Merge 'build-in git-mktree'

* commit '633e3556ccbc': (5835 commits)
  build-in git-mktree
  allow -t abbreviation for --track in git branch
  gitweb: Remove function prototypes (cleanup)
  Documentation: cloning to empty directory is allowed
  Clarify kind of conflict in merge-one-file helper
  git config: clarify --add and --get-color
  archive-tar.c: squelch a type mismatch warning
  Start 1.6.4 development
  Start 1.6.3.1 maintenance series.
  GIT 1.6.3
  t4029: use sh instead of bash
  t4200: convert sed expression which operates on non-text file to perl
  t4200: remove two unnecessary lines
  t/annotate-tests.sh: avoid passing a non-newline terminated file to sed
  t4118: avoid sed invocation on file without terminating newline
  t4118: add missing '&&'
  t8005: use egrep when extended regular expressions are required
  git-clean doc: the command only affects paths under $(cwd)
  improve error message in config.c
  t4018-diff-funcname: add cpp xfuncname pattern to syntax test
  ...

27 files changed, 1922 insertions, 36 deletions

diff --git a/Documentation/technical/.gitignore b/Documentation/technical/.gitignore
new file mode 100644
index 0000000000..8aa891daee
--- /dev/null
+++ b/Documentation/technical/.gitignore
@@ -0,0 +1 @@
+api-index.txt
diff --git a/Documentation/technical/api-allocation-growing.txt b/Documentation/technical/api-allocation-growing.txt
new file mode 100644
index 0000000000..43dbe09f73
--- /dev/null
+++ b/Documentation/technical/api-allocation-growing.txt
@@ -0,0 +1,34 @@
+allocation growing API
+======================
+
+Dynamically growing an array using realloc() is error prone and boring.
+
+Define your array with:
+
+* a pointer (`ary`) that points at the array, initialized to `NULL`;
+
+* an integer variable (`alloc`) that keeps track of how big the current
+  allocation is, initialized to `0`;
+
+* another integer variable (`nr`) to keep track of how many elements the
+  array currently has, initialized to `0`.
+
+Then before adding `n`th element to the array, call `ALLOC_GROW(ary, n,
+alloc)`.  This ensures that the array can hold at least `n` elements by
+calling `realloc(3)` and adjusting `alloc` variable.
+
+------------
+sometype *ary;
+size_t nr;
+size_t alloc
+
+for (i = 0; i < nr; i++)
+	if (we like ary[i] already)
+		return;
+
+/* we did not like any existing one, so add one */
+ALLOC_GROW(ary, nr + 1, alloc);
+ary[nr++] = value you like;
+------------
+
+You are responsible for updating the `nr` variable.
diff --git a/Documentation/technical/api-builtin.txt b/Documentation/technical/api-builtin.txt
new file mode 100644
index 0000000000..5cb2b0590a
--- /dev/null
+++ b/Documentation/technical/api-builtin.txt
@@ -0,0 +1,68 @@
+builtin API
+===========
+
+Adding a new built-in
+---------------------
+
+There are 4 things to do to add a built-in command implementation to
+git:
+
+. Define the implementation of the built-in command `foo` with
+  signature:
+
+	int cmd_foo(int argc, const char **argv, const char *prefix);
+
+. Add the external declaration for the function to `builtin.h`.
+
+. Add the command to `commands[]` table in `handle_internal_command()`,
+  defined in `git.c`.  The entry should look like:
+
+	{ "foo", cmd_foo, <options> },
++
+where options is the bitwise-or of:
+
+`RUN_SETUP`::
+
+	Make sure there is a git directory to work on, and if there is a
+	work tree, chdir to the top of it if the command was invoked
+	in a subdirectory.  If there is no work tree, no chdir() is
+	done.
+
+`USE_PAGER`::
+
+	If the standard output is connected to a tty, spawn a pager and
+	feed our output to it.
+
+`NEED_WORK_TREE`::
+
+	Make sure there is a work tree, i.e. the command cannot act
+	on bare repositories.
+	This only makes sense when `RUN_SETUP` is also set.
+
+. Add `builtin-foo.o` to `BUILTIN_OBJS` in `Makefile`.
+
+Additionally, if `foo` is a new command, there are 3 more things to do:
+
+. Add tests to `t/` directory.
+
+. Write documentation in `Documentation/git-foo.txt`.
+
+. Add an entry for `git-foo` to `command-list.txt`.
+
+
+How a built-in is called
+------------------------
+
+The implementation `cmd_foo()` takes three parameters, `argc`, `argv,
+and `prefix`.  The first two are similar to what `main()` of a
+standalone command would be called with.
+
+When `RUN_SETUP` is specified in the `commands[]` table, and when you
+were started from a subdirectory of the work tree, `cmd_foo()` is called
+after chdir(2) to the top of the work tree, and `prefix` gets the path
+to the subdirectory the command started from.  This allows you to
+convert a user-supplied pathname (typically relative to that directory)
+to a pathname relative to the top of the work tree.
+
+The return value from `cmd_foo()` becomes the exit status of the
+command.
diff --git a/Documentation/technical/api-decorate.txt b/Documentation/technical/api-decorate.txt
new file mode 100644
index 0000000000..1d52a6ce14
--- /dev/null
+++ b/Documentation/technical/api-decorate.txt
@@ -0,0 +1,6 @@
+decorate API
+============
+
+Talk about <decorate.h>
+
+(Linus)
diff --git a/Documentation/technical/api-diff.txt b/Documentation/technical/api-diff.txt
new file mode 100644
index 0000000000..20b0241d30
--- /dev/null
+++ b/Documentation/technical/api-diff.txt
@@ -0,0 +1,166 @@
+diff API
+========
+
+The diff API is for programs that compare two sets of files (e.g. two
+trees, one tree and the index) and present the found difference in
+various ways.  The calling program is responsible for feeding the API
+pairs of files, one from the "old" set and the corresponding one from
+"new" set, that are different.  The library called through this API is
+called diffcore, and is responsible for two things.
+
+* finding total rewrites (`-B`), renames (`-M`) and copies (`-C`), and
+  changes that touch a string (`-S`), as specified by the caller.
+
+* outputting the differences in various formats, as specified by the
+  caller.
+
+Calling sequence
+----------------
+
+* Prepare `struct diff_options` to record the set of diff options, and
+  then call `diff_setup()` to initialize this structure.  This sets up
+  the vanilla default.
+
+* Fill in the options structure to specify desired output format, rename
+  detection, etc.  `diff_opt_parse()` can be used to parse options given
+  from the command line in a way consistent with existing git-diff
+  family of programs.
+
+* Call `diff_setup_done()`; this inspects the options set up so far for
+  internal consistency and make necessary tweaking to it (e.g. if
+  textual patch output was asked, recursive behaviour is turned on).
+
+* As you find different pairs of files, call `diff_change()` to feed
+  modified files, `diff_addremove()` to feed created or deleted files,
+  or `diff_unmerged()` to feed a file whose state is 'unmerged' to the
+  API.  These are thin wrappers to a lower-level `diff_queue()` function
+  that is flexible enough to record any of these kinds of changes.
+
+* Once you finish feeding the pairs of files, call `diffcore_std()`.
+  This will tell the diffcore library to go ahead and do its work.
+
+* Calling `diff_flush()` will produce the output.
+
+
+Data structures
+---------------
+
+* `struct diff_filespec`
+
+This is the internal representation for a single file (blob).  It
+records the blob object name (if known -- for a work tree file it
+typically is a NUL SHA-1), filemode and pathname.  This is what the
+`diff_addremove()`, `diff_change()` and `diff_unmerged()` synthesize and
+feed `diff_queue()` function with.
+
+* `struct diff_filepair`
+
+This records a pair of `struct diff_filespec`; the filespec for a file
+in the "old" set (i.e. preimage) is called `one`, and the filespec for a
+file in the "new" set (i.e. postimage) is called `two`.  A change that
+represents file creation has NULL in `one`, and file deletion has NULL
+in `two`.
+
+A `filepair` starts pointing at `one` and `two` that are from the same
+filename, but `diffcore_std()` can break pairs and match component
+filespecs with other filespecs from a different filepair to form new
+filepair.  This is called 'rename detection'.
+
+* `struct diff_queue`
+
+This is a collection of filepairs.  Notable members are:
+
+`queue`::
+
+	An array of pointers to `struct diff_filepair`.  This
+	dynamically grows as you add filepairs;
+
+`alloc`::
+
+	The allocated size of the `queue` array;
+
+`nr`::
+
+	The number of elements in the `queue` array.
+
+
+* `struct diff_options`
+
+This describes the set of options the calling program wants to affect
+the operation of diffcore library with.
+
+Notable members are:
+
+`output_format`::
+	The output format used when `diff_flush()` is run.
+
+`context`::
+	Number of context lines to generate in patch output.
+
+`break_opt`, `detect_rename`, `rename-score`, `rename_limit`::
+	Affects the way detection logic for complete rewrites, renames
+	and copies.
+
+`abbrev`::
+	Number of hexdigits to abbreviate raw format output to.
+
+`pickaxe`::
+	A constant string (can and typically does contain newlines to
+	look for a block of text, not just a single line) to filter out
+	the filepairs that do not change the number of strings contained
+	in its preimage and postimage of the diff_queue.
+
+`flags`::
+	This is mostly a collection of boolean options that affects the
+	operation, but some do not have anything to do with the diffcore
