What's cooking in git.git (Sep 2018, #03; Fri, 14)

What's cooking in git.git (Sep 2018, #03; Fri, 14)

[Junio C Hamano]

Here are the topics that have been cooking.  Commits prefixed with
'-' are only in 'pu' (proposed updates) while commits prefixed with
'+' are in 'next'.  The ones marked with '.' do not appear in any of
the integration branches, but I am still holding onto them.

The tip of 'next' hasn't been rewound yet, but the states of many
topics that were marked to "Cook in 'next'" read "Will merge to
'master'" now.  I'll probably start merging a handful of topics that
have been in 'next' down to 'master' tonight (they appear at the
bottom of "log --first-parent master..pu" output).

The three GSoC "rewrite in C" topics are still unclassified in this
"What's cooking" report, but I am hoping that we can have them in
'next' sooner rather than later.  I got an impression that Dscho
wanted a chance for the final clean-up on some of them, so I am not
doing anything hasty yet at this moment, though.

You can find the changes described here in the integration branches
of the repositories listed at

    http://git-blame.blogspot.com/p/git-public-repositories.html

--------------------------------------------------
[New Topics]

* ab/fsck-skiplist (2018-09-12) 10 commits
 - fsck: support comments & empty lines in skipList
 - fsck: use oidset instead of oid_array for skipList
 - fsck: use strbuf_getline() to read skiplist file
 - fsck: add a performance test for skipList
 - fsck: add a performance test
 - fsck: document that skipList input must be unabbreviated
 - fsck: document and test commented & empty line skipList input
 - fsck: document and test sorted skipList input
 - fsck tests: add a test for no skipList input
 - fsck tests: setup of bogus commit object

 Update fsck.skipList implementation and documentation.

 Will merge to 'next'.


* bc/hash-independent-tests (2018-09-13) 12 commits
 - t5318: use test_oid for HASH_LEN
 - t1407: make hash size independent
 - t1406: make hash-size independent
 - t1405: make hash size independent
 - t1400: switch hard-coded object ID to variable
 - t1006: make hash size independent
 - t0064: make hash size independent
 - t0027: make hash size independent
 - t0002: abstract away SHA-1 specific constants
 - t0000: update tests for SHA-256
 - t0000: use hash translation table
 - t: add test functions to translate hash-related values

 Various tests have been updated to make it easier to swap the
 hash function used for object identification.

 Will merge to 'next'.


* bp/mv-submodules-with-fsmonitor (2018-09-12) 1 commit
 - git-mv: allow submodules and fsmonitor to work together

 When fsmonitor is in use, after operation on submodules updates
 .gitmodules, we lost track of the fact that we did so and relied on
 stale fsmonitor data.

 Will merge to 'next'.


* bp/read-cache-parallel (2018-09-13) 6 commits
 - SQUASH???
 - read-cache: clean up casting and byte decoding
 - read-cache.c: optimize reading index format v4
 - read-cache: load cache entries on worker threads
 - read-cache: load cache extensions on a worker thread
 - eoie: add End of Index Entry (EOIE) extension

 A new extension to the index file has been introduced, which allows
 the file to be read in parallel.

 Will merge to 'next' after squashing the fix in.


* ds/coverage-diff (2018-09-12) 1 commit
 - contrib: add coverage-diff script

 The result of coverage test can be combined with "git blame" to
 check the test coverage of code introduced recently with a new
 'coverage-diff' tool (in contrib/).

 Expecting a reroll.


* ds/format-patch-range-diff-test (2018-09-12) 1 commit
 - t3206-range-diff.sh: cover single-patch case
 (this branch uses es/format-patch-interdiff and es/format-patch-rangediff.)

 Will merge to 'next'.


* ds/multi-pack-verify (2018-09-13) 11 commits
 - fsck: verify multi-pack-index
 - multi-pack-index: report progress during 'verify'
 - multi-pack-index: verify object offsets
 - multi-pack-index: fix 32-bit vs 64-bit size check
 - multi-pack-index: verify oid lookup order
 - multi-pack-index: verify oid fanout order
 - multi-pack-index: verify missing pack
 - multi-pack-index: verify packname order
 - multi-pack-index: verify corrupt chunk lookup table
 - multi-pack-index: verify bad header
 - multi-pack-index: add 'verify' verb
 (this branch uses ds/multi-pack-index.)

 "git multi-pack-index" learned to detect corruption in the .midx
 file it uses, and this feature has been integrated into "git fsck".

 Will merge to 'next'.


* en/sequencer-empty-edit-result-aborts (2018-09-13) 1 commit
 - sequencer: fix --allow-empty-message behavior, make it smarter

 "git rebase" etc. in Git 2.19 fails to abort when given an empty
 commit log message as result of editing, which has been corrected.

 Will merge to 'next'.


* en/update-ref-no-deref-stdin (2018-09-12) 2 commits
 - update-ref: allow --no-deref with --stdin
 - update-ref: fix type of update_flags variable to match its usage

 "git update-ref" learned to make both "--no-deref" and "--stdin"
 work at the same time.

 Will merge to 'next'.


* jt/lazy-object-fetch-fix (2018-09-13) 2 commits
 - fetch-object: set exact_oid when fetching
 - fetch-object: unify fetch_object[s] functions

 The code to backfill objects in lazily cloned repository did not
 work correctly, which has been corrected.

 Will merge to 'next'.


* ms/remote-error-message-update (2018-09-14) 1 commit
 - builtin/remote: quote remote name on error to display empty name

 Update error messages given by "git remote" and make them consistent.

 Will merge to 'next'.


* nd/config-split (2018-09-12) 11 commits
 - config.txt: move submodule part out to a separate file
 - config.txt: move sequence.editor out of "core" part
 - config.txt: move sendemail part out to a separate file
 - config.txt: move receive part out to a separate file
 - config.txt: move push part out to a separate file
 - config.txt: move pull part out to a separate file
 - config.txt: move gui part out to a separate file
 - config.txt: move gitcvs part out to a separate file
 - config.txt: move format part out to a separate file
 - config.txt: move fetch part out to a separate file
 - config.txt: follow camelCase naming

 Split Documentation/config.txt for easier maintenance.

 Will merge to 'next'.


* sb/submodule-recursive-fetch-gets-the-tip (2018-09-12) 9 commits
 - builtin/fetch: check for submodule updates for non branch fetches
 - fetch: retry fetching submodules if sha1 were not fetched
 - submodule: fetch in submodules git directory instead of in worktree
 - submodule.c: do not copy around submodule list
 - submodule: move global changed_submodule_names into fetch submodule struct
 - submodule.c: sort changed_submodule_names before searching it
 - submodule.c: fix indentation
 - sha1-array: provide oid_array_filter
 - string-list: add string_list_{pop, last} functions

 "git fetch --recurse-submodules" may not fetch the necessary commit
 that is bound to the superproject, which is getting corrected.

 Expecting a reroll.
 cf. <CAGZ79kbavjVbTqXsmtjW6=jhkq47_p3mc6=92xOp4_mfhqDtvw@mail.gmail.com>
 cf. <b16af8c0-0435-0de4-ed6c-53888d6190af@ramsayjones.plus.com>
 cf. <CAGZ79kZKKf9N8yx9EuCRZhrZS_mA2218PouEG7aHDhK2bJGEdA@mail.gmail.com>


* sg/split-index-test (2018-09-12) 2 commits
 - t0090: disable GIT_TEST_SPLIT_INDEX for the test checking split index
 - t1700-split-index: drop unnecessary 'grep'

 Test updates.

 Will merge to 'next'.


* tb/void-check-attr (2018-09-12) 1 commit
 - Make git_check_attr() a void function

 Code clean-up.

 Will merge to 'next'.


* tg/range-diff-corner-case-fix (2018-09-14) 1 commit
 - linear-assignment: fix potential out of bounds memory access

 Recently added "range-diff" had a corner-case bug to cause it
 segfault, which has been corrected.

 Will merge to 'next'.


* bp/rename-test-env-var (2018-09-14) 5 commits
 - preload-index: update GIT_FORCE_PRELOAD_TEST support
 - read-cache: update TEST_GIT_INDEX_VERSION support
 - fsmonitor: update GIT_TEST_FSMONITOR support
 - preload-index: teach GIT_FORCE_PRELOAD_TEST to take a boolean
 - correct typo/spelling error in t/README

 Some environment variables that control the runtime options of Git
 used during tests are getting renamed for consistency.

 Will merge to 'next'.

--------------------------------------------------
[Stalled]

* bw/submodule-name-to-dir (2018-08-10) 2 commits
  (merged to 'next' on 2018-08-22 at c17f83be24)
 + submodule: munge paths to submodule git directories
 + submodule: create helper to build paths to submodule gitdirs

 In modern repository layout, the real body of a cloned submodule
 repository is held in .git/modules/ of the superproject, indexed by
 the submodule name.  URLencode the submodule name before computing
 the name of the directory to make sure they form a flat namespace.

 Will eject out of 'next', expecting further work on the topic.
 cf. <CAGZ79kYnbjaPoWdda0SM_-_X77mVyYC7JO61OV8nm2yj3Q1OvQ@mail.gmail.com>


* ng/status-i-short-for-ignored (2018-08-09) 1 commit
 - status: -i shorthand for --ignored command line option

 "git status --ignored" gained a shorthand "git status -i".

 Will discard, after hearing no strong support.
 What's the list opinion on this one?  It is Meh to me, but
 obviously the author cared enough to write a patch, so...


