Sign in

flagon-io/g1t

public

Where people and agents ship software together. The open-source git platform for the whole job: issues, agents, checks and deploys to the edge.

gitagentscloudflarerustforgemcp

g1t/services/repos/src/git_http.rs

1,190 lines46,099 bytesCodeBlame
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::meters;
16use crate::pack_limits::{PackSizer, Violation};
17use crate::resilience::{self, Busy, Failure};
18
19const ENDPOINTS: [&str; 3] = ["info/refs", "git-upload-pack", "git-receive-pack"];
20const FORWARDED_HEADERS: [&str; 5] = [
21 "accept",
22 "content-encoding",
23 "content-type",
24 "git-protocol",
25 "user-agent",
26];
27
28/// A git request, parsed from its URL.
29pub struct GitRequest {
30 pub path: RepoPath,
31 pub endpoint: &'static str,
32 pub service: GitService,
33}
34
35/// Parses `/<namespace>/<name>[.git]/<endpoint>`, or returns `None` if the
36/// request is not git's.
37pub fn parse(url: &Url) -> Option<GitRequest> {
38 let path = url.path().strip_prefix('/')?;
39 let endpoint = ENDPOINTS
40 .into_iter()
41 .find(|endpoint| path.ends_with(&format!("/{endpoint}")))?;
42 let repo = &path[..path.len() - endpoint.len() - 1];
43 let (namespace, name) = repo.split_once('/')?;
44 let name = name.strip_suffix(".git").unwrap_or(name);
45 if namespace.is_empty() || name.is_empty() || name.contains('/') {
46 return None;
47 }
48 let service = if endpoint == "info/refs" {
49 url.query_pairs()
50 .find(|(key, _)| key == "service")
51 .map(|(_, value)| value.into_owned())?
52 } else {
53 endpoint.to_owned()
54 };
55 let service = match service.as_str() {
56 "git-upload-pack" => GitService::UploadPack,
57 "git-receive-pack" => GitService::ReceivePack,
58 _ => return None,
59 };
60 Some(GitRequest {
61 path: RepoPath {
62 namespace: namespace.to_owned(),
63 name: name.to_owned(),
64 },
65 endpoint,
66 service,
67 })
68}
69
70/// How long each step of a git request took, sent back to git as a
71/// `Server-Timing` header so that a slow clone shows where its time went.
72/// Step names and whole milliseconds only. The Workers clock moves only
73/// while a request waits on something, so each step is the time spent
74/// waiting on the database, another service or the git store. A note
75/// says how a step went without a duration: `refs;desc=hit-colo`.
76pub struct Timing {
77 started: u64,
78 last: u64,
79 steps: Vec<(&'static str, u64)>,
80 notes: Vec<(&'static str, &'static str)>,
81}
82
83impl Timing {
84 pub fn start() -> Self {
85 let now = g1t_kit::now_ms();
86 Self {
87 started: now,
88 last: now,
89 steps: Vec::new(),
90 notes: Vec::new(),
91 }
92 }
93
94 /// Says how `name` went: where an answer or a credential came from.
95 pub fn note(&mut self, name: &'static str, description: &'static str) {
96 self.notes.push((name, description));
97 }
98
99 /// Ends a step, named `step`, that began when the last one ended.
100 pub fn mark(&mut self, step: &'static str) {
101 let now = g1t_kit::now_ms();
102 self.steps.push((step, now.saturating_sub(self.last)));
103 self.last = now;
104 }
105
106 /// `response`, with how long each step took.
107 pub fn apply(&self, response: Response) -> Result<Response> {
108 let total = g1t_kit::now_ms().saturating_sub(self.started);
109 let headers = response.headers().clone();
110 headers.set("server-timing", &server_timing(&self.steps, &self.notes, total))?;
111 Ok(response.with_headers(headers))
112 }
113}
114
115/// A `Server-Timing` value: each step with its duration, the notes, then
116/// the total.
117fn server_timing(steps: &[(&str, u64)], notes: &[(&str, &str)], total: u64) -> String {
118 steps
119 .iter()
120 .map(|(step, ms)| format!("{step};dur={ms}"))
121 .chain(notes.iter().map(|(name, description)| format!("{name};desc={description}")))
122 .chain(std::iter::once(format!("total;dur={total}")))
123 .collect::<Vec<_>>()
124 .join(", ")
125}
126
127/// The user named by an HTTP Basic `Authorization` header, as git sends it.
128pub async fn viewer(request: &Request, identity: &Fetcher) -> Result<Viewer> {
129 let Some(header) = request.headers().get("authorization")? else {
130 return Ok(None);
131 };
132 let Some((scheme, encoded)) = header.split_once(' ') else {
133 return Ok(None);
134 };
135 if !scheme.eq_ignore_ascii_case("basic") {
136 return Ok(None);
137 }
138 let Some(decoded) = decode_base64(encoded.trim()) else {
139 return Ok(None);
140 };
141 let Some((username, secret)) = decoded.split_once(':') else {
142 return Ok(None);
143 };
144 g1t_kit::call(
145 identity,
146 "user_for_git_credentials",
147 &GitCredentialsArgs {
148 username: username.to_owned(),
149 secret: secret.to_owned(),
150 },
151 )
152 .await
153}
154
155/// Standard base64 to a UTF-8 string, or `None` if either step fails.
156fn decode_base64(input: &str) -> Option<String> {
157 let mut bytes = Vec::with_capacity(input.len() * 3 / 4);
158 let mut buffer = 0u32;
159 let mut bits = 0;
160 for byte in input.bytes().filter(|byte| *byte != b'=') {
161 let value = match byte {
162 b'A'..=b'Z' => byte - b'A',
163 b'a'..=b'z' => byte - b'a' + 26,
164 b'0'..=b'9' => byte - b'0' + 52,
165 b'+' => 62,
166 b'/' => 63,
167 _ => return None,
168 };
169 buffer = (buffer << 6) | u32::from(value);
170 bits += 6;
171 if bits >= 8 {
172 bits -= 8;
173 bytes.push((buffer >> bits) as u8);
174 }
175 }
176 String::from_utf8(bytes).ok()
177}
178
179/// The response for a refused git request. Anonymous callers are asked to
180/// authenticate, which is what makes git prompt for credentials.
