Looks fine.
On 2/6/20 1:39 pm, Alexey Ivanov wrote:
Hello once again,
After trying a few things suggested by Sergey, I found serious issues in the
new approaches although the code looked a bit cleaner. Resolving those issues
requires substantial changes to the way we handle the active and passive DC
lists.
Thus, it's safer to postpone further refactoring to a later time.
So we're back to the original proposal.
I'm quoting it for your convenience below:
JBS: https://bugs.openjdk.java.net/browse/JDK-8235147
Webrev: http://cr.openjdk.java.net/~aivanov/8235147/webrev.00/
When a DC becomes unused, it is moved from activeDCList to passiveDCList to be
released later. It will stay in the passiveDCList until a new DC is requested
for the same window or the window is destroyed.
The DCs in passiveDCList will never be used again, so we can release them as
soon as another request for a DC comes in.
The test case, ManyFramesTest.java, attached to the bug creates and displays 10
frames. Then it hides the 9 frames. At this point, passiveDCList contains 9
entries. These DCs will not be released until the corresponding frame is shown
and painted again or the frame is destroyed.
The fix releases all DCs currently in passiveDCList when another DC is
requested.
With the fix, when running ManyFramesTest.java test case, the number of entries
in passiveDCList does not grow, the list contains at most 1 entry.
On 12/12/2019 20:46, Alexey Ivanov wrote:
Hi Sergey,
<SNIP>
It does not work good. I found serious issues in either approach I've tried so
far.
Thus, we're down to the original proposal:
JBS: https://bugs.openjdk.java.net/browse/JDK-8235147
Webrev: http://cr.openjdk.java.net/~aivanov/8235147/webrev.00/
--
Best regards, Sergey.
