Files
rea/tests/boundary/cli/reverseEngineeringCommands.test.ts
Kaoru0822andClaude Opus 5.5 144843ccc2 fix(debugger): let REA exit when a debugger could not be stopped (#8652)
* fix(debugger): let REA exit when a debugger could not be stopped

`rea debug-with-gdb` and `rea debug-with-rizin` printed cleanup_incomplete
and set exit status 1, but the retained child and its pipes kept the event
loop alive, so the process never exited. A stdio MCP server whose client
disconnected hung the same way.

closeOwnedDebuggerSessions is the owner both managers finish through. It now
unreferences retained sessions, as supervised providers already do after
incomplete cleanup, restores the references on retry, and names the retained
process IDs. The one-shot commands report those IDs, since no later command
can retry. Debugger ownership, signals and detach behavior are unchanged.

GDB cleanup requests also no longer hold the host until their 120 s command
deadline, because cleanup already bounds its own wait.

Co-Authored-By: Claude Opus 5.5 <[email protected]>

* fix(debugger): name retained debuggers when an MCP server shuts down

An MCP server that exits with a retained debugger logged only "REA provider
cleanup failed" and "could not close cleanly", so nothing named the process
left for the caller to check. The PIDs were on OwnedDebuggerCleanupError, but
each shutdown reporter read only the outer aggregate's message, and the debug
failure_cause description dropped the field too.

The failure type now lives in its own dependency-free module with one reader
that walks composed failures. The shutdown report, the server's onclose log,
the debug description and the one-shot CLI all use it, and the shutdown report
writes the same remediation the CLI returns.

The fake debuggers now model each request exactly, with real reply shapes, and
a request they do not model fails the test.

Co-Authored-By: Claude Opus 5.5 <[email protected]>

* test(debugger): compare retained-debugger guidance with its formatter

The new checks froze the wording "Check process ... and its target".
docs/testing.md asks guidance checks to compare with the owning formatter,
alongside independent checks of the facts. They now do, and each still proves
the process ID is named, because a formatter comparison alone would pass if the
formatter dropped it.

