What's cooking in git.git (May 2024, #02; Fri, 3)
Here are the topics that have been cooking in my tree. Commits prefixed with '+' are in 'next' (being in 'next' is a sign that a topic is stable enough to be used and are candidate to be in a future release). Commits prefixed with '-' are only in 'seen', and aren't considered "accepted" at all and may be annotated with an URL to a message that raises issues but they are no means exhaustive. A topic without enough support may be discarded after a long period of no activity (of course they can be resubmit when new interests arise).
Quite a few new topics have been added to 'seen', simply because I saw them, which does not necessarily mean I agree with them or even read them carefully ;-)
Copies of the source code to Git live in many repositories, and the following is a list of the ones I push into or their mirrors. Some repositories have only a subset of branches.
With maint, master, next, seen, todo:
git://git.kernel.org/pub/scm/git/git.git/ git://repo.or.cz/alt-git.git/ https://kernel.googlesource.com/pub/scm/git/git/ https://github.com/git/git/ https://gitlab.com/git-scm/git/
With all the integration branches and topics broken out:
https://github.com/gitster/git/
Even though the preformatted documentation in HTML and man format are not sources, they are published in these repositories for convenience (replace "htmldocs" with "manpages" for the manual pages):
git://git.kernel.org/pub/scm/git/git-htmldocs.git/
https://github.com/gitster/git-htmldocs.git/
Release tarballs are available at:
https://www.kernel.org/pub/software/scm/git/
--------------------------------------------------
[New Topics]
* jc/test-workaround-broken-mv (2024-05-02) 1 commit
- t/lib-chunk: work around broken "mv" on some vintage of macOS
Tests that try to corrupt in-repository files in chunked format did not work well on macOS due to its broken "mv", which has been worked around.
Will merge to 'next'.
source: <[email protected]>
* jt/port-ci-whitespace-check-to-gitlab (2024-05-03) 5 commits
- gitlab-ci: add whitespace error check
- ci: make the whitespace report optional
- ci: separate whitespace check script
- github-ci: fix link to whitespace error
- ci: pre-collapse GitLab CI sections
The "whitespace check" task that was enabled for GitHub Actions CI has been ported to GitLab CI.
Will merge to 'next'?
source: <[email protected]>
* jc/no-default-attr-tree-in-bare (2024-05-03) 1 commit
- stop using HEAD for attributes in bare repository by default
Git 2.43 started using the tree of HEAD as the source of attributes in a bare repository, which has severe performance implications. For now, revert the change, without ripping out a more explicit support for the attr.tree configuration variable.
Will merge to 'next'?
source: <[email protected]>
* jl/git-no-advice (2024-05-03) 3 commits
- advice: add --no-advice global option
- doc: add spacing around paginate options
- doc: clean up usage documentation for --no-* opts
A new global "--no-advice" option can be used to disable all advice messages, which is meant to be used only in scripts.
Will merge to 'next'.
source: <[email protected]>
* kn/ref-transaction-symref (2024-05-03) 7 commits
- refs: remove create_symref and associated dead code
- refs: rename refs_create_symref() to refs_update_symref()
- refs: use transaction in refs_create_symref()
- refs: add support for transactional symref updates
- refs: support symrefs in 'reference-transaction' hook
- files-backend: extract out create_symref_lock()
- refs: accept symref values in ref_transaction_update()
Updates to symbolic refs can now be made as a part of ref transaction.
Will merge to 'next'?
source: <[email protected]>
* ma/win32-unix-domain-socket (2024-05-03) 1 commit
- win32: fix building with NO_UNIX_SOCKETS
Build fix.
Will merge to 'next'.
source: <[email protected]>
* ps/config-subcommands (2024-05-03) 14 commits
- builtin/config: display subcommand help
- builtin/config: introduce "edit" subcommand
- builtin/config: introduce "remove-section" subcommand
- builtin/config: introduce "rename-section" subcommand
- builtin/config: introduce "unset" subcommand
- builtin/config: introduce "set" subcommand
- builtin/config: introduce "get" subcommand
- builtin/config: introduce "list" subcommand
- builtin/config: pull out function to handle --null
- builtin/config: pull out function to handle config location
- builtin/config: use OPT_CMDMODE() to specify modes
- builtin/config: move "fixed-value" option to correct group
- builtin/config: move option array around
- config: clarify memory ownership when preparing comment strings
The operation mode options (like "--get") the "git config" command uses have been deprecated and replaced with subcommands (like "git config get").