+	library.
+
+BINARY, TEXT;;
+	Affects the way how a file that is seemingly binary is treated.
+
+FULL_INDEX;;
+	Tells the patch output format not to use abbreviated object
+	names on the "index" lines.
+
+FIND_COPIES_HARDER;;
+	Tells the diffcore library that the caller is feeding unchanged
+	filepairs to allow copies from unmodified files be detected.
+
+COLOR_DIFF;;
+	Output should be colored.
+
+COLOR_DIFF_WORDS;;
+	Output is a colored word-diff.
+
+NO_INDEX;;
+	Tells diff-files that the input is not tracked files but files
+	in random locations on the filesystem.
+
+ALLOW_EXTERNAL;;
+	Tells output routine that it is Ok to call user specified patch
+	output routine.  Plumbing disables this to ensure stable output.
+
+QUIET;;
+	Do not show any output.
+
+REVERSE_DIFF;;
+	Tells the library that the calling program is feeding the
+	filepairs reversed; `one` is two, and `two` is one.
+
+EXIT_WITH_STATUS;;
+	For communication between the calling program and the options
+	parser; tell the calling program to signal the presence of
+	difference using program exit code.
+
+HAS_CHANGES;;
+	Internal; used for optimization to see if there is any change.
+
+SILENT_ON_REMOVE;;
+	Affects if diff-files shows removed files.
+
+RECURSIVE, TREE_IN_RECURSIVE;;
+	Tells if tree traversal done by tree-diff should recursively
+	descend into a tree object pair that are different in preimage
+	and postimage set.
+
+(JC)
diff --git a/Documentation/technical/api-directory-listing.txt b/Documentation/technical/api-directory-listing.txt
new file mode 100644
index 0000000000..5bbd18f020
--- /dev/null
+++ b/Documentation/technical/api-directory-listing.txt
@@ -0,0 +1,76 @@
+directory listing API
+=====================
+
+The directory listing API is used to enumerate paths in the work tree,
+optionally taking `.git/info/exclude` and `.gitignore` files per
+directory into account.
+
+Data structure
+--------------
+
+`struct dir_struct` structure is used to pass directory traversal
+options to the library and to record the paths discovered.  The notable
+options are:
+
+`exclude_per_dir`::
+
+	The name of the file to be read in each directory for excluded
+	files (typically `.gitignore`).
+
+`collect_ignored`::
+
+	Include paths that are to be excluded in the result.
+
+`show_ignored`::
+
+	The traversal is for finding just ignored files, not unignored
+	files.
+
+`show_other_directories`::
+
+	Include a directory that is not tracked.
+
+`hide_empty_directories`::
+
+	Do not include a directory that is not tracked and is empty.
+
+`no_gitlinks`::
+
+	If set, recurse into a directory that looks like a git
+	directory.  Otherwise it is shown as a directory.
+
+The result of the enumeration is left in these fields::
+
+`entries[]`::
+
+	An array of `struct dir_entry`, each element of which describes
+	a path.
+
+`nr`::
+
+	The number of members in `entries[]` array.
+
+`alloc`::
+
+	Internal use; keeps track of allocation of `entries[]` array.
+
+
+Calling sequence
+----------------
+
+* Prepare `struct dir_struct dir` and clear it with `memset(&dir, 0,
+  sizeof(dir))`.
+
+* Call `add_exclude()` to add single exclude pattern,
+  `add_excludes_from_file()` to add patterns from a file
+  (e.g. `.git/info/exclude`), and/or set `dir.exclude_per_dir`.  A
+  short-hand function `setup_standard_excludes()` can be used to set up
+  the standard set of exclude settings.
+
+* Set options described in the Data Structure section above.
+
+* Call `read_directory()`.
+
+* Use `dir.entries[]`.
+
+(JC)
diff --git a/Documentation/technical/api-gitattributes.txt b/Documentation/technical/api-gitattributes.txt
new file mode 100644
index 0000000000..9d97eaa9de
--- /dev/null
+++ b/Documentation/technical/api-gitattributes.txt
@@ -0,0 +1,111 @@
+gitattributes API
+=================
+
+gitattributes mechanism gives a uniform way to associate various
+attributes to set of paths.
+
+
+Data Structure
+--------------
+
+`struct git_attr`::
+
+	An attribute is an opaque object that is identified by its name.
+	Pass the name and its length to `git_attr()` function to obtain
+	the object of this type.  The internal representation of this
+	structure is of no interest to the calling programs.
+
+`struct git_attr_check`::
+
+	This structure represents a set of attributes to check in a call
+	to `git_checkattr()` function, and receives the results.
+
+
+Calling Sequence
+----------------
+
+* Prepare an array of `struct git_attr_check` to define the list of
+  attributes you would want to check.  To populate this array, you would
+  need to define necessary attributes by calling `git_attr()` function.
+
+* Call git_checkattr() to check the attributes for the path.
+
+* Inspect `git_attr_check` structure to see how each of the attribute in
+  the array is defined for the path.
+
+
+Attribute Values
+----------------
+
+An attribute for a path can be in one of four states: Set, Unset,
+Unspecified or set to a string, and `.value` member of `struct
+git_attr_check` records it.  There are three macros to check these:
+
+`ATTR_TRUE()`::
+
+	Returns true if the attribute is Set for the path.
+
+`ATTR_FALSE()`::
+
+	Returns true if the attribute is Unset for the path.
+
+`ATTR_UNSET()`::
+
+	Returns true if the attribute is Unspecified for the path.
+
+If none of the above returns true, `.value` member points at a string
+value of the attribute for the path.
+
+
+Example
+-------
+
+To see how attributes "crlf" and "indent" are set for different paths.
+
+. Prepare an array of `struct git_attr_check` with two elements (because
+  we are checking two attributes).  Initialize their `attr` member with
+  pointers to `struct git_attr` obtained by calling `git_attr()`:
+
+------------
+static struct git_attr_check check[2];
+static void setup_check(void)
+{
+	if (check[0].attr)
+		return; /* already done */
+	check[0].attr = git_attr("crlf", 4);
+	check[1].attr = git_attr("ident", 5);
+}
+------------
+
+. Call `git_checkattr()` with the prepared array of `struct git_attr_check`:
+
+------------
+	const char *path;
+
+	setup_check();
+	git_checkattr(path, ARRAY_SIZE(check), check);
+------------
+
+. Act on `.value` member of the result, left in `check[]`:
+
+------------
+	const char *value = check[0].value;
+
+	if (ATTR_TRUE(value)) {
+		The attribute is Set, by listing only the name of the
+		attribute in the gitattributes file for the path.
+	} else if (ATTR_FALSE(value)) {
+		The attribute is Unset, by listing the name of the
+		attribute prefixed with a dash - for the path.
+	} else if (ATTR_UNSET(value)) {
+		The attribute is not set nor unset for the path.
+	} else if (!strcmp(value, "input")) {
+		If none of ATTR_TRUE(), ATTR_FALSE(), or ATTR_UNSET() is
+		true, the value is a string set in the gitattributes
+		file for the path by saying "attr=value".
+	} else if (... other check using value as string ...) {
+		...
+	}
+------------
+
+(JC)
diff --git a/Documentation/technical/api-grep.txt b/Documentation/technical/api-grep.txt
new file mode 100644
index 0000000000..a69cc8964d
--- /dev/null
+++ b/Documentation/technical/api-grep.txt
@@ -0,0 +1,8 @@
+grep API
+========
+
+Talk about <grep.h>, things like:
+
+* grep_buffer()
+
+(JC)
diff --git a/Documentation/technical/api-hash.txt b/Documentation/technical/api-hash.txt
new file mode 100644
index 0000000000..c784d3edcb
--- /dev/null
+++ b/Documentation/technical/api-hash.txt
@@ -0,0 +1,6 @@
+hash API
+========
+
+Talk about <hash.h>
+
+(Linus)
diff --git a/Documentation/technical/api-history-graph.txt b/Documentation/technical/api-history-graph.txt
new file mode 100644
index 0000000000..d66e61b1ec
--- /dev/null
+++ b/Documentation/technical/api-history-graph.txt
@@ -0,0 +1,179 @@
+history graph API
+=================
+
+The graph API is used to draw a text-based representation of the commit
+history.  The API generates the graph in a line-by-line fashion.
+
+Functions
+---------
+
+Core functions:
+
+* `graph_init()` creates a new `struct git_graph`
+
+* `graph_release()` destroys a `struct git_graph`, and frees the memory
+  associated with it.
+
+* `graph_update()` moves the graph to a new commit.
+
+* `graph_next_line()` outputs the next line of the graph into a strbuf.  It
+  does not add a terminating newline.
+
+* `graph_padding_line()` outputs a line of vertical padding in the graph.  It
+  is similar to `graph_next_line()`, but is guaranteed to never print the line
+  containing the current commit.  Where `graph_next_line()` would print the
+  commit line next, `graph_padding_line()` prints a line that simply extends
+  all branch lines downwards one row, leaving their positions unchanged.
+
+* `graph_is_commit_finished()` determines if the graph has output all lines
+  necessary for the current commit.  If `graph_update()` is called before all
+  lines for the current commit have been printed, the next call to
+  `graph_next_line()` will output an ellipsis, to indicate that a portion of
+  the graph was omitted.
+
+The following utility functions are wrappers around `graph_next_line()` and
+`graph_is_commit_finished()`.  They always print the output to stdout.
+They can all be called with a NULL graph argument, in which case no graph
+output will be printed.
+
+* `graph_show_commit()` calls `graph_next_line()` until it returns non-zero.
+  This prints all graph lines up to, and including, the line containing this
