OXYGENHUB

Coverage / crates/flow/src/workspace.rs 77.5%

193 of 249 lines run, 26 of 57 functions · run #177 on main · run missed

1
//! A job's workspace: the files of the state being built, written to a
2
//! scratch directory and packed into an ext4 image for the microVM
3
//! (docs/m0.md §2). Returns each file's content hash for the scope hash.
4
5
use anyhow::{bail, Context, Result};
6
use flow_core::Hash;
7
use oxy_core::hash::ChangeHash;
8
use oxy_core::objects::NodeKind;
9
use oxy_core::repo::Repo;
10
use oxy_core::state::FileContent;
11
use std::path::Path;
12
13
pub type Files = Vec<(String, Hash)>;
14
15
/// Write the state `heads` of `repo` under `dir`.
160×
pub fn export_state(repo: &Repo, heads: &[ChangeHash], dir: &Path) -> Result<Files> {
170×
    let snapshot = repo.load_state(heads).context("load the state")?.materialize();
180×
    let mut files = Vec::new();
190×
    for (path, f) in &snapshot.files {
200×
        if path.split('/').any(|p| p.is_empty() || p == "." || p == ".." || p == ".oxy") {
210×
            continue;
220×
        }
230×
        let target = dir.join(path);
240×
        match (&f.kind, &f.content) {
250×
            (NodeKind::Dir, _) | (_, FileContent::Dir) => std::fs::create_dir_all(&target)?,
260×
            (NodeKind::Subrepo, _) => {}
270×
            (kind, content) => {
280×
                let bytes = match content {
290×
                    FileContent::Text { bytes, .. } => bytes.clone(),
300×
                    FileContent::Opaque { blobs } => match blobs.first() {
310×
                        Some(b) => repo.read_blob(b)?,
320×
                        None => Vec::new(),
33
                    },
340×
                    FileContent::Dir => unreachable!(),
35
                };
360×
                if let Some(parent) = target.parent() {
370×
                    std::fs::create_dir_all(parent)?;
380×
                }
390×
                if *kind == NodeKind::Symlink {
400×
                    let to = String::from_utf8_lossy(&bytes).into_owned();
410×
                    std::os::unix::fs::symlink(&to, &target).with_context(|| format!("symlink {path}"))?;
42
                } else {
430×
                    std::fs::write(&target, &bytes).with_context(|| format!("write {path}"))?;
440×
                    if f.mode.exec {
45
                        use std::os::unix::fs::PermissionsExt;
460×
                        std::fs::set_permissions(&target, std::fs::Permissions::from_mode(0o755))?;
470×
                    }
48
                }
490×
                files.push((path.clone(), *blake3::hash(&bytes).as_bytes()));
50
            }
51
        }
52
    }
530×
    Ok(files)
540×
}
55
56
/// Copy a working directory (without `.oxy` and what `.oxyignore` and
57
/// `.gitignore` exclude) under `dir`, for `flow run`.
5814×
pub fn copy_tree(src: &Path, dir: &Path) -> Result<Files> {
5914×
    let mut files = Vec::new();
6014×
    let walk = ignore::WalkBuilder::new(src)
6114×
        .hidden(false)
6214×
        .git_ignore(true)
6314×
        .require_git(false)
6414×
        .add_custom_ignore_filename(".oxyignore")
65154×
        .filter_entry(|e| e.file_name() != ".oxy" && e.file_name() != ".git")
6614×
        .build();
67168×
    for entry in walk {
68168×
        let entry = entry?;
69168×
        let rel = entry.path().strip_prefix(src)?;
70168×
        if rel.as_os_str().is_empty() {
7114×
            continue;
72154×
        }
73154×
        let rel_s = rel.to_string_lossy().replace('\\', "/");
74154×
        let target = dir.join(rel);
75154×
        let ft = entry.file_type().context("file type")?;
76154×
        if ft.is_dir() {
7742×
            std::fs::create_dir_all(&target)?;
78112×
        } else if ft.is_symlink() {
790×
            let to = std::fs::read_link(entry.path())?;
800×
            std::os::unix::fs::symlink(&to, &target)?;
810×
            files.push((rel_s, *blake3::hash(to.to_string_lossy().as_bytes()).as_bytes()));
82112×
        } else if ft.is_file() {
83112×
            std::fs::copy(entry.path(), &target)?;
84112×
            files.push((rel_s, *blake3::hash(&std::fs::read(entry.path())?).as_bytes()));
850×
        }
86
    }
8714×
    Ok(files)
8814×
}
89
90
/// Pack `dir` into an ext4 image whose root belongs to this user, so the
91
/// job (which runs as the owner of `/src`) can write to it.
