#include "helper_thread.h"
#include "smalloc.h"
#include "zbd.h"
+#include "oslib/asprintf.h"
+#ifdef WIN32
+#define LOG_MSEC_SLACK 2
+#else
#define LOG_MSEC_SLACK 1
+#endif
struct fio_sem *stat_sem;
* isn't a worry. Also note that this does not work for NaN values.
*/
if (len > 1)
- qsort((void *)plist, len, sizeof(plist[0]), double_cmp);
+ qsort(plist, len, sizeof(plist[0]), double_cmp);
ovals = malloc(len * sizeof(*ovals));
if (!ovals)
len = calc_clat_percentiles(io_u_plat, nr, plist, &ovals, &maxv, &minv);
if (!len || !ovals)
- goto out;
+ return;
/*
* We default to nsecs, but if the value range is such that we
log_buf(out, "\n");
}
-out:
- if (ovals)
- free(ovals);
+ free(ovals);
+}
+
+static int get_nr_prios_with_samples(struct thread_stat *ts, enum fio_ddir ddir)
+{
+ int i, nr_prios_with_samples = 0;
+
+ for (i = 0; i < ts->nr_clat_prio[ddir]; i++) {
+ if (ts->clat_prio[ddir][i].clat_stat.samples)
+ nr_prios_with_samples++;
+ }
+
+ return nr_prios_with_samples;
}
bool calc_lat(struct io_stat *is, unsigned long long *min,
return true;
}
+void show_mixed_group_stats(struct group_run_stats *rs, struct buf_output *out)
+{
+ char *io, *agg, *min, *max;
+ char *ioalt, *aggalt, *minalt, *maxalt;
+ uint64_t io_mix = 0, agg_mix = 0, min_mix = -1, max_mix = 0;
+ uint64_t min_run = -1, max_run = 0;
+ const int i2p = is_power_of_2(rs->kb_base);
+ int i;
+
+ for (i = 0; i < DDIR_RWDIR_CNT; i++) {
+ if (!rs->max_run[i])
+ continue;
+ io_mix += rs->iobytes[i];
+ agg_mix += rs->agg[i];
+ min_mix = min_mix < rs->min_bw[i] ? min_mix : rs->min_bw[i];
+ max_mix = max_mix > rs->max_bw[i] ? max_mix : rs->max_bw[i];
+ min_run = min_run < rs->min_run[i] ? min_run : rs->min_run[i];
+ max_run = max_run > rs->max_run[i] ? max_run : rs->max_run[i];
+ }
+ io = num2str(io_mix, rs->sig_figs, 1, i2p, N2S_BYTE);
+ ioalt = num2str(io_mix, rs->sig_figs, 1, !i2p, N2S_BYTE);
+ agg = num2str(agg_mix, rs->sig_figs, 1, i2p, rs->unit_base);
+ aggalt = num2str(agg_mix, rs->sig_figs, 1, !i2p, rs->unit_base);
+ min = num2str(min_mix, rs->sig_figs, 1, i2p, rs->unit_base);
+ minalt = num2str(min_mix, rs->sig_figs, 1, !i2p, rs->unit_base);
+ max = num2str(max_mix, rs->sig_figs, 1, i2p, rs->unit_base);
+ maxalt = num2str(max_mix, rs->sig_figs, 1, !i2p, rs->unit_base);
+ log_buf(out, " MIXED: bw=%s (%s), %s-%s (%s-%s), io=%s (%s), run=%llu-%llumsec\n",
+ agg, aggalt, min, max, minalt, maxalt, io, ioalt,
+ (unsigned long long) min_run,
+ (unsigned long long) max_run);
+ free(io);
+ free(agg);
+ free(min);
+ free(max);
+ free(ioalt);
+ free(aggalt);
+ free(minalt);
+ free(maxalt);
+}
+
void show_group_stats(struct group_run_stats *rs, struct buf_output *out)
{
char *io, *agg, *min, *max;
max = num2str(rs->max_bw[i], rs->sig_figs, 1, i2p, rs->unit_base);
maxalt = num2str(rs->max_bw[i], rs->sig_figs, 1, !i2p, rs->unit_base);
