Major rename detection changes

After doing further profiling, I found that a lot of time was
being spent attempting to insert hashes into the file hash
signature when using the rolling hash because the rolling hash
approach generates a hash per byte of the file instead of one
per run/line of data.

To optimize this, I decided to convert back to a run-based file
signature algorithm which would be more like core Git.

After changing this, a number of the existing tests started to
fail.  In some cases, this appears to have been because the test
was coded to be too specific to the particular results of the file
similarity metric and in some cases there appear to have been bugs
in the core rename detection code where only by the coincidence
of the file similarity scoring were the expected results being
generated.

This renames all the variables in the core rename detection code
to be more consistent and hopefully easier to follow which made it
a bit easier to reason about the behavior of that code and fix the
problems that I was seeing.  I think it's in better shape now.

There are a couple of tests now that attempt to stress test the
rename detection code and they are quite slow.  Most of the time
is spent setting up the test data on disk and in the index.  When
we roll out performance improvements for index insertion, it
should also speed up these tests I hope.

1 files changed, 95 insertions, 123 deletions

diff --git a/src/hashsig.c b/src/hashsig.c
index a11c4bee7..109f966ba 100644
--- a/src/hashsig.c
+++ b/src/hashsig.c
@@ -13,12 +13,15 @@ typedef uint64_t hashsig_state;
 
 #define HASHSIG_SCALE 100
 
-#define HASHSIG_HASH_WINDOW 32
-#define HASHSIG_HASH_START	0
+#define HASHSIG_MAX_RUN 80
+#define HASHSIG_HASH_START	0x012345678ABCDEF0LL
 #define HASHSIG_HASH_SHIFT  5
-#define HASHSIG_HASH_MASK   0x7FFFFFFF
+
+#define HASHSIG_HASH_MIX(S,CH) \
+	(S) = ((S) << HASHSIG_HASH_SHIFT) - (S) + (hashsig_state)(CH)
 
 #define HASHSIG_HEAP_SIZE ((1 << 7) - 1)
+#define HASHSIG_HEAP_MIN_SIZE 4
 
 typedef int (*hashsig_cmp)(const void *a, const void *b, void *);
 
@@ -28,14 +31,6 @@ typedef struct {
 	hashsig_t values[HASHSIG_HEAP_SIZE];
 } hashsig_heap;
 
-typedef struct {
-	hashsig_state state, shift_n;
-	char window[HASHSIG_HASH_WINDOW];
-	int win_len, win_pos, saw_lf;
-} hashsig_in_progress;
-
-#define HASHSIG_IN_PROGRESS_INIT { HASHSIG_HASH_START, 1, {0}, 0, 0, 1 }
-
 struct git_hashsig {
 	hashsig_heap mins;
 	hashsig_heap maxs;
@@ -115,142 +110,109 @@ static void hashsig_heap_sort(hashsig_heap *h)
 
 static void hashsig_heap_insert(hashsig_heap *h, hashsig_t val)
 {
-	/* if heap is full, pop top if new element should replace it */
-	if (h->size == h->asize && h->cmp(&val, &h->values[0], NULL) > 0) {
-		h->size--;
-		h->values[0] = h->values[h->size];
-		hashsig_heap_down(h, 0);
-	}
-
 	/* if heap is not full, insert new element */
 	if (h->size < h->asize) {
 		h->values[h->size++] = val;
 		hashsig_heap_up(h, h->size - 1);
 	}
-}
-
-GIT_INLINE(bool) hashsig_include_char(
-	char ch, git_hashsig_option_t opt, int *saw_lf)
-{
-	if ((opt & GIT_HASHSIG_IGNORE_WHITESPACE) && git__isspace(ch))
-		return false;
-
-	if (opt & GIT_HASHSIG_SMART_WHITESPACE) {
-		if (ch == '\r' || (*saw_lf && git__isspace(ch)))
-			return false;
 
-		*saw_lf = (ch == '\n');
+	/* if heap is full, pop top if new element should replace it */
+	else if (h->cmp(&val, &h->values[0], NULL) > 0) {
+		h->size--;
+		h->values[0] = h->values[h->size];
+		hashsig_heap_down(h, 0);
 	}
 