922×
pub fn image(dir: &Path, out: &Path) -> Result<()> {
932×
    let used: u64 = ignore::WalkBuilder::new(dir)
942×
        .hidden(false)
952×
        .ignore(false)
962×
        .git_ignore(false)
972×
        .build()
982×
        .flatten()
9913×
        .filter_map(|e| e.metadata().ok())
10013×
        .map(|m| m.len().max(4096))
1012×
        .sum();
1022×
    let mib = used * 12 / 10 / (1 << 20) + 512;
1032×
    let (uid, gid) = unsafe { (libc::getuid(), libc::getgid()) };
1042×
    let status = std::process::Command::new("mkfs.ext4")
1052×
        .args(["-q", "-F", "-L", "workspace", "-E"])
1062×
        .arg(format!("root_owner={uid}:{gid}"))
1072×
        .arg("-d")
1082×
        .arg(dir)
1092×
        .arg(out)
1102×
        .arg(format!("{mib}M"))
1112×
        .status()
1122×
        .context("run mkfs.ext4 (e2fsprogs 1.43 or later)")?;
1132×
    if !status.success() {
1140×
        bail!("mkfs.ext4 failed for the workspace image");
1152×
    }
1162×
    Ok(())
1172×
}
118
119
/// Give the files of `dir` whose content a cached disk was built from
120
/// the mtime they had then (oxygenflow#52), and describe this checkout's
121
/// sources for the disk to keep: unchanged files are fresh to a build
122
/// tool that judges by time, changed ones newer than anything it built.
123
/// Whole seconds, as the workspace image keeps them. Returns the sources
124
/// and how many files got their time back.
1255×
pub fn restore_times(dir: &Path, files: &Files, before: Option<&crate::cache::Sources>) -> (crate::cache::Sources, usize) {
1265×
    let mut out = crate::cache::Sources::new();
1275×
    let mut restored = 0;
12824×
    for (path, hash) in files {
12924×
        let p = dir.join(path);
13024×
        let Ok(md) = std::fs::symlink_metadata(&p) else { continue };
13124×
        if !md.is_file() {
1325×
            continue;
13319×
        }
13419×
        let hex = hex::encode(hash);
13519×
        let kept = before.and_then(|b| b.get(path)).filter(|(h, _)| *h == hex).map(|(_, t)| *t);
13619×
        let t = match kept {
1379×
            Some(t) => {
1389×
                let at = std::time::UNIX_EPOCH + std::time::Duration::from_secs(t.max(0) as u64);
1399×
                if std::fs::File::options().write(true).open(&p).and_then(|f| f.set_modified(at)).is_err() {
1400×
                    continue;
1419×
                }
1429×
                restored += 1;
1439×
                t
144
            }
14510×
            None => md.modified().ok().and_then(|m| m.duration_since(std::time::UNIX_EPOCH).ok()).map(|d| d.as_secs() as i64).unwrap_or(0),
146
        };
14719×
        out.insert(path.clone(), (hex, t));
148
    }
149
    // Directories too, after their files (#52): a build script's
150
    // `rerun-if-changed` on a directory looks at its time as well. A
151
    // directory whose listing is unchanged gets its old time back; one that
152
    // gained or lost an entry keeps its checkout's, so the change shows.
1535×
    restore_dir_times(dir, Path::new(""), before, &mut out, &mut restored);
1545×
    (out, restored)
1555×
}
156
157
/// The key a directory's entry has in the sources: its path and a slash.
