[11u] RFR (S) 8209055: c.s.t.javac.code.DeferredCompletionFailureHandler seems to use WeakHashMap incorrectly

Igor Ignatyev igor.ignatyev at oracle.com
Wed Feb 27 21:53:32 UTC 2019


Hi Aleksey,

I tried to find the difference b/w jdk11u and jdk repos, and still don't see why it works w/o 'jdk.compiler/<...>main' in jdk/jdk, but not in jdk11u. anyhow, the fix looks good to me.

Thanks,
-- Igor 

> On Feb 27, 2019, at 12:07 PM, Aleksey Shipilev <shade at redhat.com> wrote:
> 
> Hi Igor,
> 
> On 2/27/19 8:11 PM, Igor Ignatyev wrote:
>> I don't see why 'jdk.compiler/com.sun.tools.javac.main:+open' is needed, could you please explain?
> 
> Otherwise the test would fail to build with:
> 
> /home/shade/trunks/jdk-updates-jdk11u-dev/test/langtools/tools/lib/toolbox/JavacTask.java:365:
> error: Result.exitCode in package com.sun.tools.javac.main is not accessible
>            return taskImpl.doCall().exitCode;
>                                    ^
> The original tests does "jdk.compiler/com.sun.tools.javac.code:+open", which is not enough. I
> thought we can change "code" -> "main", but it would then fail to build with:
> 
> /home/shade/trunks/jdk-updates-jdk11u-dev/test/langtools/tools/javac/processing/model/completionfailure/SymbolsDontCumulate.java:44:
> error: package com.sun.tools.javac.code is not visible
> import com.sun.tools.javac.code.DeferredCompletionFailureHandler;
>                          ^
>  (package com.sun.tools.javac.code is declared in module jdk.compiler, which does not export it to
> the unnamed module)
> 
> Therefore all this is needed:
>  * @modules jdk.compiler/com.sun.tools.javac.api
>  *          jdk.compiler/com.sun.tools.javac.code:+open
>  *          jdk.compiler/com.sun.tools.javac.main:+open
> 
> Does that make sense?
> 
> -Aleksey
> 



More information about the jdk-updates-dev mailing list