[8u Backport] RFR: 8175345: possible null pointer dereference defects

[Rahul Raghavan]

Hi,

I am sorry that the first webrev submitted for 8175345 backport was wrong.
Missed to note that difference in code base and to run jprt test before submitting for review, approval.
Patch from 9 does not apply cleanly.
Apologies for the confusion.


Request for review approval for the following revised webrev for jdk8u.
<webrev.01> : http://cr.openjdk.java.net/~rraghavan/8177429/webrev.01/


<jbs> : https://bugs.openjdk.java.net/browse/JDK-8175345
The changes done in jdk9 were reviewed in open and pushed.
  <jdk9 RFR thread> : http://mail.openjdk.java.net/pipermail/hotspot-compiler-dev/2017-March/025798.html
  <jdk9 changeset> : http://hg.openjdk.java.net/jdk9/jdk9/hotspot/rev/55e3f1f3d0a7


Above <webrev.01> for jdk8u is same as jdk9 changes, 
   except following changes in 'src/share/vm/opto/ifnode.cpp', 
   due the difference in 'ProjNode::is_uncommon_trap_proj()' return type
   (returns 'bool' in jdk8u and 'CallStaticJavaNode*' in jdk9)
So change done for jdk9 was - 
        -  if (unc_proj->is_uncommon_trap_proj(Deoptimization::Reason_predicate) != NULL)
        -   prev_dom = idom;
        +  if ((unc_proj != NULL) && (unc_proj->is_uncommon_trap_proj(Deoptimization::Reason_predicate) != NULL)) {
        +    prev_dom = idom;
        +  }
And now the proposed change for jdk8u in above webrev.01 is - 
        -  if (unc_proj->is_uncommon_trap_proj(Deoptimization::Reason_predicate))
        -   prev_dom = idom;
        +  if ((unc_proj != NULL) && (unc_proj->is_uncommon_trap_proj(Deoptimization::Reason_predicate))) {
        +    prev_dom = idom;
        +  }


Thanks,
Rahul

> -----Original Message-----
> From: Rahul Raghavan
> Sent: Wednesday, May 17, 2017 10:14 AM
> To: Robert Mckenna
> Cc: jdk8u-dev at openjdk.java.net
> Subject: RE: [8u Backport] RFR: 8175345: possible null pointer dereference defects
> 
> 
> > -----Original Message-----
> > From: Rob McKenna > Sent: Tuesday, May 16, 2017 10:38 PM
> >
> > ...also please add a suitable noreg label to the main bug.
> 
> Done, thank you.
> -Rahul
> 
> >
> >     -Rob
> >
> > On 16/05/17 05:24, Rob McKenna wrote:
> > > Updated subject line to reflect the correct bug id.
> > >
> > > Rahul, for future requests, please only refer to the main bug id.
> > >
> > > Approved
> > >
> > >     -Rob
> > >
> > > On 15/05/17 11:50, Rahul Raghavan wrote:
> > > > Hi,
> > > >
> > > > Request for approval -
> > > > <webrev.01> - http://cr.openjdk.java.net/~rraghavan/8175345/webrev.01/
> > > >
> > > > <jbs> - '49 Null pointer dereference defect groups in 21 files' -
> > > >     https://bugs.openjdk.java.net/browse/JDK-8177429
> > > > Backport of -
> > > >     https://bugs.openjdk.java.net/browse/JDK-8175345
> > > >
> > > > With 8175345, the changes done in jdk9 were reviewed in open and committed.
> > > > <RFR thread> http://mail.openjdk.java.net/pipermail/hotspot-compiler-dev/2017-March/025798.html
> > > > <changeset> http://hg.openjdk.java.net/jdk9/jdk9/hotspot/rev/55e3f1f3d0a7
> > > >
> > > > This backport fix is same as in jdk9.
> > > >
> > > > Thanks,
> > > > Rahul