Files
selfpost/internal/postfix/postfix.go
T
mixandClaude Opus 5.5 f2ad699473
test / test (push) Canceled after 0s
supervision as supervisord does it: Postfix stopped, give-ups FATAL, reload's exit read
C-P42. postfix-script's start-fg runs master as its child and master
calls setsid() (Postfix 3.7.11, conf/postfix-script; src/master/master.c);
supervisord signals the pid alone (supervisor 4.2.5, process.py,
Subprocess.kill()). A stop of program postfix left master running - a
restart then found "already running" - and a container stop cut Postfix
off with SIGKILL. The wrapper now stays the parent of "postfix start-fg"
and on TERM/INT runs "postfix stop", asking again while start-fg is still
in its checks (master -t). stopwaitsecs=15.

C-V6. The wrapper's give-up on the milter sockets came after the default
startsecs, so it was EXITED and restarted for ever. startsecs is now
MILTER_WAIT_TIMEOUT + 10, exported by the entrypoint as POSTFIX_STARTSECS
and expanded by supervisord (options.py, processes_from_section(); a
non-number MILTER_WAIT_TIMEOUT stops the start): a give-up is "Exited too
quickly", BACKOFF, FATAL after startretries, crashexit. For those seconds
after each start the program is STARTING (the healthcheck's 90 s start
period covers it; the sender check counts no downtime then).

C-V5. logrotate had startsecs=0, so a refused config restarted it every
second for ever. startsecs=1: FATAL after startretries; crashexit reads
the event's processname and leaves logrotate's FATAL on Server > Health.

B-U2. "supervisorctl start" returns when a startsecs=0 program is RUNNING,
not when it ends, and mapped "already started" to success. The panel now
speaks supervisord's XML-RPC itself (internal/supervisor, RunOnce):
startProcess without waiting, getProcessInfo until it leaves
STARTING/RUNNING, exitstatus 0 or an error; a run in progress is waited
out and one of its own follows. A failed reload while program postfix is
not RUNNING is logged, not an error: the next master reads the files.
C-V10: postfix-reload rests EXITED after a run; the e2e process check
takes STOPPED or EXITED. C-P34: the dead "postfix set-permissions" branch
goes - postfix check's create-missing builds the queue tree.

Stand (2026-10-14, make e2e, exit 0; MILTER_WAIT_TIMEOUT=10 on the stand):
new supervision_stops_postfix_and_gives_up - supervisorctl restart postfix
writes postfix-script's "stopping the Postfix mail system" and leaves one
master; with the panel's sockets removed, start answers "postfix: ERROR
(spawn error)" (finish() sets spawnerr "Exited too quickly", onwait()
reads it first) and Postfix comes up once the panel is back; a
group-writable logrotate config ends FATAL with the container up; the
panel's Reload configuration answers only after mail.log has "refreshing
the Postfix mail system", postfix-reload EXITED. session_survives_restart
sees the stop line for the container's restart. Measured: master's own
"terminating on signal 15" never reaches mail.log (master TERMs its group,
postlogd included, before it logs). go vet and go test ./... pass on the
server (24 packages; new unit tests of RunOnce against a fake supervisord).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-14 23:49:25 +03:00