* sb/submodule-move-head-with-corruption (2018-08-28) 2 commits
 - submodule.c: warn about missing submodule git directories
 - t2013: add test for missing but active submodule

 Will discard and wait for a cleaned-up rewrite.
 cf. <20180907195349.GA103699@aiede.svl.corp.google.com>


* sl/commit-dry-run-with-short-output-fix (2018-07-30) 4 commits
 . commit: fix exit code when doing a dry run
 . wt-status: teach wt_status_collect about merges in progress
 . wt-status: rename commitable to committable
 . t7501: add coverage for flags which imply dry runs

 "git commit --dry-run" gave a correct exit status even during a
 conflict resolution toward a merge, but it did not with the
 "--short" option, which has been corrected.

 Seems to break 7512, 3404 and 7060 in 'pu'.


* ma/wrapped-info (2018-05-28) 2 commits
 - usage: prefix all lines in `vreportf()`, not just the first
 - usage: extract `prefix_suffix_lines()` from `advise()`

 An attempt to help making multi-line messages fed to warning(),
 error(), and friends more easily translatable.

 Will discard and wait for a cleaned-up rewrite.
 cf. <20180529213957.GF7964@sigill.intra.peff.net>


* hn/bisect-first-parent (2018-04-21) 1 commit
 - bisect: create 'bisect_flags' parameter in find_bisection()
 (this branch is used by tb/bisect-first-parent.)

 Preliminary code update to allow passing more flags down the
 bisection codepath in the future.

 We do not add random code that does not have real users to our
 codebase, so let's have it wait until such a real code materializes
 before too long.


* pb/bisect-helper-2 (2018-07-23) 8 commits
 - t6030: make various test to pass GETTEXT_POISON tests
 - bisect--helper: `bisect_start` shell function partially in C
 - bisect--helper: `get_terms` & `bisect_terms` shell function in C
 - bisect--helper: `bisect_next_check` shell function in C
 - bisect--helper: `check_and_set_terms` shell function in C
 - wrapper: move is_empty_file() and rename it as is_empty_or_missing_file()
 - bisect--helper: `bisect_write` shell function in C
 - bisect--helper: `bisect_reset` shell function in C

 Expecting a reroll.
 cf. <0102015f5e5ee171-f30f4868-886f-47a1-a4e4-b4936afc545d-000000@eu-west-1.amazonses.com>

 I just rebased the topic to a newer base as it did not build
 standalone with the base I originally queued the topic on, but
 otherwise there is no update to address any of the review comments
 in the thread above---we are still waiting for a reroll.


* jk/drop-ancient-curl (2017-08-09) 5 commits
 - http: #error on too-old curl
 - curl: remove ifdef'd code never used with curl >=7.19.4
 - http: drop support for curl < 7.19.4
 - http: drop support for curl < 7.16.0
 - http: drop support for curl < 7.11.1

 Some code in http.c that has bitrot is being removed.

 Expecting a reroll.


* mk/use-size-t-in-zlib (2017-08-10) 1 commit
 . zlib.c: use size_t for size

 The wrapper to call into zlib followed our long tradition to use
 "unsigned long" for sizes of regions in memory, which have been
 updated to use "size_t".

 Needs resurrecting by making sure the fix is good and still applies
 (or adjusted to today's codebase).

--------------------------------------------------
[Cooking]

* ab/commit-graph-progress (2018-09-11) 2 commits
 - commit-graph verify: add progress output
 - commit-graph write: add progress output

 Generation of (expermental) commit-graph files have so far been
 fairly silent, even though it takes noticeable amount of time in a
 meaningfully large repository.  The users will now see progress
 output.

 Will merge to 'next'.


* bw/protocol-v2 (2018-09-10) 1 commit
 - config: document value 2 for protocol.version

 Doc fix.

 Will merge to 'next'.


* en/double-semicolon-fix (2018-09-05) 1 commit
 - Remove superfluous trailing semicolons

 Code clean-up.

 Will merge to 'next'.


* en/rerere-multi-stage-1-fix (2018-09-11) 2 commits
 - rerere: avoid buffer overrun
 - t4200: demonstrate rerere segfault on specially crafted merge

 A corner case bugfix in "git rerere" code.

 Will merge to 'next'.


* jk/dev-build-format-security (2018-09-11) 1 commit
 - config.mak.dev: add -Wformat-security

 Build tweak to help developers.

 Will merge to 'next'.


* jk/reopen-tempfile-truncate (2018-09-05) 1 commit
 - reopen_tempfile(): truncate opened file

 Fix for a long-standing bug that leaves the index file corrupt when
 it shrinks during a partial commit.

 Will merge to 'next'.


* js/mingw-o-append (2018-09-11) 2 commits
 - mingw: fix mingw_open_append to work with named pipes
 - t0051: test GIT_TRACE to a windows named pipe

 Further fix for O_APPEND emulation on Windows

 Will merge to 'next'.


* mk/http-backend-content-length (2018-09-11) 1 commit
  (merged to 'next' on 2018-09-11 at e8095fc635)
 + http-backend test: make empty CONTENT_LENGTH test more realistic

 Test update.

 Will merge to 'master'.


* nd/test-tool (2018-09-11) 6 commits
 - Makefile: add a hint about TEST_BUILTINS_OBJS
 - t/helper: merge test-dump-fsmonitor into test-tool
 - t/helper: merge test-parse-options into test-tool
 - t/helper: merge test-pkt-line into test-tool
 - t/helper: merge test-dump-untracked-cache into test-tool
 - t/helper: keep test-tool command list sorted

 Test helper binaries clean-up.

 Will merge to 'next'.


* sb/diff-color-move-more (2018-09-11) 1 commit
 - diff: fix --color-moved-ws=allow-indentation-change

 Bugfix.

 Will merge to 'next'.


* sb/string-list-remove-unused (2018-09-11) 1 commit
 - string-list: remove unused function print_string_list

 Code clean-up.

 Will merge to 'next'.


* sg/t3701-tighten-trace (2018-09-11) 1 commit
 - t3701-add-interactive: tighten the check of trace output

 Test update.

 Will merge to 'next'.


* ss/wt-status-committable (2018-09-07) 4 commits
 - wt-status.c: set the committable flag in the collect phase
 - t7501: add test of "commit --dry-run --short"
 - wt-status: rename commitable to committable
 - wt-status.c: move has_unmerged earlier in the file
 (this branch is used by jc/wt-status-state-cleanup.)

 Code clean-up in the internal machinery used by "git status" and
 "git commit --dry-run".

 Will merge to 'next'.


* jc/wt-status-state-cleanup (2018-09-07) 1 commit
 - WIP: roll wt_status_state into wt_status and populate in the collect phase
 (this branch uses ss/wt-status-committable.)



* tz/t5551-with-curl-7.61.1 (2018-09-11) 1 commit
 - t5551-http-fetch-smart.sh: sort cookies before comparing

 Test fix.

 Will merge to 'next'.


* ab/fetch-tags-noclobber (2018-08-31) 9 commits
  (merged to 'next' on 2018-09-14 at 0384a7a8b4)
 + fetch: stop clobbering existing tags without --force
 + fetch: document local ref updates with/without --force
 + push doc: correct lies about how push refspecs work
 + push doc: move mention of "tag <tag>" later in the prose
 + push doc: remove confusing mention of remote merger
 + fetch tests: add a test for clobbering tag behavior
 + push tests: use spaces in interpolated string
 + push tests: make use of unused $1 in test description
 + fetch: change "branch" to "reference" in --force -h output

 The rules used by "git push" and "git fetch" to determine if a ref
 can or cannot be updated were inconsistent; specifically, fetching
 to update existing tags were allowed even though tags are supposed
 to be unmoving anchoring points.  "git fetch" was taught to forbid
 updates to existing tags without the "--force" option.

 Will merge to 'master'.
 This is a backward incompatible change but in a good way; it may
 still need to be treated carefully.


* es/worktree-forced-ops-fix (2018-09-05) 10 commits
  (merged to 'next' on 2018-09-14 at 1a0cc3204d)
 + doc-diff: force worktree add
 + worktree: delete .git/worktrees if empty after 'remove'
 + worktree: teach 'remove' to override lock when --force given twice
 + worktree: teach 'move' to override lock when --force given twice
 + worktree: teach 'add' to respect --force for registered but missing path
 + worktree: disallow adding same path multiple times
 + worktree: prepare for more checks of whether path can become worktree
 + worktree: generalize delete_git_dir() to reduce code duplication
 + worktree: move delete_git_dir() earlier in file for upcoming new callers
 + worktree: don't die() in library function find_worktree()

 Fix a bug in which the same path could be registered under multiple
 worktree entries if the path was missing (for instance, was removed
 manually).  Also, as a convenience, expand the number of cases in
 which --force is applicable.

 Will merge to 'master'.


* jk/patch-corrupted-delta-fix (2018-08-30) 6 commits
  (merged to 'next' on 2018-09-14 at 7517309cb0)
 + t5303: use printf to generate delta bases
 + patch-delta: handle truncated copy parameters
 + patch-delta: consistently report corruption
 + patch-delta: fix oob read
 + t5303: test some corrupt deltas
 + test-delta: read input into a heap buffer

 Malformed or crafted data in packstream can make our code attempt
 to read or write past the allocated buffer and abort, instead of
 reporting an error, which has been fixed.

 Will merge to 'master'.


* jk/pack-objects-with-bitmap-fix (2018-09-04) 4 commits
  (merged to 'next' on 2018-09-14 at 392eb2abb1)
 + pack-bitmap: drop "loaded" flag
 + traverse_bitmap_commit_list(): don't free result
 + t5310: test delta reuse with bitmaps
 + bitmap_has_sha1_in_uninteresting(): drop BUG check
 (this branch uses jk/pack-delta-reuse-with-bitmap.)

 Hotfix of the base topic.

 Will merge to 'master'.