181pub fn refuse<T>(outcome: Outcome<T>) -> Result<Response> {
182 let Outcome::Fail(failure) = outcome else {
183 return Response::error("Not found", 404);
184 };
185 let mut response = Response::error(failure.message, failure.code.http_status())?;
186 if failure.code == FailureCode::Unauthenticated {
187 response
188 .headers_mut()
189 .set("www-authenticate", "Basic realm=\"g1t\"")?;
190 }
191 Ok(response)
192}
193
194/// `url` with its first path segment, the workspace, replaced by `slug`.
195pub fn with_namespace(url: &Url, slug: &str) -> Option<String> {
196 let rest = url.path().strip_prefix('/')?.split_once('/')?.1;
197 let mut moved = url.clone();
198 moved.set_path(&format!("/{slug}/{rest}"));
199 Some(moved.to_string())
200}
201
202/// Where a git request for a renamed workspace's old address should go
203/// now, if its first segment is an old slug that still redirects.
204pub async fn renamed(url: &Url, identity: &Fetcher) -> Result<Option<String>> {
205 let Some(old) = url.path().strip_prefix('/').and_then(|path| path.split('/').next()) else {
206 return Ok(None);
207 };
208 let current: Option<String> = g1t_kit::call(
209 identity,
210 "resolve_slug",
211 &g1t_contracts::identity::SlugArgs {
212 slug: old.to_owned(),
213 },
214 )
215 .await?;
216 Ok(current.and_then(|slug| with_namespace(url, &slug)))
217}
218
219/// `url` with its repository, the first two path segments, replaced by
220/// `to`: where a request for a transferred repository's old path goes.
221/// Keeps whether the old address ended in `.git`.
222pub fn transferred(url: &Url, to: &RepoPath) -> Option<String> {
223 let path = url.path().strip_prefix('/')?;
224 let mut segments = path.splitn(3, '/');
225 let (_, name, rest) = (segments.next()?, segments.next()?, segments.next()?);
226 let suffix = if name.ends_with(".git") { ".git" } else { "" };
227 let mut moved = url.clone();
228 moved.set_path(&format!("/{}/{}{suffix}/{rest}", to.namespace, to.name));
229 Some(moved.to_string())
230}
231
232/// A permanent redirect: 301 for git's first request for refs, which it
233/// follows and then uses the new address for the rest; 308 for the
234/// others, so a POST stays a POST.
235pub fn moved(location: &str, get: bool) -> Result<Response> {
236 let mut response = Response::empty()?.with_status(if get { 301 } else { 308 });
237 response.headers_mut().set("location", location)?;
238 Ok(response)
239}
240
241const ZERO_ID: &str = "0000000000000000000000000000000000000000";
242const HEADS: &str = "refs/heads/";
243const TAGS: &str = "refs/tags/";
244
245/// One ref a push asks to change.
246struct Command {
247 old: String,
248 new: String,
249 name: String,
250}
251
252/// The commands at the start of a receive-pack request, and the
253/// capabilities the client sent with the first of them.
254fn commands(body: &[u8]) -> (Vec<Command>, String) {
255 let mut commands = Vec::new();
256 let mut capabilities = String::new();
257 let mut position = 0;
258 // Commands are pkt-lines; a flush packet ends them and the pack follows.
259 while let Some(length) = body
260 .get(position..position + 4)
261 .and_then(|hex| std::str::from_utf8(hex).ok())
262 .and_then(|hex| usize::from_str_radix(hex, 16).ok())
263 {
264 if length < 4 || position + length > body.len() {
265 break;
266 }
267 let line = &body[position + 4..position + length];
268 position += length;
269 // `<old> <new> <ref>`, and on the first command a NUL then capabilities.
270 let mut halves = line.splitn(2, |byte| *byte == 0);
271 let command = halves.next().unwrap_or_default();
272 if let Some(rest) = halves.next() {
273 capabilities = String::from_utf8_lossy(rest).trim().to_owned();
274 }
275 let Ok(command) = std::str::from_utf8(command) else {
276 continue;
277 };
278 let mut parts = command.trim_end().splitn(3, ' ');
279 if let (Some(old), Some(new), Some(name)) = (parts.next(), parts.next(), parts.next()) {
280 commands.push(Command {
281 old: old.to_owned(),
282 new: new.to_owned(),
283 name: name.to_owned(),
284 });
285 }
286 }
287 (commands, capabilities)
288}
289
290/// The bytes of the pack a receive-pack request carries: everything after
291/// the flush packet that ends its commands. Zero for a push that only
292/// deletes refs.
293pub(crate) fn pack_bytes(body: &[u8]) -> u64 {
294 let mut position = 0;
295 while let Some(length) = body
296 .get(position..position + 4)
297 .and_then(|hex| std::str::from_utf8(hex).ok())
298 .and_then(|hex| usize::from_str_radix(hex, 16).ok())
299 {
300 if length == 0 {
301 return (body.len() - position - 4) as u64;
302 }
303 if length < 4 || position + length > body.len() {
304 break;
305 }
306 position += length;
307 }
308 0
309}
310
311fn pkt_line(payload: &[u8]) -> Vec<u8> {
312 let mut line = format!("{:04x}", payload.len() + 4).into_bytes();
313 line.extend_from_slice(payload);
314 line
315}
316
317/// What git is told when a push would change a protected branch: every ref
318/// in it is declined, with the reason against the protected one, so that
319/// git prints it beside the branch. `None` if the push leaves the branch
320/// alone, or creates it in a repository that does not have it yet.
321fn refusal(body: &[u8], protected: &str) -> Option<Vec<u8>> {
322 let (commands, capabilities) = commands(body);
323 let reference = format!("{HEADS}{protected}");
324 if !commands
325 .iter()
326 .any(|command| command.name == reference && command.old != ZERO_ID)
327 {
328 return None;
329 }
330 let mut report = pkt_line(b"unpack ok\n");
331 for command in &commands {
332 let reason = if command.name == reference {
333 format!("{protected} is protected: push a branch and open a pull request")
334 } else {
335 format!("not pushed, because the same push would change {protected}")
336 };
337 report.extend(pkt_line(
338 format!("ng {} {reason}\n", command.name).as_bytes(),
339 ));
340 }
341 report.extend_from_slice(b"0000");
342 Some(framed(report, &capabilities, &[]))
343}
344
345/// A report-status as git expects it: inside channel 1 when the client
346/// asked for side-band, after `messages` on channel 2, which git prints as
347/// `remote:` lines. Without side-band the messages cannot be shown.
