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

Re: New semantic patches vs. in-flight topics [was: Re: [PATCH 00/19] Bring more repository handles into our code base]

From
Stefan Beller <sbeller@google.com>
Date
Oct 22, 2018, 18:54 UTC
Message-ID
<CAGZ79kboyakR=dARH60ZJmyN=kT+-X0SSDjgz3aswpkUd+bZmg@mail.gmail.com>
In-Reply-To
<20181022173935.GG30222@szeder.dev>
On Mon, Oct 22, 2018 at 10:39 AM SZEDER Gábor <szeder.dev@gmail.com> wrote:
Show 26 quoted lines
>
> On Tue, Oct 16, 2018 at 04:35:31PM -0700, Stefan Beller wrote:
> > the last patch (applying the semantic patches) has been omitted as that
> > would produce a lot of merge conflicts. Without that patch, this merges
> > cleanly to next.
> >
> > As for when to apply the semantic patches, I wondered if we would prefer a dirty merge
> > (created via "make coccicheck && git apply contrib/coccinelle/the_repository.cocci.patch")
> > of the semantic patches, or if we'd actually trickle in the changes over some time?
>
> >     This series takes another approach as it doesn't change the signature of
> >     functions, but introduces new functions that can deal with arbitrary
> >     repositories, keeping the old function signature around using a shallow wrapper.
> >
> >     Additionally each patch adds a semantic patch, that would port from the old to
> >     the new function. These semantic patches are all applied in the very last patch,
> >     but we could omit applying the last patch if it causes too many merge conflicts
> >     and trickl in the semantic patches over time when there are no merge conflicts.
>
> I don't really like how this or the previous RFC patch series deal
> with semantic patches (or how some past patch series dealt with them,
> for that matter), for various reasons:
>
>   - Applying the transformations from several semantic patches in one
>     single patch makes it harder to review it, because we won't know
>     which change came from which semantic patch.

Good point, so to improve the series sb/more-repo-in-api, I could send the application of the semantic patch just after each patch that adds another semantic patching rule? I personally dislike applying the semantic patch in the same patch as where the semantic rule was introduced, as then the mechanical conversions (from the semantic patch) drown out reviewers attention to the manual changes.

>     For comparison, see the patch series adding hasheq()/oideq(),
>     merged in 769af0fd9e (Merge branch 'jk/cocci', 2018-09-17), in
>     particular the four "convert <this> to <that>" patches.

The four patches "convert <this> to <that> only contain the semantic patch as well as its effects, the manual changes are separated out to later patches, which is quite nice.

Show 9 quoted lines
>   - 'make coccicheck' won't run clean (and the static analysis build
>     job on Travis CI will fail) for all commits following adding the
>     new semantic patches but before applying the resulting
>     transformations.
>
>   - These semantic patches interact badly with 'pu' and 'next',
>     because those integration branches can contain topic branches
>     adding new code that should be transformed by these semanic
>     patches.

And I thought of this as a feature. There is no merge conflict and it still compiles, which makes Junios work easy.

Of course it put the load elsewhere. :/

For the sake of a good history, I would think running 'make coccicheck' and applying the resulting patches would be best as part of the (dirty) merge of any topic that proposes new semantic patches, but that would add load to Junio as it would be an extra step during the merge.

One could argue that the step of applying such transformations into the dirty merge is cheaper than resolving merge conflicts that are had when the topic includes the transformation.

Show 10 quoted lines
>     Consequently, 'make coccicheck' won't run clean and the
>     static analysis build job will fail until all those topics reach
>     'master', and the remaining transformations are applied on top.
>
>     This was (and still is!) an issue with the hasheq()/oideq() series
>     as well: that series was added on 2018-08-28, and the static
>     analysis build job is red on 'pu' ever since.  See the follow-up
>     patch e43d2dcce1 (more oideq/hasheq conversions, 2018-10-02), and
>     one more follow-up will be necessary after the builtin stash topic
>     is merged to 'master'.

In my understanding this follow up is a feature, as it helps to avoid merge conflicts with other topics in flight.

