+1 to open a new JIRA of "make sure the CHECK NULL before monitorEnter/monitorExit will not be eliminated simply". The NPE reported in HARMONY-6013 is found since r659128 which removed some checknull in VM helpers. In a seperated e-mail Mikhail think that the checknull must be kept. http://mail-archives.apache.org/mod_mbox/harmony-dev/200811.mbox/[EMAIL PROTECTED] So I think r713673 (re-enable the checknulls in VM helpers) did fix/resolve HARMONY-6013.
On Mon, Nov 17, 2008 at 11:18 AM, Xiao-Feng Li <[EMAIL PROTECTED]>wrote: > On Sat, Nov 15, 2008 at 9:24 AM, bu qi cheng <[EMAIL PROTECTED]> wrote: > > Currently, to meet M8. ChunRong has done the patch unrolling to walk > around > > the bug. We are waiting for the Mikhail's direction on the "magic for > > monitorEnter/monitorExit". If there are "magic for > > monitorEnter/monitorExit", the unrolling will be ok. If there are no > these > > kind magic, we will use our solution (To make sure the CHECK NULL before > > monitorEnter/monitorExit will not be eliminated simply, or eliminate the > > mechanism: "catch the un-catched throw with hardware") for the problem. > > In this case, we probably should open a new JIRA or just re-open this JIRA. > > Btw, I guess "catch the un-catched throw with hardware" is still a good > idea. > > Thanks, > xiaofeng > > > > > Thanks! > > > > Buqi > > > > > > > > > > On Fri, Nov 14, 2008 at 10:01 AM, Xiao-Feng Li <[EMAIL PROTECTED] > >wrote: > > > >> Is this bug worked around or really fixed? > >> > >> On Fri, Nov 14, 2008 at 9:54 AM, Chunrong Lai (JIRA) <[EMAIL PROTECTED]> > >> wrote: > >> > > >> > [ > >> > https://issues.apache.org/jira/browse/HARMONY-6013?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel > ] > >> > > >> > Chunrong Lai resolved HARMONY-6013. > >> > ----------------------------------- > >> > > >> > Resolution: Fixed > >> > > >> > > >> > Fixed in r713673. > >> > > >> >> Problems with null pointer exception catch in server or opt mode for > >> Java applications > >> >> > >> > --------------------------------------------------------------------------------------- > >> >> > >> >> Key: HARMONY-6013 > >> >> URL: > https://issues.apache.org/jira/browse/HARMONY-6013 > >> >> Project: Harmony > >> >> Issue Type: Bug > >> >> Components: DRLVM > >> >> Affects Versions: 5.0M7 > >> >> Environment: Windows XP OS, Intel Core(TM)2 Duo CPU > >> >> Reporter: Zhiguo Ge > >> >> Assignee: Chunrong Lai > >> >> > >> >> Harmony has problems with null pointer exception catch under server > or > >> opt mode for Java applications. > >> >> The following shows a test example: > >> >> public class Btest5710 { > >> >> > >> >> Object obj; > >> >> public static void main(String [] args) { > >> >> (new Btest5710()).test(); > >> >> } > >> >> public void test() { > >> >> System.err.println("Start Btest5710 test..."); > >> >> try { > >> >> sync(); > >> >> System.err.println("FAILED: NullPointerException was > >> expected"); > >> >> System.exit(-99); > >> >> } catch (NullPointerException e) { > >> >> System.err.println("PASSED: " + e); > >> >> } catch (Throwable e) { > >> >> System.err.println("Unexpected exception was thrown:"); > >> >> e.printStackTrace(); > >> >> System.err.println("FAILED"); > >> >> System.exit(-99); > >> >> } > >> >> } > >> >> void sync() { > >> >> synchronized (obj) { > >> >> obj = new Object(); > >> >> } > >> >> } > >> >> } > >> >> When running in server or opt mode, error will happen and the > following > >> is the error message. > >> >> Signal reported: GENERAL_PROTECTION_FAULT > >> >> Registers: > >> >> EAX: 0x0003f700, EBX: 0x208667ac, ECX: 0x00000004, EDX: > 0x00000003 > >> >> ESI: 0x02fed15c, EDI: 0x0013f8c0, ESP: 0x0013f6f8, EBP: > 0x0086e740 > >> >> EIP: 0x00820b8f > >> > > >> > -- > >> > This message is automatically generated by JIRA. > >> > - > >> > You can reply to this email to add a comment to the issue online. > >> > > >> > > >> > >> > >> > >> -- > >> http://xiao-feng.blogspot.com > >> > > > > > > -- > http://xiao-feng.blogspot.com >