348fn framed(report: Vec<u8>, capabilities: &str, messages: &[String]) -> Vec<u8> {
349 let sideband = capabilities
350 .split(' ')
351 .any(|capability| capability.starts_with("side-band"));
352 if !sideband {
353 return report;
354 }
355 let mut body = Vec::new();
356 for message in messages {
357 let mut packet = vec![2u8];
358 packet.extend_from_slice(message.as_bytes());
359 packet.push(b'\n');
360 body.extend(pkt_line(&packet));
361 }
362 // side-band (not -64k) packets carry at most 1000 bytes.
363 for chunk in report.chunks(990) {
364 let mut packet = vec![1u8];
365 packet.extend_from_slice(chunk);
366 body.extend(pkt_line(&packet));
367 }
368 body.extend_from_slice(b"0000");
369 body
370}
371
372/// Declines every ref in a push with `reason`, explaining why in
373/// `messages`: what push protection answers when a push adds a secret.
374pub fn declined(body: &[u8], reason: &str, messages: &[String]) -> Result<Response> {
375 let (commands, capabilities) = commands(body);
376 let mut report = pkt_line(b"unpack ok\n");
377 for command in &commands {
378 report.extend(pkt_line(format!("ng {} {reason}\n", command.name).as_bytes()));
379 }
380 report.extend_from_slice(b"0000");
381 let headers = Headers::new();
382 headers.set("content-type", "application/x-git-receive-pack-result")?;
383 headers.set("cache-control", "no-cache")?;
384 Ok(Response::from_bytes(framed(report, &capabilities, messages))?.with_headers(headers))
385}
386
387/// A branch or tag a push asks to move.
388#[derive(Debug, PartialEq, Eq)]
389pub struct Pushed {
390 /// The full ref: `refs/heads/main`, `refs/tags/v1`.
391 pub git_ref: String,
392 /// Where it pointed before; `None` for a new ref.
393 pub before: Option<String>,
394 pub after: String,
395}
396
397impl Pushed {
398 pub fn branch(&self) -> Option<&str> {
399 self.git_ref.strip_prefix(HEADS)
400 }
401}
402
403/// The branches and tags a push asks to move, read from the commands at the
404/// start of a receive-pack request. Deletions and other refs are left out.
405fn pushed_branches(body: &[u8]) -> Vec<Pushed> {
406 commands(body)
407 .0
408 .into_iter()
409 .filter(|command| command.new != ZERO_ID)
410 .filter(|command| command.name.starts_with(HEADS) || command.name.starts_with(TAGS))
411 .map(|Command { old, new, name }| Pushed {
412 git_ref: name,
413 before: (old != ZERO_ID).then_some(old),
414 after: new,
415 })
416 .collect()
417}
418
419/// The git store's answer, and what the request asked it to change.
420pub struct Forwarded {
421 pub response: Response,
422 /// For a push: the branches and tags it asks to move, and the commits
423 /// to move them to. Whether each moved is for the caller to confirm.
424 pub pushed: Vec<Pushed>,
425 /// For a push: the size of the pack it sent, for the storage meter.
426 pub pack_bytes: u64,
427 /// The bytes sent to the store.
428 pub sent: u64,
429 /// Whether the answer is the store's own (not g1t's, for a store that
430 /// was busy).
431 pub from_store: bool,
432 /// For a push: whether it was too large to scan for secrets first and
433 /// was streamed to the store unscanned (`LargePushes::Unscanned`). Its
434 /// `git.push` events say so, and security scans it after it lands.
435 pub unscanned: bool,
436}
437
438/// What became of a git request.
439pub enum Push {
440 Forwarded(Forwarded),
441 /// A push to a protected branch, answered here without reaching the store.
442 Refused(Response),
443 /// A push that adds a secret nobody allowed, answered the same way.
444 Blocked(Response),
445 /// A push the store could not hold: an object or the repository too
446 /// large, or too large to check. With the reason, for the audit log.
447 Declined(Response, String),
448}
449
450/// What a push may bring, checked as it arrives (pack_limits.rs).
451#[derive(Clone, Copy, Debug)]
452pub struct PushLimits {
453 /// The largest object the store holds.
454 pub max_object: u64,
455 /// What the repository holds now, as g1t counts it.
456 pub held: u64,
457 /// The most a repository may hold.
458 pub repo_limit: u64,
459 /// The largest push that is read whole and scanned for secrets.
460 pub scan_cap: usize,
461 /// What happens to a larger one.
462 pub large: LargePushes,
463}
464
465impl Default for PushLimits {
466 fn default() -> Self {
467 PushLimits {
468 max_object: crate::pack_limits::MAX_OBJECT_BYTES,
469 held: 0,
470 repo_limit: crate::pack_limits::DEFAULT_REPO_LIMIT_BYTES,
471 scan_cap: crate::secret_scan::MAX_SCANNED_PUSH,
472 large: LargePushes::Refuse,
473 }
474 }
475}
476
477/// What happens to a push larger than [`PushLimits::scan_cap`]: set by
478/// `LARGE_PUSHES`.
479#[derive(Clone, Copy, Debug, PartialEq, Eq)]
480pub enum LargePushes {
481 /// Declined (the default): push protection cannot read it, so it does
482 /// not let it in.
483 Refuse,
484 /// Streamed to the store without a scan for secrets; the size limits are
485 /// still checked as it passes.
486 Unscanned,
487}
488
489impl LargePushes {
490 pub fn from_var(value: Option<&str>) -> Self {
491 match value.map(str::trim) {
492 Some("unscanned") => LargePushes::Unscanned,
493 _ => LargePushes::Refuse,
494 }
495 }
496}
497
498/// A push the store could not hold.