* js/rebase-i-autosquash-fix (2018-09-04) 2 commits
 - rebase -i: be careful to wrap up fixup/squash chains
 - rebase -i --autosquash: demonstrate a problem skipping the last squash

 "git rebase -i" did not clear the state files correctly when a run
 of "squash/fixup" is aborted and then the user manually amended the
 commit instead, which has been corrected.

 Will merge to 'next'.


* nd/bisect-show-list-fix (2018-09-04) 1 commit
  (merged to 'next' on 2018-09-14 at 18242da7ef)
 + bisect.c: make show_list() build again

 Debugging aid update.

 Will merge to 'master'.


* sg/doc-trace-appends (2018-09-04) 1 commit
  (merged to 'next' on 2018-09-14 at 6d82abb8bf)
 + Documentation/git.txt: clarify that GIT_TRACE=/path appends

 Docfix.

 Will merge to 'master'.


* ds/format-commit-graph-docs (2018-08-21) 2 commits
 - commit-graph.txt: improve formatting for asciidoc
 - Docs: Add commit-graph tech docs to Makefile

 Design docs for the commit-graph machinery is now made into HTML as
 well as text.

 Will discard.
 I am inclined to drop these, as I do not see much clarity in HTML
 output over the text source.  Opinions?


* jk/diff-rendered-docs (2018-08-31) 5 commits
  (merged to 'next' on 2018-09-14 at 9b43d5a568)
 + doc/Makefile: drop doc-diff worktree and temporary files on "make clean"
 + doc-diff: add --clean mode to remove temporary working gunk
 + doc-diff: fix non-portable 'man' invocation
 + doc-diff: always use oids inside worktree
  (merged to 'next' on 2018-08-22 at dd7a2b71cd)
 + SubmittingPatches: mention doc-diff

 Dev doc update.

 Will merge to 'master'.


* jk/pack-delta-reuse-with-bitmap (2018-08-21) 6 commits
  (merged to 'next' on 2018-08-22 at fc50b59dab)
 + pack-objects: reuse on-disk deltas for thin "have" objects
 + pack-bitmap: save "have" bitmap from walk
 + t/perf: add perf tests for fetches from a bitmapped server
 + t/perf: add infrastructure for measuring sizes
 + t/perf: factor out percent calculations
 + t/perf: factor boilerplate out of test_perf
 (this branch is used by jk/pack-objects-with-bitmap-fix.)

 When creating a thin pack, which allows objects to be made into a
 delta against another object that is not in the resulting pack but
 is known to be present on the receiving end, the code learned to
 take advantage of the reachability bitmap; this allows the server
 to send a delta against a base beyond the "boundary" commit.

 Will merge to 'master'.


* jk/rev-list-stdin-noop-is-ok (2018-08-22) 1 commit
  (merged to 'next' on 2018-08-27 at d5916f7bc1)
 + rev-list: make empty --stdin not an error

 "git rev-list --stdin </dev/null" used to be an error; it now shows
 no output without an error.  "git rev-list --stdin --default HEAD"
 still falls back to the given default when nothing is given on the
 standard input.

 Will merge to 'master'.


* js/rebase-in-c-5.5-work-with-rebase-i-in-c (2018-09-06) 2 commits
 - builtin rebase: prepare for builtin rebase -i
 - Merge branch 'ag/rebase-i-in-c' into js/rebase-in-c-5.5-work-with-rebase-i-in-c
 (this branch is used by pk/rebase-in-c-6-final; uses ag/rebase-i-in-c, pk/rebase-in-c, pk/rebase-in-c-2-basic, pk/rebase-in-c-3-acts, pk/rebase-in-c-4-opts and pk/rebase-in-c-5-test.)

 "rebase" that has been rewritten learns the new calling convention
 used by "rebase -i" that was rewritten in C, tying the loose end
 between two GSoC topics that stomped on each other's toes.


* jk/trailer-fixes (2018-08-23) 8 commits
  (merged to 'next' on 2018-08-27 at 93b671b8c6)
 + append_signoff: use size_t for string offsets
 + sequencer: ignore "---" divider when parsing trailers
 + pretty, ref-filter: format %(trailers) with no_divider option
 + interpret-trailers: allow suppressing "---" divider
 + interpret-trailers: tighten check for "---" patch boundary
 + trailer: pass process_trailer_opts to trailer_info_get()
 + trailer: use size_t for iterating trailer list
 + trailer: use size_t for string offsets

 "git interpret-trailers" and its underlying machinery had a buggy
 code that attempted to ignore patch text after commit log message,
 which triggered in various codepaths that will always get the log
 message alone and never get such an input.

 Will merge to 'master'.


* tg/rerere-doc-updates (2018-08-29) 2 commits
  (merged to 'next' on 2018-08-31 at ce4fef1a97)
 + rerere: add note about files with existing conflict markers
 + rerere: mention caveat about unmatched conflict markers
 (this branch uses tg/rerere.)

 Clarify a part of technical documentation for rerere.

 Will merge to 'master'.


* ds/commit-graph-tests (2018-08-29) 1 commit
  (merged to 'next' on 2018-09-14 at d072a0ee3e)
 + commit-graph: define GIT_TEST_COMMIT_GRAPH

 We can now optionally run tests with commit-graph enabled.

 Will merge to 'master'.


* jk/cocci (2018-08-29) 9 commits
  (merged to 'next' on 2018-08-31 at 914b4f17ce)
 + show_dirstat: simplify same-content check
 + read-cache: use oideq() in ce_compare functions
 + convert hashmap comparison functions to oideq()
 + convert "hashcmp() != 0" to "!hasheq()"
 + convert "oidcmp() != 0" to "!oideq()"
 + convert "hashcmp() == 0" to hasheq()
 + convert "oidcmp() == 0" to oideq()
 + introduce hasheq() and oideq()
 + coccinelle: use <...> for function exclusion

 spatch transformation to replace boolean uses of !hashcmp() to
 newly introduced oideq() is added, and applied, to regain
 performance lost due to support of multiple hash algorithms.

 Will merge to 'master'.


* js/add-i-coalesce-after-editing-hunk (2018-08-28) 1 commit
 - add -p: coalesce hunks before testing applicability

 Applicability check after a patch is edited in a "git add -i/p"
 session has been improved.

 Will hold.
 cf. <e5b2900a-0558-d3bf-8ea1-d526b078bbc2@talktalk.net>


* rs/mailinfo-format-flowed (2018-08-29) 1 commit
  (merged to 'next' on 2018-08-31 at 9e8b92176c)
 + mailinfo: support format=flowed

 "git mailinfo" used in "git am" learned to make a best-effort
 recovery of a patch corrupted by MUA that sends text/plain with
 format=flawed option.

 Will merge to 'master'.


* tg/conflict-marker-size (2018-08-29) 1 commit
  (merged to 'next' on 2018-08-31 at 12099161f0)
 + .gitattributes: add conflict-marker-size for relevant files

 Developer aid.

 Will merge to 'master'.


* ts/doc-build-manpage-xsl-quietly (2018-08-29) 1 commit
  (merged to 'next' on 2018-08-31 at 7527e0f8d3)
 + Documentation/Makefile: make manpage-base-url.xsl generation quieter

 Build tweak.

 Will merge to 'master'.


* bp/checkout-new-branch-optim (2018-08-16) 1 commit
  (merged to 'next' on 2018-08-27 at e69bfd115f)
 + checkout: optimize "git checkout -b <new_branch>"

 "git checkout -b newbranch [HEAD]" should not have to do as much as
 checking out a commit different from HEAD.  An attempt is made to
 optimize this special case.

 Will merge to 'master'.


* ao/submodule-wo-gitmodules-checked-out (2018-08-14) 7 commits
 - submodule: support reading .gitmodules even when it's not checked out
 - t7506: clean up .gitmodules properly before setting up new scenario
 - submodule: use the 'submodule--helper config' command
 - submodule--helper: add a new 'config' subcommand
 - t7411: be nicer to future tests and really clean things up
 - submodule: factor out a config_set_in_gitmodules_file_gently function
 - submodule: add a print_config_from_gitmodules() helper

 The submodule support has been updated to read from the blob at
 HEAD:.gitmodules when the .gitmodules file is missing from the
 working tree.

 Expecting a reroll.

 I find the design a bit iffy in that our usual "missing in the
 working tree?  let's use the latest blob" fallback is to take it
 from the index, not from the HEAD.


* cc/delta-islands (2018-08-16) 7 commits
  (merged to 'next' on 2018-08-27 at cf3d7bd93f)
 + pack-objects: move 'layer' into 'struct packing_data'
 + pack-objects: move tree_depth into 'struct packing_data'
 + t5320: tests for delta islands
 + repack: add delta-islands support
 + pack-objects: add delta-islands support
 + pack-objects: refactor code into compute_layer_order()
 + Add delta-islands.{c,h}

 Lift code from GitHub to restrict delta computation so that an
 object that exists in one fork is not made into a delta against
 another object that does not appear in the same forked repository.

 Will merge to 'master'.


* md/filter-trees (2018-09-14) 7 commits
 - list-objects-filter: implement filter tree:0
 - list-objects-filter: use BUG rather than die
 - revision: mark non-user-given objects instead
 - rev-list: handle missing tree objects properly
 - list-objects: always parse trees gently
 - list-objects: refactor to process_tree_contents
 - list-objects: store common func args in struct

 The "rev-list --filter" feature learned to exclude all trees via
 "tree:0" filter.

 Will merge to 'next'.
 The test scripts need to be cleaned up.


