RFR 8170289: Re-examine entry point support in jlink

Alan Bateman Alan.Bateman at oracle.com
Mon Dec 19 11:26:31 UTC 2016

On 19/12/2016 04:08, Sundararajan Athijegannathan wrote:

> Fixed. Updated webrev for the record: 
> http://cr.openjdk.java.net/~sundar/8170289/webrev.03
> I'm pushing this as you suggested..
cr.openjdk.java.net is currently down so I'm looking at the changes that 
were pushed to jdk9/dev/jdk. A few comments:

- I assume the usage command should say "Generate launcher ...." so that 
it's clear what the option does.

- The IAE "doesn't contain main class & main not specified in command 
line" probably could be improved to say that it doesn't have a 
ModuleMainClass class attribute or the main class wasn't specified on 
the command line.

- If the input is something like `--launcher 
keytool=java.base/sun/security/tools/keytool/Main` then it looks like it 
will get through the check that tests if the 
sun/security/tools/keytool/Main.class rather than generate a script to 
run `java.base/sun.security.tools.keytool.Main`

The rest looks okay. I think we will need a second phase get rid of the 
scripts and have it use a native launcher instead (the script generation 
was meant to be temporary, including in early prototyping to get 
something working).


More information about the jigsaw-dev mailing list