499#[derive(Clone, Debug, PartialEq, Eq)]
500pub enum SizeViolation {
501 Object { size: u64 },
502 Repository { held: u64, incoming: u64, limit: u64 },
503 Unscannable { size: u64, cap: usize },
504}
505
506/// Why a push is declined for its size, as git shows it: the `ng` reason,
507/// and the lines printed as `remote:`.
508pub fn size_refusal(violation: &SizeViolation) -> (String, Vec<String>) {
509 use crate::pack_limits::{MAX_OBJECT_BYTES, PLATFORM_BODY_LIMIT_BYTES, megabytes};
510 match violation {
511 SizeViolation::Object { size } => (
512 format!("a file of {} is over the {} limit", megabytes(*size), megabytes(MAX_OBJECT_BYTES)),
513 vec![
514 format!("g1t stores files of up to {} each; this push has one of {}.", megabytes(MAX_OBJECT_BYTES), megabytes(*size)),
515 "Take it out of the commits (git rm --cached, then amend or rebase), and keep large".to_owned(),
516 "files elsewhere: https://docs.g1t.sh/guides/git/#size-limits. Nothing was pushed.".to_owned(),
517 ],
518 ),
519 SizeViolation::Repository { held, incoming, limit } => (
520 "the repository would be over its size limit".to_owned(),
521 vec![
522 format!(
523 "This repository holds about {} and the push adds {}, past the {} a repository may hold.",
524 megabytes(*held),
525 megabytes(*incoming),
526 megabytes(*limit)
527 ),
528 "Delete what you no longer need, or split it: https://docs.g1t.sh/guides/git/#size-limits.".to_owned(),
529 "Nothing was pushed.".to_owned(),
530 ],
531 ),
532 SizeViolation::Unscannable { size, cap } => (
533 "the push is too large to check for secrets".to_owned(),
534 vec![
535 format!(
536 "g1t checks every push for secrets and reads up to {} at once; this one is {}.",
537 megabytes(*cap as u64),
538 megabytes(*size)
539 ),
540 "Push in parts, oldest commits first, then push as usual:".to_owned(),
541 " git rev-list --reverse HEAD | awk 'NR % 500 == 0' | xargs -I{} git push origin {}:refs/heads/main".to_owned(),
542 format!("A push over {} is refused by the network before it reaches g1t (HTTP 413).", megabytes(PLATFORM_BODY_LIMIT_BYTES)),
543 "See https://docs.g1t.sh/guides/git/#size-limits. Nothing was pushed.".to_owned(),
544 ],
545 ),
546 }
547}
548
549/// Feeds the next chunk of a push to the size check; the first violation.
550fn check_size(sizer: &mut Option<PackSizer>, chunk: &[u8], limits: &PushLimits) -> Option<SizeViolation> {
551 let walker = sizer.as_mut()?;
552 match walker.feed(chunk) {
553 Ok(()) => {}
554 Err(Violation::ObjectTooLarge { size }) => return Some(SizeViolation::Object { size }),
555 Err(Violation::Malformed(why)) => {
556 // Not for g1t to judge: the store will say.
557 worker::console_error!("push not checked for size: {why}");
558 *sizer = None;
559 return None;
560 }
561 }
562 let incoming = walker.pack_bytes();
563 crate::pack_limits::over_repo_limit(limits.held, incoming, limits.repo_limit).then_some(SizeViolation::Repository {
564 held: limits.held,
565 incoming,
566 limit: limits.repo_limit,
567 })
568}
569
570/// A request body that streams from `stream`, for `fetch`.
571pub(crate) fn stream_body<S>(stream: S) -> Result<JsValue>
572where
573 S: futures_util::TryStream + 'static,
574 S::Ok: Into<Vec<u8>>,
575 S::Error: Into<worker::Error>,
576{
577 let response: worker::web_sys::Response = Response::from_stream(stream)?.into();
578 Ok(response.body().map_or(JsValue::NULL, Into::into))
579}
580
581/// What git is told when the store is busy: 429 or 503, with when to try
582/// again (resilience.rs).
583pub fn busy_response(busy: Busy) -> Result<Response> {
584 let response = Response::error(busy.message(), busy.status())?;
585 response.headers().set("retry-after", &busy.retry_after.to_string())?;
586 Ok(response)
587}
588
589/// The rest of a request's body, read and thrown away so that git hears
590/// the answer; how many bytes it was.
591async fn drain(stream: &mut worker::ByteStream) -> Result<u64> {
592 let mut size = 0;
593 while let Some(chunk) = stream.next().await {
594 size += chunk?.len() as u64;
595 }
596 Ok(size)
597}
598
599/// One packet of a pkt-line stream: data, or a flush (`0000`), delimiter
600/// (`0001`) or response-end (`0002`) packet, kept as its four bytes.
601#[derive(Debug, PartialEq, Eq)]
602enum Packet {
603 Data(Vec<u8>),
604 Special([u8; 4]),
605}
606
607/// The packets in `bytes`, or `None` if it is not a whole pkt-line stream.
608fn packets(bytes: &[u8]) -> Option<Vec<Packet>> {
609 let mut out = Vec::new();
610 let mut position = 0;
611 while position < bytes.len() {
612 let header = bytes.get(position..position + 4)?;
613 let length = usize::from_str_radix(std::str::from_utf8(header).ok()?, 16).ok()?;
614 if length < 4 {
615 out.push(Packet::Special(header.try_into().ok()?));
616 position += 4;
617 continue;
618 }
619 out.push(Packet::Data(bytes.get(position + 4..position + length)?.to_vec()));
620 position += length;
621 }
622 Some(out)
623}
624
625fn encode(packets: &[Packet]) -> Vec<u8> {
626 let mut out = Vec::new();
627 for packet in packets {
628 match packet {
629 Packet::Data(data) => out.extend(pkt_line(data)),
630 Packet::Special(bytes) => out.extend_from_slice(bytes),
631 }
632 }
633 out
634}
635
636/// A ref advertisement (`info/refs` for upload-pack) or a protocol v2
637/// `ls-refs` answer with `HEAD` pointing at `branch`, the repository's
638/// default branch as g1t keeps it, so a clone checks it out. The git store
639/// holds the HEAD it was created with; g1t can change the default branch
640/// since. `None` when there is nothing to change: no `HEAD` line, `HEAD`
641/// already names `branch`, or `branch` is not advertised.
