deepseek-harness/packages/subprocess/subprocess-local/src/spawn.ts

466 lines
18 KiB
TypeScript
Raw Normal View History

/**
* Process plumbing for the local subprocess service: detached process-tree
* spawn with per-stream stdio dispositions, tail-keep collection with spill
* files, tree-scoped signalling (POSIX groups; Windows taskkill), and the
* SIGTERM→SIGKILL escalation. This layer reacts to an abort signal; callers
* own deadlines, teardown ladders, and cause classification.
* @module dsh-subprocess-local/spawn
*/
import { type ChildProcess, spawn, spawnSync } from 'node:child_process'
import type { Readable } from 'node:stream'
import { randomBytes } from 'node:crypto'
import { closeSync, mkdtempSync, openSync, unlinkSync, writeSync } from 'node:fs'
import { tmpdir } from 'node:os'
import { join } from 'node:path'
2026-07-26 16:50:36 +08:00
import { setTimeout as sleepMs } from 'node:timers/promises'
subprocess: one explicit env channel on the spawn spec Drop SubprocessSpawnSpec.dshEnv and splitEnvChannels(); childEnv() is now scrubbed-base + explicit entries with no namespace validation. The invariant dropped is the reserved-namespace check on explicit entries (DSH_* rejected from env, non-DSH_* rejected from dshEnv). Explicit-entry trust already covers it: an explicit credential-shaped entry has always merged after the scrub as a deliberate caller opt-in, and an explicit DSH_* entry is the same deliberate act — the staleness invariant lives entirely in scrubbedParentEnv dropping AMBIENT credential-shaped and DSH_* names, which stays. The validation's only observed effect was rejecting legitimate explicit entries: both recent CI breakages (DSH_GATE_CONCURRENCY exported into every job crashing lsp specs, DSH_PERMISSION_MODE in acp config.env crashing the child spawn) were this check firing on values a caller meant to pass, each fixed by routing around the bureaucracy the seam itself imposed. The bash seam keeps its own request/spec dshEnv field: that is bash-owned trusted-plugin vocabulary (the ctx.bashEnv collected overlay) whose merge-last position guarantees a caller env entry cannot displace a managed fact; bash-local now flattens ENV_OVERRIDES -> spec.env -> spec.dshEnv into the seam's one env map. subagent-acp and lsp-local pass their single config env map straight through. DshEnvironment/DshEnvironmentKey/DSH_ENV_PREFIX stay on the subprocess seam as the namespace vocabulary (bash re-exports them; scrubbedParentEnv filters on the prefix). Tests: the two channel-rejection specs and the splitEnvChannels partition spec are deleted; one spawn spec now proves an explicit DSH_* env entry reaches the child while an ambient one is scrubbed; the acp/lsp forwarding specs keep their MOCK_ECHO_ENV / LSP_FAKE_ECHO_ENV assertions with the split comments rewritten to merge-after-scrub. Docs (en+zh, re-recorded) and the owning Agent Notes updated; cordis api/services catalogs regenerated.
2026-07-27 04:14:51 +08:00
import { scrubbedParentEnv } from '@deepseek-ai/dsh-subprocess'
import type {
CollectedOutput,
SubprocessCollect,
SubprocessHandle,
SubprocessOutcome,
SubprocessOutputMode,
SubprocessSpawnSpec,
} from '@deepseek-ai/dsh-subprocess'
/**
subprocess: one explicit env channel on the spawn spec Drop SubprocessSpawnSpec.dshEnv and splitEnvChannels(); childEnv() is now scrubbed-base + explicit entries with no namespace validation. The invariant dropped is the reserved-namespace check on explicit entries (DSH_* rejected from env, non-DSH_* rejected from dshEnv). Explicit-entry trust already covers it: an explicit credential-shaped entry has always merged after the scrub as a deliberate caller opt-in, and an explicit DSH_* entry is the same deliberate act — the staleness invariant lives entirely in scrubbedParentEnv dropping AMBIENT credential-shaped and DSH_* names, which stays. The validation's only observed effect was rejecting legitimate explicit entries: both recent CI breakages (DSH_GATE_CONCURRENCY exported into every job crashing lsp specs, DSH_PERMISSION_MODE in acp config.env crashing the child spawn) were this check firing on values a caller meant to pass, each fixed by routing around the bureaucracy the seam itself imposed. The bash seam keeps its own request/spec dshEnv field: that is bash-owned trusted-plugin vocabulary (the ctx.bashEnv collected overlay) whose merge-last position guarantees a caller env entry cannot displace a managed fact; bash-local now flattens ENV_OVERRIDES -> spec.env -> spec.dshEnv into the seam's one env map. subagent-acp and lsp-local pass their single config env map straight through. DshEnvironment/DshEnvironmentKey/DSH_ENV_PREFIX stay on the subprocess seam as the namespace vocabulary (bash re-exports them; scrubbedParentEnv filters on the prefix). Tests: the two channel-rejection specs and the splitEnvChannels partition spec are deleted; one spawn spec now proves an explicit DSH_* env entry reaches the child while an ambient one is scrubbed; the acp/lsp forwarding specs keep their MOCK_ECHO_ENV / LSP_FAKE_ECHO_ENV assertions with the split comments rewritten to merge-after-scrub. Docs (en+zh, re-recorded) and the owning Agent Notes updated; cordis api/services catalogs regenerated.