Will merge to 'next'?
source: <[email protected]>
* ps/refs-without-the-repository (2024-05-03) 5 commits
- refs: remove functions without ref store
- cocci: apply rules to rewrite callers of "refs" interfaces
- cocci: introduce rules to transform "refs" to pass ref store
- refs: add exclude_patterns parameter to for_each_fullref_in()
- refs: introduce missing functions that accept a struct ref_store
The refs API lost functions that implicitly assumes to work on the primary ref_store by forcing the callers to pass a ref_store as an argument.
Will merge to 'next'?
source: <[email protected]>
* ps/reftable-write-options (2024-05-03) 11 commits
- refs/reftable: allow configuring geometric factor
- reftable: make the compaction factor configurable
- refs/reftable: allow disabling writing the object index
- refs/reftable: allow configuring restart interval
- reftable: use uint16_t to track restart interval
- refs/reftable: allow configuring block size
- reftable/dump: support dumping a table's block structure
- reftable/writer: improve error when passed an invalid block size
- reftable/writer: drop static variable used to initialize strbuf
- reftable: consistently pass write opts as value
- reftable: consistently refer to reftable_write_options as opts
The knobs to tweak how reftable files are written have been made available as configuration variables.
Needs review.
source: <[email protected]>
* tb/attr-limits (2024-05-03) 1 commit
- attr.c: move ATTR_MAX_FILE_SIZE check into read_attr_from_buf()
The maximum size of attribute files is enforced more consistently.
Will merge to 'next'.
source: <28f6267709db78ba526d7ed9fc4a734674697c70.1714763555.git.me@ttaylorr.com>
--------------------------------------------------
[Graduated to 'master']
* kn/update-ref-symrefs (2024-04-26) 7 commits
- ref: support symrefs in 'reference-transaction' hook
- update-ref: add support for 'symref-update' command
- update-ref: add support for 'symref-create' command
- update-ref: add support for 'symref-delete' command
- update-ref: add support for 'symref-verify' command
- files-backend: extract out create_symref_lock
- refs: accept symref values in ref_transaction[_add]_update
"update-ref" learns to also handle symbolic refs.
Expecting a reroll.
cf. <CAOLa=ZT4yVEuZXmiTVB2tf0qaTPCPn=0TcJRc89knZQWZBbFTg@mail.gmail.com>
source: <[email protected]>
--------------------------------------------------
[Cooking]
* bb/rgb-12-bit-colors (2024-05-02) 3 commits (merged to 'next' on 2024-05-02 at 7fe29c98d7)
+ color: add support for 12-bit RGB colors
+ t/t4026-color: add test coverage for invalid RGB colors
+ t/t4026-color: remove an extra double quote character
The color parsing code learned to handle 12-bit RGB colors, spelled as "#RGB" (in addition to "#RRGGBB" that is already supported).
Will merge to 'master'.
source: <[email protected]>
* ds/scalar-reconfigure-all-fix (2024-04-30) 1 commit
- scalar: avoid segfault in reconfigure --all
Scalar fix.
Comments?
source: <[email protected]>
* ps/pseudo-ref-terminology (2024-05-02) 10 commits
- refs: refuse to write pseudorefs
- ref-filter: properly distinuish pseudo and root refs
- refs: pseudorefs are no refs
- refs: root refs can be symbolic refs
- refs: classify HEAD as a root ref
- refs: refname is_special_ref() to is_pseudo_ref()
- refs: rename is_pseudoref() to is_root_ref()
- Documentation/glossary: define root refs as refs
- Documentation/glossary: clarify limitations of pseudorefs
- Documentation/glossary: redefine pseudorefs as special refs
Terminology to call various ref-like things are getting straightened out.
Comments?
source: <[email protected]>
* dk/zsh-git-repo-path-fix (2024-04-30) 1 commit (merged to 'next' on 2024-05-02 at 0446d8e63a)
+ completion: zsh: stop leaking local cache variable
Command line completion support for zsh (in contrib/) has been updated to stop exposing internal state to end-user shell interaction.
Will merge to 'master'.
source: <[email protected]>
* rs/diff-parseopts-cleanup (2024-05-01) 1 commit (merged to 'next' on 2024-05-02 at cd9779c175)
+ diff-lib: stop calling diff_setup_done() in do_diff_cache()
Code clean-up to remove code that is now a noop.
Will merge to 'master'.
source: <[email protected]>
* rh/complete-symbolic-ref (2024-04-25) 3 commits (merged to 'next' on 2024-05-01 at 7020ecebe7)
+ completion: add docs on how to add subcommand completions
+ completion: improve docs for using __git_complete
+ completion: add 'symbolic-ref'
Command line completion script (in contrib/) learned to complete "git symbolic-ref" a bit better (you need to enable plumbing commands to be completed with GIT_COMPLETION_SHOW_ALL_COMMANDS).