1588×
fn dir_key(rel: &Path) -> String {
1598×
    format!("{}/", rel.display())
1608×
}
161
1628×
fn restore_dir_times(
1638×
    root: &Path,
1648×
    rel: &Path,
1658×
    before: Option<&crate::cache::Sources>,
1668×
    out: &mut crate::cache::Sources,
1678×
    restored: &mut usize,
1688×
) {
1698×
    let here = root.join(rel);
1708×
    let Ok(entries) = std::fs::read_dir(&here) else { return };
1718×
    let mut names: Vec<(String, bool)> = Vec::new();
17227×
    for e in entries.flatten() {
17327×
        let is_dir = e.file_type().is_ok_and(|t| t.is_dir());
17427×
        names.push((e.file_name().to_string_lossy().into_owned(), is_dir));
175
    }
1768×
    names.sort();
17727×
    for (name, is_dir) in &names {
17827×
        if *is_dir {
1793×
            restore_dir_times(root, &rel.join(name), before, out, restored);
18024×
        }
181
    }
1828×
    let mut h = blake3::Hasher::new();
18327×
    for (name, is_dir) in &names {
18427×
        h.update(name.as_bytes());
18527×
        h.update(if *is_dir { b"/\n".as_slice() } else { b"\n".as_slice() });
186
    }
1878×
    let hex = h.finalize().to_hex().to_string();
1888×
    let key = dir_key(rel);
1898×
    let kept = before.and_then(|b| b.get(&key)).filter(|(h, _)| *h == hex).map(|(_, t)| *t);
1908×
    let t = match kept {
1913×
        Some(t) => {
1923×
            let at = std::time::UNIX_EPOCH + std::time::Duration::from_secs(t.max(0) as u64);
1933×
            if std::fs::File::open(&here).and_then(|f| f.set_modified(at)).is_err() {
1940×
                return;
1953×
            }
1963×
            *restored += 1;
1973×
            t
198
        }
1995×
        None => std::fs::metadata(&here)
2005×
            .ok()
2015×
            .and_then(|m| m.modified().ok())
2025×
            .and_then(|m| m.duration_since(std::time::UNIX_EPOCH).ok())
2035×
            .map(|d| d.as_secs() as i64)
2045×
            .unwrap_or(0),
205
    };
2068×
    out.insert(key, (hex, t));
2078×
}
208
209
/// An empty ext4 filesystem of `mib` MiB at `out`, owned by this user like
210
/// the workspace. mkfs.ext4 creates the file sparse, so only what the job
211
/// writes takes space on the runner.
2121×
pub fn empty_ext4(out: &Path, mib: u32, label: &str) -> Result<()> {
2131×
    let (uid, gid) = unsafe { (libc::getuid(), libc::getgid()) };
2141×
    let status = std::process::Command::new("mkfs.ext4")
2151×
        .args(["-q", "-F", "-L", label, "-E"])
2161×
        .arg(format!("root_owner={uid}:{gid}"))
2171×
        .arg(out)
2181×
        .arg(format!("{mib}M"))
2191×
        .status()
2201×
        .context("run mkfs.ext4 (e2fsprogs 1.43 or later)")?;
2211×
    if !status.success() {
2220×
        bail!("mkfs.ext4 failed for a scratch disk");
2231×
    }
2241×
    Ok(())
2251×
}
226
227
/// Remove `dir` and everything in it, even where a job took away its own
228
/// write or search permission on a directory; links are never followed.
22928×
pub fn remove_tree(dir: &Path) -> std::io::Result<()> {
23028×
    match std::fs::remove_dir_all(dir) {
23117×
        Ok(()) => Ok(()),
23211×
        Err(e) if e.kind() == std::io::ErrorKind::NotFound => Ok(()),
233
        Err(_) => {
2340×
            open_up(dir);
2350×
            std::fs::remove_dir_all(dir)
236
        }
237
    }
23828×
}
239
2400×
fn open_up(dir: &Path) {
241
    use std::os::unix::fs::PermissionsExt;
2420×
    let Ok(meta) = std::fs::symlink_metadata(dir) else { return };
2430×
    if !meta.is_dir() {
2440×
        return;
2450×
    }
2460×
    let _ = std::fs::set_permissions(dir, std::fs::Permissions::from_mode(0o700));
2470×
    if let Ok(entries) = std::fs::read_dir(dir) {
2480×
        for e in entries.flatten() {
2490×
            open_up(&e.path());
2500×
        }
2510×
    }
2520×
}
253
254
#[cfg(test)]
255
mod tests {
256
    use super::*;
257
    use std::time::UNIX_EPOCH;
258
2595×
    fn secs(p: &Path) -> i64 {
2605×
        std::fs::metadata(p).unwrap().modified().unwrap().duration_since(UNIX_EPOCH).unwrap().as_secs() as i64
2615×
    }
262
263
    /// oxygenflow#52: an unchanged file gets the time it had when the cached
264
    /// disk was built; a changed or new one keeps its checkout's, newer.