2026-07-27 04:14:51 +08:00
* Build a child environment: explicit caller entries merge after the scrubbed
* parent base, so a deliberately supplied credential or current `DSH_*` fact
* wins over the scrub that dropped its ambient namesake.
* @param extra - explicit caller entries, merged verbatim after the scrub.
* @returns the environment to hand to `spawn` for the child process.
feat(bash): add stdin + extra env to the executor seam as a trusted-plugin surface The hooks subsystem runs external hook commands the Claude Code / Codex way: JSON payload on stdin, context in CLAUDE_PROJECT_DIR / CLAUDE_PLUGIN_ROOT env. Reusing the ctx.bash seam for that needs two new inputs — but stdin and arbitrary env are exactly what dsh-bash-local's credential scrub exists to keep away from model-driven commands. So this adds them as a TRUSTED-PLUGIN surface: - BashExecRequest + BashExecSpec gain optional `stdin` and `env`. They are plain optionals on the resolved spec (not required-but-nullable like `owner`): a missing one means "none", the safe default, not a security footgun. - dsh-bash-local threads them through resolve/run/start. `env` merges AFTER the credential scrub, so a trusted caller's explicit entry wins even on a credential-shaped name — the scrub guards the harness's OWN ambient creds from model-driven commands, not a trusted plugin. stdin is always a pipe, closed immediately (with bytes when supplied, empty otherwise — EOF as before); an EPIPE from a child that exits without reading is swallowed. - The model-facing dsh-tool-bash NEVER forwards model input into stdin/env (its request is command/workdir/timeoutMs/signal/owner only). A regression guard drives the real tool with adversarial args and asserts the request carries neither field — proven to go red if the consumer ever forwards them. Configurable scrub (in an earlier sketch) is dropped as speculative: the explicit `env` field already gives a trusted caller full control, and no caller needs to broaden the ambient scrub. Documented in a new architecture RFC, the bash.md type-equiv blocks, and the three bash READMEs.
2026-06-30 13:52:25 +08:00
*/
subprocess: one explicit env channel on the spawn spec Drop SubprocessSpawnSpec.dshEnv and splitEnvChannels(); childEnv() is now scrubbed-base + explicit entries with no namespace validation. The invariant dropped is the reserved-namespace check on explicit entries (DSH_* rejected from env, non-DSH_* rejected from dshEnv). Explicit-entry trust already covers it: an explicit credential-shaped entry has always merged after the scrub as a deliberate caller opt-in, and an explicit DSH_* entry is the same deliberate act — the staleness invariant lives entirely in scrubbedParentEnv dropping AMBIENT credential-shaped and DSH_* names, which stays. The validation's only observed effect was rejecting legitimate explicit entries: both recent CI breakages (DSH_GATE_CONCURRENCY exported into every job crashing lsp specs, DSH_PERMISSION_MODE in acp config.env crashing the child spawn) were this check firing on values a caller meant to pass, each fixed by routing around the bureaucracy the seam itself imposed. The bash seam keeps its own request/spec dshEnv field: that is bash-owned trusted-plugin vocabulary (the ctx.bashEnv collected overlay) whose merge-last position guarantees a caller env entry cannot displace a managed fact; bash-local now flattens ENV_OVERRIDES -> spec.env -> spec.dshEnv into the seam's one env map. subagent-acp and lsp-local pass their single config env map straight through. DshEnvironment/DshEnvironmentKey/DSH_ENV_PREFIX stay on the subprocess seam as the namespace vocabulary (bash re-exports them; scrubbedParentEnv filters on the prefix). Tests: the two channel-rejection specs and the splitEnvChannels partition spec are deleted; one spawn spec now proves an explicit DSH_* env entry reaches the child while an ambient one is scrubbed; the acp/lsp forwarding specs keep their MOCK_ECHO_ENV / LSP_FAKE_ECHO_ENV assertions with the split comments rewritten to merge-after-scrub. Docs (en+zh, re-recorded) and the owning Agent Notes updated; cordis api/services catalogs regenerated.
2026-07-27 04:14:51 +08:00
export function childEnv(extra?: Readonly<Record<string, string>>): NodeJS.ProcessEnv {
return { ...scrubbedParentEnv(), ...extra }
}
/** Injectable knobs so tests can exercise spill and platform behavior deterministically. */
export interface SpawnInternals {
/** Directory for spill files (defaults to the OS temp dir). */
spillDir?: string
/** Windows tree-termination runner (defaults to `taskkill /PID <pid> /T /F`). */
taskkill?: (pid: number) => void
/** Host platform override for signalling decisions. */
platform?: NodeJS.Platform
}
/**
* Liveness-poll cadence for tree-exit waits. The timer stays ref'd: an
* awaited teardown must keep the event loop alive until the tree really
* exits, or the parent can exit while claiming quiescence and orphan the
* survivors it promised to reap.
*/
2026-07-26 16:50:36 +08:00
function sleepTick(): Promise<void> {
return sleepMs(15)
2026-07-26 16:50:36 +08:00
}
let spillCounter = 0
let defaultSpillDir: string | undefined
/**
* The default spill location: a private (0700) per-process directory under
* the OS tmpdir, created lazily. Predictable world-readable paths would let
* other local users read command output or pre-create symlinks.
