[foreign-memaccess+abi] RFR: 8264386: LibraryLookup should be more friendly with implicit unloading [v2]

[Maurizio Cimadamore]

On Mon, 29 Mar 2021 18:03:36 GMT, Jorn Vernee <jvernee at openjdk.org> wrote:

>> Maurizio Cimadamore has updated the pull request with a new target base due to a merge or a rebase. The incremental webrev excludes the unrelated changes brought in by the merge/rebase. The pull request contains three additional commits since the last revision:
>> 
>>  - Add @NativeAccess annotations and @CallerSensitive check on LibraryLookup::lookup(String, MemoryLayout)
>>  - Merge branch 'foreign-memaccess+abi' into resourceScope+libraries
>>  - Initial push
>
> src/jdk.incubator.foreign/share/classes/jdk/internal/foreign/LibrariesHelper.java line 119:
> 
>> 117:         @Override
>> 118:         public Optional<MemoryAddress> lookup(String name) {
>> 119:             try {
> 
> I think this is missing the restricted access check?

in that case there's nothing that can go wrong, as we're not assigning fictional bounds to the result.

> src/jdk.incubator.foreign/share/classes/jdk/internal/foreign/LibrariesHelper.java line 131:
> 
>> 129:         @CallerSensitive
>> 130:         public Optional<MemorySegment> lookup(String name, MemoryLayout layout) {
>> 131:             Reflection.ensureNativeAccess(Reflection.getCallerClass());
> 
> Should this method and the other overload have the `@NativeAccess` annotation?

@NativeAccess is for API only (e.g. think javadoc). So the annotations go in `LibraryLookup`

> src/jdk.incubator.foreign/share/classes/jdk/internal/foreign/LibrariesHelper.java line 140:
> 
>> 138:                 }
>> 139:                 return Optional.of(librarySegment.asSlice(addr)
>> 140:                         .asSlice(0L, layout.byteSize()));
> 
> Why not simply one `asSlice` call with `addr` and `layout.byteSize()` ?

my bad, when writing the code the IDE wasn't giving me that choice, so I assumed we forgot to add that combination, and moved on (with a view to file a bug). But now I see the overload is there...

-------------

PR: https://git.openjdk.java.net/panama-foreign/pull/480