[8u] Request for approval: 8087120: [GCC5] java.lang.StackOverflowError on Zero JVM initialization on non x86 platforms

[Andrew Hughes]

----- Original Message -----
> Related 8074552 and 8075533 are not backported to 8u.
> So I would suggest to push your point fix with new bug id (and link it to
> 8075967)
> 
> https://bugs.openjdk.java.net/browse/JDK-8074552
> https://bugs.openjdk.java.net/browse/JDK-8075533
> 

Yes, I agree that sounds like the safest option. My impression of the original
changeset is that this really should been a separate bug to begin with; the
change is unrelated to the rest. I'll prepare such a bug tomorrow.

> Thanks,
> Vladimir
> 
> On 11/19/15 7:18 PM, Andrew Hughes wrote:
> > ----- Original Message -----
> >> http://hg.openjdk.java.net/jdk8u/jdk8u-dev/hotspot
> >>
> >> On 11/19/15 9:21 AM, Severin Gehwolf wrote:
> >>> On Thu, 2015-11-19 at 15:43 +0000, Rob McKenna wrote:
> >>>> Correct. You need a sponsor. Perhaps Andrew would be willing to help?
> >>>> Failing that please work with the hotspot group.
> >>>
> >>> Thanks, Rob.
> >>>
> >>> Hotspot people: To which JDK 8 tree should this get pushed? Andrew,
> >>> tells me that he can push it for me once we know where :)
> >>>
> >>> Thanks,
> >>> Severin
> >>>
> >>>> 	-Rob
> >>>>
> >>>> On 19/11/15 09:41, Severin Gehwolf wrote:
> >>>>> On Wed, 2015-11-18 at 21:08 +0000, Rob McKenna wrote:
> >>>>>> Approved. Please add a suitable noreg label to the bug.
> >>>>>>
> >>>>>> 	-Rob
> >>>>>
> >>>>> Another question: Wouldn't I need a sponsor to get this integrated?
> >>>>> Is
> >>>>> this being taken care of automatically?
> >>>>>
> >>>>> Thanks for your help!
> >>>>>
> >>>>> Cheers,
> >>>>> Severin
> >>>>>
> >>>>>> On 18/11/15 10:50, Severin Gehwolf wrote:
> >>>>>>> Hi,
> >>>>>>>
> >>>>>>> I'd like to ask for approval of a Zero-specific hotspot 8
> >>>>>>> backport.
> >>>>>>> Currently, the jdk8u forest fails to build the Zero variant
> >>>>>>> with
> >>>>>>> GCC 5
> >>>>>>> due to this:
> >>>>>>>
> >>>>>>> src/os_cpu/linux_zero/vm/os_linux_zero.cpp:59:10: error:
> >>>>>>> function
> >>>>>>> returns address of local variable [-Werror=return-local-addr]
> >>>>>>>        return dummy;
> >>>>>>>               ^
> >>>>>>> src/os_cpu/linux_zero/vm/os_linux_zero.cpp:58:11: note:
> >>>>>>> declared
> >>>>>>> here
> >>>>>>>        address dummy = (address) &dummy;
> >>>>>>>                ^
> >>>>>>> Note: Fedora 22+ have GCC 5. The above problem and the actual
> >>>>>>> overflow
> >>>>>>> bug (if warning is ignored at build time + on relevant arch)
> >>>>>>> goes
> >>>>>>> away
> >>>>>>> when JDK-8087120 gets backported to 8u.
> >>>>>>>
> >>>>>>> Original Bug: https://bugs.openjdk.java.net/browse/JDK-8087120
> >>>>>>> 9 webrev: http://cr.openjdk.java.net/~sgehwolf/webrevs/JDK-8087
> >>>>>>> 120/
> >>>>>>> webrev.01/
> >>>>>>> 9 review thread: http://mail.openjdk.java.net/pipermail/hotspot
> >>>>>>> -dev
> >>>>>>> /2015-June/018884.html
> >>>>>>>
> >>>>>>> The 9 webrev applies with no change to the 8 forest. I'd need
> >>>>>>> somebody
> >>>>>>> to sponsor this for me.
> >>>>>>>
> >>>>>>> Thanks,
> >>>>>>> Severin
> >>>>>>>
> >>>>>
> >>>
> >>
> >
> > Built:
> >
> > $ /mnt/builder/jdk8u-dev/images/j2sdk-image/bin/java  -version
> > openjdk version "1.8.0-internal"
> > OpenJDK Runtime Environment (build
> > 1.8.0-internal-andre_2015_11_20_01_48-b00)
> > OpenJDK 64-Bit Zero VM (build 25.66-b00, interpreted mode)
> >
> > with GCC 5.1.0 and pushed the change:
> >
> > http://hg.openjdk.java.net/jdk8u/jdk8u-dev/hotspot/rev/c6ef40024aa2
> >
> > An extra fix was needed to make it build, taken from 8075967:
> >
> > diff -r c6ef40024aa2 src/cpu/zero/vm/frame_zero.cpp
> > --- a/src/cpu/zero/vm/frame_zero.cpp	Fri Jun 12 16:09:45 2015 +0100
> > +++ b/src/cpu/zero/vm/frame_zero.cpp	Fri Nov 20 03:17:01 2015 +0000
> > @@ -1,5 +1,5 @@
> >   /*
> > - * Copyright (c) 2003, 2013, Oracle and/or its affiliates. All rights
> > reserved.
> > + * Copyright (c) 2003, 2015, Oracle and/or its affiliates. All rights
> > reserved.
> >    * Copyright 2007, 2008, 2009, 2010, 2011 Red Hat, Inc.
> >    * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
> >    *
> > @@ -216,7 +216,7 @@
> >       valuebuf[buflen - 1] = '\0';
> >
> >       // Print the result
> > -    st->print_cr(" " PTR_FORMAT ": %-21s = %s", addr, fieldbuf, valuebuf);
> > +    st->print_cr(" " PTR_FORMAT ": %-21s = %s", p2i(addr), fieldbuf,
> > valuebuf);
> >     }
> >   }
> >
> > I don't know if 8075967 should be backported in full (it contains other
> > changes,
> > and this seems to have been thrown in there to make it build):
> >
> > 8075967: Zero interpreter asserts for SafeFetch<32,N> calls in
> > ObjectMonitor
> > Summary: Implement SafeFetchX unsafely and make CanUseSafeFetchX false for
> > Zero
> > Reviewed-by: sgehwolf, dholmes
> >
> > or we should backport just the above under a new bug ID. Thoughts?
> >
> > Thanks,
> >
> 

-- 
Andrew :)

Senior Free Java Software Engineer
Red Hat, Inc. (http://www.redhat.com)

PGP Key: ed25519/35964222 (hkp://keys.gnupg.net)
Fingerprint = 5132 579D D154 0ED2 3E04  C5A0 CFDA 0F9B 3596 4222

PGP Key: rsa4096/248BDC07 (hkp://keys.gnupg.net)
Fingerprint = EC5A 1F5E C0AD 1D15 8F1F  8F91 3B96 A578 248B DC07