The option callback for "apply --whitespace" exits with status "1" on
error. It makes more sense for it to just return an error to
parse-options. That code will exit, too, but it will use status "129"
that is customary for option errors.

The exit() dates back to aaf6c447 (builtin/apply: make
parse_whitespace_option() return -1 instead of die()ing, 2016-08-08).
That commit gives no reason why we'd prefer the current exit status (it
looks like it was just bumping the "die" up a level in the callstack,
but did not go as far as it could have).
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

The options callback for --batch and --batch-check detects when the two
mutually incompatible options are used. But it simply returns an error
code to parse-options, meaning the program will quit without any kind of
message to the user.

Instead, let's use error() to print something and return -1. Note that
this flips the error return from 1 to -1, but negative values are more
idiomatic here (and parse-options treats them the same).
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

We do not allow "--no-message" to work now, as the option callback
returns "-1" when it sees a NULL arg. However, that will cause
parse-options to exit(129) without printing anything further, leaving
the user confused about what happened.

Instead, let's explicitly mark it as PARSE_OPT_NONEG, which will give a
useful error message (and print the usual -h output).

In theory this could be used to override an earlier "-m", but it's not
clear how it would interact with other message options (e.g., would it
also clear data read for "-F"?). Since it's already disabled and nobody
is asking for it, let's punt on that and just improve the error message.
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Running "git show-branch --no-reflog" will behave as if "--reflog" was
given with no options, which makes no sense.

In theory this option might be used to cancel an earlier "--reflog"
option, but the semantics are not clear. Let's punt on it and just
disallow the broken option.
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

We have separate parse-options entries for "numbered" and "no-numbered",
which means that we accept "--no-no-numbered". It does not behave
sensibly, though (it ignores the "unset" flag and acts like
"--no-numbered").

We could fix that, but obviously this is silly and unintentional. Let's
just disallow it.
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

If you run "git status --no-find-renames", it will behave the same as
"--find-renames", because we ignore the "unset" parameter (we see a NULL
"arg", but since the score argument is optional, we just think that the
user did not provide a score).

We already have a separate "--no-renames" to disable renames, so there's
not much point in supporting "--no-find-renames". Let's just flag it as
an error.
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Running "cat-file --no-batch" will behave as if "--batch" was given,
since the option callback does not handle the "unset" flag (likewise for
"--no-batch-check").

In theory this might be used to cancel an earlier --batch, but it's not
immediately obvious how that would interact with --batch-check. Let's
just disallow the negated form of both options.
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Running "git pack-objects --no-index-version" will segfault, since the
callback is not prepared to handle the "unset" flag.

In theory this might be used to counteract an earlier "--index-version",
or override a pack.indexversion config setting. But the semantics aren't
immediately obvious, and it's unlikely anybody wants this. Let's just
disable the broken option for now.
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Running "git ls-files --no-exclude" will currently segfault, as its
option callback does not handle the "unset" parameter.

In theory this could be used to clear the exclude list, but it is not
clear how that would interact with the other exclude options, nor is the
current code capable of clearing the list. Let's just disable the broken
option.

Note that --no-exclude-from will similarly segfault, but
--no-exclude-standard will not. It just silently does the wrong thing
(pretending as if --exclude-standard was specified).
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Running "git am --no-patch-format" will currently segfault, since it
tries to parse a NULL argument. Instead, let's have it cancel any
previous --patch-format option.
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

The options callback for "git apply --no-include" is not ready to handle
the "unset" parameter, and as a result will segfault when it adds a NULL
argument to the include list (likewise for "--no-exclude").

In theory this might be used to clear the list, but since both
"--include" and "--exclude" add to the same list, it's not immediately
obvious what the semantics should be. Let's punt on that for now and
just disallow the broken options.
Signed-off-by: NJeff King <peff@peff.net>
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Warning message fix.

* jc/http-curlver-warnings:
  http: give curl version warnings consistently

On platforms with recent cURL library, http.sslBackend configuration
variable can be used to choose a different SSL backend at runtime.
The Windows port uses this mechanism to switch between OpenSSL and
Secure Channel while talking over the HTTPS protocol.

* js/mingw-http-ssl:
  http: when using Secure Channel, ignore sslCAInfo by default
  http: add support for disabling SSL revocation checks in cURL
  http: add support for selecting SSL backends at runtime

New "--pretty=format:" placeholders %GF and %GP that show the GPG
key fingerprints have been invented.