265
    #[test]
2661×
    fn unchanged_sources_get_their_cached_times_back() {
2671×
        let dir = tempfile::tempdir().unwrap();
2683×
        for (name, text) in [("same.rs", "a"), ("changed.rs", "b2"), ("new.rs", "c")] {
2693×
            std::fs::write(dir.path().join(name), text).unwrap();
2703×
        }
2711×
        std::os::unix::fs::symlink("same.rs", dir.path().join("link.rs")).unwrap();
27210×
        let h = |t: &str| *blake3::hash(t.as_bytes()).as_bytes();
2731×
        let files: Files =
2741×
            vec![("same.rs".into(), h("a")), ("changed.rs".into(), h("b2")), ("new.rs".into(), h("c")), ("link.rs".into(), h("same.rs"))];
2751×
        let now = secs(&dir.path().join("same.rs"));
2761×
        let before: crate::cache::Sources =
2771×
            [("same.rs".to_string(), (hex::encode(h("a")), 1_000_000)), ("changed.rs".to_string(), (hex::encode(h("b1")), 1_000_000))]
2781×
                .into();
279
2801×
        let (sources, restored) = restore_times(dir.path(), &files, Some(&before));
2811×
        assert_eq!(restored, 1);
2821×
        assert_eq!(secs(&dir.path().join("same.rs")), 1_000_000, "unchanged: fresh to cargo");
2831×
        assert!(secs(&dir.path().join("changed.rs")) >= now, "changed: newer than the cached build");
2841×
        assert_eq!(sources["same.rs"], (hex::encode(h("a")), 1_000_000));
2851×
        assert_eq!(sources["changed.rs"].0, hex::encode(h("b2")), "kept with its new content");
2861×
        assert!(sources.contains_key("new.rs") && !sources.contains_key("link.rs"), "files only");
287
288
        // Without a cached build nothing moves, and the sources are listed:
289
        // three files and the directory holding them.
2901×
        let (again, restored) = restore_times(dir.path(), &files, None);
2911×
        assert_eq!((restored, again.len()), (0, 4));
2921×
        assert!(again.contains_key("/"));
293
294
        // A directory whose listing is unchanged gets its time back; one
295
        // that gained an entry keeps its new time, so a build script's
296
        // `rerun-if-changed` on it sees the change.
2971×
        std::fs::create_dir(dir.path().join("migrations")).unwrap();
2981×
        std::fs::write(dir.path().join("migrations/0001.sql"), "x").unwrap();
2991×
        let mut files = files;
3001×
        files.push(("migrations/0001.sql".into(), h("x")));
3011×
        let (mut kept, _) = restore_times(dir.path(), &files, None);
3021×
        kept.get_mut("migrations/").unwrap().1 = 2_000_000;
3031×
        kept.get_mut("migrations/0001.sql").unwrap().1 = 2_000_000;
3041×
        let (_, restored) = restore_times(dir.path(), &files, Some(&kept));
3051×
        assert_eq!(secs(&dir.path().join("migrations")), 2_000_000, "unchanged listing: its old time");
3061×
        assert!(restored >= 2);
3071×
        std::fs::write(dir.path().join("migrations/0002.sql"), "y").unwrap();
3081×
        files.push(("migrations/0002.sql".into(), h("y")));
3091×
        let (now_sources, _) = restore_times(dir.path(), &files, Some(&kept));
3101×
        assert!(secs(&dir.path().join("migrations")) > 2_000_000, "a new entry: it looks changed");
3111×
        assert_ne!(now_sources["migrations/"].0, kept["migrations/"].0);
3121×
    }
313
}