+  commit.  Output is printed to stdout.  The last line printed does not contain
+  a terminating newline.  This should not be called if the commit line has
+  already been printed, or it will loop forever.
+
+* `graph_show_oneline()` calls `graph_next_line()` and prints the result to
+  stdout.  The line printed does not contain a terminating newline.
+
+* `graph_show_padding()` calls `graph_padding_line()` and prints the result to
+  stdout.  The line printed does not contain a terminating newline.
+
+* `graph_show_remainder()` calls `graph_next_line()` until
+  `graph_is_commit_finished()` returns non-zero.  Output is printed to stdout.
+  The last line printed does not contain a terminating newline.  Returns 1 if
+  output was printed, and 0 if no output was necessary.
+
+* `graph_show_strbuf()` prints the specified strbuf to stdout, prefixing all
+  lines but the first with a graph line.  The caller is responsible for
+  ensuring graph output for the first line has already been printed to stdout.
+  (This can be done with `graph_show_commit()` or `graph_show_oneline()`.)  If
+  a NULL graph is supplied, the strbuf is printed as-is.
+
+* `graph_show_commit_msg()` is similar to `graph_show_strbuf()`, but it also
+  prints the remainder of the graph, if more lines are needed after the strbuf
+  ends.  It is better than directly calling `graph_show_strbuf()` followed by
+  `graph_show_remainder()` since it properly handles buffers that do not end in
+  a terminating newline.  The output printed by `graph_show_commit_msg()` will
+  end in a newline if and only if the strbuf ends in a newline.
+
+Data structure
+--------------
+`struct git_graph` is an opaque data type used to store the current graph
+state.
+
+Calling sequence
+----------------
+
+* Create a `struct git_graph` by calling `graph_init()`.  When using the
+  revision walking API, this is done automatically by `setup_revisions()` if
+  the '--graph' option is supplied.
+
+* Use the revision walking API to walk through a group of contiguous commits.
+  The `get_revision()` function automatically calls `graph_update()` each time
+  it is invoked.
+
+* For each commit, call `graph_next_line()` repeatedly, until
+  `graph_is_commit_finished()` returns non-zero.  Each call go
+  `graph_next_line()` will output a single line of the graph.  The resulting
+  lines will not contain any newlines.  `graph_next_line()` returns 1 if the
+  resulting line contains the current commit, or 0 if this is merely a line
+  needed to adjust the graph before or after the current commit.  This return
+  value can be used to determine where to print the commit summary information
+  alongside the graph output.
+
+Limitations
+-----------
+
+* `graph_update()` must be called with commits in topological order.  It should
+  not be called on a commit if it has already been invoked with an ancestor of
+  that commit, or the graph output will be incorrect.
+
+* `graph_update()` must be called on a contiguous group of commits.  If
+  `graph_update()` is called on a particular commit, it should later be called
+  on all parents of that commit.  Parents must not be skipped, or the graph
+  output will appear incorrect.
++
+`graph_update()` may be used on a pruned set of commits only if the parent list
+has been rewritten so as to include only ancestors from the pruned set.
+
+* The graph API does not currently support reverse commit ordering.  In
+  order to implement reverse ordering, the graphing API needs an
+  (efficient) mechanism to find the children of a commit.
+
+Sample usage
+------------
+
+------------
+struct commit *commit;
+struct git_graph *graph = graph_init(opts);
+
+while ((commit = get_revision(opts)) != NULL) {
+	graph_update(graph, commit);
+	while (!graph_is_commit_finished(graph))
+	{
+		struct strbuf sb;
+		int is_commit_line;
+
+		strbuf_init(&sb, 0);
+		is_commit_line = graph_next_line(graph, &sb);
+		fputs(sb.buf, stdout);
+
+		if (is_commit_line)
+			log_tree_commit(opts, commit);
+		else
+			putchar(opts->diffopt.line_termination);
+	}
+}
+
+graph_release(graph);
+------------
+
+Sample output
+-------------
+
+The following is an example of the output from the graph API.  This output does
+not include any commit summary information--callers are responsible for
+outputting that information, if desired.
+
+------------
+*
+*
+*
+|\
+* |
+| | *
+| \ \
+|  \ \
+*-. \ \
+|\ \ \ \
+| | * | |
+| | | | | *
+| | | | | *
+| | | | | *
+| | | | | |\
+| | | | | | *
+| * | | | | |
+| | | | | *  \
+| | | | | |\  |
+| | | | * | | |
+| | | | * | | |
+* | | | | | | |
+| |/ / / / / /
+|/| / / / / /
+* | | | | | |
+|/ / / / / /
+* | | | | |
+| | | | | *
+| | | | |/
+| | | | *
+------------
diff --git a/Documentation/technical/api-in-core-index.txt b/Documentation/technical/api-in-core-index.txt
new file mode 100644
index 0000000000..adbdbf5d75
--- /dev/null
+++ b/Documentation/technical/api-in-core-index.txt
@@ -0,0 +1,21 @@
+in-core index API
+=================
+
+Talk about <read-cache.c> and <cache-tree.c>, things like:
+
+* cache -> the_index macros
+* read_index()
+* write_index()
+* ie_match_stat() and ie_modified(); how they are different and when to
+  use which.
+* index_name_pos()
+* remove_index_entry_at()
+* remove_file_from_index()
+* add_file_to_index()
+* add_index_entry()
+* refresh_index()
+* discard_index()
+* cache_tree_invalidate_path()
+* cache_tree_update()
+
+(JC, Linus)
diff --git a/Documentation/technical/api-index-skel.txt b/Documentation/technical/api-index-skel.txt
new file mode 100644
index 0000000000..af7cc2e395
--- /dev/null
+++ b/Documentation/technical/api-index-skel.txt
@@ -0,0 +1,15 @@
+GIT API Documents
+=================
+
+GIT has grown a set of internal API over time.  This collection
+documents them.
+
+////////////////////////////////////////////////////////////////
+// table of contents begin
+////////////////////////////////////////////////////////////////
+
+////////////////////////////////////////////////////////////////
+// table of contents end
+////////////////////////////////////////////////////////////////
+
+2007-11-24
diff --git a/Documentation/technical/api-index.sh b/Documentation/technical/api-index.sh
new file mode 100755
index 0000000000..9c3f4131b8
--- /dev/null
+++ b/Documentation/technical/api-index.sh
@@ -0,0 +1,28 @@
+#!/bin/sh
+
+(
+	c=////////////////////////////////////////////////////////////////
+	skel=api-index-skel.txt
+	sed -e '/^\/\/ table of contents begin/q' "$skel"
+	echo "$c"
+
+	ls api-*.txt |
+	while read filename
+	do
+		case "$filename" in
+		api-index-skel.txt | api-index.txt) continue ;;
+		esac
+		title=$(sed -e 1q "$filename")
+		html=${filename%.txt}.html
+		echo "* link:$html[$title]"
+	done
+	echo "$c"
+	sed -n -e '/^\/\/ table of contents end/,$p' "$skel"
+) >api-index.txt+
+
+if test -f api-index.txt && cmp api-index.txt api-index.txt+ >/dev/null
+then
+	rm -f api-index.txt+
+else
+	mv api-index.txt+ api-index.txt
+fi
diff --git a/Documentation/technical/api-lockfile.txt b/Documentation/technical/api-lockfile.txt
new file mode 100644
index 0000000000..dd894043ae
--- /dev/null
+++ b/Documentation/technical/api-lockfile.txt
@@ -0,0 +1,74 @@
+lockfile API
+============
+
+The lockfile API serves two purposes:
+
+* Mutual exclusion.  When we write out a new index file, first
+  we create a new file `$GIT_DIR/index.lock`, write the new
+  contents into it, and rename it to the final destination
+  `$GIT_DIR/index`.  We try to create the `$GIT_DIR/index.lock`
+  file with O_EXCL so that we can notice and fail when somebody
+  else is already trying to update the index file.
+
+* Automatic cruft removal.  After we create the "lock" file, we
+  may decide to `die()`, and we would want to make sure that we
+  remove the file that has not been committed to its final
+  destination.  This is done by remembering the lockfiles we
+  created in a linked list and cleaning them up from an
+  `atexit(3)` handler.  Outstanding lockfiles are also removed
+  when the program dies on a signal.
+
+
+The functions
+-------------
+
+hold_lock_file_for_update::
+
+	Take a pointer to `struct lock_file`, the filename of
+	the final destination (e.g. `$GIT_DIR/index`) and a flag
+	`die_on_error`.  Attempt to create a lockfile for the
+	destination and return the file descriptor for writing
+	to the file.  If `die_on_error` flag is true, it dies if
+	a lock is already taken for the file; otherwise it
+	returns a negative integer to the caller on failure.
+
+commit_lock_file::
+
+	Take a pointer to the `struct lock_file` initialized
+	with an earlier call to `hold_lock_file_for_update()`,
+	close the file descriptor and rename the lockfile to its
+	final destination.  Returns 0 upon success, a negative
+	value on failure to close(2) or rename(2).
+
+rollback_lock_file::
+
+	Take a pointer to the `struct lock_file` initialized
+	with an earlier call to `hold_lock_file_for_update()`,
+	close the file descriptor and remove the lockfile.
+
+close_lock_file::
+	Take a pointer to the `struct lock_file` initialized
+	with an earlier call to `hold_lock_file_for_update()`,
+	and close the file descriptor.  Returns 0 upon success,
+	a negative value on failure to close(2).
+
+Because the structure is used in an `atexit(3)` handler, its
