RFR (S) 8211432: [REDO] Handle JNIGlobalRefLocker.cpp

[JC Beyler]

Hi all,

When I pushed the fix for JDK-8210842
<https://bugs.openjdk.java.net/browse/JDK-8210842>, I failed to realize
there was seemingly an issue when using std::string. This provoked a
failure for Windows and Solaris builds described in the bug for this fix.

Therefore I offer a solution using const char* instead to not have any
issues:

Webrev: http://cr.openjdk.java.net/~jcbeyler/8211432/webrev.00/
Bug: https://bugs.openjdk.java.net/browse/JDK-8211432

The incremental between the previous fix and this one is here so that it
helps see the differences:
http://cr.openjdk.java.net/~jcbeyler/8211432/webrev.inc/

Finally, this has passed the submit-repo smoke-screen.

Thanks,
Jc

[Hohensee, Paul]

Is the pointer itself constant? I.e., “const char* const base_msg”?

Paul

From: serviceability-dev <serviceability-dev-***@openjdk.java.net> on behalf of JC Beyler <***@google.com>
Date: Tuesday, October 9, 2018 at 5:23 PM
To: "serviceability-***@openjdk.java.net" <serviceability-***@openjdk.java.net>
Subject: RFR (S) 8211432: [REDO] Handle JNIGlobalRefLocker.cpp

Hi all,

When I pushed the fix for JDK-8210842<https://bugs.openjdk.java.net/browse/JDK-8210842>, I failed to realize there was seemingly an issue when using std::string. This provoked a failure for Windows and Solaris builds described in the bug for this fix.

Therefore I offer a solution using const char* instead to not have any issues:

Webrev: http://cr.openjdk.java.net/~jcbeyler/8211432/webrev.00/
Bug: https://bugs.openjdk.java.net/browse/JDK-8211432

The incremental between the previous fix and this one is here so that it helps see the differences:
http://cr.openjdk.java.net/~jcbeyler/8211432/webrev.inc/

Finally, this has passed the submit-repo smoke-screen.

Thanks,
Jc

[JC Beyler]

Hi Paul,

Yes base_msg can be doubly const so I did that in my local version. I'll
push a new webrev once I get more reviews :)

Thanks!
Jc

Post by Hohensee, Paul
Is the pointer itself constant? I.e., “const char* const base_msg”?
Paul
*Date: *Tuesday, October 9, 2018 at 5:23 PM
*Subject: *RFR (S) 8211432: [REDO] Handle JNIGlobalRefLocker.cpp
Hi all,
When I pushed the fix for JDK-8210842
<https://bugs.openjdk.java.net/browse/JDK-8210842>, I failed to realize
there was seemingly an issue when using std::string. This provoked a
failure for Windows and Solaris builds described in the bug for this fix.
Webrev: http://cr.openjdk.java.net/~jcbeyler/8211432/webrev.00/
Bug: https://bugs.openjdk.java.net/browse/JDK-8211432
The incremental between the previous fix and this one is here so that it
http://cr.openjdk.java.net/~jcbeyler/8211432/webrev.inc/
Finally, this has passed the submit-repo smoke-screen.
Thanks,
Jc

[Alex Menkov]

Looks good.

--alex

Post by JC Beyler
Hi Paul,
Yes base_msg can be doubly const so I did that in my local version. I'll
push a new webrev once I get more reviews :)
Thanks!
Jc
Is the pointer itself constant? I.e., “const char* const base_msg”?____
__ __
Paul____
__ __
*From: *serviceability-dev
*Date: *Tuesday, October 9, 2018 at 5:23 PM
*Subject: *RFR (S) 8211432: [REDO] Handle JNIGlobalRefLocker.cpp____
__ __
Hi all, ____
__ __
When I pushed the fix for JDK-8210842
<https://bugs.openjdk.java.net/browse/JDK-8210842>, I failed to
realize there was seemingly an issue when using std::string. This
provoked a failure for Windows and Solaris builds described in the
bug for this fix.____
__ __
____
__ __
Webrev: http://cr.openjdk.java.net/~jcbeyler/8211432/webrev.00/
<http://cr.openjdk.java.net/%7Ejcbeyler/8211432/webrev.00/>____
Bug: https://bugs.openjdk.java.net/browse/JDK-8211432____
__ __
The incremental between the previous fix and this one is here so
that it helps see the differences:____
http://cr.openjdk.java.net/~jcbeyler/8211432/webrev.inc/
<http://cr.openjdk.java.net/%7Ejcbeyler/8211432/webrev.inc/>____
__ __
Finally, this has passed the submit-repo smoke-screen.____
__ __
Thanks,____
Jc____
--
Thanks,
Jc

[s***@oracle.com]

Hi Jc,

+1

Thanks,
Serguei

Post by Alex Menkov
Looks good.
--alex

Post by JC Beyler
Hi Paul,
Yes base_msg can be doubly const so I did that in my local version.
I'll push a new webrev once I get more reviews :)
Thanks!
Jc
    Is the pointer itself constant? I.e., “const char* const
base_msg”?____
    __ __
    Paul____
    __ __
    *From: *serviceability-dev
    *Date: *Tuesday, October 9, 2018 at 5:23 PM
    *Subject: *RFR (S) 8211432: [REDO] Handle JNIGlobalRefLocker.cpp____
    __ __
    Hi all, ____
    __ __
    When I pushed the fix for JDK-8210842
    <https://bugs.openjdk.java.net/browse/JDK-8210842>, I failed to
    realize there was seemingly an issue when using std::string. This
    provoked a failure for Windows and Solaris builds described in the
    bug for this fix.____
    __ __
    Therefore I offer a solution using const char* instead to not have
    ____
    __ __
    Webrev: http://cr.openjdk.java.net/~jcbeyler/8211432/webrev.00/
<http://cr.openjdk.java.net/%7Ejcbeyler/8211432/webrev.00/>____
    Bug: https://bugs.openjdk.java.net/browse/JDK-8211432____
    __ __
    The incremental between the previous fix and this one is here so
    that it helps see the differences:____
    http://cr.openjdk.java.net/~jcbeyler/8211432/webrev.inc/
<http://cr.openjdk.java.net/%7Ejcbeyler/8211432/webrev.inc/>____
    __ __
    Finally, this has passed the submit-repo smoke-screen.____
    __ __
    Thanks,____
    Jc____
--
Thanks,
Jc