* pk/rebase-in-c-2-basic (2018-09-06) 11 commits
 - builtin rebase: support `git rebase <upstream> <switch-to>`
 - builtin rebase: only store fully-qualified refs in `options.head_name`
 - builtin rebase: start a new rebase only if none is in progress
 - builtin rebase: support --force-rebase
 - builtin rebase: try to fast forward when possible
 - builtin rebase: require a clean worktree
 - builtin rebase: support the `verbose` and `diffstat` options
 - builtin rebase: support --quiet
 - builtin rebase: handle the pre-rebase hook and --no-verify
 - builtin rebase: support `git rebase --onto A...B`
 - builtin rebase: support --onto
 (this branch is used by js/rebase-in-c-5.5-work-with-rebase-i-in-c, pk/rebase-in-c-3-acts, pk/rebase-in-c-4-opts, pk/rebase-in-c-5-test and pk/rebase-in-c-6-final; uses pk/rebase-in-c.)


* pk/rebase-in-c-3-acts (2018-09-06) 7 commits
 - builtin rebase: stop if `git am` is in progress
 - builtin rebase: actions require a rebase in progress
 - builtin rebase: support --edit-todo and --show-current-patch
 - builtin rebase: support --quit
 - builtin rebase: support --abort
 - builtin rebase: support --skip
 - builtin rebase: support --continue
 (this branch is used by js/rebase-in-c-5.5-work-with-rebase-i-in-c, pk/rebase-in-c-4-opts, pk/rebase-in-c-5-test and pk/rebase-in-c-6-final; uses pk/rebase-in-c and pk/rebase-in-c-2-basic.)


* pk/rebase-in-c-4-opts (2018-09-06) 18 commits
 - builtin rebase: support --root
 - builtin rebase: add support for custom merge strategies
 - builtin rebase: support `fork-point` option
 - merge-base --fork-point: extract libified function
 - builtin rebase: support --rebase-merges[=[no-]rebase-cousins]
 - builtin rebase: support `--allow-empty-message` option
 - builtin rebase: support `--exec`
 - builtin rebase: support `--autostash` option
 - builtin rebase: support `-C` and `--whitespace=<type>`
 - builtin rebase: support `--gpg-sign` option
 - builtin rebase: support `--autosquash`
 - builtin rebase: support `keep-empty` option
 - builtin rebase: support `ignore-date` option
 - builtin rebase: support `ignore-whitespace` option
 - builtin rebase: support --committer-date-is-author-date
 - builtin rebase: support --rerere-autoupdate
 - builtin rebase: support --signoff
 - builtin rebase: allow selecting the rebase "backend"
 (this branch is used by js/rebase-in-c-5.5-work-with-rebase-i-in-c, pk/rebase-in-c-5-test and pk/rebase-in-c-6-final; uses pk/rebase-in-c, pk/rebase-in-c-2-basic and pk/rebase-in-c-3-acts.)


* pk/rebase-in-c-5-test (2018-09-06) 6 commits
 - builtin rebase: error out on incompatible option/mode combinations
 - builtin rebase: use no-op editor when interactive is "implied"
 - builtin rebase: show progress when connected to a terminal
 - builtin rebase: fast-forward to onto if it is a proper descendant
 - builtin rebase: optionally pass custom reflogs to reset_head()
 - builtin rebase: optionally auto-detect the upstream
 (this branch is used by js/rebase-in-c-5.5-work-with-rebase-i-in-c and pk/rebase-in-c-6-final; uses pk/rebase-in-c, pk/rebase-in-c-2-basic, pk/rebase-in-c-3-acts and pk/rebase-in-c-4-opts.)


* pk/rebase-in-c-6-final (2018-09-06) 1 commit
 - rebase: default to using the builtin rebase
 (this branch uses ag/rebase-i-in-c, js/rebase-in-c-5.5-work-with-rebase-i-in-c, pk/rebase-in-c, pk/rebase-in-c-2-basic, pk/rebase-in-c-3-acts, pk/rebase-in-c-4-opts and pk/rebase-in-c-5-test.)


* ps/stash-in-c (2018-08-31) 20 commits
 - stash: replace all `write-tree` child processes with API calls
 - stash: optimize `get_untracked_files()` and `check_changes()`
 - stash: convert `stash--helper.c` into `stash.c`
 - stash: convert save to builtin
 - stash: make push -q quiet
 - stash: convert push to builtin
 - stash: convert create to builtin
 - stash: convert store to builtin
 - stash: mention options in `show` synopsis
 - stash: convert show to builtin
 - stash: convert list to builtin
 - stash: convert pop to builtin
 - stash: convert branch to builtin
 - stash: convert drop and clear to builtin
 - stash: convert apply to builtin
 - stash: add tests for `git stash show` config
 - stash: rename test cases to be more descriptive
 - stash: update test cases conform to coding guidelines
 - stash: improve option parsing test coverage
 - sha1-name.c: add `get_oidf()` which acts like `get_oid()`


* nd/clone-case-smashing-warning (2018-08-17) 1 commit
  (merged to 'next' on 2018-08-22 at eedae40a8d)
 + clone: report duplicate entries on case-insensitive filesystems

 Running "git clone" against a project that contain two files with
 pathnames that differ only in cases on a case insensitive
 filesystem would result in one of the files lost because the
 underlying filesystem is incapable of holding both at the same
 time.  An attempt is made to detect such a case and warn.

 Will merge to 'master'.


* nd/unpack-trees-with-cache-tree (2018-08-27) 8 commits
  (merged to 'next' on 2018-08-27 at b1d841d034)
 + Document update for nd/unpack-trees-with-cache-tree
  (merged to 'next' on 2018-08-22 at 138b902673)
 + cache-tree: verify valid cache-tree in the test suite
 + unpack-trees: add missing cache invalidation
 + unpack-trees: reuse (still valid) cache-tree from src_index
 + unpack-trees: reduce malloc in cache-tree walk
 + unpack-trees: optimize walking same trees with cache-tree
 + unpack-trees: add performance tracing
 + trace.h: support nested performance tracing

 The unpack_trees() API used in checking out a branch and merging
 walks one or more trees along with the index.  When the cache-tree
 in the index tells us that we are walking a tree whose flattened
 contents is known (i.e. matches a span in the index), as linearly
 scanning a span in the index is much more efficient than having to
 open tree objects recursively and listing their entries, the walk
 can be optimized, which is done in this topic.

 Will merge to 'master'.


* sb/range-diff-colors (2018-08-20) 11 commits
  (merged to 'next' on 2018-08-22 at eb7ed4fca3)
 + range-diff: indent special lines as context
 + range-diff: make use of different output indicators
 + diff.c: add --output-indicator-{new, old, context}
 + diff.c: rewrite emit_line_0 more understandably
 + diff.c: omit check for line prefix in emit_line_0
 + diff: use emit_line_0 once per line
 + diff.c: add set_sign to emit_line_0
 + diff.c: reorder arguments for emit_line_ws_markup
 + diff.c: simplify caller of emit_line_0
 + t3206: add color test for range-diff --dual-color
 + test_decode_color: understand FAINT and ITALIC

 The color output support for recently introduced "range-diff"
 command got tweaked a bit.

 Will merge to 'master'.


* sg/t1404-update-ref-test-timeout (2018-08-01) 1 commit
  (merged to 'next' on 2018-08-22 at f3cd37b5ea)
 + t1404: increase core.packedRefsTimeout to avoid occasional test failure

 An attempt to unflake a test a bit.

 Will merge to 'master'.


* pw/add-p-select (2018-07-26) 4 commits
 - add -p: optimize line selection for short hunks
 - add -p: allow line selection to be inverted
 - add -p: select modified lines correctly
 - add -p: select individual hunk lines

 "git add -p" interactive interface learned to let users choose
 individual added/removed lines to be used in the operation, instead
 of accepting or rejecting a whole hunk.

 Will hold.
 cf. <d622a95b-7302-43d4-4ec9-b2cf3388c653@talktalk.net>
 I found the feature to be hard to explain, and may result in more
 end-user complaints, but let's see.


* ds/commit-graph-with-grafts (2018-08-21) 8 commits
 - commit-graph: close_commit_graph before shallow walk
 - commit-graph: not compatible with uninitialized repo
 - commit-graph: not compatible with grafts
 - commit-graph: not compatible with replace objects
 - test-repository: properly init repo
 - commit-graph: update design document
 - refs.c: upgrade for_each_replace_ref to be a each_repo_ref_fn callback
 - refs.c: migrate internal ref iteration to pass thru repository argument

 The recently introduced commit-graph auxiliary data is incompatible
 with mechanisms such as replace & grafts that "breaks" immutable
 nature of the object reference relationship.  Disable optimizations
 based on its use (and updating existing commit-graph) when these
 incompatible features are in use in the repository.

 Replaced with a newer version.


* ds/reachable (2018-08-28) 19 commits
  (merged to 'next' on 2018-08-28 at b1634b371d)
 + commit-reach: correct accidental #include of C file
  (merged to 'next' on 2018-08-22 at 17f3275afb)
 + commit-reach: use can_all_from_reach
 + commit-reach: make can_all_from_reach... linear
 + commit-reach: replace ref_newer logic
 + test-reach: test commit_contains
 + test-reach: test can_all_from_reach_with_flags
 + test-reach: test reduce_heads
 + test-reach: test get_merge_bases_many
 + test-reach: test is_descendant_of
 + test-reach: test in_merge_bases
 + test-reach: create new test tool for ref_newer
 + commit-reach: move can_all_from_reach_with_flags
 + upload-pack: generalize commit date cutoff
 + upload-pack: refactor ok_to_give_up()
 + upload-pack: make reachable() more generic
 + commit-reach: move commit_contains from ref-filter
 + commit-reach: move ref_newer from remote.c
 + commit.h: remove method declarations
 + commit-reach: move walk methods from commit.c

 The code for computing history reachability has been shuffled,
 obtained a bunch of new tests to cover them, and then being
 improved.

 Will merge to 'master'.