+storage has to stay throughout the life of the program.  It
+cannot be an auto variable allocated on the stack.
+
+Call `commit_lock_file()` or `rollback_lock_file()` when you are
+done writing to the file descriptor.  If you do not call either
+and simply `exit(3)` from the program, an `atexit(3)` handler
+will close and remove the lockfile.
+
+If you need to close the file descriptor you obtained from
+`hold_lock_file_for_update` function yourself, do so by calling
+`close_lock_file()`.  You should never call `close(2)` yourself!
+Otherwise the `struct
+lock_file` structure still remembers that the file descriptor
+needs to be closed, and a later call to `commit_lock_file()` or
+`rollback_lock_file()` will result in duplicate calls to
+`close(2)`.  Worse yet, if you `close(2)`, open another file
+descriptor for completely different purpose, and then call
+`commit_lock_file()` or `rollback_lock_file()`, they may close
+that unrelated file descriptor.
diff --git a/Documentation/technical/api-object-access.txt b/Documentation/technical/api-object-access.txt
new file mode 100644
index 0000000000..03bb0e950d
--- /dev/null
+++ b/Documentation/technical/api-object-access.txt
@@ -0,0 +1,15 @@
+object access API
+=================
+
+Talk about <sha1_file.c> and <object.h> family, things like
+
+* read_sha1_file()
+* read_object_with_reference()
+* has_sha1_file()
+* write_sha1_file()
+* pretend_sha1_file()
+* lookup_{object,commit,tag,blob,tree}
+* parse_{object,commit,tag,blob,tree}
+* Use of object flags
+
+(JC, Shawn, Daniel, Dscho, Linus)
diff --git a/Documentation/technical/api-parse-options.txt b/Documentation/technical/api-parse-options.txt
new file mode 100644
index 0000000000..e66ca9f70c
--- /dev/null
+++ b/Documentation/technical/api-parse-options.txt
@@ -0,0 +1,234 @@
+parse-options API
+=================
+
+The parse-options API is used to parse and massage options in git
+and to provide a usage help with consistent look.
+
+Basics
+------
+
+The argument vector `argv[]` may usually contain mandatory or optional
+'non-option arguments', e.g. a filename or a branch, and 'options'.
+Options are optional arguments that start with a dash and
+that allow to change the behavior of a command.
+
+* There are basically three types of options:
+  'boolean' options,
+  options with (mandatory) 'arguments' and
+  options with 'optional arguments'
+  (i.e. a boolean option that can be adjusted).
+
+* There are basically two forms of options:
+  'Short options' consist of one dash (`-`) and one alphanumeric
+  character.
+  'Long options' begin with two dashes (`\--`) and some
+  alphanumeric characters.
+
+* Options are case-sensitive.
+  Please define 'lower-case long options' only.
+
+The parse-options API allows:
+
+* 'sticked' and 'separate form' of options with arguments.
+  `-oArg` is sticked, `-o Arg` is separate form.
+  `\--option=Arg` is sticked, `\--option Arg` is separate form.
+
+* Long options may be 'abbreviated', as long as the abbreviation
+  is unambiguous.
+
+* Short options may be bundled, e.g. `-a -b` can be specified as `-ab`.
+
+* Boolean long options can be 'negated' (or 'unset') by prepending
+  `no-`, e.g. `\--no-abbrev` instead of `\--abbrev`.
+
+* Options and non-option arguments can clearly be separated using the `\--`
+  option, e.g. `-a -b \--option \-- \--this-is-a-file` indicates that
+  `\--this-is-a-file` must not be processed as an option.
+
+Steps to parse options
+----------------------
+
+. `#include "parse-options.h"`
+
+. define a NULL-terminated
+  `static const char * const builtin_foo_usage[]` array
+  containing alternative usage strings
+
+. define `builtin_foo_options` array as described below
+  in section 'Data Structure'.
+
+. in `cmd_foo(int argc, const char **argv, const char *prefix)`
+  call
+
+	argc = parse_options(argc, argv, builtin_foo_options, builtin_foo_usage, flags);
++
+`parse_options()` will filter out the processed options of `argv[]` and leave the
+non-option arguments in `argv[]`.
+`argc` is updated appropriately because of the assignment.
++
+You can also pass NULL instead of a usage array as fourth parameter of
+parse_options(), to avoid displaying a help screen with usage info and
+option list.  This should only be done if necessary, e.g. to implement
+a limited parser for only a subset of the options that needs to be run
+before the full parser, which in turn shows the full help message.
++
+Flags are the bitwise-or of:
+
+`PARSE_OPT_KEEP_DASHDASH`::
+	Keep the `\--` that usually separates options from
+	non-option arguments.
+
+`PARSE_OPT_STOP_AT_NON_OPTION`::
+	Usually the whole argument vector is massaged and reordered.
+	Using this flag, processing is stopped at the first non-option
+	argument.
+
+`PARSE_OPT_KEEP_ARGV0`::
+	Keep the first argument, which contains the program name.  It's
+	removed from argv[] by default.
+
+`PARSE_OPT_KEEP_UNKNOWN`::
+	Keep unknown arguments instead of erroring out.  This doesn't
+	work for all combinations of arguments as users might expect
+	it to do.  E.g. if the first argument in `--unknown --known`
+	takes a value (which we can't know), the second one is
+	mistakenly interpreted as a known option.  Similarly, if
+	`PARSE_OPT_STOP_AT_NON_OPTION` is set, the second argument in
+	`--unknown value` will be mistakenly interpreted as a
+	non-option, not as a value belonging to the unknown option,
+	the parser early.  That's why parse_options() errors out if
+	both options are set.
+
+`PARSE_OPT_NO_INTERNAL_HELP`::
+	By default, parse_options() handles `-h`, `--help` and
+	`--help-all` internally, by showing a help screen.  This option
+	turns it off and allows one to add custom handlers for these
+	options, or to just leave them unknown.
+
+Data Structure
+--------------
+
+The main data structure is an array of the `option` struct,
+say `static struct option builtin_add_options[]`.
+There are some macros to easily define options:
+
+`OPT__ABBREV(&int_var)`::
+	Add `\--abbrev[=<n>]`.
+
+`OPT__DRY_RUN(&int_var)`::
+	Add `-n, \--dry-run`.
+
+`OPT__QUIET(&int_var)`::
+	Add `-q, \--quiet`.
+
+`OPT__VERBOSE(&int_var)`::
+	Add `-v, \--verbose`.
+
+`OPT_GROUP(description)`::
+	Start an option group. `description` is a short string that
+	describes the group or an empty string.
+	Start the description with an upper-case letter.
+
+`OPT_BOOLEAN(short, long, &int_var, description)`::
+	Introduce a boolean option.
+	`int_var` is incremented on each use.
+
+`OPT_BIT(short, long, &int_var, description, mask)`::
+	Introduce a boolean option.
+	If used, `int_var` is bitwise-ored with `mask`.
+
+`OPT_SET_INT(short, long, &int_var, description, integer)`::
+	Introduce a boolean option.
+	If used, set `int_var` to `integer`.
+
+`OPT_SET_PTR(short, long, &ptr_var, description, ptr)`::
+	Introduce a boolean option.
+	If used, set `ptr_var` to `ptr`.
+
+`OPT_STRING(short, long, &str_var, arg_str, description)`::
+	Introduce an option with string argument.
+	The string argument is put into `str_var`.
+
+`OPT_INTEGER(short, long, &int_var, description)`::
+	Introduce an option with integer argument.
+	The integer is put into `int_var`.
+
+`OPT_DATE(short, long, &int_var, description)`::
+	Introduce an option with date argument, see `approxidate()`.
+	The timestamp is put into `int_var`.
+
+`OPT_CALLBACK(short, long, &var, arg_str, description, func_ptr)`::
+	Introduce an option with argument.
+	The argument will be fed into the function given by `func_ptr`
+	and the result will be put into `var`.
+	See 'Option Callbacks' below for a more elaborate description.
+
+`OPT_ARGUMENT(long, description)`::
+	Introduce a long-option argument that will be kept in `argv[]`.
+
+
+The last element of the array must be `OPT_END()`.
+
+If not stated otherwise, interpret the arguments as follows:
+
+* `short` is a character for the short option
+  (e.g. `\'e\'` for `-e`, use `0` to omit),
+
+* `long` is a string for the long option
+  (e.g. `"example"` for `\--example`, use `NULL` to omit),
+
+* `int_var` is an integer variable,
+
+* `str_var` is a string variable (`char *`),
+
+* `arg_str` is the string that is shown as argument
+  (e.g. `"branch"` will result in `<branch>`).
+  If set to `NULL`, three dots (`...`) will be displayed.
+
+* `description` is a short string to describe the effect of the option.
+  It shall begin with a lower-case letter and a full stop (`.`) shall be
+  omitted at the end.
+
+Option Callbacks
+----------------
+
+The function must be defined in this form:
+
+	int func(const struct option *opt, const char *arg, int unset)
+
+The callback mechanism is as follows:
+
+* Inside `funct`, the only interesting member of the structure
+  given by `opt` is the void pointer `opt->value`.
+  `\*opt->value` will be the value that is saved into `var`, if you
+  use `OPT_CALLBACK()`.
+  For example, do `*(unsigned long *)opt->value = 42;` to get 42
+  into an `unsigned long` variable.
+
+* Return value `0` indicates success and non-zero return
+  value will invoke `usage_with_options()` and, thus, die.
+
+* If the user negates the option, `arg` is `NULL` and `unset` is 1.
+
+Sophisticated option parsing
+----------------------------
+
+If you need, for example, option callbacks with optional arguments
+or without arguments at all, or if you need other special cases,
+that are not handled by the macros above, you need to specify the