Co-Authored-By: Claude Opus 5.5 <[email protected]>

---------

Co-authored-by: Claude Opus 5.5 <[email protected]>
2026-11-02 05:16:34 -08:00

670 lines
21 KiB
TypeScript

import pino from "pino";
import { Cli, z } from "incur";
import { afterEach, expect, it, vi } from "vitest";
import { rm } from "node:fs/promises";
import { registerReverseEngineeringCommands } from "../../../src/cli/reverseEngineeringCommands.js";
import { CLI_COMMANDS } from "../../../src/cliCommandNames.js";
import {
AnalysisCancelledError,
AnalysisTimeoutError,
AnalysisOutputError,
} from "../../../src/domain/analysisErrorCore.js";
import {
createEvidence,
evidenceSchema,
} from "../../../src/domain/evidence.js";
import { err, ok } from "../../../src/domain/result.js";
import { RizinDebugSessionManager } from "../../../src/rizin/RizinDebugSessionManager.js";
import { closeOwnedDebuggerSessions } from "../../../src/process/ownedDebuggerSession.js";
import { ReverseEngineeringService as CoreReverseEngineeringService } from "../../../src/application/reverse/ReverseEngineeringService.js";
import { createReverseEngineeringCommandPort } from "../../../src/composition/reverseEngineering.js";
import { OwnedCommandFailure } from "../../../src/process/OwnedCommand.js";
import type { ProviderProcessSupervisor } from "../../../src/process/ProviderProcess.js";
import {
createArtifact,
providerSnapshot,
} from "../../support/reverseEngineeringCompositionFixtures.js";
import type { CliInstance } from "../../../src/cli/types.js";
import type { ReverseEngineeringService } from "../../../src/application/reverse/ReverseEngineeringService.js";
const originalExitCode = process.exitCode;
afterEach(() => {
process.exitCode = originalExitCode;
});
const observation = (completionStatus: "complete" | "unknown") =>
createEvidence(
undefined,
{ id: "gnu.objdump", name: "GNU objdump", version: null },
{
predicateType: "rea.reverse-engineering.command-observation",
operation: "inspect_with_objdump",
parameters: { operation: "file_headers" },
result: { completion_status: completionStatus },
rawResult: { completion_status: completionStatus },
},
);
const runObjdump = async (
service: Partial<ReverseEngineeringService>,
path = "/tmp/sample",
selection: {
operation?: string;
symbol?: string;
commandTimeoutMs?: number;
} = {},
) => {
const commands = new Map<
string,
{
run(input: {
args: { path: string; operation: string };
options: {
followDebugLinks: boolean;
symbol?: string;
commandTimeoutMs?: number;
};
}): Promise<unknown>;
}
>();
const cli = {
command: (
name: string,
definition: {
run(input: {
args: { path: string; operation: string };
options: {
followDebugLinks: boolean;
symbol?: string;
commandTimeoutMs?: number;
};
}): Promise<unknown>;
},
) => {
commands.set(name, definition);
return cli;
},
} as unknown as CliInstance;
registerReverseEngineeringCommands(
cli,
pino({ enabled: false }),
{},
undefined,
() => service as ReverseEngineeringService,
);
const command = commands.get(CLI_COMMANDS.inspectWithObjdump);
if (command === undefined)
throw new Error("objdump CLI command was not registered");
return command.run({
args: { path, operation: selection.operation ?? "file_headers" },
options: {
followDebugLinks: false,
...(selection.symbol === undefined ? {} : { symbol: selection.symbol }),
...(selection.commandTimeoutMs === undefined
? {}
: { commandTimeoutMs: selection.commandTimeoutMs }),
},
});
};
it("forwards the selected native symbol to the shared workflow", async () => {
const inspectWithObjdump = vi.fn(async (_input: unknown) =>
ok(observation("complete")),
);
await runObjdump(
{ inspectWithObjdump, close: async () => undefined },
"/tmp/sample",
{ operation: "disassemble", symbol: "_Z3foov@@VERSION_1" },
);
expect(inspectWithObjdump.mock.calls[0]?.[0]).toEqual({
path: "/tmp/sample",
operation: "disassemble",
follow_debug_links: false,
symbol: "_Z3foov@@VERSION_1",
});
});
it("forwards a selected objdump command deadline and omits the field otherwise", async () => {
const inspectWithObjdump = vi.fn(async (_input: unknown) =>
ok(observation("complete")),
);
const service = { inspectWithObjdump, close: async () => undefined };
await runObjdump(service, "/tmp/sample", { commandTimeoutMs: 5_000 });
expect(inspectWithObjdump.mock.calls[0]?.[0]).toEqual({
path: "/tmp/sample",
operation: "file_headers",
follow_debug_links: false,
command_timeout_ms: 5_000,
});
await runObjdump(service, "/tmp/sample");
expect(inspectWithObjdump.mock.calls[1]?.[0]).toEqual({
path: "/tmp/sample",