* es/format-patch-interdiff (2018-07-23) 6 commits
  (merged to 'next' on 2018-08-31 at 63927e0227)
 + format-patch: allow --interdiff to apply to a lone-patch
 + log-tree: show_log: make commentary block delimiting reusable
 + interdiff: teach show_interdiff() to indent interdiff
 + format-patch: teach --interdiff to respect -v/--reroll-count
 + format-patch: add --interdiff option to embed diff in cover letter
 + format-patch: allow additional generated content in make_cover_letter()
 (this branch is used by ds/format-patch-range-diff-test and es/format-patch-rangediff.)

 "git format-patch" learned a new "--interdiff" option to explain
 the difference between this version and the previous atttempt in
 the cover letter (or after the tree-dashes as a comment).

 Will merge to 'master'.


* es/format-patch-rangediff (2018-08-14) 10 commits
  (merged to 'next' on 2018-08-31 at 65627afece)
 + format-patch: allow --range-diff to apply to a lone-patch
 + format-patch: add --creation-factor tweak for --range-diff
 + format-patch: teach --range-diff to respect -v/--reroll-count
 + format-patch: extend --range-diff to accept revision range
 + format-patch: add --range-diff option to embed diff in cover letter
 + range-diff: relieve callers of low-level configuration burden
 + range-diff: publish default creation factor
 + range-diff: respect diff_option.file rather than assuming 'stdout'
 + Merge branch 'es/format-patch-interdiff' into es/format-patch-rangediff
 + Merge branch 'js/range-diff' into es/format-patch-rangediff
 (this branch is used by ds/format-patch-range-diff-test; uses es/format-patch-interdiff.)

 "git format-patch" learned a new "--range-diff" option to explain
 the difference between this version and the previous attempt in
 the cover letter (or after the tree-dashes as a comment).

 Will merge to 'master'.


* jn/gc-auto (2018-07-17) 3 commits
 - gc: do not return error for prior errors in daemonized mode
 - gc: exit with status 128 on failure
 - gc: improve handling of errors reading gc.log

 "gc --auto" ended up calling exit(-1) upon error, which has been
 corrected to use exit(1).  Also the error reporting behaviour when
 daemonized has been updated to exit with zero status when stopping
 due to a previously discovered error (which implies there is no
 point running gc to improve the situation); we used to exit with
 failure in such a case.

 What's the donness of this one?
 cf. <20180717201348.GD26218@sigill.intra.peff.net>


* sb/submodule-update-in-c (2018-08-14) 7 commits
  (merged to 'next' on 2018-08-17 at 23c81e5ff7)
 + submodule--helper: introduce new update-module-mode helper
 + submodule--helper: replace connect-gitdir-workingtree by ensure-core-worktree
 + builtin/submodule--helper: factor out method to update a single submodule
 + builtin/submodule--helper: store update_clone information in a struct
 + builtin/submodule--helper: factor out submodule updating
 + git-submodule.sh: rename unused variables
 + git-submodule.sh: align error reporting for update mode to use path

 "git submodule update" is getting rewritten piece-by-piece into C.

 Will merge to 'master'.


* tg/rerere (2018-08-06) 11 commits
  (merged to 'next' on 2018-08-17 at 919a958cdc)
 + rerere: recalculate conflict ID when unresolved conflict is committed
 + rerere: teach rerere to handle nested conflicts
 + rerere: return strbuf from handle path
 + rerere: factor out handle_conflict function
 + rerere: only return whether a path has conflicts or not
 + rerere: fix crash with files rerere can't handle
 + rerere: add documentation for conflict normalization
 + rerere: mark strings for translation
 + rerere: wrap paths in output in sq
 + rerere: lowercase error messages
 + rerere: unify error messages when read_cache fails
 (this branch is used by tg/rerere-doc-updates.)

 Fixes to "git rerere" corner cases, especially when conflict
 markers cannot be parsed in the file.

 Will merge to 'master'.


* ag/rebase-i-in-c (2018-08-29) 20 commits
 - rebase -i: move rebase--helper modes to rebase--interactive
 - rebase -i: remove git-rebase--interactive.sh
 - rebase--interactive2: rewrite the submodes of interactive rebase in C
 - rebase -i: implement the main part of interactive rebase as a builtin
 - rebase -i: rewrite init_basic_state() in C
 - rebase -i: rewrite write_basic_state() in C
 - rebase -i: rewrite the rest of init_revisions_and_shortrevisions() in C
 - rebase -i: implement the logic to initialize $revisions in C
 - rebase -i: remove unused modes and functions
 - rebase -i: rewrite complete_action() in C
 - t3404: todo list with commented-out commands only aborts
 - sequencer: change the way skip_unnecessary_picks() returns its result
 - sequencer: refactor append_todo_help() to write its message to a buffer
 - rebase -i: rewrite checkout_onto() in C
 - rebase -i: rewrite setup_reflog_action() in C
 - sequencer: add a new function to silence a command, except if it fails
 - rebase -i: rewrite the edit-todo functionality in C
 - editor: add a function to launch the sequence editor
 - rebase -i: rewrite append_todo_help() in C
 - sequencer: make three functions and an enum from sequencer.c public
 (this branch is used by js/rebase-in-c-5.5-work-with-rebase-i-in-c and pk/rebase-in-c-6-final.)

 Rewrite of the remaining "rebase -i" machinery in C.


* lt/date-human (2018-07-09) 1 commit
 - Add 'human' date format

 A new date format "--date=human" that morphs its output depending
 on how far the time is from the current time has been introduced.
 "--date=auto" can be used to use this new format when the output is
 goint to the pager or to the terminal and otherwise the default
 format.


* pk/rebase-in-c (2018-08-06) 3 commits
 - builtin/rebase: support running "git rebase <upstream>"
 - rebase: refactor common shell functions into their own file
 - rebase: start implementing it as a builtin
 (this branch is used by js/rebase-in-c-5.5-work-with-rebase-i-in-c, pk/rebase-in-c-2-basic, pk/rebase-in-c-3-acts, pk/rebase-in-c-4-opts, pk/rebase-in-c-5-test and pk/rebase-in-c-6-final.)

 Rewrite of the "rebase" machinery in C.


* jk/branch-l-1-repurpose (2018-08-30) 2 commits
  (merged to 'next' on 2018-08-31 at cfa73bbfcb)
 + doc/git-branch: remove obsolete "-l" references
  (merged to 'next' on 2018-08-08 at d2a08dd08e)
 + branch: make "-l" a synonym for "--list"

 Updated plan to repurpose the "-l" option to "git branch".

 Will merge to 'master'.


* ds/multi-pack-index (2018-08-20) 33 commits
  (merged to 'next' on 2018-08-21 at d15e8cadd4)
 + pack-objects: consider packs in multi-pack-index
 + midx: test a few commands that use get_all_packs
 + treewide: use get_all_packs
 + packfile: add all_packs list
 + midx: fix bug that skips midx with alternates
 + midx: stop reporting garbage
 + midx: mark bad packed objects
 + multi-pack-index: store local property
 + multi-pack-index: provide more helpful usage info
 + Sync 'ds/multi-pack-index' to v2.19.0-rc0
  (merged to 'next' on 2018-08-08 at 1a56c52967)
 + midx: clear midx on repack
 + packfile: skip loading index if in multi-pack-index
 + midx: prevent duplicate packfile loads
 + midx: use midx in approximate_object_count
 + midx: use existing midx when writing new one
 + midx: use midx in abbreviation calculations
 + midx: read objects from multi-pack-index
 + config: create core.multiPackIndex setting
 + midx: write object offsets
 + midx: write object id fanout chunk
 + midx: write object ids in a chunk
 + midx: sort and deduplicate objects from packfiles
 + midx: read pack names into array
 + multi-pack-index: write pack names in chunk
 + multi-pack-index: read packfile list
 + packfile: generalize pack directory list
 + t5319: expand test data
 + multi-pack-index: load into memory
 + midx: write header information to lockfile
 + multi-pack-index: add 'write' verb
 + multi-pack-index: add builtin
 + multi-pack-index: add format details
 + multi-pack-index: add design document
 (this branch is used by ds/multi-pack-verify.)

 When there are too many packfiles in a repository (which is not
 recommended), looking up an object in these would require
 consulting many pack .idx files; a new mechanism to have a single
 file that consolidates all of these .idx files is introduced.

 Will merge to 'master'.

--------------------------------------------------
[Discarded]

* am/sequencer-author-script-fix (2018-07-18) 1 commit
 . sequencer.c: terminate the last line of author-script properly

 The author-script that records the author information created by
 the sequencer machinery lacked the closing single quote on the last
 entry.

 Superseded by another topic.


* jc/push-cas-opt-comment (2018-08-01) 1 commit
 . push: comment on a funny unbalanced option help

 Code clarification.

 Superseded by another topic.


* cc/remote-odb (2018-08-02) 9 commits
 . Documentation/config: add odb.<name>.promisorRemote
 . t0410: test fetching from many promisor remotes
 . Use odb.origin.partialclonefilter instead of core.partialclonefilter
 . Use remote_odb_get_direct() and has_remote_odb()
 . remote-odb: add remote_odb_reinit()
 . remote-odb: implement remote_odb_get_many_direct()
 . remote-odb: implement remote_odb_get_direct()
 . Add initial remote odb support
 . fetch-object: make functions return an error code

 Implement lazy fetches of missing objects to complement the
 experimental partial clone feature.

 Ejected; seems to break existing repositories that use partialclone
 repository extension.

 I haven't seen much interest in this topic on list.  What's the
 doneness of this thing?

 I do not particularly mind adding code to support a niche feature
 as long as it is cleanly made and it is clear that the feature
 won't negatively affect those who do not use it, so a review from
 that point of view may also be appropriate.


