Merge branch 'scalar'

This merges the upstreamable part of the Scalar patches.

Minor merge conflicts (caused by the gvfs-helper) were resolved
trivially.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>

Documentation/config/core.txt

@@ -806,3 +806,12 @@ core.abbrev::
 	If set to "no", no abbreviation is made and the object names
 	are shown in their full length.
 	The minimum length is 4.
+
+core.configWriteLockTimeoutMS::
+	When processes try to write to the config concurrently, it is likely
+	that one process "wins" and the other process(es) fail to lock the
+	config file. By configuring a timeout larger than zero, Git can be
+	told to try to lock the config again a couple times within the
+	specified timeout. If the timeout is configure to zero (which is the
+	default), Git will fail immediately when the config is already
+	locked.

config.c

@@ -3465,6 +3465,7 @@ int git_config_set_multivar_in_file_gently(const char *config_filename,
 					   const char *value_pattern,
 					   unsigned flags)
 {
+	static unsigned long timeout_ms = ULONG_MAX;
 	int fd = -1, in_fd = -1;
 	int ret;
 	struct lock_file lock = LOCK_INIT;
@@ -3485,11 +3486,16 @@ int git_config_set_multivar_in_file_gently(const char *config_filename,
 	if (!config_filename)
 		config_filename = filename_buf = git_pathdup("config");
 
+	if ((long)timeout_ms < 0 &&
+	    git_config_get_ulong("core.configWriteLockTimeoutMS", &timeout_ms))
+		timeout_ms = 0;
+
 	/*
 	 * The lock serves a purpose in addition to locking: the new
 	 * contents of .git/config will be written into it.
 	 */
-	fd = hold_lock_file_for_update(&lock, config_filename, 0);
+	fd = hold_lock_file_for_update_timeout(&lock, config_filename, 0,
+					       timeout_ms);
 	if (fd < 0) {
 		error_errno(_("could not lock config file %s"), config_filename);
 		ret = CONFIG_NO_LOCK;

contrib/scalar/docs/faq.md

@@ -0,0 +1,51 @@
+Frequently Asked Questions
+==========================
+
+Using Scalar
+------------
+
+### I don't want a sparse clone, I want every file after I clone!
+
+Run `scalar clone --full-clone <url>` to initialize your repo to include
+every file. You can switch to a sparse-checkout later by running
+`git sparse-checkout init --cone`.
+
+### I already cloned without `--full-clone`. How do I get everything?
+
+Run `git sparse-checkout disable`.
+
+Scalar Design Decisions
+-----------------------
+
+There may be many design decisions within Scalar that are confusing at first
+glance. Some of them may cause friction when you use Scalar with your existing
+repos and existing habits.
+
+> Scalar has the most benefit when users design repositories
+> with efficient patterns.
+
+For example: Scalar uses the sparse-checkout feature to limit the size of the
+working directory within a large monorepo. It is designed to work efficiently
+with monorepos that are highly componentized, allowing most developers to
+need many fewer files in their daily work.
+
+### Why does `scalar clone` create a `<repo>/src` folder?
+
+Scalar uses a file system watcher to keep track of changes under this `src` folder.
+Any activity in this folder is assumed to be important to Git operations. By
+creating the `src` folder, we are making it easy for your build system to
+create output folders outside the `src` directory. We commonly see systems
+create folders for build outputs and package downloads. Scalar itself creates
+these folders during its builds.
+
+Your build system may create build artifacts such as `.obj` or `.lib` files
+next to your source code. These are commonly "hidden" from Git using
+`.gitignore` files. Having such artifacts in your source tree creates
+additional work for Git because it needs to look at these files and match them
+against the `.gitignore` patterns.
+
+By following the `src` pattern Scalar tries to establish and placing your build
+intermediates and outputs parallel with the `src` folder and not inside it,
+you can help optimize Git command performance for developers in the repository
+by limiting the number of files Git needs to consider for many common
+operations.

