Skip to content
Sign up

flagon-io/g1t

public
0
1,662 linesCodeBlameRaw
1//! Git over HTTPS: the smart HTTP remote at `/<namespace>/<repo>.git`,
2//! proxied to the git store with a short-lived token.
3
4use std::cell::RefCell;
5use std::rc::Rc;
6
7use futures_util::StreamExt;
8use g1t_contracts::identity::GitCredentialsArgs;
9use g1t_contracts::repos::{GitAccess, GitService, RepoPath};
10use g1t_contracts::{FailureCode, Outcome, Viewer};
11use worker::js_sys::Uint8Array;
12use worker::wasm_bindgen::JsValue;
13use worker::{Fetch, Fetcher, Headers, Method, Request, RequestInit, Response, Result, Url};
14
15use crate::hidden_refs;
16use crate::meters;
17use crate::pack_limits::{PackSizer, Violation};
18use crate::push_checks::{Checks, Verdict};
19use crate::resilience::{self, Busy, Failure};
20
21const ENDPOINTS: [&str; 3] = ["info/refs", "git-upload-pack", "git-receive-pack"];
22const FORWARDED_HEADERS: [&str; 5] = [
23 "accept",
24 "content-encoding",
25 "content-type",
26 "git-protocol",
27 "user-agent",
28];
29
30/// A git request, parsed from its URL.
31pub struct GitRequest {
32 pub path: RepoPath,
33 pub endpoint: &'static str,
34 pub service: GitService,
35}
36
37impl GitRequest {
38 /// The same request for the repository of the same name under
39 /// `namespace`: where a workspace alias leads.
40 pub fn under(&self, namespace: &str) -> GitRequest {
41 GitRequest {
42 path: RepoPath {
43 namespace: namespace.to_owned(),
44 name: self.path.name.clone(),
45 },
46 endpoint: self.endpoint,
47 service: self.service,
48 }
49 }
50}
51
52/// Parses `/<namespace>/<name>[.git]/<endpoint>`, or returns `None` if the
53/// request is not git's.
54pub fn parse(url: &Url) -> Option<GitRequest> {
55 let path = url.path().strip_prefix('/')?;
56 let endpoint = ENDPOINTS
57 .into_iter()
58 .find(|endpoint| path.ends_with(&format!("/{endpoint}")))?;
59 let repo = &path[..path.len() - endpoint.len() - 1];
60 let (namespace, name) = repo.split_once('/')?;
61 let name = name.strip_suffix(".git").unwrap_or(name);
62 if namespace.is_empty() || name.is_empty() || name.contains('/') {
63 return None;
64 }
65 let service = if endpoint == "info/refs" {
66 url.query_pairs()
67 .find(|(key, _)| key == "service")
68 .map(|(_, value)| value.into_owned())?
69 } else {
70 endpoint.to_owned()
71 };
72 let service = match service.as_str() {
73 "git-upload-pack" => GitService::UploadPack,
74 "git-receive-pack" => GitService::ReceivePack,
75 _ => return None,
76 };
77 Some(GitRequest {
78 path: RepoPath {
79 namespace: namespace.to_owned(),
80 name: name.to_owned(),
81 },
82 endpoint,
83 service,
84 })
85}
86
87/// How long each step of a git request took, sent back to git as a
88/// `Server-Timing` header so that a slow clone shows where its time went.
89/// Step names and whole milliseconds only. The Workers clock moves only
90/// while a request waits on something, so each step is the time spent
91/// waiting on the database, another service or the git store. A note
92/// says how a step went without a duration: `refs;desc=hit-colo`. A part
93/// is one of several things a step did at once, with its own duration: a
94/// push's `checks` step is its `read`, `rules` and `scan` parts side by side.
95pub struct Timing {
96 started: u64,
97 last: u64,
98 steps: Vec<(&'static str, u64)>,
99 parts: Vec<(&'static str, u64)>,
100 notes: Vec<(&'static str, &'static str)>,
101}
102
103impl Timing {
104 pub fn start() -> Self {
105 let now = g1t_kit::now_ms();
106 Self {
107 started: now,
108 last: now,
109 steps: Vec::new(),
110 parts: Vec::new(),
111 notes: Vec::new(),
112 }
113 }
114
115 /// Says how long `part` of the step under way took. Parts overlap, so
116 /// they are listed after the steps and are not part of the total.
117 pub fn part(&mut self, part: &'static str, ms: u64) {
118 self.parts.push((part, ms));
119 }
120
121 /// Says how `name` went: where an answer or a credential came from.
122 pub fn note(&mut self, name: &'static str, description: &'static str) {
123 self.notes.push((name, description));
124 }
125
126 /// Ends a step, named `step`, that began when the last one ended.
127 pub fn mark(&mut self, step: &'static str) {
128 let now = g1t_kit::now_ms();
129 self.steps.push((step, now.saturating_sub(self.last)));
130 self.last = now;
131 }
132
133 /// `response`, with how long each step took.
134 pub fn apply(&self, response: Response) -> Result<Response> {
135 let total = g1t_kit::now_ms().saturating_sub(self.started);
136 let headers = response.headers().clone();
137 headers.set("server-timing", &server_timing(&self.steps, &self.parts, &self.notes, total))?;
138 Ok(response.with_headers(headers))
139 }
140}
141
142/// A `Server-Timing` value: each step with its duration, then each part,
143/// the notes, and the total.
144fn server_timing(steps: &[(&str, u64)], parts: &[(&str, u64)], notes: &[(&str, &str)], total: u64) -> String {
145 steps
146 .iter()
147 .chain(parts)
148 .map(|(step, ms)| format!("{step};dur={ms}"))
149 .chain(notes.iter().map(|(name, description)| format!("{name};desc={description}")))
150 .chain(std::iter::once(format!("total;dur={total}")))
151 .collect::<Vec<_>>()
152 .join(", ")
153}
154
155/// The user named by an HTTP Basic `Authorization` header, as git sends it.
156pub async fn viewer(request: &Request, identity: &Fetcher) -> Result<Viewer> {
157 let Some(header) = request.headers().get("authorization")? else {
158 return Ok(None);
159 };
160 let Some((scheme, encoded)) = header.split_once(' ') else {
161 return Ok(None);
162 };
163 if !scheme.eq_ignore_ascii_case("basic") {
164 return Ok(None);
165 }
166 let Some(decoded) = decode_base64(encoded.trim()) else {
167 return Ok(None);
168 };
169 let Some((username, secret)) = decoded.split_once(':') else {
170 return Ok(None);
171 };
172 g1t_kit::call(
173 identity,
174 "user_for_git_credentials",
175 &GitCredentialsArgs {
176 username: username.to_owned(),
177 secret: secret.to_owned(),
178 },
179 )
180 .await
181}
182
183/// Standard base64 to a UTF-8 string, or `None` if either step fails.
184fn decode_base64(input: &str) -> Option<String> {
185 let mut bytes = Vec::with_capacity(input.len() * 3 / 4);
186 let mut buffer = 0u32;
187 let mut bits = 0;
188 for byte in input.bytes().filter(|byte| *byte != b'=') {
189 let value = match byte {
190 b'A'..=b'Z' => byte - b'A',
191 b'a'..=b'z' => byte - b'a' + 26,
192 b'0'..=b'9' => byte - b'0' + 52,
193 b'+' => 62,
194 b'/' => 63,
195 _ => return None,
196 };
197 buffer = (buffer << 6) | u32::from(value);
198 bits += 6;
199 if bits >= 8 {
200 bits -= 8;
201 bytes.push((buffer >> bits) as u8);
202 }
203 }
204 String::from_utf8(bytes).ok()
205}
206
207/// The response for a refused git request. Anonymous callers are asked to
208/// authenticate, which is what makes git prompt for credentials.
209pub fn refuse<T>(outcome: Outcome<T>) -> Result<Response> {
210 let Outcome::Fail(failure) = outcome else {
211 return Response::error("Not found", 404);
212 };
213 let mut response = Response::error(failure.message, failure.code.http_status())?;
214 if failure.code == FailureCode::Unauthenticated {
215 response
216 .headers_mut()
217 .set("www-authenticate", "Basic realm=\"g1t\"")?;
218 }
219 Ok(response)
220}
221
222/// `url` with its first path segment, the workspace, replaced by `slug`.
223pub fn with_namespace(url: &Url, slug: &str) -> Option<String> {
224 let rest = url.path().strip_prefix('/')?.split_once('/')?.1;
225 let mut moved = url.clone();
226 moved.set_path(&format!("/{slug}/{rest}"));
227 Some(moved.to_string())
228}
229
230/// Where a git request for a renamed workspace's old address should go
231/// now, if its first segment is an old slug that still redirects.
232pub async fn renamed(url: &Url, identity: &Fetcher) -> Result<Option<String>> {
233 let Some(old) = url.path().strip_prefix('/').and_then(|path| path.split('/').next()) else {
234 return Ok(None);
235 };
236 let current: Option<String> = g1t_kit::call(
237 identity,
238 "resolve_slug",
239 &g1t_contracts::identity::SlugArgs {
240 slug: old.to_owned(),
241 },
242 )
243 .await?;
244 Ok(current.and_then(|slug| with_namespace(url, &slug)))
245}
246
247/// The request under the workspace its first segment is an alias of
248/// (identity's aliases.rs: `g1t` for `flagon-io`), if it is one. Answered
249/// in place rather than redirected: a push does not follow a redirect.