Will merge to 'master'.
source: <[email protected]>
* jt/doc-submitting-rerolled-series (2024-04-25) 1 commit (merged to 'next' on 2024-05-01 at 1c5865f11c)
+ doc: clarify practices for submitting updated patch versions
Developer doc update.
Will merge to 'master'.
source: <[email protected]>
* bc/zsh-compatibility (2024-04-26) 2 commits
(merged to 'next' on 2024-05-02 at 10222b2472)
+ vimdiff: make script and tests work with zsh
+ t4046: avoid continue in &&-chain for zsh
zsh can pretend to be a normal shell pretty well except for some glitches that we tickle in some of our scripts. Work them around so that "vimdiff" and our test suite works well enough with it.
Will merge to 'master'.
source: <[email protected]>
* jc/rev-parse-fatal-doc (2024-05-01) 1 commit
- rev-parse: document how --is-* options work outside a repository
Doc update.
Comments?
source: <[email protected]>
* ps/undecided-is-not-necessarily-sha1 (2024-04-30) 13 commits . repository: stop setting SHA1 as the default object hash . oss-fuzz/commit-graph: set up hash algorithm
. builtin/shortlog: don't set up revisions without repo
. builtin/diff: explicitly set hash algo when there is no repo . builtin/bundle: abort "verify" early when there is no repository . builtin/blame: don't access potentially unitialized the_hash_algo . builtin/rev-parse: allow shortening to more than 40 hex characters . remote-curl: fix parsing of detached SHA256 heads
. attr: fix BUG() when parsing attrs outside of repo
. attr: don't recompute default attribute source
. parse-options-cb: only abbreviate hashes when hash algo is known . path: move validate_headref() to its only user
. path: harden validation of HEAD with non-standard hashes
Before discovering the repository details, We used to assume SHA-1 as the "default" hash function, which has been corrected. Hopefully this will smoke out codepaths that rely on such an unwarranted assumptions.
Seems to break t0003 with a NULL the_repository.
Ejected out of 'seen' for now.
source: <[email protected]>
* rj/add-p-typo-reaction (2024-04-30) 2 commits
(merged to 'next' on 2024-05-01 at 3016062ebf)
+ add-patch: response to unknown command
+ add-patch: do not show UI messages on stderr
When the user responds to a prompt given by "git add -p" with an unsupported command, list of available commands were given, which was too much if the user knew what they wanted to type but merely made a typo. Now the user gets a much shorter error message.
Will merge to 'master'.
source: <[email protected]>
* ps/the-index-is-no-more (2024-04-18) 6 commits
(merged to 'next' on 2024-05-01 at 7b9a0ab760)
+ repository: drop initialize_the_repository()
+ repository: drop the_index variable
+ builtin/clone: stop using the_index
+ repository: initialize index in repo_init()
+ builtin: stop using the_index
+ t/helper: stop using the_index
The singleton index_state instance "the_index" has been eliminated by always instantiating "the_repository" and replacing references to "the_index" with references to its .index member.
Will merge to 'master'.
source: <[email protected]>
* ps/ci-test-with-jgit (2024-04-12) 13 commits
(merged to 'next' on 2024-05-01 at 35e293e618)
+ t0612: add tests to exercise Git/JGit reftable compatibility
+ t0610: fix non-portable variable assignment
+ t06xx: always execute backend-specific tests
+ ci: install JGit dependency
+ ci: make Perforce binaries executable for all users
+ ci: merge scripts which install dependencies
+ ci: fix setup of custom path for GitLab CI
+ ci: merge custom PATH directories
+ ci: convert "install-dependencies.sh" to use "/bin/sh"
+ ci: drop duplicate package installation for "linux-gcc-default"
+ ci: skip sudo when we are already root
+ ci: expose distro name in dockerized GitHub jobs
+ ci: rename "runs_on_pool" to "distro"
Tests to ensure interoperability between reftable written by jgit and our code have been added and enabled in CI.
Will merge to 'master'.
source: <[email protected]>
* pw/rebase-i-error-message (2024-04-08) 2 commits
- rebase -i: improve error message when picking merge
- rebase -i: pass struct replay_opts to parse_insn_line()
When the user adds to "git rebase -i" instruction to "pick" a merge commit, the error experience is not pleasant. Such an error is now caught earlier in the process that parses the todo list.