operation: "file_headers",
follow_debug_links: false,
});
});
it("lists objdump architectures without inventing an artifact path", async () => {
const evidence = observation("complete");
const inspectWithObjdump = vi.fn(async (_input: unknown) => ok(evidence));
const commands = new Map<
string,
{
readonly args?: { safeParse(input: unknown): { success: boolean } };
run(input: { args?: unknown; options?: unknown }): Promise<unknown>;
}
>();
const cli = {
command: (_name: string, definition: unknown) => {
commands.set(
_name,
definition as {
readonly args?: {
safeParse(input: unknown): { success: boolean };
};
run(input: { args?: unknown; options?: unknown }): Promise<unknown>;
},
);
return cli;
},
} as unknown as CliInstance;
registerReverseEngineeringCommands(
cli,
pino({ enabled: false }),
{},
undefined,
() =>
({
inspectWithObjdump,
close: vi.fn(async () => undefined),
}) as unknown as ReverseEngineeringService,
);
const command = commands.get(CLI_COMMANDS.listObjdumpArchitectures);
if (command === undefined)
throw new Error("architecture command was not registered");
const result = await command.run({ options: { detail: "complete" } });
expect(result).toEqual(evidence);
expect(inspectWithObjdump).toHaveBeenCalledWith(
{ operation: "architectures" },
expect.objectContaining({ signal: expect.any(AbortSignal) }),
);
const artifactCommand = commands.get(CLI_COMMANDS.inspectWithObjdump);
expect(
artifactCommand?.args?.safeParse({
path: "/tmp/sample",
operation: "dwarf",
}).success,
).toBe(true);
expect(
artifactCommand?.args?.safeParse({
path: "/tmp/sample",
operation: "architectures",
}).success,
).toBe(false);
});
it("promotes only complete production cleanup-only Evidence after CLI cleanup retry", async () => {
const { directory, path } = await createArtifact();
const owner = {
stop: vi.fn(async () => ({ status: "verified-cleanup" as const })),
} as unknown as ProviderProcessSupervisor;
const service = new CoreReverseEngineeringService({
commands: createReverseEngineeringCommandPort({}),
runOwned: vi.fn(async () => {
throw new OwnedCommandFailure(
"process",
"Cleanup after process completion was incomplete.",
providerSnapshot(),
"provider process still running",
undefined,
["owned-run"],
owner,
);
}),
});
const inspect = service.inspectWithObjdump.bind(service);
let producedCleanupFailure: AnalysisOutputError | undefined;
let cleanupIncompleteBeforeClose: boolean | undefined;
let completionBeforeClose: unknown;
vi.spyOn(service, "inspectWithObjdump").mockImplementation(
async (...args) => {
const result = await inspect(...args);
if (!result.ok && result.error instanceof AnalysisOutputError) {
producedCleanupFailure = result.error;
cleanupIncompleteBeforeClose = result.error.cleanupIncomplete;
const partialEvidence = evidenceSchema.safeParse(
result.error.partialObservation,
);
if (partialEvidence.success)
completionBeforeClose = partialEvidence.data.normalized_result;
}
return result;
},
);
try {
const result = await runObjdump(service, path);
expect(result).toMatchObject({
provider: { id: "gnu.objdump" },
normalized_result: { completion_status: "complete" },
});
expect(producedCleanupFailure?.cleanupOnly).toBe(true);
expect(cleanupIncompleteBeforeClose).toBe(true);
expect(completionBeforeClose).toMatchObject({
completion_status: "complete",
});
expect(producedCleanupFailure).toMatchObject({
cleanupIncomplete: false,
cleanup: { resources: ["owned-run"] },
partialObservation: {
normalized_result: { completion_status: "complete" },
},
});
expect(owner.stop).toHaveBeenCalledOnce();
} finally {
await service.close().catch(() => undefined);
await rm(directory, { recursive: true, force: true });
}
});
it("preserves an operation timeout and unknown partial Evidence after cleanup retry", async () => {
const { directory, path } = await createArtifact();
const owner = {
stop: vi.fn(async () => ({ status: "verified-cleanup" as const })),
} as unknown as ProviderProcessSupervisor;
const service = new CoreReverseEngineeringService({
commands: createReverseEngineeringCommandPort({}),
runOwned: vi.fn(async () => {
throw new OwnedCommandFailure(
"timeout",
"Provider command timed out.",
providerSnapshot({ exitCode: null }),
"provider process still running",
undefined,
["owned-run"],
owner,
);
}),
});
let timeout: AnalysisTimeoutError | undefined;
const inspect = service.inspectWithObjdump.bind(service);
vi.spyOn(service, "inspectWithObjdump").mockImplementation(
async (...args) => {
const result = await inspect(...args);