*/
function privateSpillDir(): string {
defaultSpillDir ??= mkdtempSync(join(tmpdir(), 'dsh-subprocess-'))
return defaultSpillDir
}
/**
* Collects one stream with a bounded in-memory tail. With a spill cap, on
* first overflow a spill file is created and every chunk (including those
* already collected) is appended there while the full stream remains within
* the cap; without one, only the in-memory tail is ever retained (the
* diagnostic-tail shape — a language server's stderr).
*
* Tail-keep rationale (pi/OpenCode): errors and final results cluster at the
* end of command output; the spill file covers the head.
*/
export class OutputCollector {
private chunks: Buffer[] = []
private bytes = 0
private dropped = false
private spillFd: number | undefined
private spillFile: string | undefined
private spillDisabled: boolean
/** Total bytes ever pushed (not just retained). */
private total = 0
constructor(
private readonly maxBytes: number,
private readonly maxSpillBytes: number | undefined,
private readonly label: string,
private readonly spillDir: string,
) {
this.spillDisabled = maxSpillBytes === undefined
}
/**
* Ingest one stream chunk, counting it toward the whole-stream total. On
* first overflow of the in-memory cap a spill file is opened (when spilling
* is enabled) and every chunk (already-collected ones included) is appended
* there from then on; the in-memory tail then drops whole chunks from its
* head (or the head of a single over-cap chunk) until it fits the cap again.
* @param chunk - the raw bytes from one stream 'data' event.
*/
push(chunk: Buffer): void {
this.total += chunk.length
const overflows = this.bytes + chunk.length > this.maxBytes
if (!this.spillDisabled && (overflows || this.spillFd !== undefined)) this.spillAll(chunk)
this.chunks.push(chunk)
this.bytes += chunk.length
2026-07-26 16:50:36 +08:00
while (this.bytes > this.maxBytes) {
const head = this.chunks[0] as Buffer
const excess = this.bytes - this.maxBytes
if (head.length <= excess) {
// Drop the whole head chunk (length ≥ 1 is guaranteed while over cap).
this.chunks.shift()
this.bytes -= head.length
} else {
// Trim the head so the retained window is byte-exact at the cap — a
// diagnostic tail (an LSP server's stderr) must hold the LAST
// maxBytes regardless of how the stream was chunked.
this.chunks[0] = head.subarray(excess)
this.bytes -= excess
}
this.dropped = true
}
}
/** Open the spill file lazily and append `chunk` (and any prior chunks once). */
private spillAll(chunk: Buffer): void {
if (this.maxSpillBytes !== undefined && this.total > this.maxSpillBytes) {
this.discardSpill()
return
}
if (this.spillFd === undefined) {
// Random suffix + O_EXCL + no-follow-equivalent ('wx' fails on any
// existing path, symlink or not) + owner-only mode: defeats spill-path
// prediction and symlink planting in shared tmp dirs.
this.spillFile = join(
this.spillDir,
`dsh-subprocess-${process.pid}-${++spillCounter}-${randomBytes(6).toString('hex')}-${this.label}.log`,
)
this.spillFd = openSync(this.spillFile, 'wx', 0o600)
for (const prior of this.chunks) writeSync(this.spillFd, prior)
}
writeSync(this.spillFd, chunk)
}
/** Stop spilling and remove the file once it can no longer hold the complete stream. */
private discardSpill(): void {
const fd = this.spillFd
const file = this.spillFile
this.spillFd = undefined
this.spillFile = undefined
this.spillDisabled = true
if (fd !== undefined) {
try {
closeSync(fd)
} catch {
// Retain the descriptor so finalize can retry the failed close.
this.spillFd = fd
}
}
if (file !== undefined) {
try {
unlinkSync(file)
} catch {
// A failed unlink leaves at most maxSpillBytes behind, never an unbounded file.
}
}
}
/**
* Incremental read in whole-stream byte coordinates: returns everything
* pushed since `fromByte`. When `fromByte` has already slid out of the
* in-memory tail window, the read is `lossy` — it returns the whole
* retained tail and the gap is only recoverable from the spill file.
* @param fromByte - whole-stream offset to resume from (a prior read's `nextOffset`; 0 for the first read).
* @returns the delta text, the offset for the next read, the `lossy` flag, and the spill path when one was created.
*/
readFrom(fromByte: number): { text: string; nextOffset: number; lossy: boolean; spillPath?: string } {
const windowStart = this.total - this.bytes
const buffer = Buffer.concat(this.chunks)
const lossy = fromByte < windowStart
const slice = lossy ? buffer : buffer.subarray(fromByte - windowStart)
return {
text: slice.toString('utf8'),
nextOffset: this.total,
lossy,
...this.spillFile !== undefined ? { spillPath: this.spillFile } : {},
}
}
/**
* Close the spill file once the stream has ended. A failed close (delayed
* writeback fault) stops advertising the spill path — the file may be
* missing its tail — while every in-memory read keeps working. Idempotent;
* the spawn path seals both collectors at settlement so reads after exit
* never point at a still-open file.
*/
seal(): void {
if (this.spillFd === undefined) return
try {
closeSync(this.spillFd)
} catch {
// A delayed writeback failure makes the spill unreliable; keep the
// in-memory result but stop advertising that file.
this.spillFile = undefined
}
this.spillFd = undefined
}
/**
* Seal the spill file and return the final output.