log_buf(out, "%s: bw=%s (%s), %s-%s (%s-%s), io=%s (%s), run=%llu-%llumsec\n",
- rs->unified_rw_rep ? " MIXED" : str[i],
+ (rs->unified_rw_rep == UNIFIED_MIXED) ? " MIXED" : str[i],
agg, aggalt, min, max, minalt, maxalt, io, ioalt,
(unsigned long long) rs->min_run[i],
(unsigned long long) rs->max_run[i]);
free(minalt);
free(maxalt);
}
+
+ /* Need to aggregate statisitics to show mixed values */
+ if (rs->unified_rw_rep == UNIFIED_BOTH)
+ show_mixed_group_stats(rs, out);
}
void stat_calc_dist(uint64_t *map, unsigned long total, double *io_u_dist)
free(maxp);
}
+static struct thread_stat *gen_mixed_ddir_stats_from_ts(struct thread_stat *ts)
+{
+ struct thread_stat *ts_lcl;
+
+ /*
+ * Handle aggregation of Reads (ddir = 0), Writes (ddir = 1), and
+ * Trims (ddir = 2)
+ */
+ ts_lcl = malloc(sizeof(struct thread_stat));
+ if (!ts_lcl) {
+ log_err("fio: failed to allocate local thread stat\n");
+ return NULL;
+ }
+
+ init_thread_stat(ts_lcl);
+
+ /* calculate mixed stats */
+ ts_lcl->unified_rw_rep = UNIFIED_MIXED;
+ ts_lcl->lat_percentiles = ts->lat_percentiles;
+ ts_lcl->clat_percentiles = ts->clat_percentiles;
+ ts_lcl->slat_percentiles = ts->slat_percentiles;
+ ts_lcl->percentile_precision = ts->percentile_precision;
+ memcpy(ts_lcl->percentile_list, ts->percentile_list, sizeof(ts->percentile_list));
+
+ sum_thread_stats(ts_lcl, ts);
+
+ return ts_lcl;
+}
+
+static double convert_agg_kbytes_percent(struct group_run_stats *rs,
+ enum fio_ddir ddir, int mean)
+{
+ double p_of_agg = 100.0;
+ if (rs && rs->agg[ddir] > 1024) {
+ p_of_agg = mean * 100.0 / (double) (rs->agg[ddir] / 1024.0);
+
+ if (p_of_agg > 100.0)
+ p_of_agg = 100.0;
+ }
+ return p_of_agg;
+}
+
static void show_ddir_status(struct group_run_stats *rs, struct thread_stat *ts,
- int ddir, struct buf_output *out)
+ enum fio_ddir ddir, struct buf_output *out)
{
unsigned long runt;
unsigned long long min, max, bw, iops;
double mean, dev;
char *io_p, *bw_p, *bw_p_alt, *iops_p, *post_st = NULL;
- int i2p;
+ int i2p, i;
+ const char *clat_type = ts->lat_percentiles ? "lat" : "clat";
if (ddir_sync(ddir)) {
if (calc_lat(&ts->sync_stat, &min, &max, &mean, &dev)) {
}
log_buf(out, " %s: IOPS=%s, BW=%s (%s)(%s/%llumsec)%s\n",
- rs->unified_rw_rep ? "mixed" : io_ddir_name(ddir),
+ (ts->unified_rw_rep == UNIFIED_MIXED) ? "mixed" : io_ddir_name(ddir),
iops_p, bw_p, bw_p_alt, io_p,
(unsigned long long) ts->runtime[ddir],
post_st ? : "");
if (calc_lat(&ts->lat_stat[ddir], &min, &max, &mean, &dev))
display_lat(" lat", min, max, mean, dev, out);
+ /* Only print per prio stats if there are >= 2 prios with samples */
+ if (get_nr_prios_with_samples(ts, ddir) >= 2) {
+ for (i = 0; i < ts->nr_clat_prio[ddir]; i++) {
+ if (calc_lat(&ts->clat_prio[ddir][i].clat_stat, &min,
+ &max, &mean, &dev)) {
+ char buf[64];