+members of the `option` structure manually.
+
+This is not covered in this document, but well documented
+in `parse-options.h` itself.
+
+Examples
+--------
+
+See `test-parse-options.c` and
+`builtin-add.c`,
+`builtin-clone.c`,
+`builtin-commit.c`,
+`builtin-fetch.c`,
+`builtin-fsck.c`,
+`builtin-rm.c`
+for real-world examples.
diff --git a/Documentation/technical/api-quote.txt b/Documentation/technical/api-quote.txt
new file mode 100644
index 0000000000..e8a1bce94e
--- /dev/null
+++ b/Documentation/technical/api-quote.txt
@@ -0,0 +1,10 @@
+quote API
+=========
+
+Talk about <quote.h>, things like
+
+* sq_quote and unquote
+* c_style quote and unquote
+* quoting for foreign languages
+
+(JC)
diff --git a/Documentation/technical/api-remote.txt b/Documentation/technical/api-remote.txt
new file mode 100644
index 0000000000..073b22bd83
--- /dev/null
+++ b/Documentation/technical/api-remote.txt
@@ -0,0 +1,123 @@
+Remotes configuration API
+=========================
+
+The API in remote.h gives access to the configuration related to
+remotes. It handles all three configuration mechanisms historically
+and currently used by git, and presents the information in a uniform
+fashion. Note that the code also handles plain URLs without any
+configuration, giving them just the default information.
+
+struct remote
+-------------
+
+`name`::
+
+	The user's nickname for the remote
+
+`url`::
+
+	An array of all of the url_nr URLs configured for the remote
+
+`push`::
+
+	 An array of refspecs configured for pushing, with
+	 push_refspec being the literal strings, and push_refspec_nr
+	 being the quantity.
+
+`fetch`::
+
+	An array of refspecs configured for fetching, with
+	fetch_refspec being the literal strings, and fetch_refspec_nr
+	being the quantity.
+
+`fetch_tags`::
+
+	The setting for whether to fetch tags (as a separate rule from
+	the configured refspecs); -1 means never to fetch tags, 0
+	means to auto-follow tags based on the default heuristic, 1
+	means to always auto-follow tags, and 2 means to fetch all
+	tags.
+
+`receivepack`, `uploadpack`::
+
+	The configured helper programs to run on the remote side, for
+	git-native protocols.
+
+`http_proxy`::
+
+	The proxy to use for curl (http, https, ftp, etc.) URLs.
+
+struct remotes can be found by name with remote_get(), and iterated
+through with for_each_remote(). remote_get(NULL) will return the
+default remote, given the current branch and configuration.
+
+struct refspec
+--------------
+
+A struct refspec holds the parsed interpretation of a refspec. If it
+will force updates (starts with a '+'), force is true. If it is a
+pattern (sides end with '*') pattern is true. src and dest are the two
+sides (if a pattern, only the part outside of the wildcards); if there
+is only one side, it is src, and dst is NULL; if sides exist but are
+empty (i.e., the refspec either starts or ends with ':'), the
+corresponding side is "".
+
+This parsing can be done to an array of strings to give an array of
+struct refpsecs with parse_ref_spec().
+
+remote_find_tracking(), given a remote and a struct refspec with
+either src or dst filled out, will fill out the other such that the
+result is in the "fetch" specification for the remote (note that this
+evaluates patterns and returns a single result).
+
+struct branch
+-------------
+
+Note that this may end up moving to branch.h
+
+struct branch holds the configuration for a branch. It can be looked
+up with branch_get(name) for "refs/heads/{name}", or with
+branch_get(NULL) for HEAD.
+
+It contains:
+
+`name`::
+
+	The short name of the branch.
+
+`refname`::
+
+	The full path for the branch ref.
+
+`remote_name`::
+
+	The name of the remote listed in the configuration.
+
+`remote`::
+
+	The struct remote for that remote.
+
+`merge_name`::
+
+	An array of the "merge" lines in the configuration.
+
+`merge`::
+
+	An array of the struct refspecs used for the merge lines. That
+	is, merge[i]->dst is a local tracking ref which should be
+	merged into this branch by default.
+
+`merge_nr`::
+
+	The number of merge configurations
+
+branch_has_merge_config() returns true if the given branch has merge
+configuration given.
+
+Other stuff
+-----------
+
+There is other stuff in remote.h that is related, in general, to the
+process of interacting with remotes.
+
+(Daniel Barkalow)
diff --git a/Documentation/technical/api-revision-walking.txt b/Documentation/technical/api-revision-walking.txt
new file mode 100644
index 0000000000..996da0503a
--- /dev/null
+++ b/Documentation/technical/api-revision-walking.txt
@@ -0,0 +1,67 @@
+revision walking API
+====================
+
+The revision walking API offers functions to build a list of revisions
+and then iterate over that list.
+
+Calling sequence
+----------------
+
+The walking API has a given calling sequence: first you need to
+initialize a rev_info structure, then add revisions to control what kind
+of revision list do you want to get, finally you can iterate over the
+revision list.
+
+Functions
+---------
+
+`init_revisions`::
+
+	Initialize a rev_info structure with default values. The second
+	parameter may be NULL or can be prefix path, and then the `.prefix`
+	variable will be set to it. This is typically the first function you
+	want to call when you want to deal with a revision list. After calling
+	this function, you are free to customize options, like set
+	`.ignore_merges` to 0 if you don't want to ignore merges, and so on. See
+	`revision.h` for a complete list of available options.
+
+`add_pending_object`::
+
+	This function can be used if you want to add commit objects as revision
+	information. You can use the `UNINTERESTING` object flag to indicate if
+	you want to include or exclude the given commit (and commits reachable
+	from the given commit) from the revision list.
++
+NOTE: If you have the commits as a string list then you probably want to
+use setup_revisions(), instead of parsing each string and using this
+function.
+
+`setup_revisions`::
+
+	Parse revision information, filling in the `rev_info` structure, and
+	removing the used arguments from the argument list. Returns the number
+	of arguments left that weren't recognized, which are also moved to the
+	head of the argument list. The last parameter is used in case no
+	parameter given by the first two arguments.
+
+`prepare_revision_walk`::
+
+	Prepares the rev_info structure for a walk. You should check if it
+	returns any error (non-zero return code) and if it does not, you can
+	start using get_revision() to do the iteration.
+
+`get_revision`::
+
+	Takes a pointer to a `rev_info` structure and iterates over it,
+	returning a `struct commit *` each time you call it. The end of the
+	revision list is indicated by returning a NULL pointer.
+
+Data structures
+---------------
+
+Talk about <revision.h>, things like:
+
+* two diff_options, one for path limiting, another for output;
+* remaining functions;
+
+(Linus, JC, Dscho)
diff --git a/Documentation/technical/api-run-command.txt b/Documentation/technical/api-run-command.txt
new file mode 100644
index 0000000000..2efe7a40be
--- /dev/null
+++ b/Documentation/technical/api-run-command.txt
@@ -0,0 +1,187 @@
+run-command API
+===============
+
+The run-command API offers a versatile tool to run sub-processes with
+redirected input and output as well as with a modified environment
+and an alternate current directory.
+
+A similar API offers the capability to run a function asynchronously,
+which is primarily used to capture the output that the function
+produces in the caller in order to process it.
+
+
+Functions
+---------
+
+`start_command`::
+
+	Start a sub-process. Takes a pointer to a `struct child_process`
+	that specifies the details and returns pipe FDs (if requested).
+	See below for details.
+
+`finish_command`::
+
+	Wait for the completion of a sub-process that was started with
+	start_command().
+
+`run_command`::
+
+	A convenience function that encapsulates a sequence of
+	start_command() followed by finish_command(). Takes a pointer
+	to a `struct child_process` that specifies the details.
+
+`run_command_v_opt`, `run_command_v_opt_cd_env`::
+
+	Convenience functions that encapsulate a sequence of
+	start_command() followed by finish_command(). The argument argv
+	specifies the program and its arguments. The argument opt is zero
+	or more of the flags `RUN_COMMAND_NO_STDIN`, `RUN_GIT_CMD`, or
+	`RUN_COMMAND_STDOUT_TO_STDERR` that correspond to the members
+	.no_stdin, .git_cmd, .stdout_to_stderr of `struct child_process`.
+	The argument dir corresponds the member .dir. The argument env
+	corresponds to the member .env.
+
+`start_async`::
+
+	Run a function asynchronously. Takes a pointer to a `struct
+	async` that specifies the details and returns a pipe FD
+	from which the caller reads. See below for details.
+
+`finish_async`::
+
+	Wait for the completion of an asynchronous function that was
+	started with start_async().
+
+`run_hook`::
+
+	Run a hook.
+	The first argument is a pathname to an index file, or NULL
+	if the hook uses the default index file or no index is needed.
+	The second argument is the name of the hook.
+	The further arguments correspond to the hook arguments.
+	The last argument has to be NULL to terminate the arguments list.
+	If the hook does not exist or is not executable, the return
+	value will be zero.
