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

Re: [PATCH v3 0/6] odb/files: reload alternates on retry again

From
Junio C Hamano <gitster@pobox.com>
Date
Oct 28, 2026, 16:12 UTC
Message-ID
<xmqq8q3hdcsg.fsf@gitster.g>
In-Reply-To
<20261028-pks-odb-files-reload-alternates-v3-0-011ddb64b626@pks.im>

H Patrick Steinhardt <ps@pks.im> writes:

Show 30 quoted lines
> 3:  da5a24624e8 = 3:  8882555124e odb/source-files: reload alternates on retry
> 4:  13e4e857f68 ! 4:  64d76b0ac94 odb/source-files: simplify directory duplication detection
>     @@ Commit message
>              Wall Time    266.5 ms  ±     2.8 ms     262.6 ms  …   276.7 ms            -7.8%
>              Memory         5.9 MiB ±     0.1 MiB      5.6 MiB …     6.1 MiB           +5.1%
>      
>     +    Rename `dirs_paths_icase` to `dir_paths_iscase` to match the naming of
>     +    the newly added set.
>     +
>          Signed-off-by: Patrick Steinhardt <ps@pks.im>
>      
>       ## odb/source-files.c ##
>     @@ odb/source-files.c: static bool odb_files_dir_is_usable(struct odb_source_files
>       	/*
>      -	 * Prevent the common mistake of listing the same
>      -	 * thing twice, or object directory itself.
>     -+	 * Prevent the common mistake of listing the same thing twice, or
>     -+	 * object directory itself.
>     ++	 * Prevent the common mistake of listing the same thing twice,
>     ++	 * or object directory itself.
>     ++	 *
>     ++	 * NEEDSWORK: we're using "core.ignoreCase" to deduplicate
>     ++	 * alternates that _may_ be the same. This requires quite a bit
>     ++	 * of boilerplate for dubious benefit:
>     ++	 *
>     ++	 *   - Duplicating alternates should really only lead to
>     ++	 *     regressed performance.
>     ++	 *
>     ++	 *   - We don't properly resolve symlinks or mointpoints, so we
>     ++	 *     may still end up duplicating alternates.
Good thought process.  "Mountpoint" is misspelled, by the way.
>     ++	 *
>     ++	 *   - The value may be lying, in which case we might
>     ++	 *     deduplicate alternates that are in fact not mapping to the
>     ++	 *     same directory.

This one is much more serious than the previous two. As an optimization to avoid repeating work known to produce no new results, deduplication can tolerate false negatives (e.g., we thought "DIR" and "Dir" were different and ended up scanning the same thing twice). It cannot, however, allow false positives, which would lead to "missing" objects that only the mistakenly deduplicated object store could have provided.

>     ++	 * We should investigate whether we can remove this whole
>     ++	 * mechanism outright.
>       	 */
Sure.

I guess the previous round was already almost ready with Christian's review, and this round addressing his comments, we are pretty much done with the topic and can mark it as ready for 'next'?

Thanks, all.
Previous: Patrick SteinhardtNext: Patrick Steinhardt
Message 37 of 38 in “odb/files: reload alternates on retry again”
  1. 0/4 odb/files: reload alternates on retry againPatrick Steinhardt, Oct 20, 2026
  2. 1/4 odb/source-files: rename alternates-related functionsPatrick Steinhardt, Oct 20, 2026
  3. 2/4 odb/source-files: reload alternates on retryPatrick Steinhardt, Oct 20, 2026
  4. Christian CouderOct 21, 2026
  5. Patrick SteinhardtOct 23, 2026
  6. 3/4 odb/source-files: avoid rechecking already-known alternatesPatrick Steinhardt, Oct 20, 2026
  7. 4/4 odb/source-files: only reload alternates when the file has changedPatrick Steinhardt, Oct 20, 2026
  8. Junio C HamanoOct 20, 2026
  9. Patrick SteinhardtOct 21, 2026
  10. 0/6 odb/files: reload alternates on retry againPatrick Steinhardt, Oct 23, 2026
  11. 1/6 odb/source-files: rename alternates-related functionsPatrick Steinhardt, Oct 23, 2026
  12. 2/6 odb/source-files: always load alternates via primary object directoryPatrick Steinhardt, Oct 23, 2026
  13. Christian CouderOct 27, 2026
  14. 3/6 odb/source-files: reload alternates on retryPatrick Steinhardt, Oct 23, 2026
  15. 4/6 odb/source-files: simplify directory duplication detectionPatrick Steinhardt, Oct 23, 2026
  16. Christian CouderOct 27, 2026
  17. Patrick SteinhardtOct 27, 2026
  18. 5/6 odb/source-files: avoid stat'ing already-known alternatesPatrick Steinhardt, Oct 23, 2026
  19. Christian CouderOct 27, 2026
  20. Patrick SteinhardtOct 27, 2026
  21. Christian CouderOct 27, 2026
  22. Patrick SteinhardtOct 28, 2026
  23. 6/6 odb/source-files: only reload alternates when the file has changedPatrick Steinhardt, Oct 23, 2026
  24. Christian CouderOct 27, 2026
  25. Patrick SteinhardtOct 27, 2026
  26. Christian CouderOct 27, 2026
  27. 0/6 odb/files: reload alternates on retry againPatrick Steinhardt, Oct 28, 2026
  28. 1/6 odb/source-files: rename alternates-related functionsPatrick Steinhardt, Oct 28, 2026
  29. 2/6 odb/source-files: always load alternates via primary object directoryPatrick Steinhardt, Oct 28, 2026
  30. 3/6 odb/source-files: reload alternates on retryPatrick Steinhardt, Oct 28, 2026
  31. 4/6 odb/source-files: simplify directory duplication detectionPatrick Steinhardt, Oct 28, 2026
  32. Christian CouderOct 28, 2026
  33. Patrick SteinhardtOct 28, 2026
  34. Junio C HamanoOct 28, 2026
  35. 5/6 odb/source-files: avoid stat'ing already-known alternatesPatrick Steinhardt, Oct 28, 2026
  36. 6/6 odb/source-files: only reload alternates when the file has changedPatrick Steinhardt, Oct 28, 2026
  37. Junio C HamanoOct 28, 2026
  38. Patrick SteinhardtOct 29, 2026

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.