[jdk16] RFR: 8259276: C2: Empty expression stack when reexecuting tableswitch/lookupswitch instructions after deoptimization

[Vladimir Ivanov]

During parsing of `lookupswitch` and `tableswitch` instructions C2 may insert a safepoint. Corresponding JVM state has always re-execute bit set since the interpreter will unconditionally re-execute the instruction after deoptimization is over (see `AbstractInterpreter::bytecode_should_reexecute` for the full list of instructions).  

But `Parse::do_tableswitch()`/`Parse::do_lookupswitch()` attach wrong JVM state: it describes the state **after** the instruction since the first thing they do is they pop the operand from the expression stack. Instead, the JVM state **before** the instruction should be used to respect the re-execution in the interpreter. 

The bug manifests as a stack corruption after deoptimization at a broken safepoint.

The fix is to keep the initial JVM state (before the instruction) intact until all the safepoints at the particular instruction are inserted.   

Testing:
- [ ] hs-tier1 - hs-tier7 (in progress)
- [ ] Kitchensink24h w/ -XX:+DeoptimizeALot (in progress)

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

Commit messages:
 - 8259276: C2: Empty expression stack when reexecuting tableswitch/lookupswitch instructions after deoptimization

Changes: https://git.openjdk.java.net/jdk16/pull/130/files
 Webrev: https://webrevs.openjdk.java.net/?repo=jdk16&pr=130&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8259276
  Stats: 20 lines in 2 files changed: 13 ins; 3 del; 4 mod
  Patch: https://git.openjdk.java.net/jdk16/pull/130.diff
  Fetch: git fetch https://git.openjdk.java.net/jdk16 pull/130/head:pull/130

PR: https://git.openjdk.java.net/jdk16/pull/130