* jh/structured-logging (2018-08-28) 26 commits
 . SQUASH??? spatch fix
 . structured-logging: add config data facility
 . structured-logging: t0420 tests for interacitve child_summary
 . structured-logging: t0420 tests for child process detail events
 . structured-logging: add child process classification
 . structured-logging: add detail-events for child processes
 . structured-logging: add structured logging to remote-curl
 . structured-logging: t0420 tests for aux-data
 . structured-logging: add aux-data for size of sparse-checkout file
 . structured-logging: add aux-data for index size
 . structured-logging: add aux-data facility
 . structured-logging: t0420 tests for timers
 . structured-logging: add timer around preload_index
 . structured-logging: add timer around wt-status functions
 . structured-logging: add timer around do_write_index
 . structured-logging: add timer around do_read_index
 . structured-logging: add timer facility
 . structured-logging: add detail-event for lazy_init_name_hash
 . structured-logging: add detail-event facility
 . structured-logging: t0420 basic tests
 . structured-logging: set sub_command field for checkout command
 . structured-logging: set sub_command field for branch command
 . structured-logging: add session-id to log events
 . structured-logging: add structured logging framework
 . structured-logging: add STRUCTURED_LOGGING=1 to Makefile
 . structured-logging: design document

 Being rerolled with an updated tracing API.


* av/fsmonitor-updates (2018-01-04) 6 commits
 . fsmonitor: use fsmonitor data in `git diff`
 . fsmonitor: remove debugging lines from t/t7519-status-fsmonitor.sh
 . fsmonitor: make output of test-dump-fsmonitor more concise
 . fsmonitor: update helper tool, now that flags are filled later
 . fsmonitor: stop inline'ing mark_fsmonitor_valid / _invalid
 . dir.c: update comments to match argument name

 Code clean-up on fsmonitor integration, plus optional utilization
 of the fsmonitor data in diff-files.

 Tired of waiting for an update.
 cf. <alpine.DEB.2.21.1.1801042335130.32@MININT-6BKU6QN.europe.corp.microsoft.com>

 Also there were backward incompatible API changes brought by other
 topics in flight; having to keep up with those disruptive changes
 is not worth the maintenance effort for a stale topic.


* jc/rebase-in-c-9-fixes (2018-09-04) 1 commit
 . rebase: re-add forgotten -k that stands for --keep-empty
 (this branch uses ag/rebase-i-in-c, js/rebase-in-c-5.5-work-with-rebase-i-in-c, pk/rebase-in-c, pk/rebase-in-c-2-basic, pk/rebase-in-c-3-acts, pk/rebase-in-c-4-opts, pk/rebase-in-c-5-test and pk/rebase-in-c-6-final.)

 The fix has been rolled into the original topic that introduced the
 issue, hence this topic is no longer necessary.


* jn/http-backend-content-length (2018-09-11) 2 commits
 . SQUASH???? perhaps after reflowing
 . http-backend: treat empty CONTENT_LENGTH as zero


* nd/optim-reading-index-v4 (2018-09-04) 1 commit
 . read-cache.c: optimize reading index format v4

 The v4 format of the index file uses prefix compression to store
 the pathnames to save file size.  The codepath to read such a file
 has been optimized.

 Absorbed in bp/read-cache-parallel topic.

Re: What's cooking in git.git (Sep 2018, #03; Fri, 14)

[Antonio Ospite]

On Fri, 14 Sep 2018 14:56:36 -0700
Junio C Hamano <gitster@pobox.com> wrote:

[...]
> * ao/submodule-wo-gitmodules-checked-out (2018-08-14) 7 commits
>  - submodule: support reading .gitmodules even when it's not checked out
>  - t7506: clean up .gitmodules properly before setting up new scenario
>  - submodule: use the 'submodule--helper config' command
>  - submodule--helper: add a new 'config' subcommand
>  - t7411: be nicer to future tests and really clean things up
>  - submodule: factor out a config_set_in_gitmodules_file_gently function
>  - submodule: add a print_config_from_gitmodules() helper
> 
>  The submodule support has been updated to read from the blob at
>  HEAD:.gitmodules when the .gitmodules file is missing from the
>  working tree.
> 
>  Expecting a reroll.
> 
>  I find the design a bit iffy in that our usual "missing in the
>  working tree?  let's use the latest blob" fallback is to take it
>  from the index, not from the HEAD.
> 

Hi Junio,

I did send a v4 addressing this:
https://public-inbox.org/git/20180824132951.8000-1-ao2@ao2.it/

In case you missed it I might as well send a v5 with some minor
cosmetic touches suggested by Ævar:
https://public-inbox.org/git/87wosfesxl.fsf@evledraar.gmail.com/

Ciao,
   Antonio

-- 
Antonio Ospite
https://ao2.it
https://twitter.com/ao2it

A: Because it messes up the order in which people normally read text.
   See http://en.wikipedia.org/wiki/Posting_style
Q: Why is top-posting such a bad thing?

Re: What's cooking in git.git (Sep 2018, #03; Fri, 14)

[Duy Nguyen]

On Fri, Sep 14, 2018 at 11:56 PM Junio C Hamano <gitster@pobox.com> wrote:
> * jn/gc-auto (2018-07-17) 3 commits
>  - gc: do not return error for prior errors in daemonized mode
>  - gc: exit with status 128 on failure
>  - gc: improve handling of errors reading gc.log
>
>  "gc --auto" ended up calling exit(-1) upon error, which has been
>  corrected to use exit(1).  Also the error reporting behaviour when
>  daemonized has been updated to exit with zero status when stopping
>  due to a previously discovered error (which implies there is no
>  point running gc to improve the situation); we used to exit with
>  failure in such a case.
>
>  What's the donness of this one?
>  cf. <20180717201348.GD26218@sigill.intra.peff.net>

This topic has stayed in 'pu' for a long time. I thought it was
concluded that this was a good change? Jeff, Jonathan?
-- 
Duy

Re: What's cooking in git.git (Sep 2018, #03; Fri, 14)

[brian m. carlson]

[-- Attachment #1: Type: text/plain, Size: 795 bytes --]

On Fri, Sep 14, 2018 at 02:56:36PM -0700, Junio C Hamano wrote:
> * bc/hash-independent-tests (2018-09-13) 12 commits
>  - t5318: use test_oid for HASH_LEN
>  - t1407: make hash size independent
>  - t1406: make hash-size independent
>  - t1405: make hash size independent
>  - t1400: switch hard-coded object ID to variable
>  - t1006: make hash size independent
>  - t0064: make hash size independent
>  - t0027: make hash size independent

Could you drop this particular patch (the t0027 fix)?  I found an issue
with it that I'd like to address, and it would be easier for me to just
send a completely fixed version in the future since it doesn't have any
dependencies with the rest of the series.
-- 
brian m. carlson: Houston, Texas, US
OpenPGP: https://keybase.io/bk2204

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 868 bytes --]

Re: What's cooking in git.git (Sep 2018, #03; Fri, 14)

[Ævar Arnfjörð Bjarmason]

On Sun, Sep 16 2018, Duy Nguyen wrote:

> On Fri, Sep 14, 2018 at 11:56 PM Junio C Hamano <gitster@pobox.com> wrote:
>> * jn/gc-auto (2018-07-17) 3 commits
>>  - gc: do not return error for prior errors in daemonized mode
>>  - gc: exit with status 128 on failure
>>  - gc: improve handling of errors reading gc.log
>>
>>  "gc --auto" ended up calling exit(-1) upon error, which has been
>>  corrected to use exit(1).  Also the error reporting behaviour when
>>  daemonized has been updated to exit with zero status when stopping
>>  due to a previously discovered error (which implies there is no
>>  point running gc to improve the situation); we used to exit with
>>  failure in such a case.
>>
>>  What's the donness of this one?
>>  cf. <20180717201348.GD26218@sigill.intra.peff.net>
>
> This topic has stayed in 'pu' for a long time. I thought it was
> concluded that this was a good change? Jeff, Jonathan?

There's still outstanding feedback on it, including my:
https://public-inbox.org/git/878t69dgvx.fsf@evledraar.gmail.com/

I think it would be great to merge [12]/3 down, and have a re-submission
of 3/3 stand-alone, once that's merged down, so we can untangle
obviously correct bugfixes ([12]/3) with meaningful changes in exit code
behavior (3/3).

Re: What's cooking in git.git (Sep 2018, #03; Fri, 14)

[Jeff King]

On Sun, Sep 16, 2018 at 08:39:03AM +0200, Duy Nguyen wrote:

> On Fri, Sep 14, 2018 at 11:56 PM Junio C Hamano <gitster@pobox.com> wrote:
> > * jn/gc-auto (2018-07-17) 3 commits
> >  - gc: do not return error for prior errors in daemonized mode
> >  - gc: exit with status 128 on failure
> >  - gc: improve handling of errors reading gc.log
> >
> >  "gc --auto" ended up calling exit(-1) upon error, which has been
> >  corrected to use exit(1).  Also the error reporting behaviour when
> >  daemonized has been updated to exit with zero status when stopping
> >  due to a previously discovered error (which implies there is no
> >  point running gc to improve the situation); we used to exit with
> >  failure in such a case.
> >
> >  What's the donness of this one?
> >  cf. <20180717201348.GD26218@sigill.intra.peff.net>
> 
> This topic has stayed in 'pu' for a long time. I thought it was
> concluded that this was a good change? Jeff, Jonathan?

I read over the thread again. I don't think I actually have any
complaints about the patches as-is. There was some discussion from Junio
and Ævar about the third one. I don't have a strong opinion. My
experience has been that "gc --auto" is garbage anyway on the server
side, but I think Ævar's experience is that it's reasonable for small to
medium sites (which seems plausible to me).