642pub fn with_head(body: &[u8], branch: &str) -> Option<Vec<u8>> {
643 let mut packets = packets(body)?;
644 let target = format!("{HEADS}{branch}");
645 let oid = packets.iter().find_map(|packet| {
646 let Packet::Data(data) = packet else { return None };
647 let line = data.split(|byte| *byte == 0).next()?;
648 let line = std::str::from_utf8(line).ok()?.trim_end();
649 let (oid, name) = line.split_once(' ')?;
650 // v2 lines may carry attributes after the name.
651 let name = name.split(' ').next()?;
652 (name == target).then(|| oid.to_owned())
653 })?;
654 let mut changed = false;
655 for packet in &mut packets {
656 let Packet::Data(data) = packet else { continue };
657 let text = String::from_utf8_lossy(data).into_owned();
658 let Some((_, rest)) = text.split_once(' ') else { continue };
659 if !(rest.starts_with("HEAD\0") || rest.starts_with("HEAD\n") || rest.starts_with("HEAD ") || rest == "HEAD") {
660 continue;
661 }
662 let mut line = format!("{oid} {rest}");
663 // v0: `symref=HEAD:refs/heads/<old>` among the capabilities.
664 // v2: `symref-target:refs/heads/<old>` after the name.
665 for marker in ["symref=HEAD:", "symref-target:"] {
666 if let Some(at) = line.find(marker) {
667 let start = at + marker.len();
668 let end = line[start..]
669 .find([' ', '\n', '\0'])
670 .map_or(line.len(), |offset| start + offset);
671 line.replace_range(start..end, &target);
672 }
673 }
674 changed = line != text;
675 if changed {
676 *data = line.into_bytes();
677 }
678 break;
679 }
680 changed.then(|| encode(&packets))
681}
682
683/// Whether a request to the git store is one whose answer names `HEAD`:
684/// the ref advertisement for a fetch, or a protocol v2 `ls-refs`.
685fn names_head(git: &GitRequest, body: Option<&[u8]>) -> bool {
686 if git.service != GitService::UploadPack {
687 return false;
688 }
689 match body {
690 None => git.endpoint == "info/refs",
691 Some(body) => {
692 git.endpoint == "git-upload-pack"
693 && body.windows(b"command=ls-refs".len()).any(|window| window == b"command=ls-refs")
694 }
695 }
696}
697
698/// Sends the request on to the git store and returns its response as is,
699/// unless it is a push that would change the `protected` branch, one the
700/// store could not hold (`limits`, pack_limits.rs), or one that `scan`
701/// (push protection) answers itself. A fetch's ref listing has its `HEAD`
702/// pointed at `default_branch` (see [`with_head`]). A POST's body is
703/// `read` when the caller has read it already.
704///
705/// A push is read as it arrives: up to `limits.scan_cap` is kept, to be
706/// scanned and sent on whole; past it, the push is declined, or streamed
707/// to the store unscanned (`LargePushes`), never held. Reads the store
708/// fails for a moment (429, 5xx) are tried again with backoff; a push never
709/// is. A store still busy after that is answered 429 or 503 with
710/// `Retry-After`.
711#[allow(clippy::too_many_arguments)]
712pub async fn forward(
713 mut request: Request,
714 read: Option<Vec<u8>>,
715 git: &GitRequest,
716 access: &GitAccess,
717 protected: Option<&str>,
718 default_branch: Option<&str>,
719 limits: PushLimits,
720 scan: impl AsyncFnOnce(&[u8]) -> Result<Option<Response>>,
721) -> Result<Push> {
722 let headers = Headers::new();
723 headers.set("authorization", &format!("Bearer {}", access.token))?;
724 for name in FORWARDED_HEADERS {
725 if let Some(value) = request.headers().get(name)? {
726 headers.set(name, &value)?;
727 }
728 }
729 let query = request.url()?.query().map(|query| format!("?{query}")).unwrap_or_default();
730 let url = format!("{}/{}{query}", access.remote, git.endpoint);
731 let method = request.method();
732 let (namespace, _) = crate::store::locate(&crate::store::key_from_remote(&access.remote).unwrap_or_default());
733
734 if method == Method::Post && git.endpoint == "git-receive-pack" {
735 return push(request, &url, headers, protected, limits, scan, &namespace).await;
736 }
737
738 // A read: the ref advertisement, `ls-refs`, or a fetch of objects.
739 let body = match (&method, read) {
740 (Method::Post, Some(body)) => Some(body),
741 (Method::Post, None) => Some(request.bytes().await?),
742 _ => None,
743 };
744 let lists_head = match &body {
745 None => method == Method::Get && names_head(git, None),
746 Some(body) => names_head(git, Some(body)),
747 };
748 let sent = body.as_ref().map_or(0, |body| body.len() as u64);
749 let mut attempt = 0;
750 let mut response = loop {
751 let mut init = RequestInit::new();
752 init.with_method(method.clone()).with_headers(headers.clone());
753 if let Some(body) = &body {
754 init.with_body(Some(Uint8Array::from(body.as_slice()).into()));
755 }
756 let started = g1t_kit::now_ms();
757 let answered = Fetch::Request(Request::new_with_init(&url, &init)?).send().await;
758 let ms = g1t_kit::now_ms().saturating_sub(started);
759 let failure = match &answered {
760 Ok(response) => resilience::classify_status(response.status_code()),
761 Err(_) => Some(Failure::Transient),
762 };
763 let outcome = match failure {
764 None => meters::Outcome::Ok,
765 Some(Failure::RateLimited) => meters::Outcome::RateLimited,
766 Some(_) => meters::Outcome::Failed,
767 };
768 meters::record_health(&namespace, outcome, ms);
769 match (answered, failure) {
770 (Ok(response), None) => break response,
771 (answered, Some(failure)) if resilience::retry(failure, attempt) => {
772 drop(answered);
773 let wait = resilience::backoff_ms(failure, attempt, worker::js_sys::Math::random());
774 worker::Delay::from(std::time::Duration::from_millis(wait)).await;
775 attempt += 1;
776 }
777 (_, Some(failure)) => {
778 let busy = Busy { rate_limited: failure == Failure::RateLimited, retry_after: 5 };
779 return Ok(Push::Forwarded(Forwarded {
780 response: busy_response(busy)?,
781 pushed: Vec::new(),
782 pack_bytes: 0,
783 sent,
784 from_store: false,
785 unscanned: false,
786 }));
787 }
788 (Err(error), None) => return Err(error),
789 }
790 };
791 if let (true, Some(branch)) = (lists_head, default_branch)
792 && response.status_code() == 200
793 {
794 let headers = response.headers().clone();
795 headers.delete("content-length")?;
796 let body = response.bytes().await?;
797 let body = with_head(&body, branch).unwrap_or(body);
798 response = Response::from_bytes(body)?.with_headers(headers);
799 }
800 Ok(Push::Forwarded(Forwarded {
801 response,
802 pushed: Vec::new(),
803 pack_bytes: 0,
804 sent,
805 from_store: true,
806 unscanned: false,
807 }))
808}
809
810/// A receive-pack request; see [`forward`].