250pub async fn aliased(git: &GitRequest, identity: &Fetcher) -> Result<Option<GitRequest>> {
251 let slug: Option<String> = g1t_kit::call(
252 identity,
253 "resolve_alias",
254 &g1t_contracts::identity::SlugArgs {
255 slug: git.path.namespace.clone(),
256 },
257 )
258 .await?;
259 Ok(slug.map(|slug| git.under(&slug)))
260}
261
262/// `url` with its repository, the first two path segments, replaced by
263/// `to`: where a request for a transferred repository's old path goes.
264/// Keeps whether the old address ended in `.git`.
265pub fn transferred(url: &Url, to: &RepoPath) -> Option<String> {
266 let path = url.path().strip_prefix('/')?;
267 let mut segments = path.splitn(3, '/');
268 let (_, name, rest) = (segments.next()?, segments.next()?, segments.next()?);
269 let suffix = if name.ends_with(".git") { ".git" } else { "" };
270 let mut moved = url.clone();
271 moved.set_path(&format!("/{}/{}{suffix}/{rest}", to.namespace, to.name));
272 Some(moved.to_string())
273}
274
275/// A permanent redirect: 301 for git's first request for refs, which it
276/// follows and then uses the new address for the rest; 308 for the
277/// others, so a POST stays a POST.
278pub fn moved(location: &str, get: bool) -> Result<Response> {
279 let mut response = Response::empty()?.with_status(if get { 301 } else { 308 });
280 response.headers_mut().set("location", location)?;
281 Ok(response)
282}
283
284const ZERO_ID: &str = "0000000000000000000000000000000000000000";
285const HEADS: &str = "refs/heads/";
286const TAGS: &str = "refs/tags/";
287
288/// One ref a push asks to change.
289struct Command {
290 old: String,
291 new: String,
292 name: String,
293}
294
295/// The commands at the start of a receive-pack request, and the
296/// capabilities the client sent with the first of them.
297fn commands(body: &[u8]) -> (Vec<Command>, String) {
298 let mut commands = Vec::new();
299 let mut capabilities = String::new();
300 let mut position = 0;
301 // Commands are pkt-lines; a flush packet ends them and the pack follows.
302 while let Some(length) = body
303 .get(position..position + 4)
304 .and_then(|hex| std::str::from_utf8(hex).ok())
305 .and_then(|hex| usize::from_str_radix(hex, 16).ok())
306 {
307 if length < 4 || position + length > body.len() {
308 break;
309 }
310 let line = &body[position + 4..position + length];
311 position += length;
312 // `<old> <new> <ref>`, and on the first command a NUL then capabilities.
313 let mut halves = line.splitn(2, |byte| *byte == 0);
314 let command = halves.next().unwrap_or_default();
315 if let Some(rest) = halves.next() {
316 capabilities = String::from_utf8_lossy(rest).trim().to_owned();
317 }
318 let Ok(command) = std::str::from_utf8(command) else {
319 continue;
320 };
321 let mut parts = command.trim_end().splitn(3, ' ');
322 if let (Some(old), Some(new), Some(name)) = (parts.next(), parts.next(), parts.next()) {
323 commands.push(Command {
324 old: old.to_owned(),
325 new: new.to_owned(),
326 name: name.to_owned(),
327 });
328 }
329 }
330 (commands, capabilities)
331}
332
333/// The bytes of the pack a receive-pack request carries: everything after
334/// the flush packet that ends its commands. Zero for a push that only
335/// deletes refs.
336pub(crate) fn pack_bytes(body: &[u8]) -> u64 {
337 let mut position = 0;
338 while let Some(length) = body
339 .get(position..position + 4)
340 .and_then(|hex| std::str::from_utf8(hex).ok())
341 .and_then(|hex| usize::from_str_radix(hex, 16).ok())
342 {
343 if length == 0 {
344 return (body.len() - position - 4) as u64;
345 }
346 if length < 4 || position + length > body.len() {
347 break;
348 }
349 position += length;
350 }
351 0
352}
353
354pub(crate) fn pkt_line(payload: &[u8]) -> Vec<u8> {
355 let mut line = format!("{:04x}", payload.len() + 4).into_bytes();
356 line.extend_from_slice(payload);
357 line
358}
359
360/// The branches and tags a push asks to change, as rules see them: the
361/// full ref, where it pointed (`None`: it is created) and where it will
362/// (`None`: it is deleted).
363pub(crate) fn ref_updates(body: &[u8]) -> Vec<(String, Option<String>, Option<String>)> {
364 commands(body)
365 .0
366 .into_iter()
367 .filter(|command| command.name.starts_with(HEADS) || command.name.starts_with(TAGS))
368 .map(|Command { old, new, name }| (name, (old != ZERO_ID).then_some(old), (new != ZERO_ID).then_some(new)))
369 .collect()
370}
371
372/// A report-status answer, as git expects it.
373pub(crate) fn report_response(report: Vec<u8>) -> Result<Response> {
374 let headers = Headers::new();
375 headers.set("content-type", "application/x-git-receive-pack-result")?;
376 headers.set("cache-control", "no-cache")?;
377 Ok(Response::from_bytes(report)?.with_headers(headers))
378}
379
380/// What git is told when a push would change a protected branch: every ref
381/// in it is declined, with the reason against the protected one, so that
382/// git prints it beside the branch. `None` if the push leaves the branch
383/// alone, or creates it in a repository that does not have it yet. Only
384/// where rulesets cannot be read (an installation without the work
385/// service); rulesets decide everywhere else (rules.rs).
386pub(crate) fn refusal(body: &[u8], protected: &str) -> Option<Vec<u8>> {
387 let (commands, capabilities) = commands(body);
388 let reference = format!("{HEADS}{protected}");
389 if !commands
390 .iter()
391 .any(|command| command.name == reference && command.old != ZERO_ID)
392 {
393 return None;
394 }
395 let mut report = pkt_line(b"unpack ok\n");
396 for command in &commands {
397 let reason = if command.name == reference {
398 format!("{protected} is protected: push a branch and open a pull request")
399 } else {
400 format!("not pushed, because the same push would change {protected}")
401 };
402 report.extend(pkt_line(
403 format!("ng {} {reason}\n", command.name).as_bytes(),
404 ));
405 }
406 report.extend_from_slice(b"0000");
407 Some(framed(report, &capabilities, &[]))
408}
409
410/// A report-status as git expects it: inside channel 1 when the client
411/// asked for side-band, after `messages` on channel 2, which git prints as
412/// `remote:` lines. Without side-band the messages cannot be shown.
413pub(crate) fn framed(report: Vec<u8>, capabilities: &str, messages: &[String]) -> Vec<u8> {
414 let sideband = capabilities
415 .split(' ')
416 .any(|capability| capability.starts_with("side-band"));
417 if !sideband {
418 return report;
419 }
420 let mut body = Vec::new();
421 for message in messages {
422 let mut packet = vec![2u8];
423 packet.extend_from_slice(message.as_bytes());
424 packet.push(b'\n');
425 body.extend(pkt_line(&packet));
426 }
427 // side-band (not -64k) packets carry at most 1000 bytes.
428 for chunk in report.chunks(990) {
429 let mut packet = vec![1u8];
430 packet.extend_from_slice(chunk);
431 body.extend(pkt_line(&packet));
432 }
433 body.extend_from_slice(b"0000");
434 body
435}
436
437/// Declines every ref in a push with `reason`, explaining why in
438/// `messages`: what push protection answers when a push adds a secret.
439pub fn declined(body: &[u8], reason: &str, messages: &[String]) -> Result<Response> {
440 let (commands, capabilities) = commands(body);
441 let mut report = pkt_line(b"unpack ok\n");
442 for command in &commands {
443 report.extend(pkt_line(format!("ng {} {reason}\n", command.name).as_bytes()));
444 }
445 report.extend_from_slice(b"0000");
446 let headers = Headers::new();
447 headers.set("content-type", "application/x-git-receive-pack-result")?;
448 headers.set("cache-control", "no-cache")?;
449 Ok(Response::from_bytes(framed(report, &capabilities, messages))?.with_headers(headers))
450}
451
452/// A branch or tag a push asks to move.
453#[derive(Debug, PartialEq, Eq)]
454pub struct Pushed {
455 /// The full ref: `refs/heads/main`, `refs/tags/v1`.
456 pub git_ref: String,
457 /// Where it pointed before; `None` for a new ref.
458 pub before: Option<String>,
459 pub after: String,
460}
461
462impl Pushed {
463 pub fn branch(&self) -> Option<&str> {
464 self.git_ref.strip_prefix(HEADS)
465 }
466}
467
468/// The branches and tags a push asks to move, read from the commands at the
469/// start of a receive-pack request. Deletions and other refs are left out.
470fn pushed_branches(body: &[u8]) -> Vec<Pushed> {
471 commands(body)
472 .0
473 .into_iter()
474 .filter(|command| command.new != ZERO_ID)
475 .filter(|command| command.name.starts_with(HEADS) || command.name.starts_with(TAGS))
476 .map(|Command { old, new, name }| Pushed {
477 git_ref: name,
478 before: (old != ZERO_ID).then_some(old),
479 after: new,
480 })
481 .collect()
482}
483
484/// The branches and tags a push asks to delete: each full ref, with the
485/// commit it pointed to.
