git/list[1] front-page[2] threads[3] people[4] search[5] about
 

[PATCH] refs: sync loose refs to disk before committing them

From
Patrick Steinhardt <ps@pks.im>
Date
Nov 4, 2021, 12:38 UTC
Message-ID
<dd65718814011eb93ccc4428f9882e0f025224a6.1636029491.git.ps@pks.im>

When writing loose refs, we first create a lockfile, write the new ref into that lockfile, close it and then rename the lockfile into place such that the actual update is atomic for that single ref. While this works as intended under normal circumstences, at GitLab we infrequently encounter corrupt loose refs in repositories after a machine encountered a hard reset. The corruption is always of the same type: the ref has been committed into place, but it is completely empty.

The root cause of this is likely that we don't sync contents of the lockfile to disk before renaming it into place. As a result, it's not guaranteed that the contents are properly persisted and one may observe weird in-between states on hard resets. Quoting ext4 documentation [1]:

    Many broken applications don't use fsync() when replacing existing
    files via patterns such as fd =
    open("foo.new")/write(fd,..)/close(fd)/ rename("foo.new", "foo"), or
    worse yet, fd = open("foo", O_TRUNC)/write(fd,..)/close(fd). If
    auto_da_alloc is enabled, ext4 will detect the replace-via-rename
    and replace-via-truncate patterns and force that any delayed
    allocation blocks are allocated such that at the next journal
    commit, in the default data=ordered mode, the data blocks of the new
    file are forced to disk before the rename() operation is committed.
    This provides roughly the same level of guarantees as ext3, and
    avoids the "zero-length" problem that can happen when a system
    crashes before the delayed allocation blocks are forced to disk.

This explicitly points out that one must call fsync(3P) before doing the rename(3P) call, or otherwise data may not be correctly persisted to disk.

Fix this by always flushing refs to disk before committing them into place to avoid this class of corruption.

[1]: https://www.kernel.org/doc/Documentation/filesystems/ext4.txt
Signed-off-by: Patrick Steinhardt <ps@pks.im>
---
 refs/files-backend.c | 1 +
 1 file changed, 1 insertion(+)
diff --git a/refs/files-backend.c b/refs/files-backend.c
index 151b0056fe..06a3f0bdea 100644
--- a/refs/files-backend.c
+++ b/refs/files-backend.c
@@ -1749,6 +1749,7 @@ static int write_ref_to_lockfile(struct ref_lock *lock,
 	fd = get_lock_file_fd(&lock->lk);
 	if (write_in_full(fd, oid_to_hex(oid), the_hash_algo->hexsz) < 0 ||
 	    write_in_full(fd, &term, 1) < 0 ||
+	    fsync(fd) < 0 ||
 	    close_ref_gently(lock) < 0) {
 		strbuf_addf(err,
 			    "couldn't write '%s'", get_lock_file_path(&lock->lk));
-- 
2.33.1
Next: Ævar Arnfjörð Bjarmason
Message 1 of 34 in “refs: sync loose refs to disk before committing them”
  1. refs: sync loose refs to disk before committing themPatrick Steinhardt, Nov 4, 2021
  2. Ævar Arnfjörð BjarmasonNov 4, 2021
  3. Patrick SteinhardtNov 4, 2021
  4. Junio C HamanoNov 4, 2021
  5. Neeraj SinghNov 4, 2021
  6. Junio C HamanoNov 5, 2021
  7. Jeff KingNov 5, 2021
  8. Ævar Arnfjörð BjarmasonNov 5, 2021
  9. Jeff KingNov 5, 2021
  10. Jeff KingNov 5, 2021
  11. Jeff KingNov 5, 2021
  12. Johannes SchindelinNov 5, 2021
  13. Patrick SteinhardtNov 5, 2021
  14. Jeff KingNov 5, 2021
  15. Patrick SteinhardtNov 9, 2021
  16. Jeff KingNov 10, 2021
  17. Patrick SteinhardtNov 10, 2021
  18. 0/3 refs: sync loose refs to disk before committing themPatrick Steinhardt, Nov 10, 2021
  19. 1/3 wrapper: handle EINTR in `git_fsync()`Patrick Steinhardt, Nov 10, 2021
  20. Johannes SchindelinNov 10, 2021
  21. Ævar Arnfjörð BjarmasonNov 10, 2021
  22. 2/3 wrapper: provide function to sync directoriesPatrick Steinhardt, Nov 10, 2021
  23. Ævar Arnfjörð BjarmasonNov 10, 2021
  24. 3/3 refs: add configuration to enable flushing of refsPatrick Steinhardt, Nov 10, 2021
  25. Ævar Arnfjörð BjarmasonNov 10, 2021
  26. Neeraj SinghNov 10, 2021
  27. Ævar Arnfjörð BjarmasonNov 10, 2021
  28. Neeraj SinghNov 11, 2021
  29. Patrick SteinhardtNov 11, 2021
  30. Patrick SteinhardtNov 11, 2021
  31. Neeraj SinghNov 11, 2021
  32. Johannes SchindelinNov 10, 2021
  33. Jeff KingNov 10, 2021
  34. Patrick SteinhardtNov 11, 2021

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.