X-Git-Url: https://git.kernel.dk/?p=fio.git;a=blobdiff_plain;f=steadystate.c;h=bd2f70dd3af1b6084994b2409ef6d7f862e2aac4;hp=94d1f5e24da99882885b7ec14e807a92856d75eb;hb=4e8c82b4e9804c52bf2c78327cc5bfca9d8aedfc;hpb=bb49c8bd53e2b3fd7abe31fde742b5f2b1563ef5

diff --git a/steadystate.c b/steadystate.c
index 94d1f5e2..bd2f70dd 100644
--- a/steadystate.c
+++ b/steadystate.c
@@ -2,27 +2,29 @@
 
 #include "fio.h"
 #include "steadystate.h"
-#include "helper_thread.h"
 
 bool steadystate_enabled = false;
 
-static void steadystate_alloc(struct thread_data *td)
+void steadystate_free(struct thread_data *td)
 {
-	int i;
+	free(td->ss.iops_data);
+	free(td->ss.bw_data);
+	td->ss.iops_data = NULL;
+	td->ss.bw_data = NULL;
+}
 
-	td->ss.bw_data = malloc(td->ss.dur * sizeof(uint64_t));
-	td->ss.iops_data = malloc(td->ss.dur * sizeof(uint64_t));
-	/* initialize so that it is obvious if the cache is not full in the output */
-	for (i = 0; i < td->ss.dur; i++)
-		td->ss.iops_data[i] = td->ss.bw_data[i] = 0;
+static void steadystate_alloc(struct thread_data *td)
+{
+	td->ss.bw_data = calloc(td->ss.dur, sizeof(uint64_t));
+	td->ss.iops_data = calloc(td->ss.dur, sizeof(uint64_t));
 
-	td->ss.state |= __FIO_SS_DATA;
+	td->ss.state |= FIO_SS_DATA;
 }
 
 void steadystate_setup(void)
 {
-	int i, prev_groupid;
 	struct thread_data *td, *prev_td;
+	int i, prev_groupid;
 
 	if (!steadystate_enabled)
 		return;
@@ -44,57 +46,55 @@ void steadystate_setup(void)
 		}
 
 		if (prev_groupid != td->groupid) {
-			if (prev_td != NULL) {
+			if (prev_td)
 				steadystate_alloc(prev_td);
-			}
 			prev_groupid = td->groupid;
 		}
 		prev_td = td;
 	}
 
-	if (prev_td != NULL && prev_td->o.group_reporting) {
+	if (prev_td && prev_td->o.group_reporting)
 		steadystate_alloc(prev_td);
-	}
 }
 
-static bool steadystate_slope(unsigned long iops, unsigned long bw,
+static bool steadystate_slope(uint64_t iops, uint64_t bw,
 			      struct thread_data *td)
 {
 	int i, j;
 	double result;
 	struct steadystate_data *ss = &td->ss;
-	unsigned long new_val;
+	uint64_t new_val;
 
 	ss->bw_data[ss->tail] = bw;
 	ss->iops_data[ss->tail] = iops;
 
-	if (ss->state & __FIO_SS_IOPS)
+	if (ss->state & FIO_SS_IOPS)
 		new_val = iops;
 	else
 		new_val = bw;
 
-	if (ss->tail < ss->head || (ss->tail - ss->head == ss->dur - 1)) {
-		if (ss->sum_y == 0) {	/* first time through */
-			for(i = 0; i < ss->dur; i++) {
-				if (ss->state & __FIO_SS_IOPS)
+	if (ss->state & FIO_SS_BUFFER_FULL || ss->tail - ss->head == ss->dur - 1) {
+		if (!(ss->state & FIO_SS_BUFFER_FULL)) {
+			/* first time through */
+			for(i = 0, ss->sum_y = 0; i < ss->dur; i++) {
+				if (ss->state & FIO_SS_IOPS)
 					ss->sum_y += ss->iops_data[i];
 				else
 					ss->sum_y += ss->bw_data[i];
-				j = ss->head + i;
-				if (j >= ss->dur)
-					j -= ss->dur;
-				if (ss->state & __FIO_SS_IOPS)
+				j = (ss->head + i) % ss->dur;
+				if (ss->state & FIO_SS_IOPS)
 					ss->sum_xy += i * ss->iops_data[j];
 				else
 					ss->sum_xy += i * ss->bw_data[j];
 			}
+			ss->state |= FIO_SS_BUFFER_FULL;
 		} else {		/* easy to update the sums */
 			ss->sum_y -= ss->oldest_y;
 			ss->sum_y += new_val;
 			ss->sum_xy = ss->sum_xy - ss->sum_y + ss->dur * new_val;
 		}
 