-	return true;
 }
 
-static void hashsig_initial_window(
-	git_hashsig *sig,
-	const char **data,
-	size_t size,
-	hashsig_in_progress *prog)
-{
-	hashsig_state state, shift_n;
-	int win_len, saw_lf = prog->saw_lf;
-	const char *scan, *end;
-	char *window = &prog->window[0];
-
-	/* init until we have processed at least HASHSIG_HASH_WINDOW data */
-
-	if (prog->win_len >= HASHSIG_HASH_WINDOW)
-		return;
-
-	state   = prog->state;
-	win_len = prog->win_len;
-	shift_n = prog->shift_n;
-
-	scan = *data;
-	end  = scan + size;
-
-	while (scan < end && win_len < HASHSIG_HASH_WINDOW) {
-		char ch = *scan++;
-
-		if (!hashsig_include_char(ch, sig->opt, &saw_lf))
-			continue;
-
-		state = (state * HASHSIG_HASH_SHIFT + ch) & HASHSIG_HASH_MASK;
-
-		if (!win_len)
-			shift_n = 1;
-		else
-			shift_n = (shift_n * HASHSIG_HASH_SHIFT) & HASHSIG_HASH_MASK;
-
-		window[win_len++] = ch;
-	}
-
-	/* insert initial hash if we just finished */
+typedef struct {
+	int use_ignores;
+	uint8_t ignore_ch[256];
+} hashsig_in_progress;
 
-	if (win_len == HASHSIG_HASH_WINDOW) {
-		hashsig_heap_insert(&sig->mins, (hashsig_t)state);
-		hashsig_heap_insert(&sig->maxs, (hashsig_t)state);
-		sig->considered = 1;
+static void hashsig_in_progress_init(
+	hashsig_in_progress *prog, git_hashsig *sig)
+{
+	int i;
+
+	switch (sig->opt) {
+	case GIT_HASHSIG_IGNORE_WHITESPACE:
+		for (i = 0; i < 256; ++i)
+			prog->ignore_ch[i] = git__isspace_nonlf(i);
+		prog->use_ignores = 1;
+		break;
+	case GIT_HASHSIG_SMART_WHITESPACE:
+		for (i = 0; i < 256; ++i)
+			prog->ignore_ch[i] = git__isspace(i);
+		prog->use_ignores = 1;
+		break;
+	default:
+		memset(prog, 0, sizeof(*prog));
+		break;
 	}
-
-	prog->state   = state;
-	prog->win_len = win_len;
-	prog->shift_n = shift_n;
-	prog->saw_lf  = saw_lf;
-
-	*data = scan;
 }
 
+#define HASHSIG_IN_PROGRESS_INIT { 1 }
+
 static int hashsig_add_hashes(
 	git_hashsig *sig,
-	const char *data,
+	const uint8_t *data,
 	size_t size,
 	hashsig_in_progress *prog)
 {
-	const char *scan = data, *end = data + size;
-	hashsig_state state, shift_n, rmv;
-	int win_pos, saw_lf;
-	char *window = &prog->window[0];
-
-	if (prog->win_len < HASHSIG_HASH_WINDOW)
-		hashsig_initial_window(sig, &scan, size, prog);
-
-	state   = prog->state;
-	shift_n = prog->shift_n;
-	saw_lf  = prog->saw_lf;
-	win_pos = prog->win_pos;
-
-	/* advance window, adding new chars and removing old */
-
-	for (; scan < end; ++scan) {
-		char ch = *scan;
-
-		if (!hashsig_include_char(ch, sig->opt, &saw_lf))
-			continue;
-
-		rmv = shift_n * window[win_pos];
+	const uint8_t *scan = data, *end = data + size;
+	hashsig_state state = HASHSIG_HASH_START;
+	int use_ignores = prog->use_ignores, len;
+	uint8_t ch;
+
+	while (scan < end) {
+		state = HASHSIG_HASH_START;
+
+		for (len = 0; scan < end && len < HASHSIG_MAX_RUN; ) {
+			ch = *scan;
+
+			if (use_ignores)
+				for (; scan < end && git__isspace_nonlf(ch); ch = *scan)
+					++scan;
+			else if (sig->opt != GIT_HASHSIG_NORMAL)
+				for (; scan < end && ch == '\r'; ch = *scan)
+					++scan;
+
+			/* peek at next character to decide what to do next */
+			if (sig->opt == GIT_HASHSIG_SMART_WHITESPACE)
+				use_ignores = (ch == '\n');
+
+			if (scan >= end)
+				break;
+			++scan;
+
+			/* check run terminator */
+			if (ch == '\n' || ch == '\0')
+				break;
+
+			++len;
+			HASHSIG_HASH_MIX(state, ch);
+		}
 