+
+ snprintf(buf, sizeof(buf),
+ "%s prio %u/%u",
+ clat_type,
+ ts->clat_prio[ddir][i].ioprio >> 13,
+ ts->clat_prio[ddir][i].ioprio & 7);
+ display_lat(buf, min, max, mean, dev, out);
+ }
+ }
+ }
+
+ if (ts->slat_percentiles && ts->slat_stat[ddir].samples > 0)
+ show_clat_percentiles(ts->io_u_plat[FIO_SLAT][ddir],
+ ts->slat_stat[ddir].samples,
+ ts->percentile_list,
+ ts->percentile_precision, "slat", out);
+ if (ts->clat_percentiles && ts->clat_stat[ddir].samples > 0)
+ show_clat_percentiles(ts->io_u_plat[FIO_CLAT][ddir],
+ ts->clat_stat[ddir].samples,
+ ts->percentile_list,
+ ts->percentile_precision, "clat", out);
+ if (ts->lat_percentiles && ts->lat_stat[ddir].samples > 0)
+ show_clat_percentiles(ts->io_u_plat[FIO_LAT][ddir],
+ ts->lat_stat[ddir].samples,
+ ts->percentile_list,
+ ts->percentile_precision, "lat", out);
+
if (ts->clat_percentiles || ts->lat_percentiles) {
- const char *name = ts->clat_percentiles ? "clat" : " lat";
+ char prio_name[64];
uint64_t samples;
- if (ts->clat_percentiles)
- samples = ts->clat_stat[ddir].samples;
- else
+ if (ts->lat_percentiles)
samples = ts->lat_stat[ddir].samples;
+ else
+ samples = ts->clat_stat[ddir].samples;
- show_clat_percentiles(ts->io_u_plat[ddir],
- samples,
- ts->percentile_list,
- ts->percentile_precision, name, out);
+ /* Only print per prio stats if there are >= 2 prios with samples */
+ if (get_nr_prios_with_samples(ts, ddir) >= 2) {
+ for (i = 0; i < ts->nr_clat_prio[ddir]; i++) {
+ uint64_t prio_samples = ts->clat_prio[ddir][i].clat_stat.samples;
+
+ if (prio_samples > 0) {
+ snprintf(prio_name, sizeof(prio_name),
+ "%s prio %u/%u (%.2f%% of IOs)",
+ clat_type,
+ ts->clat_prio[ddir][i].ioprio >> 13,
+ ts->clat_prio[ddir][i].ioprio & 7,
+ 100. * (double) prio_samples / (double) samples);
+ show_clat_percentiles(ts->clat_prio[ddir][i].io_u_plat,
+ prio_samples, ts->percentile_list,
+ ts->percentile_precision,
+ prio_name, out);
+ }
+ }
+ }
}
+
if (calc_lat(&ts->bw_stat[ddir], &min, &max, &mean, &dev)) {
double p_of_agg = 100.0, fkb_base = (double)rs->kb_base;
const char *bw_str;
else
bw_str = "kB";
- if (rs->agg[ddir]) {
- p_of_agg = mean * 100 / (double) (rs->agg[ddir] / 1024);
- if (p_of_agg > 100.0)
- p_of_agg = 100.0;
- }
+ p_of_agg = convert_agg_kbytes_percent(rs, ddir, mean);
if (rs->unit_base == 1) {
min *= 8.0;
}
}
+static void show_mixed_ddir_status(struct group_run_stats *rs,
+ struct thread_stat *ts,
+ struct buf_output *out)
+{
+ struct thread_stat *ts_lcl = gen_mixed_ddir_stats_from_ts(ts);
+
+ if (ts_lcl)
+ show_ddir_status(rs, ts_lcl, DDIR_READ, out);
+
+ free_clat_prio_stats(ts_lcl);
+ free(ts_lcl);
+}
+
static bool show_lat(double *io_u_lat, int nr, const char **ranges,
const char *msg, struct buf_output *out)
{
* isn't a worry. Also note that this does not work for NaN values.