* @returns the final collected output: tail text, truncation flag, and the spill path when intact.
*/
finalize(): CollectedOutput {
this.seal()
return {
text: Buffer.concat(this.chunks).toString('utf8'),
truncated: this.dropped,
...this.spillFile !== undefined ? { spillPath: this.spillFile } : {},
}
}
}
/**
* Send `sig` to a detached POSIX process group. Never throws: delivery races
* process exit and may run in a timer callback, so failures are contained and
* a non-positive pid is a no-op.
* @param pid - the group leader's pid; non-positive means the spawn failed and the call is a no-op.
* @param sig - the signal to deliver to the whole group.
*/
export function killGroup(pid: number, sig: NodeJS.Signals): void {
if (pid <= 0) return
try {
process.kill(-pid, sig)
} catch {
// Swallow: see contract above.
}
}
/**
* Terminate one Windows process tree with `taskkill /T /F`. Contained like
* POSIX group signalling — delivery races tree exit, so an absent tree, a
* nonzero status, or a missing taskkill binary must not break idempotent
* teardown.
* @param pid - root process id; non-positive is a no-op.
*/
export function taskkillProcessTree(pid: number): void {
if (pid <= 0) return
feat(subprocess): migrate lsp-local, subagent-acp, and the env scrubs onto the seam Review direction (tianyicui, PR #660): in a stacked PR, change all other process-running places to use the new service. - lsp-local: LspConnection spawns through ctx.subprocess (piped protocol streams + a no-spill collected stderr tail); its private process-tree helpers (POSIX group signalling, Windows taskkill, liveness polling) are deleted in favor of the seam's handle verbs, and its buildChildEnv now rides scrubbedParentEnv (LSP children also stop inheriting stale DSH_*). The plugin injects 'subprocess'; compositions/tests mount dsh-subprocess-local. - subagent-acp: the ACP child spawns through the seam (piped ndjson streams, inherited stderr); spawn failure surfaces through done-rejection into the same startup race; disposal is handle.dispose with the plugin's configured graces. dsh-subagent-subprocess is DELETED — its dispose ladder and scrub are the seam's, and the isolated-config-dir helper had no consumer. - mcp-client, pty-local, sdk-helper: adopt scrubbedParentEnv as the one scrub definition (their spawns stay put by ownership: the MCP SDK and node-pty own those calls; the SDK wizard runs outside any composition). - Coverage: per-file 100% over every touched src file, with each v8 ignore carrying a platform or contract reason; new suites cover stdio dispositions, the dispose ladder tiers, injected-win32 tree semantics, waitForExit, settled-kill/terminate no-ops, and spawn-failure disposal. - Docs: consumer-migration Agent Note (en; zh follows in this PR), seam note updated in place, subprocess.md rewritten for the reshaped vocabulary (type-equiv re-registered), READMEs and SERVICE_ROLES updated, taskkill added to knip ignoreBinaries.
2026-07-26 15:27:59 +08:00
// Outcome deliberately unchecked: an already-absent tree (status 128), exit
// races, and a missing taskkill binary (spawnSync reports, never throws) are
// as tolerable here as ESRCH is for a POSIX group signal.
spawnSync('taskkill', ['/PID', String(pid), '/T', '/F'], { stdio: 'ignore' })
}
/**
* Signal a detached process tree with platform-correct semantics: POSIX
* signals the negative process-group id and falls back to the direct child
* when the group is gone; Windows terminates the tree via taskkill (any
* signal value force-terminates — Node maps signals to TerminateProcess).
*/
function signalTree(
platform: NodeJS.Platform,
pid: number,
sig: NodeJS.Signals,
child: ChildProcess,
taskkill: (pid: number) => void,
): void {
if (platform === 'win32') {
taskkill(pid)
return
}
2026-07-26 16:50:36 +08:00
/* v8 ignore next -- kill/terminate gate on treeAlive(), which is false for pid -1; this guard protects direct callers only. */
if (pid <= 0) return
try {
process.kill(-pid, sig)
} catch {
feat(subprocess): migrate lsp-local, subagent-acp, and the env scrubs onto the seam Review direction (tianyicui, PR #660): in a stacked PR, change all other process-running places to use the new service. - lsp-local: LspConnection spawns through ctx.subprocess (piped protocol streams + a no-spill collected stderr tail); its private process-tree helpers (POSIX group signalling, Windows taskkill, liveness polling) are deleted in favor of the seam's handle verbs, and its buildChildEnv now rides scrubbedParentEnv (LSP children also stop inheriting stale DSH_*). The plugin injects 'subprocess'; compositions/tests mount dsh-subprocess-local. - subagent-acp: the ACP child spawns through the seam (piped ndjson streams, inherited stderr); spawn failure surfaces through done-rejection into the same startup race; disposal is handle.dispose with the plugin's configured graces. dsh-subagent-subprocess is DELETED — its dispose ladder and scrub are the seam's, and the isolated-config-dir helper had no consumer. - mcp-client, pty-local, sdk-helper: adopt scrubbedParentEnv as the one scrub definition (their spawns stay put by ownership: the MCP SDK and node-pty own those calls; the SDK wizard runs outside any composition). - Coverage: per-file 100% over every touched src file, with each v8 ignore carrying a platform or contract reason; new suites cover stdio dispositions, the dispose ladder tiers, injected-win32 tree semantics, waitForExit, settled-kill/terminate no-ops, and spawn-failure disposal. - Docs: consumer-migration Agent Note (en; zh follows in this PR), seam note updated in place, subprocess.md rewritten for the reshaped vocabulary (type-equiv re-registered), READMEs and SERVICE_ROLES updated, taskkill added to knip ignoreBinaries.