if (!result.ok && result.error instanceof AnalysisTimeoutError)
timeout = result.error;
return result;
},
);
try {
const result = await runObjdump(service, path);
expect(result).toMatchObject({
code: "provider_timeout",
details: {
partial_observation: {
normalized_result: { completion_status: "unknown" },
},
},
});
expect(timeout).toBeInstanceOf(AnalysisTimeoutError);
expect(timeout?.cleanupIncomplete).toBe(false);
expect(owner.stop).toHaveBeenCalledOnce();
} finally {
await service.close().catch(() => undefined);
await rm(directory, { recursive: true, force: true });
}
});
it("preserves an independent timeout after cleanup succeeds", async () => {
const evidence = observation("unknown");
const error = new AnalysisTimeoutError("inspect_with_objdump", 1_000, {
cleanup: { reason: "first stop incomplete", resources: ["owned-run"] },
partialObservation: evidence,
});
const service = {
inspectWithObjdump: vi.fn(async () => err(error)),
close: vi.fn(async () => {
error.resolveCleanup();
}),
};
const result = await runObjdump(service);
expect(result).toMatchObject({
code: "provider_timeout",
details: {
partial_observation: {
normalized_result: { completion_status: "unknown" },
},
},
});
expect(error.cleanupIncomplete).toBe(false);
expect((result as { details: unknown }).details).not.toHaveProperty(
"cleanup",
);
});
it("does not project cancellation cleanup as incomplete after a successful retry", async () => {
const error = new AnalysisCancelledError("inspect_with_objdump", {
cleanup: { reason: "first stop incomplete", resources: ["owned-run"] },
partialObservation: observation("unknown"),
});
const service = {
inspectWithObjdump: vi.fn(async () => err(error)),
close: vi.fn(async () => {
error.resolveCleanup();
}),
};
const result = await runObjdump(service);
expect(error.cleanupIncomplete).toBe(false);
expect(result).toMatchObject({
code: "cancelled",
details: { cleanup: "complete" },
});
});
it("returns cancellation with the completed observation when the caller interrupts service close", async () => {
const evidence = observation("complete");
const service = {
inspectWithObjdump: vi.fn(async () => ok(evidence)),
close: vi.fn(async () => {
process.emit("SIGINT");
}),
};
const result = await runObjdump(service);
expect(result).toMatchObject({
code: "cancelled",
details: { partial_observation: evidence },
});
expect(process.exitCode).toBe(130);
});
it.each([false, true])(
"passes the Rizin TTY and uses final cleanup outcome (retry=%s)",
async (retryClose) => {
process.exitCode = 0;
const commands = new Map<
string,
{
run(input: {
args: Record<string, unknown>;
options: Record<string, unknown>;
}): Promise<unknown>;
}
>();
const cli = {
command: (
name: string,
definition: {
run(input: {
args: Record<string, unknown>;
options: Record<string, unknown>;
}): Promise<unknown>;
},
) => {
commands.set(name, definition);
return cli;
},
} as unknown as CliInstance;
const evidence = createEvidence(
undefined,
{ id: "rizin.debugger", name: "Rizin debugger", version: null },
{
predicateType: "rea.debugger.command-observation",
operation: "rizin_debug_command",
parameters: {
session_id: "session-1",
command: "dr",
inferior_tty: "/dev/pts/7",
},
result: { output: "registers" },
rawResult: { output: "registers" },
},
);
const start = vi
.spyOn(RizinDebugSessionManager.prototype, "start")
.mockResolvedValue(ok({ session_id: "session-1", backend: null }));
const execute = vi
.spyOn(RizinDebugSessionManager.prototype, "execute")
.mockResolvedValue(
ok({
evidence,
result: {
command: "dr",
output: "registers",
output_scope: "command_and_interleaved_session_output",
output_truncated: false,
completion_status: "complete",
backend: null,
},
}),
);
vi.spyOn(RizinDebugSessionManager.prototype, "close").mockResolvedValue(
retryClose
? err(
new AnalysisOutputError(
"close_rizin_debug_session",
"first close incomplete",
{
cleanup: { reason: "retry pending", resources: ["session-1"] },
},
),
)
: ok({ target_state: "unknown" }),
);
vi.spyOn(RizinDebugSessionManager.prototype, "closeAll").mockResolvedValue(
undefined,
);
registerReverseEngineeringCommands(cli, pino({ enabled: false }), {});
const command = commands.get(CLI_COMMANDS.debugWithRizin);
if (command === undefined)
throw new Error("Rizin CLI command was not registered");
await command.run({
args: { path: "./sample.bin", command: "dr" },
options: { inferiorTty: "/dev/pts/7" },
});
expect(process.exitCode).toBe(0);
expect(start).toHaveBeenCalledWith(
{ path: "./sample.bin", command: "dr", inferior_tty: "/dev/pts/7" },