>     This makes it harder to review other patch series.
as 'make coccicheck' is an integral part of your review?
Show 77 quoted lines
>   - Is it really necessary to carry these semantic patches in-tree?
>     Let me ellaborate.  There are basically two main use cases for
>     semantic patches:
>
>       - To avoid undesirable code patterns, e.g. we should not use
>         sha1_to_hex(oid.hash) or strbuf_addf(&sb, "fixed string"), but
>         use oid_to_hex(&oid) or strbuf_addstr(&sb, "fixed string")
>         instead.  Note that in these cases we don't remove the
>         functions sha1_to_hex() or strbuf_addf(), because there are
>         good reasons to use them in other scenarios.
>
>         Our semantic patches under 'contrib/coccinelle/' fall into
>         this category, and we have 'make coccicheck' and the static
>         analysis build job on Travis CI to catch these undesirable
>         code patterns preferably early, and to prevent them from
>         entering our codebase.
>
>       - To perform one-off code transformations, e.g. to modify a
>         function's name and/or signature and convert all its
>         callsites; see e.g. commits abef9020e3 (sha1_file: convert
>         sha1_object_info* to object_id, 2018-03-12) and b4f5aca40e
>         (sha1_file: convert read_sha1_file to struct object_id,
>         2018-03-12).
>
>         As far as I understand this patch series falls into this
>         category: once the conversion is complete the old functions
>         will be removed.  After that there will be no use for these
>         semanic patches.
>
>         Having said that, it's certainly easier to double-check the
>         resulting transformations when one can apply the semantic
>         patches locally, and doing so is easier when the semantic
>         patches are in tree than when they must be copy-pasted from a
>         commit message.
>
> OK, that was already long.  Now, can we do better?
>
> How about introducing the concept of "pending" semantic patches,
> stored in 'contrib/coccinelle/<name>.pending.cocci' files, modifying
> 'make coccicheck' to skip them, and adding the new 'make
> coccicheck-pending' target to make it convenient to apply them, e.g.
> something like the simple patch at the end.
>
> So the process would go something like this:
>
>   - A new semantic patch should be added as "pending", e.g. to the
>     file 'the_repository.pending.cocci', together with the resulting
>     transformations in the same commit.
>
>     This way neither 'make coccicheck' nor the static analysis build
>     job would complain in the topic branch or in the two integration
>     branches.  And if they do complain, then we would know right away
>     that they complain because of a well-established semantic patch.
>     Yet, anyone interested could run 'make coccicheck-pending' to see
>     where are we heading.
>
>   - The author of the "pending" semanting patch should then keep an
>     eye on already cooking topics: whether any of them contain new
>     code that should be transformed, and how they progress to
>     'master', and sending followup patch(es) with the remaining
>     transformations when applicable.
>
>     Futhermore, the author should also pay attention to any new topics
>     that branch off after the "pending" semantic patch, and whether
>     any of them introduce code to be transformed, warning their
>     authors as necessary.
>
>   - Finally, after all the dust settled, the dev should follow up with
>     a patch to:
>
>       - promote the "penging" patch to '<name>.cocci', if its purpose
>         is to avoid undesirable code patterns in the future, or
>
>       - remove the semantic patch, if it was used in a one-off
>         transformation.
>
> Thoughts?
I like the approach of having separate classes of semantic patches:
(a) the regular "we need to keep checking these" as they address
    undesirable code patterns, which is what we currently have,
    and what 'make coccicheck' would complain about.
(b) The pending patches as you propose. However I would
    argue that we'd not want to include the transformation into
    the same patch as then the patch will have merge conflicts.
    Ideally we'd have an automated process/bot that would apply
    all pending semantic patches onto master and then checks for
    conflicts in HEAD..pu, and only sends off the non-conflicting
    diffs as a topic.
    Then after a couple integration cycles we'd have all pending
    changes in, with no conflicts on Junios side.

So I think we should add a patch like you post, but we would need to discuss the exact approach how to deal with pending patches. Is it the original dev who should push forward on their own pending patches, or does it become a pooled effort?