2026-07-26 15:27:59 +08:00
/* v8 ignore start -- the fallback needs a live child whose group signal fails
(EPERM-style), which POSIX CI cannot stage; the swallow keeps teardown idempotent. */
try {
child.kill(sig)
} catch {
// The direct child already exited; teardown remains idempotent.
}
feat(subprocess): migrate lsp-local, subagent-acp, and the env scrubs onto the seam Review direction (tianyicui, PR #660): in a stacked PR, change all other process-running places to use the new service. - lsp-local: LspConnection spawns through ctx.subprocess (piped protocol streams + a no-spill collected stderr tail); its private process-tree helpers (POSIX group signalling, Windows taskkill, liveness polling) are deleted in favor of the seam's handle verbs, and its buildChildEnv now rides scrubbedParentEnv (LSP children also stop inheriting stale DSH_*). The plugin injects 'subprocess'; compositions/tests mount dsh-subprocess-local. - subagent-acp: the ACP child spawns through the seam (piped ndjson streams, inherited stderr); spawn failure surfaces through done-rejection into the same startup race; disposal is handle.dispose with the plugin's configured graces. dsh-subagent-subprocess is DELETED — its dispose ladder and scrub are the seam's, and the isolated-config-dir helper had no consumer. - mcp-client, pty-local, sdk-helper: adopt scrubbedParentEnv as the one scrub definition (their spawns stay put by ownership: the MCP SDK and node-pty own those calls; the SDK wizard runs outside any composition). - Coverage: per-file 100% over every touched src file, with each v8 ignore carrying a platform or contract reason; new suites cover stdio dispositions, the dispose ladder tiers, injected-win32 tree semantics, waitForExit, settled-kill/terminate no-ops, and spawn-failure disposal. - Docs: consumer-migration Agent Note (en; zh follows in this PR), seam note updated in place, subprocess.md rewritten for the reshaped vocabulary (type-equiv re-registered), READMEs and SERVICE_ROLES updated, taskkill added to knip ignoreBinaries.
2026-07-26 15:27:59 +08:00
/* v8 ignore stop */
}
}
/**
* Spawn one isolated detached process tree with the spec's per-stream stdio
* dispositions. Runtime exits resolve `done` as {@link SubprocessOutcome};
* only spawn failures reject.
* @param spec - fully resolved argv, cwd, stdio, grace, cancellation, environment.
* @param internals - test-only spill-directory, platform, and taskkill overrides.
* @returns live subprocess handle.
*/
export function spawnSubprocess(spec: SubprocessSpawnSpec, internals: SpawnInternals = {}): SubprocessHandle {
const spillDir = internals.spillDir ?? privateSpillDir()
const platform = internals.platform ?? process.platform
const taskkill = internals.taskkill ?? taskkillProcessTree
if (spec.signal?.aborted) {
throw new Error(`aborted before spawn: ${String(spec.signal.reason ?? 'aborted')}`)
}
const [program, ...args] = spec.argv
if (program === undefined || program.length === 0) {
throw new Error('invalid argv: expected a non-empty program name at argv[0]')
}
const isCollect = (mode: SubprocessOutputMode): mode is SubprocessCollect =>
mode !== 'pipe' && mode !== 'inherit'
const outMode = spec.stdio.stdout
const errMode = spec.stdio.stderr
const stdinMode = spec.stdio.stdin
subprocess: one explicit env channel on the spawn spec Drop SubprocessSpawnSpec.dshEnv and splitEnvChannels(); childEnv() is now scrubbed-base + explicit entries with no namespace validation. The invariant dropped is the reserved-namespace check on explicit entries (DSH_* rejected from env, non-DSH_* rejected from dshEnv). Explicit-entry trust already covers it: an explicit credential-shaped entry has always merged after the scrub as a deliberate caller opt-in, and an explicit DSH_* entry is the same deliberate act — the staleness invariant lives entirely in scrubbedParentEnv dropping AMBIENT credential-shaped and DSH_* names, which stays. The validation's only observed effect was rejecting legitimate explicit entries: both recent CI breakages (DSH_GATE_CONCURRENCY exported into every job crashing lsp specs, DSH_PERMISSION_MODE in acp config.env crashing the child spawn) were this check firing on values a caller meant to pass, each fixed by routing around the bureaucracy the seam itself imposed. The bash seam keeps its own request/spec dshEnv field: that is bash-owned trusted-plugin vocabulary (the ctx.bashEnv collected overlay) whose merge-last position guarantees a caller env entry cannot displace a managed fact; bash-local now flattens ENV_OVERRIDES -> spec.env -> spec.dshEnv into the seam's one env map. subagent-acp and lsp-local pass their single config env map straight through. DshEnvironment/DshEnvironmentKey/DSH_ENV_PREFIX stay on the subprocess seam as the namespace vocabulary (bash re-exports them; scrubbedParentEnv filters on the prefix). Tests: the two channel-rejection specs and the splitEnvChannels partition spec are deleted; one spawn spec now proves an explicit DSH_* env entry reaches the child while an ambient one is scrubbed; the acp/lsp forwarding specs keep their MOCK_ECHO_ENV / LSP_FAKE_ECHO_ENV assertions with the split comments rewritten to merge-after-scrub. Docs (en+zh, re-recorded) and the owning Agent Notes updated; cordis api/services catalogs regenerated.