486fn deleted_refs(body: &[u8]) -> Vec<(String, String)> {
487 commands(body)
488 .0
489 .into_iter()
490 .filter(|command| command.new == ZERO_ID && command.old != ZERO_ID)
491 .filter(|command| command.name.starts_with(HEADS) || command.name.starts_with(TAGS))
492 .map(|Command { old, name, .. }| (name, old))
493 .collect()
494}
495
496/// The git store's answer, and what the request asked it to change.
497pub struct Forwarded {
498 pub response: Response,
499 /// For a push: the branches and tags it asks to move, and the commits
500 /// to move them to. Whether each moved is for the caller to confirm.
501 pub pushed: Vec<Pushed>,
502 /// For a push: the branches and tags it asks to delete, each with the
503 /// commit it pointed to. Whether each went is for the caller to confirm.
504 pub deleted: Vec<(String, String)>,
505 /// For a push: the size of the pack it sent, for the storage meter.
506 pub pack_bytes: u64,
507 /// The bytes sent to the store.
508 pub sent: u64,
509 /// Whether the answer is the store's own (not g1t's, for a store that
510 /// was busy).
511 pub from_store: bool,
512 /// For a push: whether it was too large to scan for secrets first and
513 /// was streamed to the store unscanned (`LargePushes::Unscanned`). Its
514 /// `git.push` events say so, and security scans it after it lands.
515 pub unscanned: bool,
516}
517
518/// What became of a git request.
519pub enum Push {
520 Forwarded(Forwarded),
521 /// A push the rules of its branches or tags refuse (rules.rs), answered
522 /// here without reaching the store.
523 Refused(Response),
524 /// A push that adds a secret nobody allowed, answered the same way.
525 Blocked(Response),
526 /// A push the store could not hold: an object or the repository too
527 /// large, or too large to check. With the reason, for the audit log.
528 Declined(Response, String),
529}
530
531/// What a push may bring, checked as it arrives (pack_limits.rs).
532#[derive(Clone, Copy, Debug)]
533pub struct PushLimits {
534 /// The largest object the store holds.
535 pub max_object: u64,
536 /// What the repository holds now, as g1t counts it.
537 pub held: u64,
538 /// The most a repository may hold.
539 pub repo_limit: u64,
540 /// The largest push that is read whole and scanned for secrets.
541 pub scan_cap: usize,
542 /// What happens to a larger one.
543 pub large: LargePushes,
544}
545
546impl Default for PushLimits {
547 fn default() -> Self {
548 PushLimits {
549 max_object: crate::pack_limits::MAX_OBJECT_BYTES,
550 held: 0,
551 repo_limit: crate::pack_limits::DEFAULT_REPO_LIMIT_BYTES,
552 scan_cap: crate::secret_scan::MAX_SCANNED_PUSH,
553 large: LargePushes::Refuse,
554 }
555 }
556}
557
558/// What happens to a push larger than [`PushLimits::scan_cap`]: set by
559/// `LARGE_PUSHES`.
560#[derive(Clone, Copy, Debug, PartialEq, Eq)]
561pub enum LargePushes {
562 /// Declined (the default): push protection cannot read it, so it does
563 /// not let it in.
564 Refuse,
565 /// Streamed to the store without a scan for secrets; the size limits are
566 /// still checked as it passes.
567 Unscanned,
568}
569
570impl LargePushes {
571 pub fn from_var(value: Option<&str>) -> Self {
572 match value.map(str::trim) {
573 Some("unscanned") => LargePushes::Unscanned,
574 _ => LargePushes::Refuse,
575 }
576 }
577}
578
579/// A push the store could not hold.
580#[derive(Clone, Debug, PartialEq, Eq)]
581pub enum SizeViolation {
582 Object { size: u64 },
583 Repository { held: u64, incoming: u64, limit: u64 },
584 Unscannable { size: u64, cap: usize },
585}
586
587/// Why a push is declined for its size, as git shows it: the `ng` reason,
588/// and the lines printed as `remote:`.
589pub fn size_refusal(violation: &SizeViolation) -> (String, Vec<String>) {
590 use crate::pack_limits::{MAX_OBJECT_BYTES, PLATFORM_BODY_LIMIT_BYTES, megabytes};
591 match violation {
592 SizeViolation::Object { size } => (
593 format!("a file of {} is over the {} limit", megabytes(*size), megabytes(MAX_OBJECT_BYTES)),
594 vec![
595 format!("g1t stores files of up to {} each; this push has one of {}.", megabytes(MAX_OBJECT_BYTES), megabytes(*size)),
596 "Take it out of the commits (git rm --cached, then amend or rebase), and keep large".to_owned(),
597 "files elsewhere: https://docs.g1t.sh/guides/git/#size-limits. Nothing was pushed.".to_owned(),
598 ],
599 ),
600 SizeViolation::Repository { held, incoming, limit } => (
601 "the repository would be over its size limit".to_owned(),
602 vec![
603 format!(
604 "This repository holds about {} and the push adds {}, past the {} a repository may hold.",
605 megabytes(*held),
606 megabytes(*incoming),
607 megabytes(*limit)
608 ),
609 "Delete what you no longer need, or split it: https://docs.g1t.sh/guides/git/#size-limits.".to_owned(),
610 "Nothing was pushed.".to_owned(),
611 ],
612 ),
613 SizeViolation::Unscannable { size, cap } => (
614 "the push is too large to check for secrets".to_owned(),
615 vec![
616 format!(
617 "g1t checks every push for secrets and reads up to {} at once; this one is {}.",
618 megabytes(*cap as u64),
619 megabytes(*size)
620 ),
621 "Push in parts, oldest commits first, then push as usual:".to_owned(),
622 " git rev-list --reverse HEAD | awk 'NR % 500 == 0' | xargs -I{} git push origin {}:refs/heads/main".to_owned(),
623 format!("A push over {} is refused by the network before it reaches g1t (HTTP 413).", megabytes(PLATFORM_BODY_LIMIT_BYTES)),
624 "See https://docs.g1t.sh/guides/git/#size-limits. Nothing was pushed.".to_owned(),
625 ],
626 ),
627 }
628}
629
630/// Feeds the next chunk of a push to the size check; the first violation.
631fn check_size(sizer: &mut Option<PackSizer>, chunk: &[u8], limits: &PushLimits) -> Option<SizeViolation> {
632 let walker = sizer.as_mut()?;
633 match walker.feed(chunk) {
634 Ok(()) => {}
635 Err(Violation::ObjectTooLarge { size }) => return Some(SizeViolation::Object { size }),
636 Err(Violation::Malformed(why)) => {
637 // Not for g1t to judge: the store will say.
638 worker::console_error!("push not checked for size: {why}");
639 *sizer = None;
640 return None;
641 }
642 }
643 let incoming = walker.pack_bytes();
644 crate::pack_limits::over_repo_limit(limits.held, incoming, limits.repo_limit).then_some(SizeViolation::Repository {
645 held: limits.held,
646 incoming,
647 limit: limits.repo_limit,
648 })
649}
650
651/// A request body that streams from `stream`, for `fetch`.
652pub(crate) fn stream_body<S>(stream: S) -> Result<JsValue>
653where
654 S: futures_util::TryStream + 'static,
655 S::Ok: Into<Vec<u8>>,
656 S::Error: Into<worker::Error>,
657{
658 let response: worker::web_sys::Response = Response::from_stream(stream)?.into();
659 Ok(response.body().map_or(JsValue::NULL, Into::into))
660}
661
662/// What git is told when the store is busy: 429 or 503, with when to try
663/// again (resilience.rs).
664pub fn busy_response(busy: Busy) -> Result<Response> {
665 let response = Response::error(busy.message(), busy.status())?;
666 response.headers().set("retry-after", &busy.retry_after.to_string())?;
667 Ok(response)
668}
669
670/// The rest of a request's body, read and thrown away so that git hears
671/// the answer; how many bytes it was.
672async fn drain(stream: &mut worker::ByteStream) -> Result<u64> {
673 let mut size = 0;
674 while let Some(chunk) = stream.next().await {
675 size += chunk?.len() as u64;
676 }
677 Ok(size)
678}
679
680/// One packet of a pkt-line stream: data, or a flush (`0000`), delimiter
681/// (`0001`) or response-end (`0002`) packet, kept as its four bytes.
682#[derive(Debug, PartialEq, Eq)]
683pub(crate) enum Packet {
684 Data(Vec<u8>),
685 Special([u8; 4]),
686}
687
688/// The packets in `bytes`, or `None` if it is not a whole pkt-line stream.
