for jdk6u18-b02.

It is GO with with required Action.

There is one serious regression found in this PIT:
  6879572: SA fails with \"RuntimeException: field 
\"_is_marked_dependent\" not found in type instanceKlass\"

Consequences: some troubleshooting tools (for most of them using SA is 
not default mode) are not usable, SA tests will fail.

This bug is caused by the following fix:
  6830542 \"3/3 Performance: new JVM_DefineClass with conditional verify\"

Required action: resolve the regression for hs16-b10/jdk6u18-b03.

We decided give GO for integration because
  1. SA is not primarily functionality
  2. HS-dev team confirmed that nothing except SA is broken
      (the testing results also show only failed SA tests)
  3. integrating hs16b09 in 6u18b02  will allow other changes to get 
more exposure and
    they  will be better tested (6u18b03 is month from now and it is 
even later in release cycle).

Note, all failures in this PIT reported as new ones are actually known 
ones and are not related to tested hs16-b09.
Unverified fixes are because no tests exist or the fixes are build/jprt 
or source comment/typo ones.


-------------------------------
 From VM SQE




-- 
Erik Trimble
Java System Support
Mailstop:  usca22-123
Phone:  x17195
Santa Clara, CA