expect.any(AbortSignal),
);
expect(execute).toHaveBeenCalledWith(
"session-1",
"dr",
expect.any(AbortSignal),
);
const parserCli = Cli.create("rea-test", { version: "1.0.0", sync: false });
registerReverseEngineeringCommands(
parserCli as unknown as CliInstance,
pino({ enabled: false }),
{},
);
const commandContract = Cli.toCommands
.get(parserCli)
?.get(CLI_COMMANDS.debugWithRizin);
expect(commandContract).toBeDefined();
const argsSchema =
commandContract === undefined
? undefined
: Reflect.get(commandContract, "args");
const optionsSchema =
commandContract === undefined
? undefined
: Reflect.get(commandContract, "options");
if (
!(argsSchema instanceof z.ZodObject) ||
!(optionsSchema instanceof z.ZodObject)
)
throw new Error("Rizin CLI args/options schemas were not registered");
expect(argsSchema.parse({ path: "./sample.bin", command: "dr" })).toEqual({
path: "./sample.bin",
command: "dr",
});
expect(optionsSchema.parse({ inferiorTty: "/dev/pts/7" })).toMatchObject({
inferiorTty: "/dev/pts/7",
});
},
);
it.each(["incomplete", "interrupt", "control"] as const)(
"reports final Rizin cleanup %s as a typed outcome",
async (mode) => {
const commands = new Map<
string,
{ run(input: unknown): Promise<unknown> }
>();
const cli = {
command(
name: string,
definition: { run(input: unknown): Promise<unknown> },
) {
commands.set(name, definition);
return cli;
},
} as unknown as CliInstance;
const previous = process.exitCode;
const evidence = createEvidence(
undefined,
{ id: "rizin.debugger", name: "Rizin debugger", version: null },
{
predicateType: "rea.debugger.command-observation",
operation: "rizin_debug_command",
parameters: {
session_id: "fixture-session",
command: "dr",
inferior_tty: "/dev/pts/7",
},
result: { output: "registers" },
rawResult: { output: "registers" },
},
);
const sessions = new Map([
[
"fixture-session",
{
id: "fixture-session",
launched: { process: { pid: 4242 } },
supervisor: { ref: vi.fn(), unref: vi.fn() },
},
],
]);
vi.spyOn(RizinDebugSessionManager.prototype, "start").mockResolvedValue(
ok({ session_id: "fixture-session", backend: null }),
);
const execute = vi
.spyOn(RizinDebugSessionManager.prototype, "execute")
.mockResolvedValue(
ok({
evidence,
result: {
command: "dr",
output: "registers",
output_scope: "command_and_interleaved_session_output",
output_truncated: false,
completion_status: "complete",
backend: null,
},
}),
);
vi.spyOn(RizinDebugSessionManager.prototype, "close").mockImplementation(
async () => {
if (mode === "incomplete")
return err(
new AnalysisOutputError(
"close_rizin_debug_session",
"stop unconfirmed",
{
cleanup: {
reason: "stop unconfirmed",
resources: ["fixture-session"],
},
},
),
);
sessions.clear();
return ok({ target_state: "unknown" });
},
);
let abortedAtCleanupEnd = false;
let signal: AbortSignal | undefined;
const closeAll = vi
.spyOn(RizinDebugSessionManager.prototype, "closeAll")
.mockImplementation(async () => {
signal = execute.mock.calls[0]?.[2];
await new Promise<void>((resolve) =>
setImmediate(() => {
if (mode === "interrupt") process.emit("SIGINT");
resolve();
}),
);
abortedAtCleanupEnd = signal?.aborted === true;
await closeOwnedDebuggerSessions(
sessions,
async () => ({ status: "incomplete", reason: "stop unconfirmed" }),
"Rizin",
);
});
registerReverseEngineeringCommands(cli, pino({ enabled: false }), {});
const command = commands.get(CLI_COMMANDS.debugWithRizin);
if (command === undefined)
throw new Error("Rizin CLI command was not registered");
try {
process.exitCode = 0;
const result = await command.run({
args: { path: "./sample.bin", command: "dr" },
options: { inferiorTty: "/dev/pts/7" },
});
expect(closeAll).toHaveBeenCalledOnce();
if (mode === "control") {
expect(result).toBe(evidence);
expect(process.exitCode).toBe(0);
expect(abortedAtCleanupEnd).toBe(false);
} else if (mode === "interrupt") {
expect(result).toMatchObject({
code: "cancelled",
details: {
partial_observation: {
operation: "rizin_debug_command",
parameters: { session_id: "fixture-session" },
},
},
});
expect(process.exitCode).toBe(130);
expect(abortedAtCleanupEnd).toBe(true);
} else {
expect(result).toMatchObject({
code: "cleanup_incomplete",
details: {
cleanup: "incomplete",
resources: ["fixture-session"],
diagnostics: { retained_debugger_pids: [4242] },
partial_observation: {
operation: "rizin_debug_command",
parameters: { session_id: "fixture-session" },
},
},
});
expect(sessions.size).toBe(1);
expect(process.exitCode).toBe(1);
}
} finally {
process.exitCode = previous;
vi.restoreAllMocks();
}
},
);