*/
if (len > 1)
- qsort((void *)plist, len, sizeof(plist[0]), double_cmp);
+ qsort(plist, len, sizeof(plist[0]), double_cmp);
/* Start only after the uninit entries end */
for (nr_uninit = 0;
free(p2);
}
+static void show_agg_stats(struct disk_util_agg *agg, int terse,
+ struct buf_output *out)
+{
+ if (!agg->slavecount)
+ return;
+
+ if (!terse) {
+ log_buf(out, ", aggrios=%llu/%llu, aggrmerge=%llu/%llu, "
+ "aggrticks=%llu/%llu, aggrin_queue=%llu, "
+ "aggrutil=%3.2f%%",
+ (unsigned long long) agg->ios[0] / agg->slavecount,
+ (unsigned long long) agg->ios[1] / agg->slavecount,
+ (unsigned long long) agg->merges[0] / agg->slavecount,
+ (unsigned long long) agg->merges[1] / agg->slavecount,
+ (unsigned long long) agg->ticks[0] / agg->slavecount,
+ (unsigned long long) agg->ticks[1] / agg->slavecount,
+ (unsigned long long) agg->time_in_queue / agg->slavecount,
+ agg->max_util.u.f);
+ } else {
+ log_buf(out, ";slaves;%llu;%llu;%llu;%llu;%llu;%llu;%llu;%3.2f%%",
+ (unsigned long long) agg->ios[0] / agg->slavecount,
+ (unsigned long long) agg->ios[1] / agg->slavecount,
+ (unsigned long long) agg->merges[0] / agg->slavecount,
+ (unsigned long long) agg->merges[1] / agg->slavecount,
+ (unsigned long long) agg->ticks[0] / agg->slavecount,
+ (unsigned long long) agg->ticks[1] / agg->slavecount,
+ (unsigned long long) agg->time_in_queue / agg->slavecount,
+ agg->max_util.u.f);
+ }
+}
+
+static void aggregate_slaves_stats(struct disk_util *masterdu)
+{
+ struct disk_util_agg *agg = &masterdu->agg;
+ struct disk_util_stat *dus;
+ struct flist_head *entry;
+ struct disk_util *slavedu;
+ double util;
+
+ flist_for_each(entry, &masterdu->slaves) {
+ slavedu = flist_entry(entry, struct disk_util, slavelist);
+ dus = &slavedu->dus;
+ agg->ios[0] += dus->s.ios[0];
+ agg->ios[1] += dus->s.ios[1];
+ agg->merges[0] += dus->s.merges[0];
+ agg->merges[1] += dus->s.merges[1];
+ agg->sectors[0] += dus->s.sectors[0];
+ agg->sectors[1] += dus->s.sectors[1];
+ agg->ticks[0] += dus->s.ticks[0];
+ agg->ticks[1] += dus->s.ticks[1];
+ agg->time_in_queue += dus->s.time_in_queue;
+ agg->slavecount++;
+
+ util = (double) (100 * dus->s.io_ticks / (double) slavedu->dus.s.msec);
+ /* System utilization is the utilization of the
+ * component with the highest utilization.
