So ... the original change that removed the mapfiles broke the 32 bit build
because of inconsistency between declarations + definitions of some
functions.
It did not affect 64 bit build because JNICALL is a no-op there.
The next change (8201226) added JNICALL to make it consistent, but
was not reviewed on 2d and was pushed without such review
and may have fixed build issues but did NOT fix the product.
This change now removes JNICALL where it is not needed and fixes the
product and does not break
any builds!
Please confirm that you've run 64+32 bit builds through the build system
AND run imaging tests
on both of those.
After that is confirmed you have my OK.
-phil.
On 05/02/2018 05:03 AM, Magnus Ihse Bursie wrote:
Looks good to me, but you should have a reviewer from the client team as well.
/Magnus
2 maj 2018 kl. 11:52 skrev Alexey Ivanov <alexey.iva...@oracle.com>:
Hi,
Could you please review the following fix for jdk11?
bug: https://bugs.openjdk.java.net/browse/JDK-8202476
webrev: http://cr.openjdk.java.net/~aivanov/8202476/jdk11/webrev.0/
This is a follow-up fix for JDK-8201226 which enabled building JDK for 32 bit
Windows, its code review:
http://mail.openjdk.java.net/pipermail/build-dev/2018-April/021553.html
I found this issue:
http://mail.openjdk.java.net/pipermail/2d-dev/2018-April/009150.html
This fix removes JNICALL modifiers from exported functions in medialib.
After the fix, the failing tests in test/jdk/java/awt/image pass.
Thank you in advance.
Regards,
Alexey