811#[allow(clippy::too_many_arguments)]
812async fn push(
813 mut request: Request,
814 url: &str,
815 headers: Headers,
816 protected: Option<&str>,
817 limits: PushLimits,
818 scan: impl AsyncFnOnce(&[u8]) -> Result<Option<Response>>,
819 namespace: &str,
820) -> Result<Push> {
821 let mut stream = request.stream()?;
822 let mut head: Vec<u8> = Vec::new();
823 let mut sizer = Some(PackSizer::new(limits.max_object));
824 let mut violation = None;
825 let mut ended = false;
826 while head.len() <= limits.scan_cap {
827 match stream.next().await {
828 Some(chunk) => {
829 let chunk = chunk?;
830 if violation.is_none() {
831 violation = check_size(&mut sizer, &chunk, &limits);
832 }
833 head.extend_from_slice(&chunk);
834 }
835 None => {
836 ended = true;
837 break;
838 }
839 }
840 }
841 let report_headers = || -> Result<Headers> {
842 let headers = Headers::new();
843 headers.set("content-type", "application/x-git-receive-pack-result")?;
844 headers.set("cache-control", "no-cache")?;
845 Ok(headers)
846 };
847 if let Some(report) = protected.and_then(|branch| refusal(&head, branch)) {
848 if !ended {
849 drain(&mut stream).await?;
850 }
851 return Ok(Push::Refused(Response::from_bytes(report)?.with_headers(report_headers()?)));
852 }
853 if !ended && limits.large == LargePushes::Refuse && violation.is_none() {
854 let size = head.len() as u64 + drain(&mut stream).await?;
855 violation = Some(SizeViolation::Unscannable { size, cap: limits.scan_cap });
856 ended = true;
857 }
858 if let Some(violation) = violation {
859 if !ended {
860 drain(&mut stream).await?;
861 }
862 let (reason, messages) = size_refusal(&violation);
863 return Ok(Push::Declined(declined(&head, &reason, &messages)?, reason));
864 }
865 let pushed = pushed_branches(&head);
866 let mut init = RequestInit::new();
867 init.with_method(Method::Post).with_headers(headers);
868 let started = g1t_kit::now_ms();
869 let (answered, pack_bytes, sent, unscanned) = if ended {
870 if let Some(response) = scan(&head).await? {
871 return Ok(Push::Blocked(response));
872 }
873 let pack = pack_bytes(&head);
874 let sent = head.len() as u64;
875 init.with_body(Some(Uint8Array::from(head.as_slice()).into()));
876 drop(head);
877 (Fetch::Request(Request::new_with_init(url, &init)?).send().await, pack, sent, false)
878 } else {
879 // Larger than can be scanned, and let through unscanned: streamed,
880 // with the size limits checked as it passes. A violation ends the
881 // stream before the pack does, so the store refuses it whole.
882 worker::console_warn!("a push of more than {} bytes goes to the store unscanned", limits.scan_cap);
883 let commands = head.iter().take(64 * 1024).copied().collect::<Vec<u8>>();
884 let found: Rc<RefCell<Option<SizeViolation>>> = Rc::default();
885 let walked = Rc::new(RefCell::new((sizer, 0u64)));
886 let rest = {
887 let found = found.clone();
888 let walked = walked.clone();
889 stream.map(move |chunk| {
890 let chunk = chunk?;
891 let mut walked = walked.borrow_mut();
892 walked.1 += chunk.len() as u64;
893 if let Some(violation) = check_size(&mut walked.0, &chunk, &limits) {
894 *found.borrow_mut() = Some(violation);
895 return Err(worker::Error::RustError("push over the size limit".into()));
896 }
897 Ok(chunk)
898 })
899 };
900 let first = head.len() as u64;
901 let body = futures_util::stream::once(async move { Ok::<Vec<u8>, worker::Error>(head) }).chain(rest);
902 init.with_body(Some(stream_body(body)?));
903 let answered = Fetch::Request(Request::new_with_init(url, &init)?).send().await;
904 if let Some(violation) = found.borrow_mut().take() {
905 let (reason, messages) = size_refusal(&violation);
906 return Ok(Push::Declined(declined(&commands, &reason, &messages)?, reason));
907 }
908 let walked = walked.borrow();
909 let pack = walked.0.as_ref().map_or_else(|| pack_bytes(&commands), PackSizer::pack_bytes);
910 (answered, pack, first + walked.1, true)
911 };
912 let ms = g1t_kit::now_ms().saturating_sub(started);
913 let failure = match &answered {
914 Ok(response) => resilience::classify_status(response.status_code()),
915 Err(_) => Some(Failure::Transient),
916 };
917 meters::record_health(
918 namespace,
919 match failure {
920 None => meters::Outcome::Ok,
921 Some(Failure::RateLimited) => meters::Outcome::RateLimited,
922 Some(_) => meters::Outcome::Failed,
923 },
924 ms,
925 );
926 // A push is never tried again: the store may have taken it.