contrib/scalar/docs/getting-started.md

@@ -0,0 +1,93 @@
+Getting Started
+===============
+
+Registering existing Git repos
+------------------------------
+
+To add a repository to the list of registered repos, run `scalar register [<path>]`.
+If `<path>` is not provided, then the "current repository" is discovered from
+the working directory by scanning the parent paths for a path containing a `.git`
+folder, possibly inside a `src` folder.
+
+To see which repositories are currently tracked by the service, run
+`scalar list`.
+
+Run `scalar unregister [<path>]` to remove the repo from this list.
+
+Creating a new Scalar clone
+---------------------------------------------------
+
+The `clone` verb creates a local enlistment of a remote repository using the
+partial clone feature available e.g. on GitHub.
+
+
+```
+scalar clone [options] <url> [<dir>]
+```
+
+Create a local copy of the repository at `<url>`. If specified, create the `<dir>`
+directory and place the repository there. Otherwise, the last section of the `<url>`
+will be used for `<dir>`.
+
+At the end, the repo is located at `<dir>/src`. By default, the sparse-checkout
+feature is enabled and the only files present are those in the root of your
+Git repository. Use `git sparse-checkout set` to expand the set of directories
+you want to see, or `git sparse-checkout disable` to expand to all files. You
+can explore the subdirectories outside your sparse-checkout specification using
+`git ls-tree HEAD`.
+
+### Sparse Repo Mode
+
+By default, Scalar reduces your working directory to only the files at the
+root of the repository. You need to add the folders you care about to build up
+to your working set.
+
+* `scalar clone <url>`
+  * Please choose the **Clone with HTTPS** option in the `Clone Repository` dialog in Azure Repos, not **Clone with SSH**.
+* `cd <root>\src`
+* At this point, your `src` directory only contains files that appear in your root
+  tree. No folders are populated.
+* Set the directory list for your sparse-checkout using:
+	1. `git sparse-checkout set <dir1> <dir2> ...`
+	2. `git sparse-checkout set --stdin < dir-list.txt`
+* Run git commands as you normally would.
+* To fully populate your working directory, run `git sparse-checkout disable`.
+
+If instead you want to start with all files on-disk, you can clone with the
+`--full-clone` option. To enable sparse-checkout after the fact, run
+`git sparse-checkout init --cone`. This will initialize your sparse-checkout
+patterns to only match the files at root.
+
+If you are unfamiliar with what directories are available in the repository,
+then you can run `git ls-tree -d --name-only HEAD` to discover the directories
+at root, or `git ls-tree -d --name-only HEAD <path>` to discover the directories
+in `<path>`.
+
+### Options
+
+These options allow a user to customize their initial enlistment.
+
+* `--full-clone`: If specified, do not initialize the sparse-checkout feature.
+  All files will be present in your `src` directory. This uses a Git partial
+  clone: blobs are downloaded on demand.
+
+* `--branch=<ref>`: Specify the branch to checkout after clone.
+
+### Advanced Options
+
+The options below are not intended for use by a typical user. These are
+usually used by build machines to create a temporary enlistment that
+operates on a single commit.
+
+* `--single-branch`: Use this option to only download metadata for the branch
+  that will be checked out. This is helpful for build machines that target
+  a remote with many branches. Any `git fetch` commands after the clone will
+  still ask for all branches.
+
+Removing a Scalar Clone
+-----------------------
+
+Since the `scalar clone` command sets up a file-system watcher (when available),
+that watcher could prevent deleting the enlistment. Run `scalar delete <path>`
+from outside of your enlistment to unregister the enlistment from the filesystem
+watcher and delete the enlistment at `<path>`.

contrib/scalar/docs/index.md

@@ -0,0 +1,50 @@
+Scalar: Enabling Git at Scale
+=============================
+
+Scalar is a tool that helps Git scale to some of the largest Git repositories.
+It achieves this by enabling some advanced Git features, such as:
+
+* *Partial clone:* reduces time to get a working repository by not
+  downloading all Git objects right away.
+
+* *Background prefetch:* downloads Git object data from all remotes every
+  hour, reducing the amount of time for foreground `git fetch` calls.
+
+* *Sparse-checkout:* limits the size of your working directory.
+
+* *File system monitor:* tracks the recently modified files and eliminates
+  the need for Git to scan the entire worktree.
+
+* *Commit-graph:* accelerates commit walks and reachability calculations,
+   speeding up commands like `git log`.
+
+* *Multi-pack-index:* enables fast object lookups across many pack-files.
+
+* *Incremental repack:* Repacks the packed Git data into fewer pack-file
+  without disrupting concurrent commands by using the multi-pack-index.
+
+By running `scalar register` in any Git repo, Scalar will automatically enable
+these features for that repo (except partial clone) and start running suggested
+maintenance in the background using
+[the `git maintenance` feature](https://git-scm.com/docs/git-maintenance).
+
+Repos cloned with the `scalar clone` command use partial clone to significantly
+reduce the amount of data required to get started using a repository. By
+delaying all blob downloads until they are required, Scalar allows you to work
+with very large repositories quickly.
+
+Documentation
+-------------
+
+* [Getting Started](getting-started.md): Get started with Scalar.
+  Includes `scalar register`, `scalar unregister`, `scalar clone`, and
+  `scalar delete`.
+
+* [Troubleshooting](troubleshooting.md):
+  Collect diagnostic information or update custom settings. Includes
+  `scalar diagnose`.
+
+* [The Philosophy of Scalar](philosophy.md): Why does Scalar work the way
+  it does, and how do we make decisions about its future?
+
+* [Frequently Asked Questions](faq.md)

