[lworld+vector] RFR: 8320815: [lworld+vector] Fix vector API jtreg crashes with "-XX:InlineFieldMaxFlatSize=0"

[Xiaohong Gong]

On Thu, 7 Dec 2023 03:39:44 GMT, Jatin Bhateja <jbhateja at openjdk.org> wrote:

> Hi @XiaohongGong ,
> 
> I am seeing following assertion failures in Vector API JTREG tests with flattening disabled, not yet spent time in root causing it.
> 
> ```
> # A fatal error has been detected by the Java Runtime Environment:
> #
> #  Internal Error (/home/jatinbha/sandboxes/lworld+vector/valhalla/src/hotspot/share/opto/narrowptrnode.cpp:45), pid=342594, tid=342623
> #  assert(t->isa_narrowoop()) failed: only narrowoop here
> #
> # JRE version: OpenJDK Runtime Environment (22.0) (fastdebug build 22-internal-adhoc.root.valhalla)
> # Java VM: OpenJDK 64-Bit Server VM (fastdebug 22-internal-adhoc.root.valhalla, mixed mode, compressed oops, compressed class ptrs, g1 gc, linux-amd64)
> # Problematic frame:
> # V  [libjvm.so+0x14a9044]  DecodeNNode::Value(PhaseGVN*) const+0x124
> #
> # Core dump will be written. Default location: Core dumps may be processed with "/usr/share/apport/apport -p%p -s%s -c%c -d%d -P%P -u%u -g%g -- %E" (or dumping to /home/jatinbha/sandboxes/lworld+vector/valhalla/build/linux-x86_64-server-fastdebug/test-support/jtreg_test_jdk_jdk_incubator_vector/scratch/1/core.342594)
> #
> # An error report file with more information is saved as:
> # /home/jatinbha/sandboxes/lworld+vector/valhalla/build/linux-x86_64-server-fastdebug/test-support/jtreg_test_jdk_jdk_incubator_vector/scratch/1/hs_err_pid342594.log
> #
> # Compiler replay data is saved as:
> # /home/jatinbha/sandboxes/lworld+vector/valhalla/build/linux-x86_64-server-fastdebug/test-support/jtreg_test_jdk_jdk_incubator_vector/scratch/1/replay_pid342594.log
> #
> # If you would like to submit a bug report, please visit:
> #   https://bugreport.java.com/bugreport/crash.jsp
> #
> ```

Thanks for reporting this issue. Seems the workaround for `VectorBox` in `LoadNode::Identity()` is not right, that if the value is `InlineType` field of `VectorBox`, it returned the `InlineType` not rightly. We should do nothing for such cases. Could you please help check whether it can be fixed by applying following change:

diff --git a/src/hotspot/share/opto/memnode.cpp b/src/hotspot/share/opto/memnode.cpp
index d88ff398e..3ecf780a7 100644
--- a/src/hotspot/share/opto/memnode.cpp
+++ b/src/hotspot/share/opto/memnode.cpp
@@ -1251,12 +1251,13 @@ Node* LoadNode::Identity(PhaseGVN* phase) {
       offset > oopDesc::klass_offset_in_bytes()) {
     Node* value = base->as_InlineType()->field_value_by_offset((int)offset, true);
     if (value != nullptr) {
-      if (Opcode() == Op_LoadN && !base->is_VectorBox()) {
+      if (Opcode() != Op_LoadN) {
+        return value;
+      } else if (!base->is_VectorBox()) {
         // Encode oop value if we are loading a narrow oop
         assert(!phase->type(value)->isa_narrowoop(), "should already be decoded");
-        value = phase->transform(new EncodePNode(value, bottom_type()));
+        return phase->transform(new EncodePNode(value, bottom_type()));
       }
-      return value;
     }
   }


Thanks so much in advance!

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

PR Comment: https://git.openjdk.org/valhalla/pull/958#issuecomment-1844223541