The goal here is to avoid duplicate code.
I assume the checking of the flag is a minimal overhead since it only happens when caching, and on the fallback path. Both are not optimized.

Signed-off-by: Stefan Marr <git@stefan-marr.de>

Signed-off-by: Stefan Marr <git@stefan-marr.de>

Signed-off-by: Stefan Marr <git@stefan-marr.de>

The NPE is triggered when accessing an unresolvable outer name.

Signed-off-by: Stefan Marr <git@stefan-marr.de>

Signed-off-by: Stefan Marr <git@stefan-marr.de>

Fabio Niephaus authored 5 years ago and Stefan Marr committed 5 years ago

SubstrateVM can intercept Unsafe access in order to recalculate field offsets before native image generation. Therefore, it is no longer necessary to maintain a feature and a substitution for this purpose.

Closes #327

- update JaCoCo to 0.8.5
- fix class path for extension
- update JVMCI
- update MX

Signed-off-by: Stefan Marr <git@stefan-marr.de>

Signed-off-by: Stefan Marr <git@stefan-marr.de>

- also add JAVA_HOME to clean truffle

Signed-off-by: Stefan Marr <git@stefan-marr.de>

Signed-off-by: Stefan Marr <git@stefan-marr.de>

Signed-off-by: Stefan Marr <git@stefan-marr.de>

- focus methods on main purpose
- use a static trace file to avoid opening trace multiple times
- fold TraceRecord ids into enum and move to own file
- handle statistics in EventParseContext, too
- record subtrace length during scanning to avoid over-reading
- make TraceParser an object and instantiate in VM
  - this change requires some adaptation of primitives
  - for instance, TicksPrim needs to be explicitly a
    BasicPrimitiveOperation now
- make buffer size testable
- move initial trace scan to initialization that’s done explicitly
- don’t swallow exceptions
- store offset and length for context subtrace
- if Subtrace length is not set, also use standardBufferSize
  this should only be relevant for the last subtrace

Signed-off-by: Stefan Marr <git@stefan-marr.de>

Signed-off-by: Stefan Marr <git@stefan-marr.de>

Signed-off-by: Stefan Marr <git@stefan-marr.de>

Signed-off-by: Stefan Marr <git@stefan-marr.de>

Signed-off-by: Stefan Marr <git@stefan-marr.de>

Signed-off-by: Stefan Marr <git@stefan-marr.de>

Signed-off-by: Stefan Marr <git@stefan-marr.de>