212 lines
8.5 KiB
Go

// Package postfix owns the Postfix configuration files the panel edits at
// runtime and the privileged reload that applies them (architecture.md § Mail
// path, security.md): the smtpd_sender_login_maps table binding each
// application's SASL login to the sender addresses it may use, plus the relay
// configuration in main.cf.
package postfix
import (
"fmt"
"log"
"path/filepath"
"sort"
"strings"
"github.com/mixeme/selfpost/internal/configsafe"
"github.com/mixeme/selfpost/internal/supervisor"
)
// Postfix manages the on-disk Postfix state the panel is responsible for. After
// rewriting a map it asks Postfix to reload.
type Postfix struct {
senderLoginMapsPath string
// reload asks the running Postfix to re-read its configuration. It is a
// field so tests can substitute a no-op; the default drives supervisord.
reload func() error
}
// New builds a manager rooted at dir (typically /data/postfix), the same layout
// entrypoint.sh prepares. The default reload path signals Postfix through
// supervisord.
func New(dir string) *Postfix {
return &Postfix{
senderLoginMapsPath: filepath.Join(dir, "sender_login_maps"),
reload: reloadViaSupervisor,
}
}
// SenderLoginMapsPath is the absolute path of the generated map, so main.cf
// can point smtpd_sender_login_maps at it.
func (p *Postfix) SenderLoginMapsPath() string {
return p.senderLoginMapsPath
}
// Binding is one sender-address → login pair (architecture.md § Mail path).
// Address is either a domain wildcard "@example.com" or a specific address
// "alerts@example.com".
type Binding struct {
Address string
Login string
}
// RebuildSenderLoginMaps regenerates the sender_login_maps file from the full
// set of bindings and reloads Postfix (architecture.md § Mail path). Full
// regeneration (rather than incremental edits) keeps the file a pure function
// of the registry, so add, edit and delete share one idempotent path. The file
// is written atomically before the reload.
//
// Several applications may be authorised for the same address (many-to-one,
// architecture.md § Mail path) — their logins are merged onto a single line as
// a comma-separated list, which is how Postfix expects multiple owners of one
// sender.
//
// Postfix 3.7.11, src/smtpd/smtpd_check.c, reject_auth_sender_login_mismatch():
// the value is split into owners with mystrtok(&cp, CHARS_COMMA_SP).
func (p *Postfix) RebuildSenderLoginMaps(bindings []Binding) error {
content, err := renderSenderLoginMaps(bindings)
if err != nil {
return err
}
if err := writeFileAtomic(p.senderLoginMapsPath, content, 0o640); err != nil {
return err
}
return p.reload()
}
// Reload asks Postfix to re-read its configuration without regenerating any
// file. It backs the panel's manual reload button (architecture.md § Panel
// HTTP surface).
func (p *Postfix) Reload() error {
return p.reload()
}
// SetReloadHook replaces how configuration is applied after a rebuild. Tests
// that cannot reach supervisord use this to verify file regeneration alone.
func (p *Postfix) SetReloadHook(fn func() error) {
if fn != nil {
p.reload = fn
}
}
// renderSenderLoginMaps builds the sender_login_maps file contents. Keys are
// sorted for deterministic output and the logins under each key are sorted and
// de-duplicated. Every address and login is re-checked for injection safety
// before being written (security.md) — upstream validation already guarantees
// this, but the writer refuses to emit anything unsafe as a hard backstop.
func renderSenderLoginMaps(bindings []Binding) ([]byte, error) {
byAddr := make(map[string][]string)
order := make([]string, 0)
for _, b := range bindings {
if err := assertMapSafe(b.Address, b.Login); err != nil {
return nil, err
}
// One line per key as Postfix folds it: smtpd opens the map with
// DICT_FLAG_FOLD_FIX and no DICT_FLAG_DUP_* flag, so two keys that
// differ only in case would make the whole table unusable (Postfix
// 3.7.11, src/util/dict_thash.c, dict_thash_open()). The panel stores
// addresses in lower case already; this keeps the file safe if a row
// ever is not.
addr := strings.ToLower(b.Address)
if _, seen := byAddr[addr]; !seen {
order = append(order, addr)
}
byAddr[addr] = appendUnique(byAddr[addr], b.Login)
}
sort.Strings(order)
var sb strings.Builder
for _, addr := range order {
logins := byAddr[addr]
sort.Strings(logins)
// texthash format: <key><whitespace><value>. A comma-separated value lists
// every login permitted to use this sender (architecture.md § Mail path).
fmt.Fprintf(&sb, "%s %s\n", addr, strings.Join(logins, ","))
}
return []byte(sb.String()), nil
}
func appendUnique(list []string, v string) []string {
for _, x := range list {
if x == v {
return list
}
}
return append(list, v)
}
// What may never reach a sender-map line: whitespace and the newline that would
// end it, the comma that separates values, the colon a lookup key cannot carry,
// and the backslash. A login additionally may not carry "@" — that would make it
// look like an address.
//
// Postfix 3.7.11, src/util/dict_thash.c, dict_thash_open(): the key ends at
// the first blank outside quotes, a backslash escapes the next character;
// src/util/readlline.c: a newline ends the line (one that starts with a blank
// continues the one before); the owners split on comma and blank (see
// RebuildSenderLoginMaps).
const (
senderMapAddressForbidden = " \t\r\n,:\\"
senderMapLoginForbidden = " \t\r\n,:@\\"
)
// assertMapSafe rejects any address/login value that could break out of a single
// map line or inject a directive. Addresses are validated to a strict whitelist
// (letters, digits, '@', '.', '-', '_', '+') and logins to an even stricter one
// upstream (security.md); this is defence in depth against a validation gap ever
// letting whitespace, a newline or a comma (the value separator) through into
// the file (security.md).
func assertMapSafe(address, login string) error {
if err := configsafe.Token("address", address, senderMapAddressForbidden); err != nil {
return fmt.Errorf("postfix: %w", err)
}
if err := configsafe.Token("login", login, senderMapLoginForbidden); err != nil {
return fmt.Errorf("postfix: %w", err)
}
return nil
}
// reloadViaSupervisor asks supervisord (PID 1, running as root) to run the
// one-shot `postfix-reload` program, which executes the canonical `postfix
// reload` and re-reads main.cf/master.cf and the lookup tables they reference.
// The panel runs unprivileged: it cannot run `postfix reload` itself, and it
// cannot signal the Postfix master directly because `postfix start-fg` forks a
// separate master whose PID supervisord does not track (a SIGHUP to the
// supervised process would never reach it). Going through supervisord's
// group-accessible control socket runs the reload as root without any panel
// privilege (architecture.md § Mail path, security.md).
//
// Arguments are fixed literals — no user input is interpolated into the command,
// and it never goes through a shell (security.md).
//
// Postfix 3.7.11, conf/postfix-script, start-fg: a process that is not PID 1
// runs "$daemon_directory/master -s" as its child, and src/master/master.c
// puts that master in a session of its own (setsid()). "reload" sends the
// master SIGHUP; src/master/master.c calls master_refresh(), and
// master_conf.c, master_config(), lets the running children "commit suicide
// whenever it is convenient" while new ones get the new configuration; a
// texthash table is read into memory once per process
// (README_FILES/DATABASE_README).
//
// The call returns when "postfix reload" has ended, with an error unless it
// exited 0 (a postsuper failure, a main.cf postfix(1) cannot read:
// conf/postfix-script, "reload"; src/postfix/postfix.c), which the page that
// saved shows. Before, supervisorctl's start returned as the reload was
// spawned, and a reload already running counted as done though it may have
// read the files before this change (supervisor.RunOnce; assumptions B-U2).
// A reload that fails while program postfix is not RUNNING - stopped, or
// still in its start, the wrapper waiting for the milter sockets - is no
// error: "the Postfix mail system is not running", and the master that starts
// next reads the files as they are now.
func reloadViaSupervisor() error {
err := supervisor.RunOnce("postfix-reload")
if err == nil {
return nil
}
if state, serr := supervisor.State("postfix"); serr == nil && state != "RUNNING" {
log.Printf("postfix: reload while program postfix is %s, left to its start: %v", state, err)
return nil
}
return err
}