oracle/graal · error · IOException

unknown root :

Error message

unknown root : 

What it means

parseRoot switches on the root-entry type byte (BEGIN_GROUP, CLOSE_GROUP, GRAPH, STREAM_PROPERTIES, ...). The default branch reports a loading error and throws IOException('unknown root : N') when the tag matches no known root entry, meaning the reader has desynchronized from the stream or the producer used an extension this reader does not know.

Source

Thrown at compiler/src/jdk.graal.compiler/src/jdk/graal/compiler/graphio/parsing/BinaryReader.java:907

                case BEGIN_GRAPH: {
                    parseGraph();
                    break;
                }
                case BEGIN_GROUP: {
                    beginGroup();
                    break;
                }
                case CLOSE_GROUP: {
                    doCloseGroup();
                    break;
                }
                case STREAM_PROPERTIES: {
                    loadDocumentProperties();
                    break;
                }
                default:
                    reporter.reportLoadingError(ContextStrings.unknownObjectType(type), builder);
                    throw new IOException("unknown root : " + type);
            }
        } catch (SkipRootException ex) {
            long s = ex.getStart();
            long e = ex.getEnd();
            long pos = dataSource.getMark();
            ConstantPool pool = ex.getConstantPool();
            if (e == -1) {
                // skip rest of the stream
                throw new EOFException();
            }
            instLog.log(Level.FINE, "Skipping to offset {0}, {1} bytes skipped, using cpool", new Object[]{e, e - pos, Integer.toHexString(pool == null ? 0 : System.identityHashCode(pool))});

            assert s < pos && e >= pos;
            if (pos < e) {
                long count = e - pos;
                byte[] scratch = new byte[(int) Math.min(count, 1024 * 1024 * 50)];
                while (count > 0) {
                    int l = (int) Math.min(scratch.length, count);

View on GitHub (pinned to a66e9ccd1d)

Solutions

1. Check the numeric tag in the message against the writer's root-entry constants to see if it is a known-but-newer entry type.
2. Upgrade the reader to a graphio version >= the writer's.
3. Regenerate the dump if corruption is suspected; validate the file digest/size.

try { reader.parse(); } catch (IOException e) { if (e.getMessage().startsWith("unknown root")) { /* parse desync or newer entry type: upgrade reader / re-dump */ } }

When it happens

Trigger: Corruption or truncation earlier in the stream shifting the position so a data byte is read as a root tag; reading a file with root-entry types added by a newer/custom writer; wrong stream offset after skip handling.

Common situations: Mixed-version tooling (new writer, old reader/IGV), damaged files, or spliced streams that lost alignment.

Related errors

• Document properties unexpected in version < 7
• Invalid constant pool index :
• unknown klass type :
• unknown pool type
• Unknown type

AI-assisted analysis of oracle/graal@a66e9ccd1d (2026-08-14). Data as JSON: /api/errors/d1fd9f836fc3e9af. Report an issue: GitHub.