+	If it is executable, the hook will be executed and the exit
+	status of the hook is returned.
+	On execution, .stdout_to_stderr and .no_stdin will be set.
+	(See below.)
+
+
+Data structures
+---------------
+
+* `struct child_process`
+
+This describes the arguments, redirections, and environment of a
+command to run in a sub-process.
+
+The caller:
+
+1. allocates and clears (memset(&chld, 0, sizeof(chld));) a
+   struct child_process variable;
+2. initializes the members;
+3. calls start_command();
+4. processes the data;
+5. closes file descriptors (if necessary; see below);
+6. calls finish_command().
+
+The .argv member is set up as an array of string pointers (NULL
+terminated), of which .argv[0] is the program name to run (usually
+without a path). If the command to run is a git command, set argv[0] to
+the command name without the 'git-' prefix and set .git_cmd = 1.
+
+The members .in, .out, .err are used to redirect stdin, stdout,
+stderr as follows:
+
+. Specify 0 to request no special redirection. No new file descriptor
+  is allocated. The child process simply inherits the channel from the
+  parent.
+
+. Specify -1 to have a pipe allocated; start_command() replaces -1
+  by the pipe FD in the following way:
+
+	.in: Returns the writable pipe end into which the caller writes;
+		the readable end of the pipe becomes the child's stdin.
+
+	.out, .err: Returns the readable pipe end from which the caller
+		reads; the writable end of the pipe end becomes child's
+		stdout/stderr.
+
+  The caller of start_command() must close the so returned FDs
+  after it has completed reading from/writing to it!
+
+. Specify a file descriptor > 0 to be used by the child:
+
+	.in: The FD must be readable; it becomes child's stdin.
+	.out: The FD must be writable; it becomes child's stdout.
+	.err > 0 is not supported.
+
+  The specified FD is closed by start_command(), even if it fails to
+  run the sub-process!
+
+. Special forms of redirection are available by setting these members
+  to 1:
+
+	.no_stdin, .no_stdout, .no_stderr: The respective channel is
+		redirected to /dev/null.
+
+	.stdout_to_stderr: stdout of the child is redirected to its
+		stderr. This happens after stderr is itself redirected.
+		So stdout will follow stderr to wherever it is
+		redirected.
+
+To modify the environment of the sub-process, specify an array of
+string pointers (NULL terminated) in .env:
+
+. If the string is of the form "VAR=value", i.e. it contains '='
+  the variable is added to the child process's environment.
+
+. If the string does not contain '=', it names an environment
+  variable that will be removed from the child process's environment.
+
+To specify a new initial working directory for the sub-process,
+specify it in the .dir member.
+
+
+* `struct async`
+
+This describes a function to run asynchronously, whose purpose is
+to produce output that the caller reads.
+
+The caller:
+
+1. allocates and clears (memset(&asy, 0, sizeof(asy));) a
+   struct async variable;
+2. initializes .proc and .data;
+3. calls start_async();
+4. processes the data by reading from the fd in .out;
+5. closes .out;
+6. calls finish_async().
+
+The function pointer in .proc has the following signature:
+
+	int proc(int fd, void *data);
+
+. fd specifies a writable file descriptor to which the function must
+  write the data that it produces. The function *must* close this
+  descriptor before it returns.
+
+. data is the value that the caller has specified in the .data member
+  of struct async.
+
+. The return value of the function is 0 on success and non-zero
+  on failure. If the function indicates failure, finish_async() will
+  report failure as well.
+
+
+There are serious restrictions on what the asynchronous function can do
+because this facility is implemented by a pipe to a forked process on
+UNIX, but by a thread in the same address space on Windows:
+
+. It cannot change the program's state (global variables, environment,
+  etc.) in a way that the caller notices; in other words, .out is the
+  only communication channel to the caller.
+
+. It must not change the program's state that the caller of the
+  facility also uses.
diff --git a/Documentation/technical/api-setup.txt b/Documentation/technical/api-setup.txt
new file mode 100644
index 0000000000..4f63a04d7d
--- /dev/null
+++ b/Documentation/technical/api-setup.txt
@@ -0,0 +1,13 @@
+setup API
+=========
+
+Talk about
+
+* setup_git_directory()
+* setup_git_directory_gently()
+* is_inside_git_dir()
+* is_inside_work_tree()
+* setup_work_tree()
+* get_pathspec()
+
+(Dscho)
diff --git a/Documentation/technical/api-strbuf.txt b/Documentation/technical/api-strbuf.txt
new file mode 100644
index 0000000000..7438149249
--- /dev/null
+++ b/Documentation/technical/api-strbuf.txt
@@ -0,0 +1,255 @@
+strbuf API
+==========
+
+strbuf's are meant to be used with all the usual C string and memory
+APIs. Given that the length of the buffer is known, it's often better to
+use the mem* functions than a str* one (memchr vs. strchr e.g.).
+Though, one has to be careful about the fact that str* functions often
+stop on NULs and that strbufs may have embedded NULs.
+
+An strbuf is NUL terminated for convenience, but no function in the
+strbuf API actually relies on the string being free of NULs.
+
+strbufs has some invariants that are very important to keep in mind:
+
+. The `buf` member is never NULL, so you it can be used in any usual C
+string operations safely. strbuf's _have_ to be initialized either by
+`strbuf_init()` or by `= STRBUF_INIT` before the invariants, though.
++
+Do *not* assume anything on what `buf` really is (e.g. if it is
+allocated memory or not), use `strbuf_detach()` to unwrap a memory
+buffer from its strbuf shell in a safe way. That is the sole supported
+way. This will give you a malloced buffer that you can later `free()`.
++
+However, it is totally safe to modify anything in the string pointed by
+the `buf` member, between the indices `0` and `len-1` (inclusive).
+
+. The `buf` member is a byte array that has at least `len + 1` bytes
+  allocated. The extra byte is used to store a `'\0'`, allowing the
+  `buf` member to be a valid C-string. Every strbuf function ensure this
+  invariant is preserved.
++
+NOTE: It is OK to "play" with the buffer directly if you work it this
+      way:
++
+----
+strbuf_grow(sb, SOME_SIZE); <1>
+strbuf_setlen(sb, sb->len + SOME_OTHER_SIZE);
+----
+<1> Here, the memory array starting at `sb->buf`, and of length
+`strbuf_avail(sb)` is all yours, and you can be sure that
+`strbuf_avail(sb)` is at least `SOME_SIZE`.
++
+NOTE: `SOME_OTHER_SIZE` must be smaller or equal to `strbuf_avail(sb)`.
++
+Doing so is safe, though if it has to be done in many places, adding the
+missing API to the strbuf module is the way to go.
++
+WARNING: Do _not_ assume that the area that is yours is of size `alloc
+- 1` even if it's true in the current implementation. Alloc is somehow a
+"private" member that should not be messed with. Use `strbuf_avail()`
+instead.
+
+Data structures
+---------------
+
+* `struct strbuf`
+
+This is string buffer structure. The `len` member can be used to
+determine the current length of the string, and `buf` member provides access to
+the string itself.
+
+Functions
+---------
+
+* Life cycle
+
+`strbuf_init`::
+
+	Initialize the structure. The second parameter can be zero or a bigger
+	number to allocate memory, in case you want to prevent further reallocs.
+
+`strbuf_release`::
+
+	Release a string buffer and the memory it used. You should not use the
+	string buffer after using this function, unless you initialize it again.
+
+`strbuf_detach`::
+
+	Detach the string from the strbuf and returns it; you now own the
+	storage the string occupies and it is your responsibility from then on
+	to release it with `free(3)` when you are done with it.
+
+`strbuf_attach`::
+
+	Attach a string to a buffer. You should specify the string to attach,
+	the current length of the string and the amount of allocated memory.
+	The amount must be larger than the string length, because the string you
+	pass is supposed to be a NUL-terminated string.  This string _must_ be
+	malloc()ed, and after attaching, the pointer cannot be relied upon
+	anymore, and neither be free()d directly.
+
+`strbuf_swap`::
+
+	Swap the contents of two string buffers.
+
+* Related to the size of the buffer
+
+`strbuf_avail`::
+
+	Determine the amount of allocated but unused memory.
+
+`strbuf_grow`::
+
+	Ensure that at least this amount of unused memory is available after
+	`len`. This is used when you know a typical size for what you will add
+	and want to avoid repetitive automatic resizing of the underlying buffer.
+	This is never a needed operation, but can be critical for performance in
+	some cases.
+
+`strbuf_setlen`::
+
+	Set the length of the buffer to a given value. This function does *not*
+	allocate new memory, so you should not perform a `strbuf_setlen()` to a
+	length that is larger than `len + strbuf_avail()`. `strbuf_setlen()` is
+	just meant as a 'please fix invariants from this strbuf I just messed
+	with'.
+
+`strbuf_reset`::
+
+	Empty the buffer by setting the size of it to zero.
+
+* Related to the contents of the buffer
+
+`strbuf_rtrim`::
+
+	Strip whitespace from the end of a string.
+
+`strbuf_cmp`::
+