689pub(crate) fn packets(bytes: &[u8]) -> Option<Vec<Packet>> {
690 let mut out = Vec::new();
691 let mut position = 0;
692 while position < bytes.len() {
693 let header = bytes.get(position..position + 4)?;
694 let length = usize::from_str_radix(std::str::from_utf8(header).ok()?, 16).ok()?;
695 if length < 4 {
696 out.push(Packet::Special(header.try_into().ok()?));
697 position += 4;
698 continue;
699 }
700 out.push(Packet::Data(bytes.get(position + 4..position + length)?.to_vec()));
701 position += length;
702 }
703 Some(out)
704}
705
706pub(crate) fn encode(packets: &[Packet]) -> Vec<u8> {
707 let mut out = Vec::new();
708 for packet in packets {
709 match packet {
710 Packet::Data(data) => out.extend(pkt_line(data)),
711 Packet::Special(bytes) => out.extend_from_slice(bytes),
712 }
713 }
714 out
715}
716
717/// A ref advertisement (`info/refs` for upload-pack) or a protocol v2
718/// `ls-refs` answer with `HEAD` pointing at `branch`, the repository's
719/// default branch as g1t keeps it, so a clone checks it out. The git store
720/// holds the HEAD it was created with; g1t can change the default branch
721/// since. `None` when there is nothing to change: no `HEAD` line, `HEAD`
722/// already names `branch`, or `branch` is not advertised.
723pub fn with_head(body: &[u8], branch: &str) -> Option<Vec<u8>> {
724 let mut packets = packets(body)?;
725 let target = format!("{HEADS}{branch}");
726 let oid = packets.iter().find_map(|packet| {
727 let Packet::Data(data) = packet else { return None };
728 let line = data.split(|byte| *byte == 0).next()?;
729 let line = std::str::from_utf8(line).ok()?.trim_end();
730 let (oid, name) = line.split_once(' ')?;
731 // v2 lines may carry attributes after the name.
732 let name = name.split(' ').next()?;
733 (name == target).then(|| oid.to_owned())
734 })?;
735 let mut changed = false;
736 for packet in &mut packets {
737 let Packet::Data(data) = packet else { continue };
738 let text = String::from_utf8_lossy(data).into_owned();
739 let Some((_, rest)) = text.split_once(' ') else { continue };
740 if !(rest.starts_with("HEAD\0") || rest.starts_with("HEAD\n") || rest.starts_with("HEAD ") || rest == "HEAD") {
741 continue;
742 }
743 let mut line = format!("{oid} {rest}");
744 // v0: `symref=HEAD:refs/heads/<old>` among the capabilities.
745 // v2: `symref-target:refs/heads/<old>` after the name.
746 for marker in ["symref=HEAD:", "symref-target:"] {
747 if let Some(at) = line.find(marker) {
748 let start = at + marker.len();
749 let end = line[start..]
750 .find([' ', '\n', '\0'])
751 .map_or(line.len(), |offset| start + offset);
752 line.replace_range(start..end, &target);
753 }
754 }
755 changed = line != text;
756 if changed {
757 *data = line.into_bytes();
758 }
759 break;
760 }
761 changed.then(|| encode(&packets))
762}
763
764/// The capability that asks git to send a push's pack whole: every delta's
765/// base inside it, none left for the receiving end to find (a "thin" pack).
766const NO_THIN: &[u8] = b"no-thin";
767
768/// The receive-pack ref advertisement (`info/refs?service=git-receive-pack`,
769/// protocol v0 or v1) with `no-thin` among its capabilities, so that git
770/// sends a pack whose deltas have their bases in it (git's `send-pack`
771/// turns thin packs off when the server says `no-thin`). Push protection
772/// and the rules then read every object from the pack, and none from the
773/// store (secret_scan.rs `supply_bases`). The capabilities follow the NUL
774/// on the first ref line, or on the `capabilities^{}` line of an empty
775/// repository. `None` when there is nothing to change or it cannot be
776/// changed safely: `no-thin` is there already, or the answer is not a whole
777/// pkt-line advertisement in that shape. Never for upload-pack.
778pub fn with_no_thin(body: &[u8]) -> Option<Vec<u8>> {
779 let mut packets = packets(body)?;
780 // Past the `# service=` line and its flush, and v1's `version 1`: the
781 // first ref line, which carries the capabilities.
782 let line = packets.iter_mut().find_map(|packet| match packet {
783 Packet::Data(data) if !data.starts_with(b"# service=") && !data.starts_with(b"version ") => Some(data),
784 _ => None,
785 })?;
786 let nul = line.iter().position(|byte| *byte == 0)?;
787 // `<oid> <ref>` before the NUL: 40 (SHA-1) or 64 (SHA-256) hex digits.
788 let (oid, name) = std::str::from_utf8(&line[..nul]).ok()?.split_once(' ')?;
789 if !matches!(oid.len(), 40 | 64) || !oid.bytes().all(|byte| byte.is_ascii_hexdigit()) || name.is_empty() {
790 return None;
791 }
792 let end = if line.ends_with(b"\n") { line.len() - 1 } else { line.len() };
793 let capabilities = &line[nul + 1..end];
794 if capabilities.split(|byte| *byte == b' ').any(|capability| capability == NO_THIN) {
795 return None;
796 }
797 let mut added = Vec::with_capacity(NO_THIN.len() + 1);
798 if !capabilities.is_empty() && !capabilities.ends_with(b" ") {
799 added.push(b' ');
800 }
801 added.extend_from_slice(NO_THIN);
802 // A pkt-line holds at most 65516 bytes of data.
803 if line.len() + added.len() > 65516 {
804 return None;
805 }
806 line.splice(end..end, added);
807 Some(encode(&packets))
808}
809
810/// Whether a request to the git store is one whose answer names `HEAD`:
811/// the ref advertisement for a fetch, or a protocol v2 `ls-refs`.
812fn names_head(git: &GitRequest, body: Option<&[u8]>) -> bool {
813 if git.service != GitService::UploadPack {
814 return false;
815 }
816 match body {
817 None => git.endpoint == "info/refs",
818 Some(body) => {
819 git.endpoint == "git-upload-pack"
820 && body.windows(b"command=ls-refs".len()).any(|window| window == b"command=ls-refs")
821 }
822 }
823}
824
825/// Whether a request is a protocol v2 `fetch` still negotiating: it sends
826/// `have` lines and no `done`, so the answer may be acknowledgments only.
827fn negotiating(body: &[u8]) -> bool {
828 let Some(packets) = packets(body) else { return false };
829 let lines: Vec<&[u8]> = packets
830 .iter()
831 .filter_map(|packet| match packet {
832 Packet::Data(data) => Some(data.strip_suffix(b"\n").unwrap_or(data)),
833 Packet::Special(_) => None,
834 })
835 .collect();
836 lines.contains(&b"command=fetch".as_slice())
837 && lines.iter().any(|line| line.starts_with(b"have "))
838 && !lines.contains(&b"done".as_slice())
839}
840
841/// What to do with the start of a store's answer to a negotiating fetch.
842#[derive(Debug, PartialEq, Eq)]
843enum Acknowledged {
844 /// Not enough of it yet to tell.
845 NeedMore,
846 /// Send it on as it is.
847 Whole,
848 /// Acknowledgments without `ready`, followed by more sections: the
849 /// store's answer to keep is these first bytes, ended by a flush.
850 CutAt(usize),
851}
852
853/// How much of the answer to keep. The git store answers a fetch whose
854/// `have`s it does not know with `acknowledgments`, `NAK`, then a pack
855/// anyway; git refuses that ("expected no other sections to be sent after
856/// no 'ready'"), since a server that is not ready must end the response
857/// there and let the client negotiate again. Lines may be `sideband-all`
858/// framed (band 1, `\x01`).
859fn acknowledged(head: &[u8]) -> Acknowledged {
860 let mut position = 0;
861 let mut first = true;
862 loop {
863 let Some(header) = head.get(position..position + 4) else { return Acknowledged::NeedMore };
864 let Some(length) = std::str::from_utf8(header).ok().and_then(|hex| usize::from_str_radix(hex, 16).ok()) else {
865 return Acknowledged::Whole;
866 };
867 if length < 4 {
868 // The acknowledgments section's end: a delimiter means more
869 // sections follow, which only `ready` allows.
870 return match (first, header) {
871 (false, b"0001") => Acknowledged::CutAt(position),
872 _ => Acknowledged::Whole,
873 };
874 }
875 let Some(payload) = head.get(position + 4..position + length) else { return Acknowledged::NeedMore };
876 let line = payload.strip_prefix(b"\x01").unwrap_or(payload);
877 let line = line.strip_suffix(b"\n").unwrap_or(line);
878 if first && line != b"acknowledgments" {
879 return Acknowledged::Whole;
880 }
881 if line == b"ready" {
882 return Acknowledged::Whole;
883 }
884 first = false;
885 position += length;
886 }
887}
888
889/// The answer to a negotiating fetch, with the sections the store sent
890/// after acknowledgments without `ready` left off (see [`acknowledged`]).
891/// Reads only the start of the answer; the rest streams through.
892async fn without_early_pack(mut response: Response) -> Result<Response> {
893 const LOOK: usize = 64 * 1024;
894 let headers = response.headers().clone();
895 headers.delete("content-length")?;
896 let mut stream = response.stream()?;
897 let mut head = Vec::new();
898 loop {
899 match acknowledged(&head) {
900 Acknowledged::CutAt(at) => {
901 head.truncate(at);
902 // A flush ends the acknowledgments and the answer. No
903 // response-end packet: git's HTTP transport adds its own
904 // and refuses one from the server.