Expecting a reroll.
cf. <[email protected]>
source: <[email protected]>
* ps/reftable-write-optim (2024-04-08) 11 commits
(merged to 'next' on 2024-04-30 at 0667e3c05b)
+ reftable/block: reuse compressed array
+ reftable/block: reuse zstream when writing log blocks
+ reftable/writer: reset last_key instead of releasing it
+ reftable/writer: unify releasing memory
+ reftable/writer: refactorings for writer_flush_nonempty_block()
+ reftable/writer: refactorings for writer_add_record()
+ refs/reftable: don't recompute committer ident
+ reftable: remove name checks
+ refs/reftable: skip duplicate name checks
+ refs/reftable: perform explicit D/F check when writing symrefs
+ refs/reftable: fix D/F conflict error message on ref copy
Code to write out reftable has seen some optimization and simplification.
Will merge to 'master'.
source: <[email protected]>
* ds/send-email-per-message-block (2024-04-10) 2 commits
- send-email: make it easy to discern the messages for each patch
- send-email: move newline characters out of a few translatable strings
"git send-email" learned to separate its reports on each message it sends out with an extra blank line in between.
Comments?
source: <[email protected]>
* ew/khash-to-khashl (2024-03-28) 3 commits
- khashl: fix ensemble lookups on empty table
- treewide: switch to khashl for memory savings
- list-objects-filter: use kh_size API
The hashtable library "khash.h" has been replaced with "khashl.h" that has better memory usage characteristics.
Needs review.
cf. <[email protected]>
source: <[email protected]>
* bc/credential-scheme-enhancement (2024-04-16) 16 commits (merged to 'next' on 2024-05-01 at 789ec5bd35)
+ credential: add method for querying capabilities
+ credential-cache: implement authtype capability
+ t: add credential tests for authtype
+ credential: add support for multistage credential rounds
+ t5563: refactor for multi-stage authentication
+ docs: set a limit on credential line length
+ credential: enable state capability
+ credential: add an argument to keep state
+ http: add support for authtype and credential
+ docs: indicate new credential protocol fields
+ credential: add a field called "ephemeral"
+ credential: gate new fields on capability
+ credential: add a field for pre-encoded credentials
+ http: use new headers for each object request
+ remote-curl: reset headers on new request
+ credential: add an authtype field
The credential helper protocol, together with the HTTP layer, have been enhanced to support authentication schemes different from username & password pair, like Bearer and NTLM.
Will merge to 'master'.
source: <[email protected]>
* tb/pseudo-merge-reachability-bitmap (2024-04-30) 23 commits
- t/perf: implement performace tests for pseudo-merge bitmaps
- pseudo-merge: implement support for finding existing merges
- ewah: bitmap_equals_ewah()
- pack-bitmap: extra trace2 information
- pack-bitmap.c: use pseudo-merges during traversal
- t/test-lib-functions.sh: support --date in test_commit_bulk()
- pack-bitmap: implement test helpers for pseudo-merge
- ewah: implement ewah_bitmap_popcount()
- pseudo-merge: implement support for reading pseudo-merge commits
- pack-bitmap.c: read pseudo-merge extension
- pseudo-merge: scaffolding for reads
- pack-bitmap: extract read_bitmap() function
- pack-bitmap-write.c: write pseudo-merge table
- pack-bitmap-write.c: select pseudo-merge commits
- pseudo-merge: implement support for selecting pseudo-merge commits
- pack-bitmap: make bitmap_writer_push_bitmapped_commit() public
- pack-bitmap: implement bitmap_writer_has_bitmapped_object_id()
- pack-bitmap-write: support storing pseudo-merge commits
- pseudo-merge.ch: initial commit
- pack-bitmap: move some initialization to bitmap_writer_init()
- pack-bitmap: drop unused max_bitmaps parameter
- ewah: implement ewah_bitmap_is_subset()
- Documentation/technical: describe pseudo-merge bitmaps format
The pack-bitmap machinery learned to write pseudo-merge bitmaps, which act as imaginary octopus merges covering un-bitmapped reference tips. This enhances bitmap coverage, and thus, performance, for repositories with many references using bitmaps.
Needs review.
source: <[email protected]>
* la/hide-trailer-info (2024-05-02) 11 commits
- trailer unit tests: inspect iterator contents
- trailer: document parse_trailers() usage
- trailer: retire trailer_info_get() from API
- trailer: make trailer_info struct private
- trailer: make parse_trailers() return trailer_info pointer
- interpret-trailers: access trailer_info with new helpers
- sequencer: use the trailer iterator
- trailer: teach iterator about non-trailer lines
- trailer: add unit tests for trailer iterator
- Makefile: sort UNIT_TEST_PROGRAMS
- Merge branch 'la/format-trailer-info' into la/hide-trailer-info
The trailer API has been reshuffled a bit.
