Hello,
I like the consolidation of all module information. I don't like that
every time Modules.gmk is included, it calculates the IMPORTED_MODULES.
It also looks like that can't work in the current patch since the call
is before the definition of FindImportedModules.
I can see if I can modify this to have the calculation done lazily or at
least only on demand.
/Erik
On 2016-03-22 01:42, Mandy Chung wrote:
On Mar 21, 2016, at 5:36 PM, Mandy Chung <[email protected]> wrote:
On Mar 21, 2016, at 11:57 AM, Alan Bateman <[email protected]> wrote:
On 21/03/2016 17:32, Mandy Chung wrote:
Erik,
This cleans up several makefiles and now have one single file
(make/common/Modules.gmk) to specify the module-specific information for the
build:
http://cr.openjdk.java.net/~mchung/jdk9/webrevs/8152197/webrev.00/index.html
This looks a good clean-up.
One thing that isn't clear is how to get tool modules into the JRE. I suspect
that JRE_MODULES needs to be $(BOOT_MODULES) + $(PLATFORM_MODULES) +
jdk.policytool + jdk.scripting.nashorn.shell to be consistent with JRE 8 builds.
Good catch. I compared the modules listed in make/Images.gmk before the change
with make/common/Modules.gmk and found some existing issues - jdk.policytool
has not been linked in JRE and I can’t quite tell why jdk.jvmstat and
jdk.jvmstat.rmi are linked in JRE. I believe we only need the following tools
be included in JRE to be consistent with JRE 8 builds:
JRE_TOOL_MODULES += \
jdk.pack200 \
jdk.policytool \
jdk.scripting.nashorn.shell \
Updated webrev:
http://cr.openjdk.java.net/~mchung/jdk9/webrevs/8152197/webrev.01/index.html
In fact, JDK-8071338 has moved policytool from JRE to JDK in JDK 9. webrev.01
updated in place.
Mandy