+ */
+ if (util > agg->max_util.u.f)
+ agg->max_util.u.f = util;
+
+ }
+
+ if (agg->max_util.u.f > 100.0)
+ agg->max_util.u.f = 100.0;
+}
+
+void print_disk_util(struct disk_util_stat *dus, struct disk_util_agg *agg,
+ int terse, struct buf_output *out)
+{
+ double util = 0;
+
+ if (dus->s.msec)
+ util = (double) 100 * dus->s.io_ticks / (double) dus->s.msec;
+ if (util > 100.0)
+ util = 100.0;
+
+ if (!terse) {
+ if (agg->slavecount)
+ log_buf(out, " ");
+
+ log_buf(out, " %s: ios=%llu/%llu, merge=%llu/%llu, "
+ "ticks=%llu/%llu, in_queue=%llu, util=%3.2f%%",
+ dus->name,
+ (unsigned long long) dus->s.ios[0],
+ (unsigned long long) dus->s.ios[1],
+ (unsigned long long) dus->s.merges[0],
+ (unsigned long long) dus->s.merges[1],
+ (unsigned long long) dus->s.ticks[0],
+ (unsigned long long) dus->s.ticks[1],
+ (unsigned long long) dus->s.time_in_queue,
+ util);
+ } else {
+ log_buf(out, ";%s;%llu;%llu;%llu;%llu;%llu;%llu;%llu;%3.2f%%",
+ dus->name,
+ (unsigned long long) dus->s.ios[0],
+ (unsigned long long) dus->s.ios[1],
+ (unsigned long long) dus->s.merges[0],
+ (unsigned long long) dus->s.merges[1],
+ (unsigned long long) dus->s.ticks[0],
+ (unsigned long long) dus->s.ticks[1],
+ (unsigned long long) dus->s.time_in_queue,
+ util);
+ }
+
+ /*
+ * If the device has slaves, aggregate the stats for
+ * those slave devices also.
+ */
+ show_agg_stats(agg, terse, out);
+
+ if (!terse)
+ log_buf(out, "\n");
+}
+
+void json_array_add_disk_util(struct disk_util_stat *dus,
+ struct disk_util_agg *agg, struct json_array *array)
+{
+ struct json_object *obj;
+ double util = 0;
+
+ if (dus->s.msec)
+ util = (double) 100 * dus->s.io_ticks / (double) dus->s.msec;
+ if (util > 100.0)
+ util = 100.0;
+
+ obj = json_create_object();
+ json_array_add_value_object(array, obj);
+
+ json_object_add_value_string(obj, "name", (const char *)dus->name);
+ json_object_add_value_int(obj, "read_ios", dus->s.ios[0]);
+ json_object_add_value_int(obj, "write_ios", dus->s.ios[1]);
+ json_object_add_value_int(obj, "read_merges", dus->s.merges[0]);
+ json_object_add_value_int(obj, "write_merges", dus->s.merges[1]);
+ json_object_add_value_int(obj, "read_ticks", dus->s.ticks[0]);
+ json_object_add_value_int(obj, "write_ticks", dus->s.ticks[1]);
+ json_object_add_value_int(obj, "in_queue", dus->s.time_in_queue);
+ json_object_add_value_float(obj, "util", util);
+
+ /*
+ * If the device has slaves, aggregate the stats for
+ * those slave devices also.