927 let response = match (answered, failure) {
928 (Ok(response), None) => response,
929 (Ok(response), Some(Failure::RateLimited)) => {
930 drop(response);
931 busy_response(Busy { rate_limited: true, retry_after: 5 })?
932 }
933 (Ok(response), Some(_)) => response,
934 (Err(error), _) => {
935 worker::console_error!("a push did not reach the store: {error}");
936 busy_response(Busy { rate_limited: false, retry_after: 5 })?
937 }
938 };
939 Ok(Push::Forwarded(Forwarded {
940 response,
941 pushed,
942 pack_bytes,
943 sent,
944 from_store: true,
945 unscanned,
946 }))
947}
948
949#[cfg(test)]
950mod tests {
951 use super::{Pushed, RepoPath, Url, ZERO_ID, framed, pack_bytes, pushed_branches, refusal, server_timing, transferred, with_head, with_namespace};
952
953 #[test]
954 fn server_timing_names_each_step_and_the_total() {
955 assert_eq!(
956 server_timing(&[("repo", 12), ("token", 0), ("store", 140)], &[], 153),
957 "repo;dur=12, token;dur=0, store;dur=140, total;dur=153"
958 );
959 assert_eq!(server_timing(&[], &[], 3), "total;dur=3");
960 assert_eq!(
961 server_timing(&[("repo", 1), ("cache", 2)], &[("refs", "hit-colo")], 4),
962 "repo;dur=1, cache;dur=2, refs;desc=hit-colo, total;dur=4"
963 );
964 }
965
966 #[test]
967 fn a_renamed_repository_redirects_to_its_new_name() {
968 // A rename keeps the old path in the same table as a transfer, so
969 // the old remote is sent to the new name the same way.
970 let to = RepoPath {
971 namespace: "acme".into(),
972 name: "booster".into(),
973 };
974 let url = Url::parse("https://g1t.sh/acme/rocket.git/info/refs?service=git-upload-pack").unwrap();
975 assert_eq!(
976 transferred(&url, &to).as_deref(),
977 Some("https://g1t.sh/acme/booster.git/info/refs?service=git-upload-pack")
978 );
979 }
980
981 #[test]
982 fn head_follows_the_default_branch_in_a_v0_advertisement() {
983 let main = "1111111111111111111111111111111111111111";
984 let trunk = "2222222222222222222222222222222222222222";
985 let body = [
986 pkt("# service=git-upload-pack\n"),
987 b"0000".to_vec(),
988 pkt(&format!("{main} HEAD\0multi_ack symref=HEAD:refs/heads/main agent=git/2\n")),
989 pkt(&format!("{main} refs/heads/main\n")),
990 pkt(&format!("{trunk} refs/heads/trunk\n")),
991 b"0000".to_vec(),
992 ]
993 .concat();
994 let changed = String::from_utf8(with_head(&body, "trunk").unwrap()).unwrap();
995 assert!(changed.contains(&format!("{trunk} HEAD\0multi_ack symref=HEAD:refs/heads/trunk agent=git/2\n")));
996 assert!(changed.contains(&format!("{main} refs/heads/main\n")));
997 assert!(changed.starts_with("001e# service=git-upload-pack\n0000"));
998 // Already right, or a branch it does not have: left alone.
999 assert!(with_head(&body, "main").is_none());
1000 assert!(with_head(&body, "gone").is_none());
1001 }
1002
1003 #[test]
1004 fn head_follows_the_default_branch_in_a_v2_listing() {
1005 let main = "1111111111111111111111111111111111111111";
1006 let trunk = "2222222222222222222222222222222222222222";
1007 let body = [
1008 pkt(&format!("{main} HEAD symref-target:refs/heads/main\n")),
1009 pkt(&format!("{main} refs/heads/main\n")),
1010 pkt(&format!("{trunk} refs/heads/trunk\n")),
1011 b"0000".to_vec(),
1012 ]
1013 .concat();
1014 let changed = String::from_utf8(with_head(&body, "trunk").unwrap()).unwrap();
1015 assert!(changed.starts_with(&String::from_utf8(pkt(&format!("{trunk} HEAD symref-target:refs/heads/trunk\n"))).unwrap()));
1016 assert!(changed.ends_with("0000"));
1017 // Without symrefs asked for, only the commit changes.
1018 let plain = [pkt(&format!("{main} HEAD\n")), pkt(&format!("{trunk} refs/heads/trunk\n")), b"0000".to_vec()].concat();
1019 let changed = String::from_utf8(with_head(&plain, "trunk").unwrap()).unwrap();
1020 assert!(changed.starts_with(&format!("0032{trunk} HEAD\n")));
1021 }
1022
1023 #[test]
1024 fn a_push_is_measured_by_the_pack_after_its_commands() {
1025 let old = "c71546fcd893ef8b0f57388b65e620d759705dda";
1026 let new = "4807077b296e6edbf410d55e72749d3e1170c291";
1027 let pack = b"PACK\0\0\0\x02\0\0\0\0rest-of-pack";
1028 let body = [
1029 pkt(&format!("{old} {new} refs/heads/main\0 report-status\n")),
1030 b"0000".to_vec(),
1031 pack.to_vec(),
1032 ]
1033 .concat();
1034 assert_eq!(pack_bytes(&body), pack.len() as u64);
1035 // Only deletions: no pack.