-		if (ss->state & __FIO_SS_IOPS)
+		if (ss->state & FIO_SS_IOPS)
 			ss->oldest_y = ss->iops_data[ss->head];
 		else
 			ss->oldest_y = ss->bw_data[ss->head];
@@ -107,15 +107,16 @@ static bool steadystate_slope(unsigned long iops, unsigned long bw,
 		 */
 		ss->slope = (ss->sum_xy - (double) ss->sum_x * ss->sum_y / ss->dur) /
 				(ss->sum_x_sq - (double) ss->sum_x * ss->sum_x / ss->dur);
-		if (ss->state & __FIO_SS_PCT)
+		if (ss->state & FIO_SS_PCT)
 			ss->criterion = 100.0 * ss->slope / (ss->sum_y / ss->dur);
 		else
 			ss->criterion = ss->slope;
 
 		dprint(FD_STEADYSTATE, "sum_y: %llu, sum_xy: %llu, slope: %f, "
 					"criterion: %f, limit: %f\n",
-					ss->sum_y, ss->sum_xy, ss->slope,
-					ss->criterion, ss->limit);
+					(unsigned long long) ss->sum_y,
+					(unsigned long long) ss->sum_xy,
+					ss->slope, ss->criterion, ss->limit);
 
 		result = ss->criterion * (ss->criterion < 0.0 ? -1.0 : 1.0);
 		if (result < ss->limit)
@@ -129,7 +130,7 @@ static bool steadystate_slope(unsigned long iops, unsigned long bw,
 	return false;
 }
 
-static bool steadystate_deviation(unsigned long iops, unsigned long bw,
+static bool steadystate_deviation(uint64_t iops, uint64_t bw,
 				  struct thread_data *td)
 {
 	int i;
@@ -141,22 +142,24 @@ static bool steadystate_deviation(unsigned long iops, unsigned long bw,
 	ss->bw_data[ss->tail] = bw;
 	ss->iops_data[ss->tail] = iops;
 
-	if (ss->tail < ss->head || (ss->tail - ss->head == ss->dur - 1)) {
-		if (ss->sum_y == 0) {	/* first time through */
-			for(i = 0; i < ss->dur; i++)
-				if (ss->state & __FIO_SS_IOPS)
+	if (ss->state & FIO_SS_BUFFER_FULL || ss->tail - ss->head == ss->dur - 1) {
+		if (!(ss->state & FIO_SS_BUFFER_FULL)) {
+			/* first time through */
+			for(i = 0, ss->sum_y = 0; i < ss->dur; i++)
+				if (ss->state & FIO_SS_IOPS)
 					ss->sum_y += ss->iops_data[i];
 				else
 					ss->sum_y += ss->bw_data[i];
+			ss->state |= FIO_SS_BUFFER_FULL;
 		} else {		/* easy to update the sum */
 			ss->sum_y -= ss->oldest_y;
-			if (ss->state & __FIO_SS_IOPS)
+			if (ss->state & FIO_SS_IOPS)
 				ss->sum_y += ss->iops_data[ss->tail];
 			else
 				ss->sum_y += ss->bw_data[ss->tail];
 		}
 