The message-id quoted there is my "this looks good". I mentioned a few
possible nits, but I think it would be OK with or without them
addressed.

-Peff

Re: What's cooking in git.git (Sep 2018, #03; Fri, 14)

[Junio C Hamano]

"brian m. carlson" <sandals@crustytoothpaste.net> writes:

> On Fri, Sep 14, 2018 at 02:56:36PM -0700, Junio C Hamano wrote:
>> * bc/hash-independent-tests (2018-09-13) 12 commits
>>  - t5318: use test_oid for HASH_LEN
>>  - t1407: make hash size independent
>>  - t1406: make hash-size independent
>>  - t1405: make hash size independent
>>  - t1400: switch hard-coded object ID to variable
>>  - t1006: make hash size independent
>>  - t0064: make hash size independent
>>  - t0027: make hash size independent
>
> Could you drop this particular patch (the t0027 fix)?  I found an issue
> with it that I'd like to address, and it would be easier for me to just
> send a completely fixed version in the future since it doesn't have any
> dependencies with the rest of the series.

Thanks, will do.

Re: What's cooking in git.git (Sep 2018, #03; Fri, 14)

[Junio C Hamano]

Antonio Ospite <ao2@ao2.it> writes:

> I did send a v4 addressing this:
> https://public-inbox.org/git/20180824132951.8000-1-ao2@ao2.it/
>
> In case you missed it I might as well send a v5 with some minor
> cosmetic touches suggested by Ævar:
> https://public-inbox.org/git/87wosfesxl.fsf@evledraar.gmail.com/

Thanks, will take a look.

Re: What's cooking in git.git (Sep 2018, #03; Fri, 14)

[Junio C Hamano]

Jeff King <peff@peff.net> writes:

>> >  What's the donness of this one?
>> >  cf. <20180717201348.GD26218@sigill.intra.peff.net>
>> 
>> This topic has stayed in 'pu' for a long time. I thought it was
>> concluded that this was a good change? Jeff, Jonathan?
>
> I read over the thread again. I don't think I actually have any
> complaints about the patches as-is. There was some discussion from Junio
> and Ævar about the third one. I don't have a strong opinion. My
> experience has been that "gc --auto" is garbage anyway on the server
> side, but I think Ævar's experience is that it's reasonable for small to
> medium sites (which seems plausible to me).
>
> The message-id quoted there is my "this looks good". I mentioned a few
> possible nits, but I think it would be OK with or without them
> addressed.

That matches my reading of your position.  I tend to agree with
Ævar's recommendation to postpone 3/3 and use 1 & 2 for now.

By the way, people shouldn't read too much into the messages
referred to in "What's cooking"; they are not "these messages point
out all the issues that block the topic".  It's just "I am aware of
this thread, which readers would hopefully find a good starting
point to form their opinions".

Thanks.

Re: What's cooking in git.git (Sep 2018, #03; Fri, 14)

[Jeff King]

On Mon, Sep 17, 2018 at 10:51:35AM -0700, Junio C Hamano wrote:

> Jeff King <peff@peff.net> writes:
> 
> >> >  What's the donness of this one?
> >> >  cf. <20180717201348.GD26218@sigill.intra.peff.net>
> >> 
> >> This topic has stayed in 'pu' for a long time. I thought it was
> >> concluded that this was a good change? Jeff, Jonathan?
> >
> > I read over the thread again. I don't think I actually have any
> > complaints about the patches as-is. There was some discussion from Junio
> > and Ævar about the third one. I don't have a strong opinion. My
> > experience has been that "gc --auto" is garbage anyway on the server
> > side, but I think Ævar's experience is that it's reasonable for small to
> > medium sites (which seems plausible to me).
> >
> > The message-id quoted there is my "this looks good". I mentioned a few
> > possible nits, but I think it would be OK with or without them
> > addressed.
> 
> That matches my reading of your position.  I tend to agree with
> Ævar's recommendation to postpone 3/3 and use 1 & 2 for now.

Let me inject some more uncertainty, then. ;)

If we are not going to do 3/3, then should 2/3 simply avoid passing "-1"
back via return from main? I guess I don't have a strong opinion, but
one of the things I noted was that we converted those die() calls
introduced in 2/3 back into returns in 3/3. Do we want to leave it in
the state where we are calling die() a lot more?

-Peff

Re: What's cooking in git.git (Sep 2018, #03; Fri, 14)

[Jonathan Nieder]

Jeff King wrote:

> Let me inject some more uncertainty, then. ;)
>
> If we are not going to do 3/3, then should 2/3 simply avoid passing "-1"
> back via return from main? I guess I don't have a strong opinion, but
> one of the things I noted was that we converted those die() calls
> introduced in 2/3 back into returns in 3/3. Do we want to leave it in
> the state where we are calling die() a lot more?

Would you mind replying in the patch thread instead of this what's
cooking email?

That way, I can understand your suggestion better in context, I can
find it more easily later, I would feel less bad about adding noise by
replying, etc.

Thanks,
Jonathan

Re: What's cooking in git.git (Sep 2018, #03; Fri, 14)

[Derrick Stolee]

On 9/14/2018 5:56 PM, Junio C Hamano wrote:
> * ds/format-commit-graph-docs (2018-08-21) 2 commits
>   - commit-graph.txt: improve formatting for asciidoc
>   - Docs: Add commit-graph tech docs to Makefile
>
>   Design docs for the commit-graph machinery is now made into HTML as
>   well as text.
>
>   Will discard.
>   I am inclined to drop these, as I do not see much clarity in HTML
>   output over the text source.  Opinions?

Discarding is fine. I originally created it because I thought we were 
supposed to mark all documents for HTML generation. You're right, the 
HTML doesn't add anything.

> * ds/commit-graph-with-grafts (2018-08-21) 8 commits
>   - commit-graph: close_commit_graph before shallow walk
>   - commit-graph: not compatible with uninitialized repo
>   - commit-graph: not compatible with grafts
>   - commit-graph: not compatible with replace objects
>   - test-repository: properly init repo
>   - commit-graph: update design document
>   - refs.c: upgrade for_each_replace_ref to be a each_repo_ref_fn callback
>   - refs.c: migrate internal ref iteration to pass thru repository argument
>
>   The recently introduced commit-graph auxiliary data is incompatible
>   with mechanisms such as replace & grafts that "breaks" immutable
>   nature of the object reference relationship.  Disable optimizations
>   based on its use (and updating existing commit-graph) when these
>   incompatible features are in use in the repository.
>
>   Replaced with a newer version.

I think this has been the same note for a few weeks now. What does it 
mean? Did I send a new version out that you haven't picked up?

Also, my "git log --topo-order" series was never picked up, but I see it 
has conflicts with 'next' now, so I'll rebase to resolve conflicts and 
send a v2.

Thanks,
-Stolee

Re: What's cooking in git.git (Sep 2018, #03; Fri, 14)

[Junio C Hamano]

Derrick Stolee <stolee@gmail.com> writes:

>>   Replaced with a newer version.
>
> I think this has been the same note for a few weeks now. What does it
> mean? Did I send a new version out that you haven't picked up?

It just means that I didn't bother to look at the current status
message for the topic that had no activity X-<.

> Also, my "git log --topo-order" series was never picked up, but I see
> it has conflicts with 'next' now, so I'll rebase to resolve conflicts
> and send a v2.

Good idea, as I do not recall anything about that series.

Re: What's cooking in git.git (Sep 2018, #03; Fri, 14)

[Junio C Hamano]

Junio C Hamano <gitster@pobox.com> writes:

> The tip of 'next' hasn't been rewound yet, but the states of many
> topics that were marked to "Cook in 'next'" read "Will merge to
> 'master'" now.  I'll probably start merging a handful of topics that
> have been in 'next' down to 'master' tonight (they appear at the
> bottom of "log --first-parent master..pu" output).

A bunch of topics have been merged to 'master' and also to 'next'
with today's update.  The tip of 'next' hasn't been rewound.  I am
hoping that we can thin down the difference between 'master' and
'next' by merging a bit more topic from 'next' to 'master' without
adding more topics to 'next' for a week or so, and then rebuild
'next' at the end of that period, and then start to accept more
topics to 'next'.

[PATCH] pack-objects: handle island check for "external" delta base

[Jeff King]

On Fri, Sep 14, 2018 at 02:56:36PM -0700, Junio C Hamano wrote:

> * cc/delta-islands (2018-08-16) 7 commits
>   (merged to 'next' on 2018-08-27 at cf3d7bd93f)
>  + pack-objects: move 'layer' into 'struct packing_data'
>  + pack-objects: move tree_depth into 'struct packing_data'
>  + t5320: tests for delta islands
>  + repack: add delta-islands support
>  + pack-objects: add delta-islands support
>  + pack-objects: refactor code into compute_layer_order()
>  + Add delta-islands.{c,h}
> 
>  Lift code from GitHub to restrict delta computation so that an
>  object that exists in one fork is not made into a delta against
>  another object that does not appear in the same forked repository.
> 
>  Will merge to 'master'.

This needed some conflict resolution with my pack-bitmap-reuse-delta
topic, but there's a subtle bug in the result that went to 'master'.
Details and a fix below.

As a side note, I did this same resolution myself at least twice (for my
personal build and for porting the refreshed delta-reuse series to our
GitHub build), and I wrote the exact same resolution you did both times.
So I think it was an easy mistake to make. :)

-Peff

-- >8 --
Subject: pack-objects: handle island check for "external" delta base

Two recent topics, jk/pack-delta-reuse-with-bitmap and
cc/delta-islands, can have a funny interaction. When
checking if we can reuse an on-disk delta, the first topic
allows base_entry to be NULL when we find an object that's
not in the packing list. But the latter topic introduces a
call to in_same_island(), which needs to look at
base_entry->idx.oid. When these two features are used
together, we might try to dereference a NULL base_entry.