1036 let body = [pkt(&format!("{old} {ZERO_ID} refs/heads/gone\n")), b"0000".to_vec()].concat();
1037 assert_eq!(pack_bytes(&body), 0);
1038 assert_eq!(pack_bytes(b"garbage"), 0);
1039 }
1040
1041 #[test]
1042 fn a_transferred_repository_keeps_the_rest_of_the_address() {
1043 let to = RepoPath {
1044 namespace: "flagon-io".into(),
1045 name: "g1t".into(),
1046 };
1047 let url = Url::parse("https://g1t.sh/syntaqx/g1t.git/info/refs?service=git-receive-pack").unwrap();
1048 assert_eq!(
1049 transferred(&url, &to).as_deref(),
1050 Some("https://g1t.sh/flagon-io/g1t.git/info/refs?service=git-receive-pack")
1051 );
1052 let url = Url::parse("https://g1t.sh/syntaqx/g1t/git-upload-pack").unwrap();
1053 assert_eq!(
1054 transferred(&url, &to).as_deref(),
1055 Some("https://g1t.sh/flagon-io/g1t/git-upload-pack")
1056 );
1057 }
1058
1059 #[test]
1060 fn a_renamed_workspace_keeps_the_rest_of_the_address() {
1061 let url = worker::Url::parse(
1062 "https://g1t.sh/acme/rocket.git/info/refs?service=git-upload-pack",
1063 )
1064 .unwrap();
1065 assert_eq!(
1066 with_namespace(&url, "acme-inc").as_deref(),
1067 Some("https://g1t.sh/acme-inc/rocket.git/info/refs?service=git-upload-pack")
1068 );
1069 let bare = worker::Url::parse("https://g1t.sh/acme").unwrap();
1070 assert_eq!(with_namespace(&bare, "acme-inc"), None);
1071 }
1072
1073 fn pkt(payload: &str) -> Vec<u8> {
1074 format!("{:04x}{payload}", payload.len() + 4).into_bytes()
1075 }
1076
1077 #[test]
1078 fn pushed_branches_are_read_from_the_commands() {
1079 let old = "c71546fcd893ef8b0f57388b65e620d759705dda";
1080 let new = "4807077b296e6edbf410d55e72749d3e1170c291";
1081 let body = [
1082 pkt(&format!(
1083 "{old} {new} refs/heads/main\0 report-status side-band-64k\n"
1084 )),
1085 pkt(&format!("{ZERO_ID} {new} refs/heads/feature/x\n")),
1086 pkt(&format!("{old} {ZERO_ID} refs/heads/gone\n")),
1087 pkt(&format!("{ZERO_ID} {new} refs/tags/v1\n")),
1088 b"0000".to_vec(),
1089 b"PACK\0\0\0\x02\0\0\0\0".to_vec(),
1090 ]
1091 .concat();
1092 assert_eq!(
1093 pushed_branches(&body),
1094 [
1095 Pushed {
1096 git_ref: "refs/heads/main".to_owned(),
1097 before: Some(old.to_owned()),
1098 after: new.to_owned()
1099 },
1100 Pushed {
1101 git_ref: "refs/heads/feature/x".to_owned(),
1102 before: None,
1103 after: new.to_owned()
1104 },
1105 Pushed {
1106 git_ref: "refs/tags/v1".to_owned(),
1107 before: None,
1108 after: new.to_owned()
1109 },
1110 ]
1111 );
1112 }
1113
1114 #[test]
1115 fn a_push_to_a_protected_branch_is_declined_with_the_reason() {
1116 let old = "c71546fcd893ef8b0f57388b65e620d759705dda";
1117 let new = "4807077b296e6edbf410d55e72749d3e1170c291";
1118 let body = [
1119 pkt(&format!("{old} {new} refs/heads/main\0 report-status\n")),
1120 pkt(&format!("{ZERO_ID} {new} refs/heads/feature\n")),
1121 b"0000".to_vec(),
1122 ]
1123 .concat();
1124 let report = String::from_utf8(refusal(&body, "main").unwrap()).unwrap();
1125 assert!(report.starts_with("000eunpack ok\n"));
1126 assert!(report.contains("ng refs/heads/main main is protected"));
1127 assert!(report.contains("ng refs/heads/feature not pushed"));
1128 assert!(report.ends_with("0000"));
1129 }
1130
1131 #[test]
1132 fn the_report_is_framed_for_a_client_that_asked_for_side_band() {
1133 let old = "c71546fcd893ef8b0f57388b65e620d759705dda";
1134 let body = [
1135 pkt(&format!(
1136 "{old} {ZERO_ID} refs/heads/main\0 report-status side-band-64k\n"
1137 )),
1138 b"0000".to_vec(),
1139 ]
1140 .concat();
1141 let report = refusal(&body, "main").unwrap();
1142 // A length, then channel 1, then the report itself.
1143 assert_eq!(report[4], 1);
1144 assert_eq!(&report[5..18], b"000eunpack ok");
1145 assert!(report.ends_with(b"00000000"));
1146 }
1147
1148 #[test]
1149 fn other_branches_and_a_first_push_are_let_through() {
1150 let old = "c71546fcd893ef8b0f57388b65e620d759705dda";
1151 let new = "4807077b296e6edbf410d55e72749d3e1170c291";
1152 let feature = [
1153 pkt(&format!("{old} {new} refs/heads/feature\0 report-status\n")),
1154 b"0000".to_vec(),
1155 ]
1156 .concat();
1157 assert!(refusal(&feature, "main").is_none());
1158 // An empty repository has to be able to receive its first commits.
1159 let first = [
1160 pkt(&format!(
1161 "{ZERO_ID} {new} refs/heads/main\0 report-status\n"
1162 )),
1163 b"0000".to_vec(),
1164 ]
1165 .concat();
1166 assert!(refusal(&first, "main").is_none());
1167 }
1168
1169 #[test]
1170 fn a_blocked_push_explains_itself_on_the_progress_channel() {
1171 let report = b"000eunpack ok\n0000".to_vec();
1172 let messages = vec!["g1t found a secret in this push, so nothing was pushed.".to_owned()];
1173 let body = framed(report.clone(), "report-status side-band-64k", &messages);
1174 // Channel 2 first, which git prints as `remote:` lines.
1175 assert_eq!(body[4], 2);
1176 assert!(String::from_utf8_lossy(&body).contains("so nothing was pushed.\n"));
1177 let at = body.windows(5).position(|w| w == b"000eu").unwrap();
1178 assert_eq!(body[at - 1], 1);
1179 assert!(body.ends_with(b"0000"));
1180 // A client without side-band gets the bare report.
1181 assert_eq!(framed(report.clone(), "report-status", &messages), report);
1182 }
1183
1184 #[test]
1185 fn a_fetch_request_names_no_branches() {
1186 assert!(
1187 pushed_branches(b"0032want c71546fcd893ef8b0f57388b65e620d759705dda\n0000").is_empty()
1188 );
1189 }
1190}