-		if (ss->state & __FIO_SS_IOPS)
+		if (ss->state & FIO_SS_IOPS)
 			ss->oldest_y = ss->iops_data[ss->head];
 		else
 			ss->oldest_y = ss->bw_data[ss->head];
@@ -165,22 +168,22 @@ static bool steadystate_deviation(unsigned long iops, unsigned long bw,
 		ss->deviation = 0.0;
 
 		for (i = 0; i < ss->dur; i++) {
-			if (ss->state & __FIO_SS_IOPS)
+			if (ss->state & FIO_SS_IOPS)
 				diff = ss->iops_data[i] - mean;
 			else
 				diff = ss->bw_data[i] - mean;
 			ss->deviation = max(ss->deviation, diff * (diff < 0.0 ? -1.0 : 1.0));
 		}
 
-		if (ss->state & __FIO_SS_PCT)
+		if (ss->state & FIO_SS_PCT)
 			ss->criterion = 100.0 * ss->deviation / mean;
 		else
 			ss->criterion = ss->deviation;
 
 		dprint(FD_STEADYSTATE, "sum_y: %llu, mean: %f, max diff: %f, "
 					"objective: %f, limit: %f\n",
-					ss->sum_y, mean, ss->deviation,
-					ss->criterion, ss->limit);
+					(unsigned long long) ss->sum_y, mean,
+					ss->deviation, ss->criterion, ss->limit);
 
 		if (ss->criterion < ss->limit)
 			return true;
@@ -198,18 +201,19 @@ void steadystate_check(void)
 	int i, j, ddir, prev_groupid, group_ramp_time_over = 0;
 	unsigned long rate_time;
 	struct thread_data *td, *td2;
-	struct timeval now;
-	unsigned long group_bw = 0, group_iops = 0;
-	unsigned long long td_iops;
-	unsigned long long td_bytes;
+	struct timespec now;
+	uint64_t group_bw = 0, group_iops = 0;
+	uint64_t td_iops, td_bytes;
 	bool ret;
 
 	prev_groupid = -1;
 	for_each_td(td, i) {
+		const bool needs_lock = td_async_processing(td);
 		struct steadystate_data *ss = &td->ss;
 
 		if (!ss->dur || td->runstate <= TD_SETTING_UP ||
-		    td->runstate >= TD_EXITED || (ss->state & __FIO_SS_ATTAINED))
+		    td->runstate >= TD_EXITED || !ss->state ||
+		    ss->state & FIO_SS_ATTAINED)
 			continue;
 
 		td_iops = 0;
@@ -223,21 +227,25 @@ void steadystate_check(void)
 		prev_groupid = td->groupid;
 
 		fio_gettime(&now, NULL);
-		if (ss->ramp_time && !(ss->state & __FIO_SS_RAMP_OVER)) {
+		if (ss->ramp_time && !(ss->state & FIO_SS_RAMP_OVER)) {
 			/*
 			 * Begin recording data one second after ss->ramp_time
 			 * has elapsed
 			 */
 			if (utime_since(&td->epoch, &now) >= (ss->ramp_time + 1000000L))
-				ss->state |= __FIO_SS_RAMP_OVER;
+				ss->state |= FIO_SS_RAMP_OVER;
 		}
 
-		td_io_u_lock(td);
-		for (ddir = DDIR_READ; ddir < DDIR_RWDIR_CNT; ddir++) {
+		if (needs_lock)
+			__td_io_u_lock(td);
+
+		for (ddir = 0; ddir < DDIR_RWDIR_CNT; ddir++) {
 			td_iops += td->io_blocks[ddir];
 			td_bytes += td->io_bytes[ddir];
 		}
-		td_io_u_unlock(td);
+
+		if (needs_lock)
+			__td_io_u_unlock(td);
 
 		rate_time = mtime_since(&ss->prev_time, &now);
 		memcpy(&ss->prev_time, &now, sizeof(now));
@@ -249,7 +257,7 @@ void steadystate_check(void)
 		 * prev_iops/bw the first time through after ss->ramp_time
 		 * is done.
 		 */
-		if (ss->state & __FIO_SS_RAMP_OVER) {
+		if (ss->state & FIO_SS_RAMP_OVER) {
 			group_bw += 1000 * (td_bytes - ss->prev_bytes) / rate_time;
 			group_iops += 1000 * (td_iops - ss->prev_iops) / rate_time;
 			++group_ramp_time_over;
@@ -257,7 +265,7 @@ void steadystate_check(void)
 		ss->prev_iops = td_iops;
 		ss->prev_bytes = td_bytes;
 