-		state = (state - rmv) & HASHSIG_HASH_MASK;
-		state = (state * HASHSIG_HASH_SHIFT) & HASHSIG_HASH_MASK;
-		state = (state + ch) & HASHSIG_HASH_MASK;
+		if (len > 0) {
+			hashsig_heap_insert(&sig->mins, (hashsig_t)state);
+			hashsig_heap_insert(&sig->maxs, (hashsig_t)state);
 
-		hashsig_heap_insert(&sig->mins, (hashsig_t)state);
-		hashsig_heap_insert(&sig->maxs, (hashsig_t)state);
-		sig->considered++;
+			sig->considered++;
 
-		window[win_pos] = ch;
-		win_pos = (win_pos + 1) % HASHSIG_HASH_WINDOW;
+			while (scan < end && (*scan == '\n' || !*scan))
+				++scan;
+		}
 	}
 
-	prog->state   = state;
-	prog->saw_lf  = saw_lf;
-	prog->win_pos = win_pos;
+	prog->use_ignores = use_ignores;
 
 	return 0;
 }
 
 static int hashsig_finalize_hashes(git_hashsig *sig)
 {
-	if (sig->mins.size < HASHSIG_HEAP_SIZE) {
+	if (sig->mins.size < HASHSIG_HEAP_MIN_SIZE) {
 		giterr_set(GITERR_INVALID,
 			"File too small for similarity signature calculation");
 		return GIT_EBUFS;
@@ -282,11 +244,13 @@ int git_hashsig_create(
 	git_hashsig_option_t opts)
 {
 	int error;
-	hashsig_in_progress prog = HASHSIG_IN_PROGRESS_INIT;
+	hashsig_in_progress prog;
 	git_hashsig *sig = hashsig_alloc(opts);
 	GITERR_CHECK_ALLOC(sig);
 
-	error = hashsig_add_hashes(sig, buf, buflen, &prog);
+	hashsig_in_progress_init(&prog, sig);
+
+	error = hashsig_add_hashes(sig, (const uint8_t *)buf, buflen, &prog);
 
 	if (!error)
 		error = hashsig_finalize_hashes(sig);
@@ -304,10 +268,10 @@ int git_hashsig_create_fromfile(
 	const char *path,
 	git_hashsig_option_t opts)
 {
-	char buf[0x1000];
+	uint8_t buf[0x1000];
 	ssize_t buflen = 0;
 	int error = 0, fd;
-	hashsig_in_progress prog = HASHSIG_IN_PROGRESS_INIT;
+	hashsig_in_progress prog;
 	git_hashsig *sig = hashsig_alloc(opts);
 	GITERR_CHECK_ALLOC(sig);
 
@@ -316,6 +280,8 @@ int git_hashsig_create_fromfile(
 		return fd;
 	}
 
+	hashsig_in_progress_init(&prog, sig);
+
 	while (!error) {
 		if ((buflen = p_read(fd, buf, sizeof(buf))) <= 0) {
 			if ((error = (int)buflen) < 0)
@@ -370,6 +336,12 @@ static int hashsig_heap_compare(const hashsig_heap *a, const hashsig_heap *b)
 
 int git_hashsig_compare(const git_hashsig *a, const git_hashsig *b)
 {
-	return (hashsig_heap_compare(&a->mins, &b->mins) +
-			hashsig_heap_compare(&a->maxs, &b->maxs)) / 2;
+	/* if we have fewer than the maximum number of elements, then just use
+	 * one array since the two arrays will be the same
+	 */
+	if (a->mins.size < HASHSIG_HEAP_SIZE)
+		return hashsig_heap_compare(&a->mins, &b->mins);
+	else
+		return (hashsig_heap_compare(&a->mins, &b->mins) +
+				hashsig_heap_compare(&a->maxs, &b->maxs)) / 2;
 }