905 head.extend_from_slice(b"0000");
906 return Ok(Response::from_bytes(head)?.with_headers(headers));
907 }
908 Acknowledged::Whole => break,
909 Acknowledged::NeedMore if head.len() >= LOOK => break,
910 Acknowledged::NeedMore => match stream.next().await {
911 Some(chunk) => head.extend_from_slice(&chunk?),
912 None => break,
913 },
914 }
915 }
916 let rest = futures_util::stream::once(async move { Ok::<Vec<u8>, worker::Error>(head) }).chain(stream);
917 Ok(Response::from_stream(rest)?.with_headers(headers))
918}
919
920/// Sends the request on to the git store and returns its response as is,
921/// unless it is a push `checks` refuse (the workflow gate, the rules, push
922/// protection: push_checks.rs), or one the store could not hold (`limits`,
923/// pack_limits.rs). `checks` is given as much of the push as was read,
924/// whether that is all of it, and whether to scan it for secrets. A
925/// fetch's ref listing has its `HEAD` pointed at `default_branch` (see
926/// [`with_head`]). A POST's body is `read` when the caller has read it
927/// already.
928///
929/// A push is read as it arrives: up to `limits.scan_cap` is kept, to be
930/// scanned and sent on whole; past it, the push is declined, or streamed
931/// to the store unscanned (`LargePushes`), never held. Reads the store
932/// fails for a moment (429, 5xx) are tried again with backoff; a push never
933/// is. A store still busy after that is answered 429 or 503 with
934/// `Retry-After`.
935#[allow(clippy::too_many_arguments)]
936pub async fn forward(
937 mut request: Request,
938 read: Option<Vec<u8>>,
939 git: &GitRequest,
940 access: &GitAccess,
941 checks: impl AsyncFnOnce(&[u8], bool, bool) -> Result<Checks>,
942 default_branch: Option<&str>,
943 limits: PushLimits,
944 timing: &mut Timing,
945) -> Result<Push> {
946 let headers = Headers::new();
947 headers.set("authorization", &format!("Bearer {}", access.token))?;
948 for name in FORWARDED_HEADERS {
949 if let Some(value) = request.headers().get(name)? {
950 headers.set(name, &value)?;
951 }
952 }
953 let query = request.url()?.query().map(|query| format!("?{query}")).unwrap_or_default();
954 let url = format!("{}/{}{query}", access.remote, git.endpoint);
955 let method = request.method();
956 // Its own health and breaker: the fallback store's apart from Artifacts'.
957 let namespace = crate::store::health_namespace(&access.remote);
958
959 if method == Method::Post && git.endpoint == "git-receive-pack" {
960 return push(request, &url, headers, checks, limits, &namespace, timing).await;
961 }
962
963 // A read: the ref advertisement, `ls-refs`, or a fetch of objects.
964 let body = match (&method, read) {
965 (Method::Post, Some(body)) => Some(body),
966 (Method::Post, None) => Some(request.bytes().await?),
967 _ => None,
968 };
969 let lists_head = match &body {
970 None => method == Method::Get && names_head(git, None),
971 Some(body) => names_head(git, Some(body)),
972 };
973 let sent = body.as_ref().map_or(0, |body| body.len() as u64);
974 let mut attempt = 0;
975 let mut response = loop {
976 let mut init = RequestInit::new();
977 init.with_method(method.clone()).with_headers(headers.clone());
978 if let Some(body) = &body {
979 init.with_body(Some(Uint8Array::from(body.as_slice()).into()));
980 }
981 let started = g1t_kit::now_ms();
982 let answered = Fetch::Request(Request::new_with_init(&url, &init)?).send().await;
983 let ms = g1t_kit::now_ms().saturating_sub(started);
984 let failure = match &answered {
985 Ok(response) => resilience::classify_status(response.status_code()),
986 Err(_) => Some(Failure::Transient),
987 };
988 let outcome = match failure {
989 None => meters::Outcome::Ok,
990 Some(Failure::RateLimited) => meters::Outcome::RateLimited,
991 Some(_) => meters::Outcome::Failed,
992 };
993 meters::record_health(&namespace, outcome, ms);
994 match (answered, failure) {
995 (Ok(response), None) => break response,
996 (answered, Some(failure)) if resilience::retry(failure, attempt) => {
997 drop(answered);
998 let wait = resilience::backoff_ms(failure, attempt, worker::js_sys::Math::random());
999 worker::Delay::from(std::time::Duration::from_millis(wait)).await;
1000 attempt += 1;
1001 }
1002 (_, Some(failure)) => {
1003 let busy = Busy { rate_limited: failure == Failure::RateLimited, retry_after: 5, read_only: false };
1004 return Ok(Push::Forwarded(Forwarded {
1005 response: busy_response(busy)?,
1006 pushed: Vec::new(),
1007 deleted: Vec::new(),
1008 pack_bytes: 0,
1009 sent,
1010 from_store: false,
1011 unscanned: false,
1012 }));
1013 }
1014 (Err(error), None) => return Err(error),
1015 }
1016 };
1017 if let (true, Some(branch)) = (lists_head, default_branch)
1018 && response.status_code() == 200
1019 {
1020 let headers = response.headers().clone();
1021 headers.delete("content-length")?;
1022 let body = response.bytes().await?;
1023 let body = with_head(&body, branch).unwrap_or(body);
1024 response = Response::from_bytes(body)?.with_headers(headers);
1025 }
1026 // A push's ref advertisement asks for a pack without outside bases.
1027 if method == Method::Get
1028 && git.service == GitService::ReceivePack
1029 && git.endpoint == "info/refs"
1030 && response.status_code() == 200
1031 {
1032 let headers = response.headers().clone();
1033 headers.delete("content-length")?;
1034 let body = response.bytes().await?;
1035 let body = with_no_thin(&body).unwrap_or(body);
1036 // Without the refs g1t keeps, as GitHub hides them from a push.
1037 let body = hidden_refs::advertisement_without_hidden(&body).unwrap_or(body);
1038 response = Response::from_bytes(body)?.with_headers(headers);
1039 }
1040 if git.endpoint == "git-upload-pack"
1041 && response.status_code() == 200
1042 && body.as_deref().is_some_and(negotiating)
1043 {
1044 response = without_early_pack(response).await?;
1045 }
1046 Ok(Push::Forwarded(Forwarded {
1047 response,
1048 pushed: Vec::new(),
1049 deleted: Vec::new(),
1050 pack_bytes: 0,
1051 sent,
1052 from_store: true,
1053 unscanned: false,
1054 }))
1055}
1056
1057/// A receive-pack request; see [`forward`]. Its steps: `recv` (the push
1058/// read), `checks`, `upload` (the store's answer).
1059#[allow(clippy::too_many_arguments)]
1060async fn push(
1061 mut request: Request,
1062 url: &str,
1063 headers: Headers,
1064 checks: impl AsyncFnOnce(&[u8], bool, bool) -> Result<Checks>,
1065 limits: PushLimits,
1066 namespace: &str,
1067 timing: &mut Timing,
1068) -> Result<Push> {
1069 let mut stream = request.stream()?;
1070 let mut head: Vec<u8> = Vec::new();
1071 let mut sizer = Some(PackSizer::new(limits.max_object));
1072 let mut violation = None;
1073 let mut ended = false;
1074 while head.len() <= limits.scan_cap {
1075 match stream.next().await {
1076 Some(chunk) => {
1077 let chunk = chunk?;
1078 if violation.is_none() {
1079 violation = check_size(&mut sizer, &chunk, &limits);
1080 }
1081 head.extend_from_slice(&chunk);
1082 }
1083 None => {
1084 ended = true;
1085 break;
1086 }
1087 }
1088 }
1089 timing.mark("recv");
1090 // Commands to the refs g1t keeps for itself (`refs/pull/*`) are taken
1091 // out and refused ref by ref once the store has answered for the rest
1092 // (hidden_refs.rs). One that leaves nothing to send is answered here.
1093 let hidden = match hidden_refs::split(&head) {
1094 hidden_refs::Split::Untouched => Vec::new(),
1095 hidden_refs::Split::Without { body, hidden } => {
1096 head = body;
1097 hidden
1098 }
1099 hidden_refs::Split::Refused { report, .. } => {
1100 if !ended {
1101 drain(&mut stream).await?;
1102 }
1103 return Ok(Push::Declined(report_response(report)?, hidden_refs::HIDDEN_REASON.to_owned()));
1104 }
1105 };
1106 // The workflow gate and the rules of the branches and tags it changes,
1107 // with push protection alongside for a push read whole that is within
1108 // the size limits (push_checks.rs). What the rules refuse is refused
1109 // whatever else is wrong with it.
1110 let checked = checks(&head, ended, ended && violation.is_none()).await?;
1111 for (part, ms) in checked.spans {
1112 timing.part(part, ms);
1113 }
1114 // Whether the pack came thin: g1t asks for whole ones (`no-thin`, see
1115 // [`with_no_thin`]), so a thin one is a client that ignored it, and its
1116 // bases were read from the store (`read`).
