git/fsmonitor.c

#define USE_THE_REPOSITORY_VARIABLE

#include "git-compat-util.h"
#include "config.h"
#include "dir.h"
#include "environment.h"
#include "ewah/ewok.h"
#include "fsmonitor.h"
#include "fsmonitor-ipc.h"
#include "name-hash.h"
#include "repository.h"
#include "run-command.h"
#include "strbuf.h"
#include "trace2.h"

#define INDEX_EXTENSION_VERSION1 …
#define INDEX_EXTENSION_VERSION2 …
#define HOOK_INTERFACE_VERSION1 …
#define HOOK_INTERFACE_VERSION2 …

struct trace_key trace_fsmonitor = …;

static void assert_index_minimum(struct index_state *istate, size_t pos)
{ … }

static void fsmonitor_ewah_callback(size_t pos, void *is)
{ … }

static int fsmonitor_hook_version(void)
{ … }

int read_fsmonitor_extension(struct index_state *istate, const void *data,
	unsigned long sz)
{ … }

void fill_fsmonitor_bitmap(struct index_state *istate)
{ … }

void write_fsmonitor_extension(struct strbuf *sb, struct index_state *istate)
{ … }

/*
 * Call the query-fsmonitor hook passing the last update token of the saved results.
 */
static int query_fsmonitor_hook(struct repository *r,
				int version,
				const char *last_update,
				struct strbuf *query_result)
{ … }

/*
 * Invalidate the FSM bit on this CE.  This is like mark_fsmonitor_invalid()
 * but we've already handled the untracked-cache, so let's not repeat that
 * work.  This also lets us have a different trace message so that we can
 * see everything that was done as part of the refresh-callback.
 */
static void invalidate_ce_fsm(struct cache_entry *ce)
{ … }

static size_t handle_path_with_trailing_slash(
	struct index_state *istate, const char *name, int pos);

/*
 * Use the name-hash to do a case-insensitive cache-entry lookup with
 * the pathname and invalidate the cache-entry.
 *
 * Returns the number of cache-entries that we invalidated.
 */
static size_t handle_using_name_hash_icase(
	struct index_state *istate, const char *name)
{ … }

/*
 * Use the dir-name-hash to find the correct-case spelling of the
 * directory.  Use the canonical spelling to invalidate all of the
 * cache-entries within the matching cone.
 *
 * Returns the number of cache-entries that we invalidated.
 */
static size_t handle_using_dir_name_hash_icase(
	struct index_state *istate, const char *name)
{ … }

/*
 * The daemon sent an observed pathname without a trailing slash.
 * (This is the normal case.)  We do not know if it is a tracked or
 * untracked file, a sparse-directory, or a populated directory (on a
 * platform such as Windows where FSEvents are not qualified).
 *
 * The pathname contains the observed case reported by the FS. We
 * do not know it is case-correct or -incorrect.
 *
 * Assume it is case-correct and try an exact match.
 *
 * Return the number of cache-entries that we invalidated.
 */
static size_t handle_path_without_trailing_slash(
	struct index_state *istate, const char *name, int pos)
{ … }

/*
 * The daemon can decorate directory events, such as a move or rename,
 * by adding a trailing slash to the observed name.  Use this to
 * explicitly invalidate the entire cone under that directory.
 *
 * The daemon can only reliably do that if the OS FSEvent contains
 * sufficient information in the event.
 *
 * macOS FSEvents have enough information.
 *
 * Other platforms may or may not be able to do it (and it might
 * depend on the type of event (for example, a daemon could lstat() an
 * observed pathname after a rename, but not after a delete)).
 *
 * If we find an exact match in the index for a path with a trailing
 * slash, it means that we matched a sparse-index directory in a
 * cone-mode sparse-checkout (since that's the only time we have
 * directories in the index).  We should never see this in practice
 * (because sparse directories should not be present and therefore
 * not generating FS events).  Either way, we can treat them in the
 * same way and just invalidate the cache-entry and the untracked
 * cache (and in this case, the forward cache-entry scan won't find
 * anything and it doesn't hurt to let it run).
 *
 * Return the number of cache-entries that we invalidated.  We will
 * use this later to determine if we need to attempt a second
 * case-insensitive search on case-insensitive file systems.  That is,
 * if the search using the observed-case in the FSEvent yields any
 * results, we assume the prefix is case-correct.  If there are no
 * matches, we still don't know if the observed path is simply
 * untracked or case-incorrect.
 */
static size_t handle_path_with_trailing_slash(
	struct index_state *istate, const char *name, int pos)
{ … }

static void fsmonitor_refresh_callback(struct index_state *istate, char *name)
{ … }

/*
 * The number of pathnames that we need to receive from FSMonitor
 * before we force the index to be updated.
 *
 * Note that any pathname within the set of received paths MAY cause
 * cache-entry or istate flag bits to be updated and thus cause the
 * index to be updated on disk.
 *
 * However, the response may contain many paths (such as ignored
 * paths) that will not update any flag bits.  And thus not force the
 * index to be updated.  (This is fine and normal.)  It also means
 * that the token will not be updated in the FSMonitor index
 * extension.  So the next Git command will find the same token in the
 * index, make the same token-relative request, and receive the same
 * response (plus any newly changed paths).  If this response is large
 * (and continues to grow), performance could be impacted.
 *
 * For example, if the user runs a build and it writes 100K object
 * files but doesn't modify any source files, the index would not need
 * to be updated.  The FSMonitor response (after the build and
 * relative to a pre-build token) might be 5MB.  Each subsequent Git
 * command will receive that same 100K/5MB response until something
 * causes the index to be updated.  And `refresh_fsmonitor()` will
 * have to iterate over those 100K paths each time.
 *
 * Performance could be improved if we optionally force update the
 * index after a very large response and get an updated token into
 * the FSMonitor index extension.  This should allow subsequent
 * commands to get smaller and more current responses.
 *
 * The value chosen here does not need to be precise.  The index
 * will be updated automatically the first time the user touches
 * a tracked file and causes a command like `git status` to
 * update an mtime to be updated and/or set a flag bit.
 */
static int fsmonitor_force_update_threshold = …;

void refresh_fsmonitor(struct index_state *istate)
{ … }

/*
 * The caller wants to turn on FSMonitor.  And when the caller writes
 * the index to disk, a FSMonitor extension should be included.  This
 * requires that `istate->fsmonitor_last_update` not be NULL.  But we
 * have not actually talked to a FSMonitor process yet, so we don't
 * have an initial value for this field.
 *
 * For a protocol V1 FSMonitor process, this field is a formatted
 * "nanoseconds since epoch" field.  However, for a protocol V2
 * FSMonitor process, this field is an opaque token.
 *
 * Historically, `add_fsmonitor()` has initialized this field to the
 * current time for protocol V1 processes.  There are lots of race
 * conditions here, but that code has shipped...
 *
 * The only true solution is to use a V2 FSMonitor and get a current
 * or default token value (that it understands), but we cannot do that
 * until we have actually talked to an instance of the FSMonitor process
 * (but the protocol requires that we send a token first...).
 *
 * For simplicity, just initialize like we have a V1 process and require
 * that V2 processes adapt.
 */
static void initialize_fsmonitor_last_update(struct index_state *istate)
{ … }

void add_fsmonitor(struct index_state *istate)
{ … }

void remove_fsmonitor(struct index_state *istate)
{ … }

void tweak_fsmonitor(struct index_state *istate)
{ … }