contrib/scalar/docs/philosophy.md

@@ -0,0 +1,66 @@
+The Philosophy of Scalar
+========================
+
+The team building Scalar has **opinions** about Git performance. Scalar
+takes out the guesswork by automatically configuring your Git repositories
+to take advantage of the latest and greatest features. It is difficult to
+say that these are the absolute best settings for every repository, but
+these settings do work for some of the largest repositories in the world.
+
+Scalar intends to do very little more than the standard Git client. We
+actively implement new features into Git instead of Scalar, then update
+Scalar only to configure those new settings. In particular, we ported
+features like background maintenance to Git to make Scalar simpler and
+make Git more powerful.
+
+Services such as GitHub support partial clone , a standard adopted by the Git
+project to download only part of the Git objects when cloning, and fetching
+further objects on demand. If your hosting service supports partial clone, then
+we absolutely recommend it as a way to greatly speed up your clone and fetch
+times and to reduce how much disk space your Git repository requires. Scalar
+will help with this!
+
+Most of the value of Scalar can be found in the core Git client. However, most
+of the advanced features that really optimize Git's performance are off by
+default for compatibility reasons. To really take advantage of Git's latest and
+greatest features, you either need to study the [`git config`
+documentation](https://git-scm.com/docs/git-config) and regularly read [the Git
+release notes](https://github.com/git/git/tree/master/Documentation/RelNotes).
+Even if you do all that work and customize your Git settings on your machines,
+you likely will want to share those settings with other team members. Or, you
+can just use Scalar!
+
+Using `scalar register` on an existing Git repository will give you these
+benefits:
+
+* Additional compression of your `.git/index` file.
+* Hourly background `git fetch` operations, keeping you in-sync with your
+  remotes.
+* Advanced data structures, such as the `commit-graph` and `multi-pack-index`
+  are updated automatically in the background.
+* If using macOS or Windows, then Scalar configures Git's builtin File System
+  Monitor, providing faster commands such as `git status` or `git add`.
+
+Additionally, if you use `scalar clone` to create a new repository, then
+you will automatically get these benefits:
+
+* Use Git's partial clone feature to only download the files you need for
+  your current checkout.
+* Use Git's [sparse-checkout feature][sparse-checkout] to minimize the
+  number of files required in your working directory.
+  [Read more about sparse-checkout here.][sparse-checkout-blog]
+* Create the Git repository inside `<repo-name>/src` to make it easy to
+  place build artifacts outside of the Git repository, such as in
+  `<repo-name>/bin` or `<repo-name>/packages`.
+
+We also admit that these **opinions** can always be improved! If you have
+an idea of how to improve our setup, consider
+[creating an issue](https://github.com/microsoft/scalar/issues/new) or
+contributing a pull request! Some [existing](https://github.com/microsoft/scalar/issues/382)
+[issues](https://github.com/microsoft/scalar/issues/388) have already
+improved our configuration settings and roadmap!
+
+[gvfs-protocol]: https://github.com/microsoft/VFSForGit/blob/HEAD/Protocol.md
+[microsoft-git]: https://github.com/microsoft/git
+[sparse-checkout]: https://git-scm.com/docs/git-sparse-checkout
+[sparse-checkout-blog]: https://github.blog/2020-01-17-bring-your-monorepo-down-to-size-with-sparse-checkout/

contrib/scalar/docs/troubleshooting.md

@@ -0,0 +1,20 @@
+Troubleshooting
+===============
+
+Diagnosing Issues
+-----------------
+
+The `scalar diagnose` command collects logs and config details for the current
+repository. The resulting zip file helps root-cause issues.
+
+When run inside your repository, creates a zip file containing several important
+files for that repository. This includes:
+
+* Configuration files from your `.git` folder, such as the `config` file,
+  `index`, `hooks`, and `refs`.
+
+* A summary of your Git object database, including the number of loose objects
+  and the names and sizes of pack-files.
+
+As the `diagnose` command completes, it provides the path of the resulting
+zip file. This zip can be attached to bug reports to make the analysis easier.