1117 if let Some(bases) = checked.bases {
1118 timing.note("thin", if bases.thin() { "yes" } else { "no" });
1119 if bases.thin() {
1120 let agent = request.headers().get("user-agent").ok().flatten().unwrap_or_default();
1121 worker::console_warn!(
1122 "a thin push from {agent}: {} bases outside the pack, {} asked of the store, {} left unresolved",
1123 bases.missing,
1124 bases.asked,
1125 bases.left
1126 );
1127 }
1128 }
1129 timing.mark("checks");
1130 let blocked = match checked.verdict {
1131 Verdict::Refused(response) => {
1132 if !ended {
1133 drain(&mut stream).await?;
1134 }
1135 return Ok(Push::Refused(response));
1136 }
1137 Verdict::Blocked(response) => Some(response),
1138 Verdict::Clear => None,
1139 };
1140 if !ended && limits.large == LargePushes::Refuse && violation.is_none() {
1141 let size = head.len() as u64 + drain(&mut stream).await?;
1142 violation = Some(SizeViolation::Unscannable { size, cap: limits.scan_cap });
1143 ended = true;
1144 }
1145 if let Some(violation) = violation {
1146 if !ended {
1147 drain(&mut stream).await?;
1148 }
1149 let (reason, messages) = size_refusal(&violation);
1150 return Ok(Push::Declined(declined(&head, &reason, &messages)?, reason));
1151 }
1152 let pushed = pushed_branches(&head);
1153 let deleted = deleted_refs(&head);
1154 // The commands as sent, for framing the refusals of hidden refs.
1155 let reported_head: Vec<u8> = if hidden.is_empty() { Vec::new() } else { head.iter().take(64 * 1024).copied().collect() };
1156 let mut init = RequestInit::new();
1157 init.with_method(Method::Post).with_headers(headers);
1158 let started = g1t_kit::now_ms();
1159 let (answered, pack_bytes, sent, unscanned) = if ended {
1160 if let Some(response) = blocked {
1161 return Ok(Push::Blocked(response));
1162 }
1163 let pack = pack_bytes(&head);
1164 let sent = head.len() as u64;
1165 init.with_body(Some(Uint8Array::from(head.as_slice()).into()));
1166 drop(head);
1167 (Fetch::Request(Request::new_with_init(url, &init)?).send().await, pack, sent, false)
1168 } else {
1169 // Larger than can be scanned, and let through unscanned: streamed,
1170 // with the size limits checked as it passes. A violation ends the
1171 // stream before the pack does, so the store refuses it whole.
1172 worker::console_warn!("a push of more than {} bytes goes to the store unscanned", limits.scan_cap);
1173 let commands = head.iter().take(64 * 1024).copied().collect::<Vec<u8>>();
1174 let found: Rc<RefCell<Option<SizeViolation>>> = Rc::default();
1175 let walked = Rc::new(RefCell::new((sizer, 0u64)));
1176 let rest = {
1177 let found = found.clone();
1178 let walked = walked.clone();
1179 stream.map(move |chunk| {
1180 let chunk = chunk?;
1181 let mut walked = walked.borrow_mut();
1182 walked.1 += chunk.len() as u64;
1183 if let Some(violation) = check_size(&mut walked.0, &chunk, &limits) {
1184 *found.borrow_mut() = Some(violation);
1185 return Err(worker::Error::RustError("push over the size limit".into()));
1186 }
1187 Ok(chunk)
1188 })
1189 };
1190 let first = head.len() as u64;
1191 let body = futures_util::stream::once(async move { Ok::<Vec<u8>, worker::Error>(head) }).chain(rest);
1192 init.with_body(Some(stream_body(body)?));
1193 let answered = Fetch::Request(Request::new_with_init(url, &init)?).send().await;
1194 if let Some(violation) = found.borrow_mut().take() {
1195 let (reason, messages) = size_refusal(&violation);
1196 return Ok(Push::Declined(declined(&commands, &reason, &messages)?, reason));
1197 }
1198 let walked = walked.borrow();
1199 let pack = walked.0.as_ref().map_or_else(|| pack_bytes(&commands), PackSizer::pack_bytes);
1200 (answered, pack, first + walked.1, true)
1201 };
1202 let ms = g1t_kit::now_ms().saturating_sub(started);
1203 timing.mark("upload");
1204 let failure = match &answered {
1205 Ok(response) => resilience::classify_status(response.status_code()),
1206 Err(_) => Some(Failure::Transient),
1207 };
1208 meters::record_health(
1209 namespace,
1210 match failure {
1211 None => meters::Outcome::Ok,
1212 Some(Failure::RateLimited) => meters::Outcome::RateLimited,
1213 Some(_) => meters::Outcome::Failed,
1214 },
1215 ms,
1216 );
1217 // A push is never tried again: the store may have taken it.
1218 let mut response = match (answered, failure) {
1219 (Ok(response), None) => response,
1220 (Ok(response), Some(Failure::RateLimited)) => {
1221 drop(response);
1222 busy_response(Busy { rate_limited: true, retry_after: 5, read_only: false })?
1223 }
1224 (Ok(response), Some(_)) => response,
1225 (Err(error), _) => {
1226 worker::console_error!("a push did not reach the store: {error}");
1227 busy_response(Busy { rate_limited: false, retry_after: 5, read_only: false })?
1228 }
1229 };
1230 if !hidden.is_empty() && response.status_code() == 200 {
1231 let sideband = commands(&reported_head).1.split(' ').any(|capability| capability.starts_with("side-band"));
1232 let headers = response.headers().clone();
1233 headers.delete("content-length")?;
1234 let answer = response.bytes().await?;
1235 let answer = hidden_refs::with_refused(&answer, &hidden, sideband).unwrap_or(answer);
1236 response = Response::from_bytes(answer)?.with_headers(headers);
1237 }
1238 Ok(Push::Forwarded(Forwarded {
1239 response,
1240 pushed,
1241 deleted,
1242 pack_bytes,
1243 sent,
1244 from_store: true,
1245 unscanned,
1246 }))
1247}
1248
1249#[cfg(test)]
1250mod tests {
1251 use super::{Acknowledged, GitService, Pushed, RepoPath, Url, ZERO_ID, acknowledged, deleted_refs, framed, negotiating, pack_bytes, parse, pushed_branches, refusal, server_timing, transferred, with_head, with_namespace, with_no_thin};
1252
1253 #[test]
1254 fn server_timing_names_each_step_and_the_total() {
1255 assert_eq!(
1256 server_timing(&[("repo", 12), ("token", 0), ("store", 140)], &[], &[], 153),
1257 "repo;dur=12, token;dur=0, store;dur=140, total;dur=153"
1258 );
1259 assert_eq!(server_timing(&[], &[], &[], 3), "total;dur=3");
1260 // A push's checks run side by side: their parts come after the steps.
1261 assert_eq!(
1262 server_timing(&[("recv", 30), ("checks", 120), ("upload", 300)], &[("read", 40), ("rules", 110), ("scan", 90)], &[], 450),
1263 "recv;dur=30, checks;dur=120, upload;dur=300, read;dur=40, rules;dur=110, scan;dur=90, total;dur=450"
1264 );
1265 assert_eq!(
1266 server_timing(&[("repo", 1), ("cache", 2)], &[], &[("refs", "hit-colo")], 4),
1267 "repo;dur=1, cache;dur=2, refs;desc=hit-colo, total;dur=4"
1268 );
1269 }
1270
1271 #[test]
1272 fn a_renamed_repository_redirects_to_its_new_name() {
1273 // A rename keeps the old path in the same table as a transfer, so
1274 // the old remote is sent to the new name the same way.
1275 let to = RepoPath {
1276 namespace: "acme".into(),
1277 name: "booster".into(),
1278 };
1279 let url = Url::parse("https://g1t.sh/acme/rocket.git/info/refs?service=git-upload-pack").unwrap();
1280 assert_eq!(
1281 transferred(&url, &to).as_deref(),
1282 Some("https://g1t.sh/acme/booster.git/info/refs?service=git-upload-pack")
1283 );
1284 }
1285
1286 #[test]
1287 fn head_follows_the_default_branch_in_a_v0_advertisement() {
1288 let main = "1111111111111111111111111111111111111111";
1289 let trunk = "2222222222222222222222222222222222222222";
1290 let body = [
1291 pkt("# service=git-upload-pack\n"),
1292 b"0000".to_vec(),
1293 pkt(&format!("{main} HEAD\0multi_ack symref=HEAD:refs/heads/main agent=git/2\n")),
1294 pkt(&format!("{main} refs/heads/main\n")),
1295 pkt(&format!("{trunk} refs/heads/trunk\n")),
1296 b"0000".to_vec(),
1297 ]
1298 .concat();
1299 let changed = String::from_utf8(with_head(&body, "trunk").unwrap()).unwrap();
1300 assert!(changed.contains(&format!("{trunk} HEAD\0multi_ack symref=HEAD:refs/heads/trunk agent=git/2\n")));
1301 assert!(changed.contains(&format!("{main} refs/heads/main\n")));
1302 assert!(changed.starts_with("001e# service=git-upload-pack\n0000"));
1303 // Already right, or a branch it does not have: left alone.