Will merge to 'next'?
source: <[email protected]>
* ds/doc-config-reflow (2024-03-14) 1 commit
- config.txt: perform some minor reformatting
Reflow a paragraph in the documentation source without any effect to the formatted text.
Will discard.
source: <97bdaf075bf5a68554cca1731eca78aff2662907.1710444774.git.dsimic@manjaro.org>
* ie/config-includeif-hostname (2024-03-19) 2 commits
- config: learn the "hostname:" includeIf condition
- t: add a test helper for getting hostname
The conditional inclusion mechanism for configuration files learned to switch on the hostname.
Expecting a reroll.
cf. <[email protected]> cf. <[email protected]>
source: <[email protected]>
* cw/git-std-lib (2024-02-28) 4 commits
- SQUASH??? get rid of apparent debugging crufts
- test-stdlib: show that git-std-lib is independent
- git-std-lib: introduce Git Standard Library
- pager: include stdint.h because uintmax_t is used
Split libgit.a out to a separate git-std-lib tor easier reuse.
Expecting a reroll.
source: <[email protected]>
* js/cmake-with-test-tool (2024-04-30) 2 commits
- cmake: let test-tool run the unit tests, too
- Merge branch 'js/unit-test-suite-runner' into js/cmake-with-test-tool (this branch uses js/unit-test-suite-runner.)
"test-tool" is now built in CMake build to also run the unit tests.
May want to roll it into the base topic.
Will merge to 'next' together with the base topic.
source: <[email protected]>
* js/unit-test-suite-runner (2024-04-30) 8 commits
- ci: use test-tool as unit test runner on Windows
- t/Makefile: run unit tests alongside shell tests
- unit tests: add rule for running with test-tool
- test-tool run-command testsuite: support unit tests
- test-tool run-command testsuite: remove hardcoded filter
- test-tool run-command testsuite: get shell from env
- t0080: turn t-basic unit test into a helper
- Merge branch 'jk/unit-tests-buildfix' into js/unit-test-suite-runner (this branch is used by js/cmake-with-test-tool.)
The "test-tool" has been taught to run testsuite tests in parallel, bypassing the need to use the "prove" tool.
Will merge to 'next' together with js/cmake-with-test-tool?
source: <[email protected]>
* bk/complete-dirname-for-am-and-format-patch (2024-01-12) 1 commit
- completion: dir-type optargs for am, format-patch
Command line completion support (in contrib/) has been updated for a few commands to complete directory names where a directory name is expected.
Expecting a reroll.
cf. <[email protected]>
cf. <[email protected]>
source: <[email protected]>
* bk/complete-send-email (2024-01-12) 1 commit
- completion: don't complete revs when --no-format-patch
Command line completion support (in contrib/) has been taught to avoid offering revision names as candidates to "git send-email" when the command is used to send pre-generated files.
Expecting a reroll.
cf. <CAC4O8c88Z3ZqxH2VVaNPpEGB3moL5dJcg3cOWuLWwQ_hLrJMtA@mail.gmail.com>
source: <[email protected]>
* tb/path-filter-fix (2024-01-31) 16 commits
- bloom: introduce deinit_bloom_filters()
- commit-graph: reuse existing Bloom filters where possible
- object.h: fix mis-aligned flag bits table
- commit-graph: new Bloom filter version that fixes murmur3
- commit-graph: unconditionally load Bloom filters
- bloom: prepare to discard incompatible Bloom filters
- bloom: annotate filters with hash version
- repo-settings: introduce commitgraph.changedPathsVersion
- t4216: test changed path filters with high bit paths
- t/helper/test-read-graph: implement bloom-filters mode
- bloom.h: make load_bloom_filter_from_graph() public
- t/helper/test-read-graph.c: extract dump_graph_info()
- gitformat-commit-graph: describe version 2 of BDAT
- commit-graph: ensure Bloom filters are read with consistent settings
- revision.c: consult Bloom filters for root commits
- t/t4216-log-bloom.sh: harden test_bloom_filters_not_used()
The Bloom filter used for path limited history traversal was broken on systems whose "char" is unsigned; update the implementation and bump the format version to 2.
Waiting for a final ack?
cf. <[email protected]>
source: <[email protected]>
* jc/rerere-cleanup (2023-08-25) 4 commits
- rerere: modernize use of empty strbuf
- rerere: try_merge() should use LL_MERGE_ERROR when it means an error
- rerere: fix comment on handle_file() helper
- rerere: simplify check_one_conflict() helper function
Code clean-up.