2026-07-27 04:14:51 +08:00
const env = childEnv(spec.env)
const child = spawn(program, args, {
cwd: spec.cwd,
env,
stdio: [
stdinMode === 'ignore' ? 'ignore' : 'pipe',
outMode === 'inherit' ? 'inherit' : 'pipe',
errMode === 'inherit' ? 'inherit' : 'pipe',
],
// `detached` gives teardown a tree root on POSIX (its own process group);
// Windows terminates by root pid through taskkill /T instead.
detached: platform !== 'win32',
})
const collectStream = (mode: SubprocessOutputMode, stream: Readable | null, label: string): OutputCollector | undefined => {
if (!isCollect(mode) || stream === null) return undefined
const collector = new OutputCollector(mode.maxBytes, mode.spill?.maxBytes, label, spillDir)
stream.on('data', (chunk: Buffer) => { collector.push(chunk) })
return collector
}
const stdoutCollector = collectStream(outMode, child.stdout, 'stdout')
const stderrCollector = collectStream(errMode, child.stderr, 'stderr')
let graceTimer: NodeJS.Timeout | undefined
let settled = false
// Failed spawns use pid -1 so signalling remains a no-op.
const pid = child.pid ?? -1
2026-07-26 16:50:36 +08:00
/** Whether the detached tree's root (or POSIX group) is still alive. */
const treeAlive = (): boolean => {
if (pid <= 0) return false
if (platform === 'win32') {
// Windows has no group-liveness probe; the direct child's exit is the
// observable boundary (taskkill /T already took the tree with it).
return child.exitCode === null && child.signalCode === null
}
try {
process.kill(-pid, 0)
return true
} catch (error) {
const code = (error as NodeJS.ErrnoException).code
/* v8 ignore next 2 -- POSIX reports an absent group as ESRCH; child-reaping timing
makes observing the other arm platform-dependent. */
if (code === 'ESRCH') return false
/* v8 ignore start -- EPERM and non-POSIX negative-pid failures are platform defenses; CI runs
tree-lifecycle tests on POSIX hosts where absence reports ESRCH. */
if (code === 'EPERM') return true
return child.exitCode === null && child.signalCode === null
/* v8 ignore stop */
}
}
// The escalation's tier primitive (not on the handle — terminate() is the
// only consumer-facing termination verb). Guards on TREE liveness, not
refactor(subprocess): keep terminate() as the seam's only termination verb Delete kill(signal?) from SubprocessHandle: consumers stop a process only through terminate()'s tree-scoped SIGTERM→graceMs→SIGKILL escalation (idempotent, also driven by the spec's abort signal, a no-op once the tree is gone). The single-signal verb had exactly one consumer family — lsp-local — and what it bought there was a private re-implementation of the same escalation. The internal kill closure stays in spawn.ts as the dispose ladder's tier primitive; terminate() now routes through it too. lsp-local collapses onto the seam's escalation: - LspConnection replaces its terminate()/kill() pair with one terminate() that delegates to handle.terminate(). Behavior change: the framing-failure path terminates instead of instant SIGKILL, so a misbehaving server now gets SIGTERM plus the killGraceMs window to flush before SIGKILL. - ConnectionSpec.pipeDrainGraceMs becomes killGraceMs: one grace, the spawn spec's graceMs, drives both the escalation window and post-exit pipe draining (the provider already passed killGraceMs for it). - LspInstance.forceTerminate() drops its hand-rolled bounded first wait (LSP_KILL_GRACE) and escalateProcessTree (deleted with its export and unit test): the seam's escalation already commits to SIGKILL after killGraceMs, so only the unbounded quiescence awaits stay load-bearing. Tests: kill()-shaped spawn specs become terminate()-shaped or fold into the terminate() suites (group-wide delivery; the settled no-op case was already pinned by 'terminate() after the tree died'); tree-survivor coverage is intact. A stderr-'inherit' disposition test completes the stdout/stderr symmetry so the scoped subprocess+lsp coverage gate stands alone instead of leaning on subagent-acp's cross-package runs. Docs: SubprocessHandle type-equiv block, seam/impl/group READMEs, and the consumer-migration Agent Note lose the kill(signal?) vocabulary (zh pairs re-recorded); cordis api/services catalogs regenerated.