scalar.c

@@ -44,6 +44,9 @@ static void setup_enlistment_directory(int argc, const char **argv,
 		die(_("need a working directory"));
 
 	strbuf_trim_trailing_dir_sep(&path);
+#ifdef GIT_WINDOWS_NATIVE
+	convert_slashes(path.buf);
+#endif
 
 	/* check if currently in enlistment root with src/ workdir */
 	len = path.len;
@@ -70,20 +73,33 @@ static void setup_enlistment_directory(int argc, const char **argv,
 	strbuf_release(&path);
 }
 
+static int git_retries = 3;
+
 static int run_git(const char *arg, ...)
 {
-	struct child_process cmd = CHILD_PROCESS_INIT;
 	va_list args;
 	const char *p;
+	struct strvec argv = STRVEC_INIT;
+	int res = 0, attempts;
 
 	va_start(args, arg);
-	strvec_push(&cmd.args, arg);
+	strvec_push(&argv, arg);
 	while ((p = va_arg(args, const char *)))
-		strvec_push(&cmd.args, p);
+		strvec_push(&argv, p);
 	va_end(args);
 
-	cmd.git_cmd = 1;
-	return run_command(&cmd);
+	for (attempts = 0, res = 1;
+	     res && attempts < git_retries;
+	     attempts++) {
+		struct child_process cmd = CHILD_PROCESS_INIT;
+
+		cmd.git_cmd = 1;
+		strvec_pushv(&cmd.args, argv.v);
+		res = run_command(&cmd);
+	}
+
+	strvec_clear(&argv);
+	return res;
 }
 
 struct scalar_config {
@@ -167,6 +183,7 @@ static int set_recommended_config(int reconfigure)
 		{ "core.autoCRLF", "false" },
 		{ "core.safeCRLF", "false" },
 		{ "fetch.showForcedUpdates", "false" },
+		{ "core.configWriteLockTimeoutMS", "150" },
 		{ NULL, NULL },
 	};
 	int i;
@@ -208,6 +225,11 @@ static int set_recommended_config(int reconfigure)
 
 static int toggle_maintenance(int enable)
 {
+	unsigned long ul;
+
+	if (git_config_get_ulong("core.configWriteLockTimeoutMS", &ul))
+		git_config_push_parameter("core.configWriteLockTimeoutMS=150");
+
 	return run_git("maintenance",
 		       enable ? "start" : "unregister",
 		       enable ? NULL : "--force",
@@ -217,10 +239,14 @@ static int toggle_maintenance(int enable)
 static int add_or_remove_enlistment(int add)
 {
 	int res;
+	unsigned long ul;
 
 	if (!the_repository->worktree)
 		die(_("Scalar enlistments require a worktree"));
 
+	if (git_config_get_ulong("core.configWriteLockTimeoutMS", &ul))
+		git_config_push_parameter("core.configWriteLockTimeoutMS=150");
+
 	res = run_git("config", "--global", "--get", "--fixed-value",
 		      "scalar.repo", the_repository->worktree, NULL);
 
@@ -559,6 +585,8 @@ static int cmd_diagnose(int argc, const char **argv)
 	setup_enlistment_directory(argc, argv, usage, options, &diagnostics_root);
 	strbuf_addstr(&diagnostics_root, "/.scalarDiagnostics");
 
+	/* Here, a failure should not repeat itself. */
+	git_retries = 1;
 	res = run_git("diagnose", "--mode=all", "-s", "%Y%m%d_%H%M%S",
 		      "-o", diagnostics_root.buf, NULL);
 
@@ -933,6 +961,9 @@ int cmd_main(int argc, const char **argv)
 		argv++;
 		argc--;
 
+		if (!strcmp(argv[0], "config"))
+			argv[0] = "reconfigure";
+
 		for (i = 0; builtins[i].name; i++)
 			if (!strcmp(builtins[i].name, argv[0]))
 				return !!builtins[i].fn(argc, argv);