On 11/29/17 4:37 AM, Alan Bateman wrote:
On 29/11/2017 11:26, David Holmes wrote:
The current approach basically prevents anyone using JVMCI from shooting themselves in the foot by setting
--limit-modules in a way that excludes the jdk.internal.vm.compiler module. In essence we want to ensure that if using
JVMCI then all the requisite pieces will be available. But perhaps we are doing this the wrong way: how do
--limit-modules and --add-modules combine? We could add jdk.internal.vm.compiler rather than expanding the limit-set.
But the end result would seem the same.
The VM shouldn't augment the value specified to --limit-modules so I think we need to do back to the original issue and
find a better solution. Is JDK-8190975 the real issue that we should be looking at? Is it just the two tests listed? In
the case of BootAppendTests then it can check if the module is observable before specifying it to --limit-modules when
creating the child VM. WithSecurityManager might need additional work or maybe it can drop the use of --limit-modules.
There are more AppCDS tests which failed too. Originally they we in closed repo
and not listed. But now they are in open:
http://hg.openjdk.java.net/jdk/hs/file/461e9c898e80/test/hotspot/jtreg/runtime/appcds/jigsaw
16 tests.
As regards using --limit-modules and --add-modules together then it is
possible, the details are in JEP 261.
I tried to add jdk.internal.vm.compiler to --add-modules instead of --limit-modules. But it fails because it will
require to add all Graal's "required" modules too:
http://hg.openjdk.java.net/jdk/hs/file/461e9c898e80/src/jdk.internal.vm.compiler/share/classes/module-info.java#l26
If it acceptable I can do that instead.
But I insist that we should do that in JVM. I don't want to skip tests on SPARC
which have nothing to do with Graal.
Thanks,
Vladimir
-Alan.