2026-07-27 04:41:04 +08:00
// outcome settlement: a TERM-trapping helper can outlive the settled direct
// child and must stay signalable, while a fully-dead tree (possible pid
// reuse) must not be re-signalled by a later tier.
const kill = (sig: NodeJS.Signals): void => {
2026-07-26 16:50:36 +08:00
if (!treeAlive()) return
signalTree(platform, pid, sig, child, taskkill)
}
const terminate = (): void => {
if (graceTimer !== undefined) return // escalation already in flight
2026-07-26 16:50:36 +08:00
if (!treeAlive()) return
refactor(subprocess): keep terminate() as the seam's only termination verb Delete kill(signal?) from SubprocessHandle: consumers stop a process only through terminate()'s tree-scoped SIGTERM→graceMs→SIGKILL escalation (idempotent, also driven by the spec's abort signal, a no-op once the tree is gone). The single-signal verb had exactly one consumer family — lsp-local — and what it bought there was a private re-implementation of the same escalation. The internal kill closure stays in spawn.ts as the dispose ladder's tier primitive; terminate() now routes through it too. lsp-local collapses onto the seam's escalation: - LspConnection replaces its terminate()/kill() pair with one terminate() that delegates to handle.terminate(). Behavior change: the framing-failure path terminates instead of instant SIGKILL, so a misbehaving server now gets SIGTERM plus the killGraceMs window to flush before SIGKILL. - ConnectionSpec.pipeDrainGraceMs becomes killGraceMs: one grace, the spawn spec's graceMs, drives both the escalation window and post-exit pipe draining (the provider already passed killGraceMs for it). - LspInstance.forceTerminate() drops its hand-rolled bounded first wait (LSP_KILL_GRACE) and escalateProcessTree (deleted with its export and unit test): the seam's escalation already commits to SIGKILL after killGraceMs, so only the unbounded quiescence awaits stay load-bearing. Tests: kill()-shaped spawn specs become terminate()-shaped or fold into the terminate() suites (group-wide delivery; the settled no-op case was already pinned by 'terminate() after the tree died'); tree-survivor coverage is intact. A stderr-'inherit' disposition test completes the stdout/stderr symmetry so the scoped subprocess+lsp coverage gate stands alone instead of leaning on subagent-acp's cross-package runs. Docs: SubprocessHandle type-equiv block, seam/impl/group READMEs, and the consumer-migration Agent Note lose the kill(signal?) vocabulary (zh pairs re-recorded); cordis api/services catalogs regenerated.
2026-07-27 04:41:04 +08:00
kill('SIGTERM')
2026-07-26 16:50:36 +08:00
// The escalation must survive direct-child settlement — the leader dying
// does not mean the tree died — so settle does not clear this timer, and
refactor(subprocess): keep terminate() as the seam's only termination verb Delete kill(signal?) from SubprocessHandle: consumers stop a process only through terminate()'s tree-scoped SIGTERM→graceMs→SIGKILL escalation (idempotent, also driven by the spec's abort signal, a no-op once the tree is gone). The single-signal verb had exactly one consumer family — lsp-local — and what it bought there was a private re-implementation of the same escalation. The internal kill closure stays in spawn.ts as the dispose ladder's tier primitive; terminate() now routes through it too. lsp-local collapses onto the seam's escalation: - LspConnection replaces its terminate()/kill() pair with one terminate() that delegates to handle.terminate(). Behavior change: the framing-failure path terminates instead of instant SIGKILL, so a misbehaving server now gets SIGTERM plus the killGraceMs window to flush before SIGKILL. - ConnectionSpec.pipeDrainGraceMs becomes killGraceMs: one grace, the spawn spec's graceMs, drives both the escalation window and post-exit pipe draining (the provider already passed killGraceMs for it). - LspInstance.forceTerminate() drops its hand-rolled bounded first wait (LSP_KILL_GRACE) and escalateProcessTree (deleted with its export and unit test): the seam's escalation already commits to SIGKILL after killGraceMs, so only the unbounded quiescence awaits stay load-bearing. Tests: kill()-shaped spawn specs become terminate()-shaped or fold into the terminate() suites (group-wide delivery; the settled no-op case was already pinned by 'terminate() after the tree died'); tree-survivor coverage is intact. A stderr-'inherit' disposition test completes the stdout/stderr symmetry so the scoped subprocess+lsp coverage gate stands alone instead of leaning on subagent-acp's cross-package runs. Docs: SubprocessHandle type-equiv block, seam/impl/group READMEs, and the consumer-migration Agent Note lose the kill(signal?) vocabulary (zh pairs re-recorded); cordis api/services catalogs regenerated.
2026-07-27 04:41:04 +08:00
// kill() re-probes tree liveness before force-killing. It stays ref'd:
// the pending SIGKILL is a commitment, and a parent exiting before it
// fires would orphan a trapped survivor. Self-bounds at graceMs.
graceTimer = setTimeout(() => { kill('SIGKILL') }, spec.graceMs)
}
// The caller owns timeout classification; this layer only reacts to abort.
const onAbort = (): void => { terminate() }
spec.signal?.addEventListener('abort', onAbort, { once: true })
// Batch stdin is written and closed up front; process exit and captured
// output remain authoritative, so write errors (EPIPE) are best-effort.
if (typeof stdinMode === 'object' && child.stdin !== null) {
child.stdin.on('error', () => { /* stdin write is best-effort; outcome rides on exit/output. */ })
child.stdin.end(stdinMode.data)
}
const done = new Promise<SubprocessOutcome>((resolve, reject) => {
let pipeDrainTimer: NodeJS.Timeout | undefined
const settle = (exitCode: number | null, signal: NodeJS.Signals | null): void => {
if (settled) return
settled = true
// Only harness-collected pipes are force-closed at the drain boundary;
// a 'pipe'-mode stream belongs to the caller and closes with the child.