+ */
+ if (!agg->slavecount)
+ return;
+ json_object_add_value_int(obj, "aggr_read_ios",
+ agg->ios[0] / agg->slavecount);
+ json_object_add_value_int(obj, "aggr_write_ios",
+ agg->ios[1] / agg->slavecount);
+ json_object_add_value_int(obj, "aggr_read_merges",
+ agg->merges[0] / agg->slavecount);
+ json_object_add_value_int(obj, "aggr_write_merge",
+ agg->merges[1] / agg->slavecount);
+ json_object_add_value_int(obj, "aggr_read_ticks",
+ agg->ticks[0] / agg->slavecount);
+ json_object_add_value_int(obj, "aggr_write_ticks",
+ agg->ticks[1] / agg->slavecount);
+ json_object_add_value_int(obj, "aggr_in_queue",
+ agg->time_in_queue / agg->slavecount);
+ json_object_add_value_float(obj, "aggr_util", agg->max_util.u.f);
+}
+
+static void json_object_add_disk_utils(struct json_object *obj,
+ struct flist_head *head)
+{
+ struct json_array *array = json_create_array();
+ struct flist_head *entry;
+ struct disk_util *du;
+
+ json_object_add_value_array(obj, "disk_util", array);
+
+ flist_for_each(entry, head) {
+ du = flist_entry(entry, struct disk_util, list);
+
+ aggregate_slaves_stats(du);
+ json_array_add_disk_util(&du->dus, &du->agg, array);
+ }
+}
+
+void show_disk_util(int terse, struct json_object *parent,
+ struct buf_output *out)
+{
+ struct flist_head *entry;
+ struct disk_util *du;
+ bool do_json;
+
+ if (!is_running_backend())
+ return;
+
+ if (flist_empty(&disk_list))
+ return;
+
+ if ((output_format & FIO_OUTPUT_JSON) && parent)
+ do_json = true;
+ else
+ do_json = false;
+
+ if (!terse && !do_json)
+ log_buf(out, "\nDisk stats (read/write):\n");
+
+ if (do_json) {
+ json_object_add_disk_utils(parent, &disk_list);
+ } else if (output_format & ~(FIO_OUTPUT_JSON | FIO_OUTPUT_JSON_PLUS)) {
+ flist_for_each(entry, &disk_list) {
+ du = flist_entry(entry, struct disk_util, list);
+
+ aggregate_slaves_stats(du);
+ print_disk_util(&du->dus, &du->agg, terse, out);
+ }
+ }
+}
+
static void show_thread_status_normal(struct thread_stat *ts,
struct group_run_stats *rs,
struct buf_output *out)
if (!ddir_rw_sum(ts->io_bytes) && !ddir_rw_sum(ts->total_io_u))
return;
-
+
memset(time_buf, 0, sizeof(time_buf));
time(&time_p);
if (strlen(ts->description))
log_buf(out, " Description : [%s]\n", ts->description);
- if (ts->io_bytes[DDIR_READ])
- show_ddir_status(rs, ts, DDIR_READ, out);
- if (ts->io_bytes[DDIR_WRITE])
- show_ddir_status(rs, ts, DDIR_WRITE, out);
- if (ts->io_bytes[DDIR_TRIM])
- show_ddir_status(rs, ts, DDIR_TRIM, out);
+ for_each_rw_ddir(ddir) {
+ if (ts->io_bytes[ddir])
+ show_ddir_status(rs, ts, ddir, out);
+ }
+
+ if (ts->unified_rw_rep == UNIFIED_BOTH)
+ show_mixed_ddir_status(rs, ts, out);
show_latencies(ts, out);
}
static void show_ddir_status_terse(struct thread_stat *ts,
- struct group_run_stats *rs, int ddir,
- int ver, struct buf_output *out)
+ struct group_run_stats *rs,
+ enum fio_ddir ddir, int ver,
+ struct buf_output *out)
{
unsigned long long min, max, minv, maxv, bw, iops;
unsigned long long *ovals = NULL;
else
log_buf(out, ";%llu;%llu;%f;%f", 0ULL, 0ULL, 0.0, 0.0);
- if (ts->clat_percentiles || ts->lat_percentiles) {
- len = calc_clat_percentiles(ts->io_u_plat[ddir],