+	Compare two buffers. Returns an integer less than, equal to, or greater
+	than zero if the first buffer is found, respectively, to be less than,
+	to match, or be greater than the second buffer.
+
+* Adding data to the buffer
+
+NOTE: All of the functions in this section will grow the buffer as necessary.
+If they fail for some reason other than memory shortage and the buffer hadn't
+been allocated before (i.e. the `struct strbuf` was set to `STRBUF_INIT`),
+then they will free() it.
+
+`strbuf_addch`::
+
+	Add a single character to the buffer.
+
+`strbuf_insert`::
+
+	Insert data to the given position of the buffer. The remaining contents
+	will be shifted, not overwritten.
+
+`strbuf_remove`::
+
+	Remove given amount of data from a given position of the buffer.
+
+`strbuf_splice`::
+
+	Remove the bytes between `pos..pos+len` and replace it with the given
+	data.
+
+`strbuf_add`::
+
+	Add data of given length to the buffer.
+
+`strbuf_addstr`::
+
+Add a NUL-terminated string to the buffer.
++
+NOTE: This function will *always* be implemented as an inline or a macro
+that expands to:
++
+----
+strbuf_add(..., s, strlen(s));
+----
++
+Meaning that this is efficient to write things like:
++
+----
+strbuf_addstr(sb, "immediate string");
+----
+
+`strbuf_addbuf`::
+
+	Copy the contents of an other buffer at the end of the current one.
+
+`strbuf_adddup`::
+
+	Copy part of the buffer from a given position till a given length to the
+	end of the buffer.
+
+`strbuf_expand`::
+
+	This function can be used to expand a format string containing
+	placeholders. To that end, it parses the string and calls the specified
+	function for every percent sign found.
++
+The callback function is given a pointer to the character after the `%`
+and a pointer to the struct strbuf.  It is expected to add the expanded
+version of the placeholder to the strbuf, e.g. to add a newline
+character if the letter `n` appears after a `%`.  The function returns
+the length of the placeholder recognized and `strbuf_expand()` skips
+over it.
++
+All other characters (non-percent and not skipped ones) are copied
+verbatim to the strbuf.  If the callback returned zero, meaning that the
+placeholder is unknown, then the percent sign is copied, too.
++
+In order to facilitate caching and to make it possible to give
+parameters to the callback, `strbuf_expand()` passes a context pointer,
+which can be used by the programmer of the callback as she sees fit.
+
+`strbuf_expand_dict_cb`::
+
+	Used as callback for `strbuf_expand()`, expects an array of
+	struct strbuf_expand_dict_entry as context, i.e. pairs of
+	placeholder and replacement string.  The array needs to be
+	terminated by an entry with placeholder set to NULL.
+
+`strbuf_addf`::
+
+	Add a formatted string to the buffer.
+
+`strbuf_fread`::
+
+	Read a given size of data from a FILE* pointer to the buffer.
++
+NOTE: The buffer is rewound if the read fails. If -1 is returned,
+`errno` must be consulted, like you would do for `read(3)`.
+`strbuf_read()`, `strbuf_read_file()` and `strbuf_getline()` has the
+same behaviour as well.
+
+`strbuf_read`::
+
+	Read the contents of a given file descriptor. The third argument can be
+	used to give a hint about the file size, to avoid reallocs.
+
+`strbuf_read_file`::
+
+	Read the contents of a file, specified by its path. The third argument
+	can be used to give a hint about the file size, to avoid reallocs.
+
+`strbuf_readlink`::
+
+	Read the target of a symbolic link, specified by its path.  The third
+	argument can be used to give a hint about the size, to avoid reallocs.
+
+`strbuf_getline`::
+
+	Read a line from a FILE* pointer. The second argument specifies the line
+	terminator character, typically `'\n'`.
+
+`stripspace`::
+
+	Strip whitespace from a buffer. The second parameter controls if
+	comments are considered contents to be removed or not.
+
+`launch_editor`::
diff --git a/Documentation/technical/api-string-list.txt b/Documentation/technical/api-string-list.txt
new file mode 100644
index 0000000000..293bb15d20
--- /dev/null
+++ b/Documentation/technical/api-string-list.txt
@@ -0,0 +1,128 @@
+string-list API
+===============
+
+The string_list API offers a data structure and functions to handle sorted
+and unsorted string lists.
+
+The 'string_list' struct used to be called 'path_list', but was renamed
+because it is not specific to paths.
+
+The caller:
+
+. Allocates and clears a `struct string_list` variable.
+
+. Initializes the members. You might want to set the flag `strdup_strings`
+  if the strings should be strdup()ed. For example, this is necessary
+  when you add something like git_path("..."), since that function returns
+  a static buffer that will change with the next call to git_path().
++
+If you need something advanced, you can manually malloc() the `items`
+member (you need this if you add things later) and you should set the
+`nr` and `alloc` members in that case, too.
+
+. Adds new items to the list, using `string_list_append` or
+  `string_list_insert`.
+
+. Can check if a string is in the list using `string_list_has_string` or
+  `unsorted_string_list_has_string` and get it from the list using
+  `string_list_lookup` for sorted lists.
+
+. Can sort an unsorted list using `sort_string_list`.
+
+. Finally it should free the list using `string_list_clear`.
+
+Example:
+
+----
+struct string_list list;
+int i;
+
+memset(&list, 0, sizeof(struct string_list));
+string_list_append("foo", &list);
+string_list_append("bar", &list);
+for (i = 0; i < list.nr; i++)
+	printf("%s\n", list.items[i].string)
+----
+
+NOTE: It is more efficient to build an unsorted list and sort it
+afterwards, instead of building a sorted list (`O(n log n)` instead of
+`O(n^2)`).
++
+However, if you use the list to check if a certain string was added
+already, you should not do that (using unsorted_string_list_has_string()),
+because the complexity would be quadratic again (but with a worse factor).
+
+Functions
+---------
+
+* General ones (works with sorted and unsorted lists as well)
+
+`print_string_list`::
+
+	Dump a string_list to stdout, useful mainly for debugging purposes. It
+	can take an optional header argument and it writes out the
+	string-pointer pairs of the string_list, each one in its own line.
+
+`string_list_clear`::
+
+	Free a string_list. The `string` pointer of the items will be freed in
+	case the `strdup_strings` member of the string_list is set. The second
+	parameter controls if the `util` pointer of the items should be freed
+	or not.
+
+* Functions for sorted lists only
+
+`string_list_has_string`::
+
+	Determine if the string_list has a given string or not.
+
+`string_list_insert`::
+
+	Insert a new element to the string_list. The returned pointer can be
+	handy if you want to write something to the `util` pointer of the
+	string_list_item containing the just added string.
++
+Since this function uses xrealloc() (which die()s if it fails) if the
+list needs to grow, it is safe not to check the pointer. I.e. you may
+write `string_list_insert(...)->util = ...;`.
+
+`string_list_lookup`::
+
+	Look up a given string in the string_list, returning the containing
+	string_list_item. If the string is not found, NULL is returned.
+
+* Functions for unsorted lists only
+
+`string_list_append`::
+
+	Append a new string to the end of the string_list.
+
+`sort_string_list`::
+
+	Make an unsorted list sorted.
+
+`unsorted_string_list_has_string`::
+
+	It's like `string_list_has_string()` but for unsorted lists.
++
+This function needs to look through all items, as opposed to its
+counterpart for sorted lists, which performs a binary search.
+
+Data structures
+---------------
+
+* `struct string_list_item`
+
+Represents an item of the list. The `string` member is a pointer to the
+string, and you may use the `util` member for any purpose, if you want.
+
+* `struct string_list`
+
+Represents the list itself.
+
+. The array of items are available via the `items` member.
+. The `nr` member contains the number of items stored in the list.
+. The `alloc` member is used to avoid reallocating at every insertion.
+  You should not tamper with it.
+. Setting the `strdup_strings` member to 1 will strdup() the strings
+  before adding them, see above.
diff --git a/Documentation/technical/api-tree-walking.txt b/Documentation/technical/api-tree-walking.txt
new file mode 100644
index 0000000000..e3ddf91284
--- /dev/null
+++ b/Documentation/technical/api-tree-walking.txt
@@ -0,0 +1,12 @@
+tree walking API
+================
+
+Talk about <tree-walk.h>, things like
+
+* struct tree_desc
+* init_tree_desc
+* tree_entry_extract
+* update_tree_entry
+* get_tree_entry
+
+(JC, Linus)
diff --git a/Documentation/technical/api-xdiff-interface.txt b/Documentation/technical/api-xdiff-interface.txt
new file mode 100644
index 0000000000..6296ecad1d
--- /dev/null
+++ b/Documentation/technical/api-xdiff-interface.txt
@@ -0,0 +1,7 @@
+xdiff interface API
+===================
+
+Talk about our calling convention to xdiff library, including
+xdiff_emit_consume_fn.
+
+(Dscho, JC)
diff --git a/Documentation/technical/pack-format.txt b/Documentation/technical/pack-format.txt
index e5b31c81fa..1803e64e46 100644
--- a/Documentation/technical/pack-format.txt
+++ b/Documentation/technical/pack-format.txt
@@ -1,9 +1,9 @@
 GIT pack format
 ===============
 