if (stdoutCollector !== undefined) child.stdout?.destroy()
if (stderrCollector !== undefined) child.stderr?.destroy()
stdoutCollector?.seal()
stderrCollector?.seal()
cleanup()
resolve({ exitCode, signal })
}
child.on('error', (error) => {
// No meaningful close outcome follows a spawn failure.
settled = true
cleanup()
reject(error)
})
child.on('exit', (exitCode, signal) => {
// A surviving descendant that inherited a pipe must not hold the
// outcome open indefinitely: after exit, the same bounded grace that
// governs kills also bounds the close wait.
pipeDrainTimer = setTimeout(() => { settle(exitCode, signal) }, spec.graceMs)
})
child.on('close', settle)
function cleanup(): void {
2026-07-26 16:50:36 +08:00
// graceTimer deliberately NOT cleared: the SIGKILL escalation must be
// able to reach tree survivors after the direct child settles.
if (pipeDrainTimer !== undefined) clearTimeout(pipeDrainTimer)
spec.signal?.removeEventListener('abort', onAbort)
}
})
const waitForExit = async (signal?: AbortSignal): Promise<boolean> => {
while (treeAlive()) {
if (signal?.aborted) return false
2026-07-26 16:50:36 +08:00
await sleepTick()
}
return true
}
return {
pid,
feat(subprocess): migrate lsp-local, subagent-acp, and the env scrubs onto the seam Review direction (tianyicui, PR #660): in a stacked PR, change all other process-running places to use the new service. - lsp-local: LspConnection spawns through ctx.subprocess (piped protocol streams + a no-spill collected stderr tail); its private process-tree helpers (POSIX group signalling, Windows taskkill, liveness polling) are deleted in favor of the seam's handle verbs, and its buildChildEnv now rides scrubbedParentEnv (LSP children also stop inheriting stale DSH_*). The plugin injects 'subprocess'; compositions/tests mount dsh-subprocess-local. - subagent-acp: the ACP child spawns through the seam (piped ndjson streams, inherited stderr); spawn failure surfaces through done-rejection into the same startup race; disposal is handle.dispose with the plugin's configured graces. dsh-subagent-subprocess is DELETED — its dispose ladder and scrub are the seam's, and the isolated-config-dir helper had no consumer. - mcp-client, pty-local, sdk-helper: adopt scrubbedParentEnv as the one scrub definition (their spawns stay put by ownership: the MCP SDK and node-pty own those calls; the SDK wizard runs outside any composition). - Coverage: per-file 100% over every touched src file, with each v8 ignore carrying a platform or contract reason; new suites cover stdio dispositions, the dispose ladder tiers, injected-win32 tree semantics, waitForExit, settled-kill/terminate no-ops, and spawn-failure disposal. - Docs: consumer-migration Agent Note (en; zh follows in this PR), seam note updated in place, subprocess.md rewritten for the reshaped vocabulary (type-equiv re-registered), READMEs and SERVICE_ROLES updated, taskkill added to knip ignoreBinaries.
2026-07-26 15:27:59 +08:00
/* v8 ignore start -- pipe-mode fds exist on every spawn Node returns; the null-coalesces guard a nonconforming ChildProcess only. */
stdin: stdinMode === 'pipe' ? child.stdin ?? undefined : undefined,
stdout: outMode === 'pipe' ? child.stdout ?? undefined : undefined,
stderr: errMode === 'pipe' ? child.stderr ?? undefined : undefined,
feat(subprocess): migrate lsp-local, subagent-acp, and the env scrubs onto the seam Review direction (tianyicui, PR #660): in a stacked PR, change all other process-running places to use the new service. - lsp-local: LspConnection spawns through ctx.subprocess (piped protocol streams + a no-spill collected stderr tail); its private process-tree helpers (POSIX group signalling, Windows taskkill, liveness polling) are deleted in favor of the seam's handle verbs, and its buildChildEnv now rides scrubbedParentEnv (LSP children also stop inheriting stale DSH_*). The plugin injects 'subprocess'; compositions/tests mount dsh-subprocess-local. - subagent-acp: the ACP child spawns through the seam (piped ndjson streams, inherited stderr); spawn failure surfaces through done-rejection into the same startup race; disposal is handle.dispose with the plugin's configured graces. dsh-subagent-subprocess is DELETED — its dispose ladder and scrub are the seam's, and the isolated-config-dir helper had no consumer. - mcp-client, pty-local, sdk-helper: adopt scrubbedParentEnv as the one scrub definition (their spawns stay put by ownership: the MCP SDK and node-pty own those calls; the SDK wizard runs outside any composition). - Coverage: per-file 100% over every touched src file, with each v8 ignore carrying a platform or contract reason; new suites cover stdio dispositions, the dispose ladder tiers, injected-win32 tree semantics, waitForExit, settled-kill/terminate no-ops, and spawn-failure disposal. - Docs: consumer-migration Agent Note (en; zh follows in this PR), seam note updated in place, subprocess.md rewritten for the reshaped vocabulary (type-equiv re-registered), READMEs and SERVICE_ROLES updated, taskkill added to knip ignoreBinaries.
2026-07-26 15:27:59 +08:00
/* v8 ignore stop */
collected: {
...stdoutCollector !== undefined ? { stdout: stdoutCollector } : {},
...stderrCollector !== undefined ? { stderr: stderrCollector } : {},
},
done,
terminate,
waitForExit,
}
}