In practice, this doesn't really happen. We'd generally only
use delta islands when packing to disk, since the whole
point is to optimize the pack for serving fetches later. And
the new delta-reuse code relies on having used reachability
bitmaps to determine the set of objects, which we would
typically only do when serving an actual fetch.

However, it is technically possible to combine these
features. And even without doing so, building with
"SANITIZE=address,undefined" will cause t5310.46 to
complain.  Even though that test does not have delta islands
enabled, we still take the address of the NULL entry to pass
to in_same_island(). That function then promptly returns
without dereferencing the value when it sees that islands
are not enabled, but it's enough to trigger a sanitizer
error.

The solution is straight-forward: when both features are
used together, we should pass the oid of the found base to
in_same_island().

This is tricky to do inside a single "if" statement. And
after the merge in f3504ea3dd (Merge branch
'cc/delta-islands', 2018-09-17), that "if" condition is
already getting pretty unwieldy. So this patch moves the
logic into a helper function, where we can easily use
multiple return paths. The result is a bit longer, but the
logic should be much easier to follow.

Signed-off-by: Jeff King <peff@peff.net>
---
 builtin/pack-objects.c | 68 ++++++++++++++++++++++++++++++++----------
 1 file changed, 52 insertions(+), 16 deletions(-)

diff --git a/builtin/pack-objects.c b/builtin/pack-objects.c
index 5041818ddf..27cb674124 100644
--- a/builtin/pack-objects.c
+++ b/builtin/pack-objects.c
@@ -1470,6 +1470,57 @@ static void cleanup_preferred_base(void)
 	done_pbase_paths_num = done_pbase_paths_alloc = 0;
 }
 
+/*
+ * Return 1 iff the object specified by "delta" can be sent
+ * literally as a delta against the base in "base_sha1". If
+ * so, then *base_out will point to the entry in our packing
+ * list, or NULL if we must use the external-base list.
+ *
+ * Depth value does not matter - find_deltas() will
+ * never consider reused delta as the base object to
+ * deltify other objects against, in order to avoid
+ * circular deltas.
+ */
+static int can_reuse_delta(const unsigned char *base_sha1,
+			   struct object_entry *delta,
+			   struct object_entry **base_out)
+{
+	struct object_entry *base;
+
+	if (!base_sha1)
+		return 0;
+
+	/*
+	 * First see if we're already sending the base (or it's explicitly in
+	 * our "excluded" list.
+	 */
+	base = packlist_find(&to_pack, base_sha1, NULL);
+	if (base) {
+		if (!in_same_island(&delta->idx.oid, &base->idx.oid))
+			return 0;
+		*base_out = base;
+		return 1;
+	}
+
+	/*
+	 * Otherwise, reachability bitmaps may tell us if the receiver has it,
+	 * even if it was buried too deep in history to make it into the
+	 * packing list.
+	 */
+	if (thin && bitmap_has_sha1_in_uninteresting(bitmap_git, base_sha1)) {
+		if (use_delta_islands) {
+			struct object_id base_oid;
+			hashcpy(base_oid.hash, base_sha1);
+			if (!in_same_island(&delta->idx.oid, &base_oid))
+				return 0;
+		}
+		*base_out = NULL;
+		return 1;
+	}
+
+	return 0;
+}
+
 static void check_object(struct object_entry *entry)
 {
 	unsigned long canonical_size;
@@ -1556,22 +1607,7 @@ static void check_object(struct object_entry *entry)
 			break;
 		}
 
-		if (base_ref && (
-		    (base_entry = packlist_find(&to_pack, base_ref, NULL)) ||
-		    (thin &&
-		     bitmap_has_sha1_in_uninteresting(bitmap_git, base_ref))) &&
-		    in_same_island(&entry->idx.oid, &base_entry->idx.oid)) {
-			/*
-			 * If base_ref was set above that means we wish to
-			 * reuse delta data, and either we found that object in
-			 * the list of objects we want to pack, or it's one we
-			 * know the receiver has.
-			 *
-			 * Depth value does not matter - find_deltas() will
-			 * never consider reused delta as the base object to
-			 * deltify other objects against, in order to avoid
-			 * circular deltas.
-			 */
+		if (can_reuse_delta(base_ref, entry, &base_entry)) {
 			oe_set_type(entry, entry->in_pack_type);
 			SET_SIZE(entry, in_pack_size); /* delta size */
 			SET_DELTA_SIZE(entry, in_pack_size);
-- 
2.19.0.745.g75ede3edf3

Re: [PATCH] pack-objects: handle island check for "external" delta base

[Martin Ågren]

On Wed, 19 Sep 2018 at 05:49, Jeff King <peff@peff.net> wrote:
> This is tricky to do inside a single "if" statement. And
> after the merge in f3504ea3dd (Merge branch
> 'cc/delta-islands', 2018-09-17), that "if" condition is
> already getting pretty unwieldy. So this patch moves the
> logic into a helper function, where we can easily use
> multiple return paths. The result is a bit longer, but the
> logic should be much easier to follow.

> +static int can_reuse_delta(const unsigned char *base_sha1,
> +                          struct object_entry *delta,
> +                          struct object_entry **base_out)
> +{
> +       struct object_entry *base;
> +
> +       if (!base_sha1)
> +               return 0;

So this corresponds to "if (base_ref &&".

> +       /*
> +        * First see if we're already sending the base (or it's explicitly in
> +        * our "excluded" list.
> +        */

Missing ')'.

> +       base = packlist_find(&to_pack, base_sha1, NULL);
> +       if (base) {
> +               if (!in_same_island(&delta->idx.oid, &base->idx.oid))
> +                       return 0;

This logic matches the removed code...

> +               *base_out = base;
> +               return 1;
> +       }
> +
> +       /*
> +        * Otherwise, reachability bitmaps may tell us if the receiver has it,
> +        * even if it was buried too deep in history to make it into the
> +        * packing list.
> +        */
> +       if (thin && bitmap_has_sha1_in_uninteresting(bitmap_git, base_sha1)) {

This matches...

> +               if (use_delta_islands) {
> +                       struct object_id base_oid;
> +                       hashcpy(base_oid.hash, base_sha1);
> +                       if (!in_same_island(&delta->idx.oid, &base_oid))
> +                               return 0;

This does some extra juggling to avoid using `base->idx.oid`, which
would have been the moral equivalent of the original code, but which
won't fly since `base` is NULL.

> +               }
> +               *base_out = NULL;
> +               return 1;
> +       }
> +
> +       return 0;
> +}
> +
>  static void check_object(struct object_entry *entry)
>  {
>         unsigned long canonical_size;
> @@ -1556,22 +1607,7 @@ static void check_object(struct object_entry *entry)
>                         break;
>                 }
>
> -               if (base_ref && (
> -                   (base_entry = packlist_find(&to_pack, base_ref, NULL)) ||
> -                   (thin &&
> -                    bitmap_has_sha1_in_uninteresting(bitmap_git, base_ref))) &&
> -                   in_same_island(&entry->idx.oid, &base_entry->idx.oid)) {

Yeah, the new function looks much simpler than this. We have

  if (A && (B1 || B2) && C) {.

Knowing what to look for, it can be seen that we can -- under the right
circumstances -- have A and B2, but not B1, and try to evalute C by
dereferencing `base_entry` which will be NULL.

> +               if (can_reuse_delta(base_ref, entry, &base_entry)) {
>                         oe_set_type(entry, entry->in_pack_type);
>                         SET_SIZE(entry, in_pack_size); /* delta size */
>                         SET_DELTA_SIZE(entry, in_pack_size);

Without being at all familiar with this code, this looks sane to me.
Just had a small nit about the missing closing ')'.

Martin

Re: [PATCH] pack-objects: handle island check for "external" delta base

[Jeff King]

On Wed, Sep 19, 2018 at 08:34:05PM +0200, Martin Ågren wrote:

> > +       /*
> > +        * First see if we're already sending the base (or it's explicitly in
> > +        * our "excluded" list.
> > +        */
> 
> Missing ')'.

Oops, yes.

> > +               if (use_delta_islands) {
> > +                       struct object_id base_oid;
> > +                       hashcpy(base_oid.hash, base_sha1);
> > +                       if (!in_same_island(&delta->idx.oid, &base_oid))
> > +                               return 0;
> 
> This does some extra juggling to avoid using `base->idx.oid`, which
> would have been the moral equivalent of the original code, but which
> won't fly since `base` is NULL.

Yeah, this is the actual bug-fix.

I wasn't happy about having to write in_same_island() twice, but writing
it the other way ended up pretty nasty, too. Something like:

  struct object_id ext_oid;
  struct object_id *base_oid;

  base = packlist_find(&to_pack, base_sha1, NULL);
  if (base) {
	base_oid = &base->idx.oid;
	*base_out = base;
  }
  else if (thin && bitmap_has_sha1_in_uninteresting(bitmap_git, base_sha1)) {
	hashcpy(ext_oid.hash, base_sha1);
	base_oid = &ext_oid;
	*base_out = NULL;
  } else {
	return 0;
  }

  if (use_island_marks && !in_same_island(&delta->idx.oid, base_oid))
	return 0;

  return 1;

That's less repetitive, but I feel like it's harder to follow which
variables are valid when.

> > +               if (can_reuse_delta(base_ref, entry, &base_entry)) {
> >                         oe_set_type(entry, entry->in_pack_type);
> >                         SET_SIZE(entry, in_pack_size); /* delta size */
> >                         SET_DELTA_SIZE(entry, in_pack_size);
> 
> Without being at all familiar with this code, this looks sane to me.
> Just had a small nit about the missing closing ')'.

Thanks for the review!

-Peff