Not ready to be reviewed yet.
source: <[email protected]>
Quoted 10 lines · Junio C Hamano
* ds/send-email-per-message-block (2024-04-10) 2 commits
- send-email: make it easy to discern the messages for each patch
- send-email: move newline characters out of a few translatable
strings"git send-email" learned to separate its reports on each message it sends out with an extra blank line in between.
Comments?
source: <[email protected]>
After thinking a bit more about these patches, I think that the second patch in the series ("send-email: make it easy to discern the messages for each patch") should be dropped, but the first patch ("send-email:
move newline characters out of a few translatable strings") should have no roadblocks that prevent it from becoming applied to "next".
The reason why I'm proposing that the second patch gets dropped is that I now think there are better ways to utilize the vertical screen space in the outputs produced by "git send-email". I'll keep thinking about it for a while, and possibly submit a clean and irresistible [1] patch at some point in the future.
ds/scalar-reconfigure-all-fix, was Re: What's cooking in git.git (May 2024, #02; Fri, 3)
In reply to Junio C HamanoHi Junio,
Quoted 7 lines · Junio C Hamano
* ds/scalar-reconfigure-all-fix (2024-04-30) 1 commit
- scalar: avoid segfault in reconfigure --allScalar fix.
Comments?
source: <[email protected]>
Patrick commented on it in the affirmative. I do not have anything in addition to Patrick's comments, except that I agree.
Ciao,
Johannes
Quoted 16 lines · Junio C Hamano
* ps/config-subcommands (2024-05-03) 14 commits
- builtin/config: display subcommand help
- builtin/config: introduce "edit" subcommand
- builtin/config: introduce "remove-section" subcommand
- builtin/config: introduce "rename-section" subcommand
- builtin/config: introduce "unset" subcommand
- builtin/config: introduce "set" subcommand
- builtin/config: introduce "get" subcommand
- builtin/config: introduce "list" subcommand
- builtin/config: pull out function to handle--null
- builtin/config: pull out function to handle config location
- builtin/config: useOPT_CMDMODE()to specify modes
- builtin/config: move "fixed-value" option to correct group
- builtin/config: move option array around
- config: clarify memory ownership when preparing comment strings
Quoted 6 lines
The operation mode options (like "--get") the "git config" command uses have been deprecated and replaced with subcommands (like "git config get").
Will merge to 'next'?
source: <[email protected]>
I have just sent out a new version [1] that fixes a few commit messages. Other than that I think this version should be ready to go.
[1]: https://lore.kernel.org/git/[email protected]/T/#u
Quoted 8 lines
* ps/refs-without-the-repository (2024-05-03) 5 commits
- refs: remove functions without ref store
- cocci: apply rules to rewrite callers of "refs" interfaces
- cocci: introduce rules to transform "refs" to pass ref store
- refs: addexclude_patternsparameter tofor_each_fullref_in()
- refs: introduce missing functions that accept astruct ref_store
Quoted 6 lines
The refs API lost functions that implicitly assumes to work on the primary
ref_storeby forcing the callers to pass aref_storeas an argument.Will merge to 'next'?
source: <[email protected]>
It depends. Personally, I'm fine to merge this as-is and make this a hard breakage. I don't really feel like it's worth it to have extra compatibility macros or something like that.
If we want to play nice with current in-flight topics I'd propose to just drop the last patch and reintroduce it in the next release cycle. The downside here is that we may end up adding new callers to the old interfaces, which may make us work against a moving target.
Quoted 20 lines
* ps/undecided-is-not-necessarily-sha1 (2024-04-30) 13 commits . repository: stop setting SHA1 as the default object hash . oss-fuzz/commit-graph: set up hash algorithm
. builtin/shortlog: don't set up revisions without repo
. builtin/diff: explicitly set hash algo when there is no repo . builtin/bundle: abort "verify" early when there is no repository . builtin/blame: don't access potentially unitializedthe_hash_algo. builtin/rev-parse: allow shortening to more than 40 hex characters . remote-curl: fix parsing of detached SHA256 heads
. attr: fixBUG()when parsing attrs outside of repo
. attr: don't recompute default attribute source
. parse-options-cb: only abbreviate hashes when hash algo is known . path: movevalidate_headref()to its only user
. path: harden validation of HEAD with non-standard hashesBefore discovering the repository details, We used to assume SHA-1 as the "default" hash function, which has been corrected. Hopefully this will smoke out codepaths that rely on such an unwarranted assumptions.