* mg/gpg-fingerprint:
  gpg-interface.c: obtain primary key fingerprint as well
  gpg-interface.c: support getting key fingerprint via %GF format
  gpg-interface.c: use flags to determine key/signer info presence

Detect and reject a signature block that has more than one GPG
signature.

* mg/gpg-parse-tighten:
  gpg-interface.c: detect and reject multiple signatures on commits

Further clean-up of merge-recursive machinery.

* en/merge-cleanup-more:
  merge-recursive: avoid showing conflicts with merge branch before HEAD
  merge-recursive: improve auto-merging messages with path collisions

Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Favor oideq() over !oidcmp() when checking for equality.
Signed-off-by: NJunio C Hamano <gitster@pobox.com>

"git rebase -i" learned to take 'b' as the short form of 'break'
option in the todo list.

* js/rebase-i-shortopt:
  rebase -i: recognize short commands without arguments

"git rebase -i" learned a new insn, 'break', that the user can
insert in the to-do list.  Upon hitting it, the command returns
control back to the user.

* js/rebase-i-break:
  rebase -i: introduce the 'break' command
  rebase -i: clarify what happens on a failed `exec`

"git rebase" that has recently been rewritten in C had a few issues
in its "--autstash" feature, which have been corrected.

* js/rebase-autostash-fix:
  rebase --autostash: fix issue with dirty submodules
  rebase --autostash: demonstrate a problem with dirty submodules
  rebase (autostash): use an explicit OID to apply the stash
  rebase (autostash): store the full OID in <state-dir>/autostash
  rebase (autostash): avoid duplicate call to state_dir_path()

Build fix for a topic in flight.

* cb/printf-empty-format:
  sequencer: cleanup for gcc warning in non developer mode

Typofix.

* jc/rebase-in-c-5-test-typofix:
  rebase: fix typoes in error messages

The final step of rewriting "rebase -i" in C.

* pk/rebase-in-c-6-final:
  rebase: default to using the builtin rebase

"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.

* js/rebase-in-c-5.5-work-with-rebase-i-in-c:
  builtin rebase: prepare for builtin rebase -i

Rewrite "git rebase" in C.

* pk/rebase-in-c-5-test:
  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

Rewrite "git rebase" in C.

* pk/rebase-in-c-4-opts:
  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"

Rewrite "git rebase" in C.

* pk/rebase-in-c-3-acts:
  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

Rewrite "git rebase" in C.

* pk/rebase-in-c-2-basic:
  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

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

* ag/rebase-i-in-c:
  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

Rewrite of the "rebase" machinery in C.

* pk/rebase-in-c:
  builtin/rebase: support running "git rebase <upstream>"
  rebase: refactor common shell functions into their own file
  rebase: start implementing it as a builtin

Signed-off-by: NJunio C Hamano <gitster@pobox.com>

Code clean-up that results in a small bugfix.

* jk/uploadpack-packobjectshook-fix:
  upload-pack: fix broken if/else chain in config callback

Belated documentation update to adjust to a new world order that
happened a yew years ago.

* uk/merge-subtree-doc-update:
  howto/using-merge-subtree: mention --allow-unrelated-histories

Code tightening.

* cb/compat-mmap-is-private-read-only:
  compat: make sure git_mmap is not expected to write

"git mergetool" learned to take the "--[no-]gui" option, just like
"git difftool" does.

* dl/mergetool-gui-option:
  doc: document diff/merge.guitool config keys
  completion: support `git mergetool --[no-]gui`
  mergetool: accept -g/--[no-]gui as arguments

The way DLLs are loaded on the Windows port has been improved.

* js/mingw-load-sys-dll:
  mingw: load system libraries the recommended way

The way the Windows port figures out the current directory has been
improved.

* js/mingw-getcwd:
  mingw: fix getcwd when the parent directory cannot be queried
  mingw: ensure `getcwd()` reports the correct case

Build fix.

* cb/khash-maybe-unused-function:
  khash: silence -Wunused-function for delta-islands
  commit-slabs: move MAYBE_UNUSED out

Code cleanup.

* jc/cocci-preincr:
  fsck: s/++i > 1/i++/
  cocci: simplify "if (++u > 1)" to "if (u++)"