-		if (td->o.group_reporting && !(ss->state & __FIO_SS_DATA))
+		if (td->o.group_reporting && !(ss->state & FIO_SS_DATA))
 			continue;
 
 		/*
@@ -269,11 +277,13 @@ void steadystate_check(void)
 
 		dprint(FD_STEADYSTATE, "steadystate_check() thread: %d, "
 					"groupid: %u, rate_msec: %ld, "
-					"iops: %lu, bw: %lu, head: %d, tail: %d\n",
-					i, td->groupid, rate_time, group_iops,
-					group_bw, ss->head, ss->tail);
+					"iops: %llu, bw: %llu, head: %d, tail: %d\n",
+					i, td->groupid, rate_time,
+					(unsigned long long) group_iops,
+					(unsigned long long) group_bw,
+					ss->head, ss->tail);
 
-		if (td->o.ss & __FIO_SS_SLOPE)
+		if (ss->state & FIO_SS_SLOPE)
 			ret = steadystate_slope(group_iops, group_bw, td);
 		else
 			ret = steadystate_deviation(group_iops, group_bw, td);
@@ -282,12 +292,12 @@ void steadystate_check(void)
 			if (td->o.group_reporting) {
 				for_each_td(td2, j) {
 					if (td2->groupid == td->groupid) {
-						td2->ss.state |= __FIO_SS_ATTAINED;
+						td2->ss.state |= FIO_SS_ATTAINED;
 						fio_mark_td_terminate(td2);
 					}
 				}
 			} else {
-				ss->state |= __FIO_SS_ATTAINED;
+				ss->state |= FIO_SS_ATTAINED;
 				fio_mark_td_terminate(td);
 			}
 		}
@@ -312,9 +322,9 @@ int td_steadystate_init(struct thread_data *td)
 		ss->limit = o->ss_limit.u.f;
 		ss->ramp_time = o->ss_ramp_time;
 
-		ss->state = o->ss;
+		ss->state = o->ss_state;
 		if (!td->ss.ramp_time)
-			ss->state |= __FIO_SS_RAMP_OVER;
+			ss->state |= FIO_SS_RAMP_OVER;
 
 		ss->sum_x = o->ss_dur * (o->ss_dur - 1) / 2;
 		ss->sum_x_sq = (o->ss_dur - 1) * (o->ss_dur) * (2*o->ss_dur - 1) / 6;
@@ -340,10 +350,13 @@ int td_steadystate_init(struct thread_data *td)
 	return 0;
 }
 
-unsigned long long steadystate_bw_mean(struct thread_stat *ts)
+uint64_t steadystate_bw_mean(struct thread_stat *ts)
 {
 	int i;
-	unsigned long long sum;
+	uint64_t sum;
+
+	if (!ts->ss_dur)
+		return 0;
 
 	for (i = 0, sum = 0; i < ts->ss_dur; i++)
 		sum += ts->ss_bw_data[i];
@@ -351,10 +364,13 @@ unsigned long long steadystate_bw_mean(struct thread_stat *ts)
 	return sum / ts->ss_dur;
 }
 
-unsigned long long steadystate_iops_mean(struct thread_stat *ts)
+uint64_t steadystate_iops_mean(struct thread_stat *ts)
 {
 	int i;
-	unsigned long long sum;
+	uint64_t sum;
+
+	if (!ts->ss_dur)
+		return 0;
 
 	for (i = 0, sum = 0; i < ts->ss_dur; i++)
 		sum += ts->ss_iops_data[i];