Thanks for starting this discussion! Stefan

Previous: SZEDER GáborNext: SZEDER Gábor
Message 35 of 46 in “Bring more repository handles into our code base”
  1. 00/19 Bring more repository handles into our code baseStefan Beller, Oct 16, 2018
  2. 01/19 sha1_file: allow read_object to read objects in arbitrary repositoriesStefan Beller, Oct 16, 2018
  3. 02/19 packfile: allow has_packed_and_bad to handle arbitrary repositoriesStefan Beller, Oct 16, 2018
  4. 03/19 object-store: allow read_object_file_extended to read from arbitrary repositoriesStefan Beller, Oct 16, 2018
  5. 04/19 object-store: prepare read_object_file to deal with arbitrary repositoriesStefan Beller, Oct 16, 2018
  6. 05/19 object-store: prepare has_{sha1, object}_file[_with_flags] to handle arbitrary repositoriesStefan Beller, Oct 16, 2018
  7. 06/19 object: parse_object to honor its repository argumentStefan Beller, Oct 16, 2018
  8. 07/19 commit: allow parse_commit* to handle arbitrary repositoriesStefan Beller, Oct 16, 2018
  9. 08/19 commit-reach.c: allow paint_down_to_common to handle arbitrary repositoriesStefan Beller, Oct 16, 2018
  10. 09/19 commit-reach.c: allow merge_bases_many to handle arbitrary repositoriesStefan Beller, Oct 16, 2018
  11. 10/19 commit-reach.c: allow remove_redundant to handle arbitrary repositoriesStefan Beller, Oct 16, 2018
  12. 11/19 commit-reach.c: allow get_merge_bases_many_0 to handle arbitrary repositoriesStefan Beller, Oct 16, 2018
  13. 12/19 commit-reach: prepare get_merge_bases to handle arbitrary repositoriesStefan Beller, Oct 16, 2018
  14. 13/19 commit-reach: prepare in_merge_bases[_many] to handle arbitrary repositoriesStefan Beller, Oct 16, 2018
  15. 14/19 commit: prepare get_commit_buffer to handle arbitrary repositoriesStefan Beller, Oct 16, 2018
  16. 15/19 commit: prepare repo_unuse_commit_buffer to handle arbitrary repositoriesStefan Beller, Oct 16, 2018
  17. 16/19 commit: prepare logmsg_reencode to handle arbitrary repositoriesStefan Beller, Oct 16, 2018
  18. 17/19 pretty: prepare format_commit_message to handle arbitrary repositoriesStefan Beller, Oct 16, 2018
  19. 18/19 submodule: use submodule repos for object lookupStefan Beller, Oct 16, 2018
  20. Jonathan TanOct 19, 2018
  21. SZEDER GáborOct 25, 2018
  22. Derrick StoleeOct 31, 2018
  23. Stefan BellerNov 1, 2018
  24. 19/19 submodule: don't add submodule as odb for pushStefan Beller, Oct 16, 2018
  25. Jonathan TanOct 19, 2018
  26. Derrick StoleeOct 17, 2018
  27. Stefan BellerOct 17, 2018
  28. 0/2 Bring the_repository into cmd_fooStefan Beller, Oct 18, 2018
  29. 1/2 repository: have get_the_repository() to remove the_repository dependencyStefan Beller, Oct 18, 2018
  30. 2/2 builtin/merge-base.c: do not rely on the_repository any moreStefan Beller, Oct 18, 2018
  31. Jonathan TanOct 18, 2018
  32. Stefan BellerOct 18, 2018
  33. Junio C HamanoOct 19, 2018
  34. New semantic patches vs. in-flight topics [was: Re: [PATCH 00/19] Bring more repository handles into our code base]SZEDER Gábor, Oct 22, 2018
  35. Stefan BellerOct 22, 2018
  36. SZEDER GáborOct 25, 2018
  37. Stefan BellerOct 25, 2018
  38. Junio C HamanoOct 22, 2018
  39. Stefan BellerOct 23, 2018
  40. Junio C HamanoOct 23, 2018
  41. Junio C HamanoOct 23, 2018
  42. Carlo ArenasOct 23, 2018
  43. Junio C HamanoOct 23, 2018
  44. Stefan BellerOct 23, 2018
  45. Junio C HamanoOct 24, 2018
  46. Jeff KingOct 25, 2018

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.