1304 assert!(with_head(&body, "main").is_none());
1305 assert!(with_head(&body, "gone").is_none());
1306 }
1307
1308 #[test]
1309 fn a_push_advertisement_asks_for_a_pack_without_outside_bases() {
1310 let main = "1111111111111111111111111111111111111111";
1311 let topic = "2222222222222222222222222222222222222222";
1312 let body = [
1313 pkt("# service=git-receive-pack\n"),
1314 b"0000".to_vec(),
1315 pkt(&format!("{main} refs/heads/main\0report-status delete-refs side-band-64k quiet ofs-delta agent=git/2.45\n")),
1316 pkt(&format!("{topic} refs/heads/topic\n")),
1317 b"0000".to_vec(),
1318 ]
1319 .concat();
1320 let changed = with_no_thin(&body).unwrap();
1321 let expected = [
1322 pkt("# service=git-receive-pack\n"),
1323 b"0000".to_vec(),
1324 pkt(&format!("{main} refs/heads/main\0report-status delete-refs side-band-64k quiet ofs-delta agent=git/2.45 no-thin\n")),
1325 pkt(&format!("{topic} refs/heads/topic\n")),
1326 b"0000".to_vec(),
1327 ]
1328 .concat();
1329 assert_eq!(String::from_utf8(changed.clone()).unwrap(), String::from_utf8(expected).unwrap());
1330 // The length of the line that grew is its new one: the whole parses.
1331 assert!(super::packets(&changed).is_some());
1332 // Said once: an answer that has it already is left alone.
1333 assert!(with_no_thin(&changed).is_none());
1334
1335 // Protocol v1 begins with `version 1`.
1336 let v1 = [
1337 pkt("# service=git-receive-pack\n"),
1338 b"0000".to_vec(),
1339 pkt("version 1\n"),
1340 pkt(&format!("{main} refs/heads/main\0report-status ofs-delta\n")),
1341 b"0000".to_vec(),
1342 ]
1343 .concat();
1344 let changed = String::from_utf8(with_no_thin(&v1).unwrap()).unwrap();
1345 assert!(changed.contains(&String::from_utf8(pkt(&format!("{main} refs/heads/main\0report-status ofs-delta no-thin\n"))).unwrap()));
1346 assert!(changed.contains("000eversion 1\n"));
1347 }
1348
1349 #[test]
1350 fn an_empty_repository_advertisement_asks_for_a_whole_pack_too() {
1351 let zero = "0000000000000000000000000000000000000000";
1352 let body = [
1353 pkt("# service=git-receive-pack\n"),
1354 b"0000".to_vec(),
1355 pkt(&format!("{zero} capabilities^{{}}\0report-status delete-refs ofs-delta\n")),
1356 b"0000".to_vec(),
1357 ]
1358 .concat();
1359 let changed = String::from_utf8(with_no_thin(&body).unwrap()).unwrap();
1360 assert!(changed.contains(&String::from_utf8(pkt(&format!("{zero} capabilities^{{}}\0report-status delete-refs ofs-delta no-thin\n"))).unwrap()));
1361 // No capabilities at all, and no newline: still one list.
1362 let bare = [pkt("# service=git-receive-pack\n"), b"0000".to_vec(), pkt(&format!("{zero} capabilities^{{}}\0")), b"0000".to_vec()].concat();
1363 let changed = String::from_utf8(with_no_thin(&bare).unwrap()).unwrap();
1364 assert!(changed.contains(&String::from_utf8(pkt(&format!("{zero} capabilities^{{}}\0no-thin"))).unwrap()));
1365 }
1366
1367 #[test]
1368 fn an_advertisement_that_cannot_be_read_goes_through_untouched() {
1369 let main = "1111111111111111111111111111111111111111";
1370 // Not pkt-lines; a length past the end; an error page.
1371 assert!(with_no_thin(b"not a git answer").is_none());
1372 assert!(with_no_thin(b"00ff1111").is_none());
1373 assert!(with_no_thin(b"<html>503 Service Unavailable</html>").is_none());
1374 assert!(with_no_thin(b"").is_none());
1375 // A first ref line without capabilities, or without an object id.
1376 let without = [pkt("# service=git-receive-pack\n"), b"0000".to_vec(), pkt(&format!("{main} refs/heads/main\n")), b"0000".to_vec()].concat();
1377 assert!(with_no_thin(&without).is_none());
1378 let unnamed = [pkt("# service=git-receive-pack\n"), b"0000".to_vec(), pkt("nothing here\0report-status\n"), b"0000".to_vec()].concat();
1379 assert!(with_no_thin(&unnamed).is_none());
1380 // `no-thin` inside another capability's value is not `no-thin`.
1381 let lookalike = [pkt("# service=git-receive-pack\n"), b"0000".to_vec(), pkt(&format!("{main} refs/heads/main\0agent=no-thin-ish\n")), b"0000".to_vec()].concat();
1382 assert!(String::from_utf8(with_no_thin(&lookalike).unwrap()).unwrap().contains("agent=no-thin-ish no-thin\n"));
1383 }
1384
1385 #[test]
1386 fn head_follows_the_default_branch_in_a_v2_listing() {
1387 let main = "1111111111111111111111111111111111111111";
1388 let trunk = "2222222222222222222222222222222222222222";
1389 let body = [
1390 pkt(&format!("{main} HEAD symref-target:refs/heads/main\n")),
1391 pkt(&format!("{main} refs/heads/main\n")),
1392 pkt(&format!("{trunk} refs/heads/trunk\n")),
1393 b"0000".to_vec(),
1394 ]
1395 .concat();
1396 let changed = String::from_utf8(with_head(&body, "trunk").unwrap()).unwrap();
1397 assert!(changed.starts_with(&String::from_utf8(pkt(&format!("{trunk} HEAD symref-target:refs/heads/trunk\n"))).unwrap()));
1398 assert!(changed.ends_with("0000"));
1399 // Without symrefs asked for, only the commit changes.
1400 let plain = [pkt(&format!("{main} HEAD\n")), pkt(&format!("{trunk} refs/heads/trunk\n")), b"0000".to_vec()].concat();
1401 let changed = String::from_utf8(with_head(&plain, "trunk").unwrap()).unwrap();
1402 assert!(changed.starts_with(&format!("0032{trunk} HEAD\n")));
1403 }
1404
1405 #[test]
1406 fn a_push_is_measured_by_the_pack_after_its_commands() {
1407 let old = "c71546fcd893ef8b0f57388b65e620d759705dda";
1408 let new = "4807077b296e6edbf410d55e72749d3e1170c291";
1409 let pack = b"PACK\0\0\0\x02\0\0\0\0rest-of-pack";
1410 let body = [
1411 pkt(&format!("{old} {new} refs/heads/main\0 report-status\n")),
1412 b"0000".to_vec(),
1413 pack.to_vec(),
1414 ]
1415 .concat();
1416 assert_eq!(pack_bytes(&body), pack.len() as u64);
1417 // Only deletions: no pack.
1418 let body = [pkt(&format!("{old} {ZERO_ID} refs/heads/gone\n")), b"0000".to_vec()].concat();
1419 assert_eq!(pack_bytes(&body), 0);
1420 assert_eq!(pack_bytes(b"garbage"), 0);
1421 }
1422
1423 #[test]
1424 fn a_transferred_repository_keeps_the_rest_of_the_address() {
1425 let to = RepoPath {
1426 namespace: "flagon-io".into(),
1427 name: "g1t".into(),
1428 };
1429 let url = Url::parse("https://g1t.sh/syntaqx/g1t.git/info/refs?service=git-receive-pack").unwrap();
1430 assert_eq!(
1431 transferred(&url, &to).as_deref(),
1432 Some("https://g1t.sh/flagon-io/g1t.git/info/refs?service=git-receive-pack")
1433 );
1434 let url = Url::parse("https://g1t.sh/syntaqx/g1t/git-upload-pack").unwrap();
1435 assert_eq!(
1436 transferred(&url, &to).as_deref(),
1437 Some("https://g1t.sh/flagon-io/g1t/git-upload-pack")
1438 );
1439 }
1440
1441 #[test]
1442 fn an_alias_is_answered_as_its_workspaces_repository() {
1443 for (address, service) in [
1444 ("https://g1t.sh/g1t/g1t.git/info/refs?service=git-upload-pack", GitService::UploadPack),
1445 ("https://g1t.sh/g1t/g1t.git/git-receive-pack", GitService::ReceivePack),
1446 ("https://g1t.sh/g1t/g1t/git-upload-pack", GitService::UploadPack),
1447 ] {
1448 let git = parse(&Url::parse(address).unwrap()).unwrap();
1449 assert_eq!(git.path.namespace, "g1t", "{address}");
1450 let canonical = git.under("flagon-io");
1451 assert_eq!(
1452 canonical.path,
1453 RepoPath {
1454 namespace: "flagon-io".into(),
1455 name: "g1t".into(),
1456 },
1457 "{address}"
1458 );
1459 assert_eq!(canonical.service, service);
1460 assert_eq!(canonical.endpoint, git.endpoint);
1461 }
1462 }
1463
1464 #[test]
1465 fn a_renamed_workspace_keeps_the_rest_of_the_address() {
1466 let url = worker::Url::parse(
1467 "https://g1t.sh/acme/rocket.git/info/refs?service=git-upload-pack",
1468 )
1469 .unwrap();
1470 assert_eq!(
1471 with_namespace(&url, "acme-inc").as_deref(),
1472 Some("https://g1t.sh/acme-inc/rocket.git/info/refs?service=git-upload-pack")
1473 );
1474 let bare = worker::Url::parse("https://g1t.sh/acme").unwrap();
1475 assert_eq!(with_namespace(&bare, "acme-inc"), None);
1476 }
1477
1478 fn pkt(payload: &str) -> Vec<u8> {
1479 format!("{:04x}{payload}", payload.len() + 4).into_bytes()
1480 }
1481
1482 fn joined(parts: &[&[u8]]) -> Vec<u8> {
1483 parts.concat()
1484 }
1485
1486 #[test]
1487 fn a_fetch_with_haves_and_no_done_is_negotiating() {
1488 let request = |lines: &[&str]| {
1489 let mut body = joined(&[&pkt("command=fetch\n"), &pkt("object-format=sha1\n"), b"0001"]);
1490 for line in lines {
1491 body.extend(pkt(&format!("{line}\n")));
1492 }
1493 body.extend(b"0000");
1494 body
1495 };
1496 let want = "want 8407eba58b925619274d012258c2b474a5dbf012";
1497 let have = "have 55cd670a89a80df4fa9d9f0244c44fbd2ed1db8b";
1498 assert!(negotiating(&request(&["deepen 1", want, have])));
1499 assert!(!negotiating(&request(&[want, have, "done"])));
1500 assert!(!negotiating(&request(&[want, "done"])));
1501 let ls_refs = joined(&[&pkt("command=ls-refs\n"), b"0001", &pkt("have nothing\n"), b"0000"]);
1502 assert!(!negotiating(&ls_refs));
1503 }
1504
1505 #[test]
1506 fn acknowledgments_without_ready_end_the_answer() {
1507 // What the store sent a shallow fetch whose only `have` it did not
1508 // know, sideband-all framed: a NAK, then a pack anyway.