Quoted 4 lines
Seems to break t0003 with a NULL
the_repository.Ejected out of 'seen' for now.
source: <[email protected]>
Interesting, I couldn't reproduce this issue when rebasing the patches onto "seen". There were merge conflicts though, both with jc/no-default-attr-tree-in-bare and ps/the-index-is-no-more. So maybe there was a mismerge involved somewhere?
The first one looks like the following. Given that this merge conflict happened in "attr.c", I strongly suspect that this one was mis-merged because t0003 is about attrs, as well.
diff --cc attr.c
index 6af7151088,9d911aeb31..0000000000
--- a/attr.c
+++ b/attr.c
@@@ -1223,8 -1224,15 +1224,8 @@@ static int compute_default_attr_source(
ignore_bad_attr_tree = 1;
}
- if (!default_attr_source_tree_object_name || !is_null_oid(attr_source))
- return;
- if (!default_attr_source_tree_object_name &&
- startup_info->have_repository &&
- is_bare_repository()) {
- default_attr_source_tree_object_name = "HEAD";
- ignore_bad_attr_tree = 1;
- }
-
+ if (!default_attr_source_tree_object_name)
+ return 0;
if (repo_get_oid_treeish(the_repository,
default_attr_source_tree_object_name,And the second conflict looks like this:
diff --cc repository.c
index 2118f563e3,b65b1a8c8b..0000000000
--- a/repository.c
+++ b/repository.c
@@@ -17,35 -22,19 +17,15 @@@
/* The main repository */
static struct repository the_repo;
-struct repository *the_repository;
-struct index_state the_index;
+struct repository *the_repository = &the_repo;
-void initialize_the_repository(void)
+void initialize_repository(struct repository *repo)
{
- the_repository = &the_repo;
-
- the_repo.index = &the_index;
- the_repo.objects = raw_object_store_new();
- the_repo.remote_state = remote_state_new();
- the_repo.parsed_objects = parsed_object_pool_new();
-
- index_state_init(&the_index, the_repository);
+ repo->objects = raw_object_store_new();
+ repo->remote_state = remote_state_new();
+ repo->parsed_objects = parsed_object_pool_new();
+ ALLOC_ARRAY(repo->index, 1);
+ index_state_init(repo->index, repo);
-
- /*
- * Unfortunately, we need to keep this hack around for the time being:
- *
- * - Not setting up the hash algorithm for `the_repository` leads to
- * crashes because `the_hash_algo` is a macro that expands to
- * `the_repository->hash_algo`. So if Git commands try to access
- * `the_hash_algo` without a Git directory we crash.
- *
- * - Setting up the hash algorithm to be SHA1 by default breaks other
- * commands when running with SHA256.
- *
- * This is another point in case why having global state is a bad idea.
- * Eventually, we should remove this hack and stop setting the hash
- * algorithm in this function altogether. Instead, it should only ever
- * be set via our repository setup procedures. But that requires more
- * work.
- */
- if (repo == the_repository)
- repo_set_hash_algo(repo, GIT_HASH_SHA1);
}
static void expand_base_dir(char **out, const char *in,I'll maybe just revise the series and pull both of these topics in as a dependency.
Quoted 16 lines
* ps/ci-test-with-jgit (2024-04-12) 13 commits
(merged to 'next' on 2024-05-01 at35e293e618)
+ t0612: add tests to exercise Git/JGit reftable compatibility
+ t0610: fix non-portable variable assignment
+ t06xx: always execute backend-specific tests
+ ci: install JGit dependency
+ ci: make Perforce binaries executable for all users
+ ci: merge scripts which install dependencies
+ ci: fix setup of custom path for GitLab CI
+ ci: merge custom PATH directories
+ ci: convert "install-dependencies.sh" to use "/bin/sh"
+ ci: drop duplicate package installation for "linux-gcc-default"
+ ci: skip sudo when we are already root
+ ci: expose distro name in dockerized GitHub jobs
+ ci: rename "runs_on_pool" to "distro"
Quoted 5 lines
Tests to ensure interoperability between reftable written by jgit and our code have been added and enabled in CI.
Will merge to 'master'.
source: <[email protected]>
I have sent a follow-up here [2] that unbreaks the "ubuntu:latest" jobs. The breakage is caused by the release of Ubuntu 24.04, where there is no more Python 2 anymore.
[2]: https://lore.kernel.org/git/cb8cefc20f373a3516695e7cbee975132553ea95.1714973381.git.ps@pks.im/T/#u
Patrick
Quoted 4 lines · Patrick Steinhardt
I have just sent out a new version [1] that fixes a few commit messages. Other than that I think this version should be ready to go.
Thanks.