-= pack-*.pack file has the following format:
+= pack-*.pack files have the following format:
 
-   - The header appears at the beginning and consists of the following:
+   - A header appears at the beginning and consists of the following:
 
      4-byte signature:
          The signature is: {'P', 'A', 'C', 'K'}
@@ -34,18 +34,14 @@ GIT pack format
 
   - The trailer records 20-byte SHA1 checksum of all of the above.
 
-= pack-*.idx file has the following format:
+= Original (version 1) pack-*.idx files have the following format:
 
   - The header consists of 256 4-byte network byte order
     integers.  N-th entry of this table records the number of
     objects in the corresponding pack, the first byte of whose
-    object name are smaller than N.  This is called the
+    object name is less than or equal to N.  This is called the
     'first-level fan-out' table.
 
-    Observation: we would need to extend this to an array of
-    8-byte integers to go beyond 4G objects per pack, but it is
-    not strictly necessary.
-
   - The header is followed by sorted 24-byte entries, one entry
     per object in the pack.  Each entry is:
 
@@ -55,10 +51,6 @@ GIT pack format
 
     20-byte object name.
 
-    Observation: we would definitely need to extend this to
-    8-byte integer plus 20-byte object name to handle a packfile
-    that is larger than 4GB.
-
   - The file is concluded with a trailer:
 
     A copy of the 20-byte SHA1 checksum at the end of
@@ -68,31 +60,30 @@ GIT pack format
 
 Pack Idx file:
 
-	idx
-	    +--------------------------------+
-	    | fanout[0] = 2                  |-.
-	    +--------------------------------+ |
+	--  +--------------------------------+
+fanout	    | fanout[0] = 2 (for example)    |-.
+table	    +--------------------------------+ |
 	    | fanout[1]                      | |
 	    +--------------------------------+ |
 	    | fanout[2]                      | |
 	    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
-	    | fanout[255]                    | |
-	    +--------------------------------+ |
-main	    | offset                         | |
-index	    | object name 00XXXXXXXXXXXXXXXX | |
-table	    +--------------------------------+ |
-	    | offset                         | |
-	    | object name 00XXXXXXXXXXXXXXXX | |
-	    +--------------------------------+ |
-	  .-| offset                         |<+
-	  | | object name 01XXXXXXXXXXXXXXXX |
-	  | +--------------------------------+
-	  | | offset                         |
-	  | | object name 01XXXXXXXXXXXXXXXX |
-	  | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-	  | | offset                         |
-	  | | object name FFXXXXXXXXXXXXXXXX |
-	  | +--------------------------------+
+	    | fanout[255] = total objects    |---.
+	--  +--------------------------------+ | |
+main	    | offset                         | | |
+index	    | object name 00XXXXXXXXXXXXXXXX | | |
+table	    +--------------------------------+ | |
+	    | offset                         | | |
+	    | object name 00XXXXXXXXXXXXXXXX | | |
+	    +--------------------------------+<+ |
+	  .-| offset                         |   |
+	  | | object name 01XXXXXXXXXXXXXXXX |   |
+	  | +--------------------------------+   |
+	  | | offset                         |   |
+	  | | object name 01XXXXXXXXXXXXXXXX |   |
+	  | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~   |
+	  | | offset                         |   |
+	  | | object name FFXXXXXXXXXXXXXXXX |   |
+	--| +--------------------------------+<--+
 trailer	  | | packfile checksum              |
 	  | +--------------------------------+
 	  | | idxfile checksum               |
@@ -112,7 +103,58 @@ Pack file entry: <+
      packed object data:
         If it is not DELTA, then deflated bytes (the size above
 		is the size before compression).
-	If it is DELTA, then
+	If it is REF_DELTA, then
 	  20-byte base object name SHA1 (the size above is the
 		size of the delta data that follows).
           delta data, deflated.
+	If it is OFS_DELTA, then
+	  n-byte offset (see below) interpreted as a negative
+		offset from the type-byte of the header of the
+		ofs-delta entry (the size above is the size of
+		the delta data that follows).
+	  delta data, deflated.
+
+     offset encoding:
+	  n bytes with MSB set in all but the last one.
+	  The offset is then the number constructed by
+	  concatenating the lower 7 bit of each byte, and
+	  for n >= 2 adding 2^7 + 2^14 + ... + 2^(7*(n-1))
+	  to the result.
+
+
+
+= Version 2 pack-*.idx files support packs larger than 4 GiB, and
+  have some other reorganizations.  They have the format:
+
+  - A 4-byte magic number '\377tOc' which is an unreasonable
+    fanout[0] value.
+
+  - A 4-byte version number (= 2)
+
+  - A 256-entry fan-out table just like v1.
+
+  - A table of sorted 20-byte SHA1 object names.  These are
+    packed together without offset values to reduce the cache
+    footprint of the binary search for a specific object name.
+
+  - A table of 4-byte CRC32 values of the packed object data.
+    This is new in v2 so compressed data can be copied directly
+    from pack to pack during repacking without undetected
+    data corruption.
+
+  - A table of 4-byte offset values (in network byte order).
+    These are usually 31-bit pack file offsets, but large
+    offsets are encoded as an index into the next table with
+    the msbit set.
+
+  - A table of 8-byte offset entries (empty for pack files less
+    than 2 GiB).  Pack files are organized with heavily used
+    objects toward the front, so most object references should
+    not need to refer to this table.
+
+  - The same trailer as a v1 pack file:
+
+    A copy of the 20-byte SHA1 checksum at the end of
+    corresponding packfile.
+
+    20-byte SHA1-checksum of all of the above.
diff --git a/Documentation/technical/racy-git.txt b/Documentation/technical/racy-git.txt
index 5030d9f2f8..48bb97f0b1 100644
--- a/Documentation/technical/racy-git.txt
+++ b/Documentation/technical/racy-git.txt
@@ -135,7 +135,7 @@ them, and give the same timestamp to the index file:
 
 This will make all index entries racily clean.  The linux-2.6
 project, for example, there are over 20,000 files in the working
-tree.  On my Athron 64X2 3800+, after the above:
+tree.  On my Athlon 64 X2 3800+, after the above:
 
   $ /usr/bin/time git diff-files
   1.68user 0.54system 0:02.22elapsed 100%CPU (0avgtext+0avgdata 0maxresident)k
@@ -184,7 +184,7 @@ In a large project where raciness avoidance cost really matters,
 however, the initial computation of all object names in the
 index takes more than one second, and the index file is written
 out after all that happens.  Therefore the timestamp of the
-index file will be more than one seconds later than the the
+index file will be more than one seconds later than the
 youngest file in the working tree.  This means that in these
 cases there actually will not be any racily clean entry in
 the resulting index.