1509 let answer = joined(&[
1510 &pkt("\x01acknowledgments\n"),
1511 &pkt("\x01NAK\n"),
1512 b"0001",
1513 &pkt("\x01shallow-info\n"),
1514 &pkt("\x01shallow 8407eba58b925619274d012258c2b474a5dbf012\n"),
1515 b"0001",
1516 &pkt("\x01packfile\n"),
1517 ]);
1518 let cut = joined(&[&pkt("\x01acknowledgments\n"), &pkt("\x01NAK\n")]).len();
1519 assert_eq!(acknowledged(&answer), Acknowledged::CutAt(cut));
1520 // Not yet at the section's end.
1521 assert_eq!(acknowledged(&answer[..cut - 2]), Acknowledged::NeedMore);
1522 assert_eq!(acknowledged(&answer[..cut]), Acknowledged::NeedMore);
1523 // Without sideband framing too.
1524 let plain = joined(&[&pkt("acknowledgments\n"), &pkt("ACK abc\n"), b"0001", &pkt("packfile\n")]);
1525 assert!(matches!(acknowledged(&plain), Acknowledged::CutAt(_)));
1526 }
1527
1528 #[test]
1529 fn a_ready_store_or_a_plain_pack_streams_through() {
1530 let ready = joined(&[
1531 &pkt("\x01acknowledgments\n"),
1532 &pkt("\x01ACK bab14ff1b6d9c4918100098009747d776759a967\n"),
1533 &pkt("\x01ready\n"),
1534 b"0001",
1535 &pkt("\x01packfile\n"),
1536 ]);
1537 assert_eq!(acknowledged(&ready), Acknowledged::Whole);
1538 // Acknowledgments only, ended by a flush: already right.
1539 let only = joined(&[&pkt("acknowledgments\n"), &pkt("NAK\n"), b"0000"]);
1540 assert_eq!(acknowledged(&only), Acknowledged::Whole);
1541 let pack = joined(&[&pkt("\x01packfile\n"), b"0000"]);
1542 assert_eq!(acknowledged(&pack), Acknowledged::Whole);
1543 assert_eq!(acknowledged(b"00"), Acknowledged::NeedMore);
1544 }
1545
1546 #[test]
1547 fn pushed_branches_are_read_from_the_commands() {
1548 let old = "c71546fcd893ef8b0f57388b65e620d759705dda";
1549 let new = "4807077b296e6edbf410d55e72749d3e1170c291";
1550 let body = [
1551 pkt(&format!(
1552 "{old} {new} refs/heads/main\0 report-status side-band-64k\n"
1553 )),
1554 pkt(&format!("{ZERO_ID} {new} refs/heads/feature/x\n")),
1555 pkt(&format!("{old} {ZERO_ID} refs/heads/gone\n")),
1556 pkt(&format!("{ZERO_ID} {new} refs/tags/v1\n")),
1557 b"0000".to_vec(),
1558 b"PACK\0\0\0\x02\0\0\0\0".to_vec(),
1559 ]
1560 .concat();
1561 assert_eq!(
1562 pushed_branches(&body),
1563 [
1564 Pushed {
1565 git_ref: "refs/heads/main".to_owned(),
1566 before: Some(old.to_owned()),
1567 after: new.to_owned()
1568 },
1569 Pushed {
1570 git_ref: "refs/heads/feature/x".to_owned(),
1571 before: None,
1572 after: new.to_owned()
1573 },
1574 Pushed {
1575 git_ref: "refs/tags/v1".to_owned(),
1576 before: None,
1577 after: new.to_owned()
1578 },
1579 ]
1580 );
1581 assert_eq!(deleted_refs(&body), [("refs/heads/gone".to_owned(), old.to_owned())]);
1582 let tag = [pkt(&format!("{old} {ZERO_ID} refs/tags/v0\0 report-status\n")), pkt(&format!("{old} {ZERO_ID} refs/notes/x\n")), b"0000".to_vec()].concat();
1583 assert_eq!(deleted_refs(&tag), [("refs/tags/v0".to_owned(), old.to_owned())]);
1584 }
1585
1586 #[test]
1587 fn a_push_to_a_protected_branch_is_declined_with_the_reason() {
1588 let old = "c71546fcd893ef8b0f57388b65e620d759705dda";
1589 let new = "4807077b296e6edbf410d55e72749d3e1170c291";
1590 let body = [
1591 pkt(&format!("{old} {new} refs/heads/main\0 report-status\n")),
1592 pkt(&format!("{ZERO_ID} {new} refs/heads/feature\n")),
1593 b"0000".to_vec(),
1594 ]
1595 .concat();
1596 let report = String::from_utf8(refusal(&body, "main").unwrap()).unwrap();
1597 assert!(report.starts_with("000eunpack ok\n"));
1598 assert!(report.contains("ng refs/heads/main main is protected"));
1599 assert!(report.contains("ng refs/heads/feature not pushed"));
1600 assert!(report.ends_with("0000"));
1601 }
1602
1603 #[test]
1604 fn the_report_is_framed_for_a_client_that_asked_for_side_band() {
1605 let old = "c71546fcd893ef8b0f57388b65e620d759705dda";
1606 let body = [
1607 pkt(&format!(
1608 "{old} {ZERO_ID} refs/heads/main\0 report-status side-band-64k\n"
1609 )),
1610 b"0000".to_vec(),
1611 ]
1612 .concat();
1613 let report = refusal(&body, "main").unwrap();
1614 // A length, then channel 1, then the report itself.
1615 assert_eq!(report[4], 1);
1616 assert_eq!(&report[5..18], b"000eunpack ok");
1617 assert!(report.ends_with(b"00000000"));
1618 }
1619
1620 #[test]
1621 fn other_branches_and_a_first_push_are_let_through() {
1622 let old = "c71546fcd893ef8b0f57388b65e620d759705dda";
1623 let new = "4807077b296e6edbf410d55e72749d3e1170c291";
1624 let feature = [
1625 pkt(&format!("{old} {new} refs/heads/feature\0 report-status\n")),
1626 b"0000".to_vec(),
1627 ]
1628 .concat();
1629 assert!(refusal(&feature, "main").is_none());
1630 // An empty repository has to be able to receive its first commits.
1631 let first = [
1632 pkt(&format!(
1633 "{ZERO_ID} {new} refs/heads/main\0 report-status\n"
1634 )),
1635 b"0000".to_vec(),
1636 ]
1637 .concat();
1638 assert!(refusal(&first, "main").is_none());
1639 }
1640
1641 #[test]
1642 fn a_blocked_push_explains_itself_on_the_progress_channel() {
1643 let report = b"000eunpack ok\n0000".to_vec();
1644 let messages = vec!["g1t found a secret in this push, so nothing was pushed.".to_owned()];
1645 let body = framed(report.clone(), "report-status side-band-64k", &messages);
1646 // Channel 2 first, which git prints as `remote:` lines.
1647 assert_eq!(body[4], 2);
1648 assert!(String::from_utf8_lossy(&body).contains("so nothing was pushed.\n"));
1649 let at = body.windows(5).position(|w| w == b"000eu").unwrap();
1650 assert_eq!(body[at - 1], 1);
1651 assert!(body.ends_with(b"0000"));
1652 // A client without side-band gets the bare report.
1653 assert_eq!(framed(report.clone(), "report-status", &messages), report);
1654 }
1655
1656 #[test]
1657 fn a_fetch_request_names_no_branches() {
1658 assert!(
1659 pushed_branches(b"0032want c71546fcd893ef8b0f57388b65e620d759705dda\n0000").is_empty()
1660 );
1661 }
1662}