Quoted 22 lines
Quoted 22 lines
* ps/undecided-is-not-necessarily-sha1 (2024-04-30) 13 commits . repository: stop setting SHA1 as the default object hash . oss-fuzz/commit-graph: set up hash algorithm
. builtin/shortlog: don't set up revisions without repo
. builtin/diff: explicitly set hash algo when there is no repo . builtin/bundle: abort "verify" early when there is no repository . builtin/blame: don't access potentially unitializedthe_hash_algo. builtin/rev-parse: allow shortening to more than 40 hex characters . remote-curl: fix parsing of detached SHA256 heads
. attr: fixBUG()when parsing attrs outside of repo
. attr: don't recompute default attribute source
. parse-options-cb: only abbreviate hashes when hash algo is known . path: movevalidate_headref()to its only user
. path: harden validation of HEAD with non-standard hashesBefore discovering the repository details, We used to assume SHA-1 as the "default" hash function, which has been corrected. Hopefully this will smoke out codepaths that rely on such an unwarranted assumptions.
Seems to break t0003 with a NULL
the_repository.
Quoted 7 lines
Quoted 2 lines
Ejected out of 'seen' for now.
source: <[email protected]>Interesting, I couldn't reproduce this issue when rebasing the patches onto "seen". There were merge conflicts though, both with jc/no-default-attr-tree-in-bare and ps/the-index-is-no-more. So maybe there was a mismerge involved somewhere?
That is very possible.
Quoted 8 lines · Junio C Hamano
Quoted 6 lines
Quoted 2 lines
* ps/undecided-is-not-necessarily-sha1 (2024-04-30) 13 commits
...Interesting, I couldn't reproduce this issue when rebasing the patches onto "seen". There were merge conflicts though, both with jc/no-default-attr-tree-in-bare and ps/the-index-is-no-more. So maybe there was a mismerge involved somewhere?
That is very possible.
Yeah, I think I botched the merge. Somehow my resolution had an unneeded "|| !is_null_oid()" in the result, but the right resolution of course should look like below.
Thanks.
attr.c +0 -22
@@ -1214,53 +1214,31 @@ void set_git_attr_source(const char *tree_object_name)static int compute_default_attr_source (struct object_id *attr_source) {int ignore_bad_attr_tree =0 ;if (!default_attr_source_tree_object_name) default_attr_source_tree_object_name =getenv (GIT_ATTR_SOURCE_ENVIRONMENT );if (!default_attr_source_tree_object_name && git_attr_tree) { default_attr_source_tree_object_name = git_attr_tree; ignore_bad_attr_tree =1 ; } -<<<<<<< HEAD - if (!default_attr_source_tree_object_name || !is_null_oid(attr_source)) -return ; -||||||| 00e10ef10e -if (!default_attr_source_tree_object_name && - startup_info->have_repository && -is_bare_repository ()) { - default_attr_source_tree_object_name ="HEAD" ; - ignore_bad_attr_tree =1 ; - } - -if (!default_attr_source_tree_object_name || !is_null_oid (attr_source)) -return ; -======= -if (!default_attr_source_tree_object_name && - startup_info->have_repository && -is_bare_repository ()) { - default_attr_source_tree_object_name ="HEAD" ; - ignore_bad_attr_tree =1 ; - } -if (!default_attr_source_tree_object_name)return 0 ;if (!startup_info->have_repository) {if (!ignore_bad_attr_tree)die (_ ("cannot use --attr-source or GIT_ATTR_SOURCE without repo" ));return 0 ; } ->>>>>>> ps/undecided-is-not-necessarily-sha1@{1}if (repo_get_oid_treeish (the_repository, default_attr_source_tree_object_name, attr_source)) {if (!ignore_bad_attr_tree)die (_ ("bad --attr-source or GIT_ATTR_SOURCE" ));return 0 ; }return 1 ; }
Quoted 9 lines · Junio C Hamano
Quoted 7 lines · Junio C Hamano
Quoted 6 lines
Quoted 2 lines
* ps/undecided-is-not-necessarily-sha1 (2024-04-30) 13 commits
...Interesting, I couldn't reproduce this issue when rebasing the patches onto "seen". There were merge conflicts though, both with jc/no-default-attr-tree-in-bare and ps/the-index-is-no-more. So maybe there was a mismerge involved somewhere?
Quoted 7 lines
Quoted 1 line
That is very possible.
Yeah, I think I botched the merge. Somehow my resolution had an unneeded "|| !is_null_oid()" in the result, but the right resolution of course should look like below.
Thanks.
Yeah, that resolution looks correct to me. I've also sent a v4 with increased context and with the two conflicting topics pulles in as dependency. Other than that it is the same to v3.
Patrick