+ if (ts->lat_percentiles) {
+ len = calc_clat_percentiles(ts->io_u_plat[FIO_LAT][ddir],
+ ts->lat_stat[ddir].samples,
+ ts->percentile_list, &ovals, &maxv,
+ &minv);
+ } else if (ts->clat_percentiles) {
+ len = calc_clat_percentiles(ts->io_u_plat[FIO_CLAT][ddir],
ts->clat_stat[ddir].samples,
ts->percentile_list, &ovals, &maxv,
&minv);
- } else
+ } else {
len = 0;
+ }
for (i = 0; i < FIO_IO_U_LIST_MAX_LEN; i++) {
if (i >= len) {
else
log_buf(out, ";%llu;%llu;%f;%f", 0ULL, 0ULL, 0.0, 0.0);
- if (ovals)
- free(ovals);
+ free(ovals);
bw_stat = calc_lat(&ts->bw_stat[ddir], &min, &max, &mean, &dev);
if (bw_stat) {
}
log_buf(out, ";%llu;%llu;%f%%;%f;%f", min, max, p_of_agg, mean, dev);
- } else
+ } else {
log_buf(out, ";%llu;%llu;%f%%;%f;%f", 0ULL, 0ULL, 0.0, 0.0, 0.0);
+ }
if (ver == 5) {
if (bw_stat)
}
}
+static void show_mixed_ddir_status_terse(struct thread_stat *ts,
+ struct group_run_stats *rs,
+ int ver, struct buf_output *out)
+{
+ struct thread_stat *ts_lcl = gen_mixed_ddir_stats_from_ts(ts);
+
+ if (ts_lcl)
+ show_ddir_status_terse(ts_lcl, rs, DDIR_READ, ver, out);
+
+ free_clat_prio_stats(ts_lcl);
+ free(ts_lcl);
+}
+
+static struct json_object *add_ddir_lat_json(struct thread_stat *ts,
+ uint32_t percentiles,
+ struct io_stat *lat_stat,
+ uint64_t *io_u_plat)
+{
+ char buf[120];
+ double mean, dev;
+ unsigned int i, len;
+ struct json_object *lat_object, *percentile_object, *clat_bins_object;
+ unsigned long long min, max, maxv, minv, *ovals = NULL;
+
+ if (!calc_lat(lat_stat, &min, &max, &mean, &dev)) {
+ min = max = 0;
+ mean = dev = 0.0;
+ }
+ lat_object = json_create_object();
+ json_object_add_value_int(lat_object, "min", min);
+ json_object_add_value_int(lat_object, "max", max);
+ json_object_add_value_float(lat_object, "mean", mean);
+ json_object_add_value_float(lat_object, "stddev", dev);
+ json_object_add_value_int(lat_object, "N", lat_stat->samples);
+
+ if (percentiles && lat_stat->samples) {
+ len = calc_clat_percentiles(io_u_plat, lat_stat->samples,
+ ts->percentile_list, &ovals, &maxv, &minv);
+
+ if (len > FIO_IO_U_LIST_MAX_LEN)
+ len = FIO_IO_U_LIST_MAX_LEN;
+
+ percentile_object = json_create_object();
+ json_object_add_value_object(lat_object, "percentile", percentile_object);
+ for (i = 0; i < len; i++) {
+ snprintf(buf, sizeof(buf), "%f", ts->percentile_list[i].u.f);
+ json_object_add_value_int(percentile_object, buf, ovals[i]);
+ }
+ free(ovals);
+
+ if (output_format & FIO_OUTPUT_JSON_PLUS) {
+ clat_bins_object = json_create_object();
+ json_object_add_value_object(lat_object, "bins", clat_bins_object);
+
+ for(i = 0; i < FIO_IO_U_PLAT_NR; i++)
+ if (io_u_plat[i]) {
+ snprintf(buf, sizeof(buf), "%llu", plat_idx_to_val(i));
+ json_object_add_value_int(clat_bins_object, buf, io_u_plat[i]);
+ }
+ }
+ }
+
+ return lat_object;
+}
+
static void add_ddir_status_json(struct thread_stat *ts,
- struct group_run_stats *rs, int ddir, struct json_object *parent)
+ struct group_run_stats *rs, enum fio_ddir ddir,
+ struct json_object *parent)
{
- unsigned long long min, max, minv, maxv;
+ unsigned long long min, max;
unsigned long long bw_bytes, bw;
- unsigned long long *ovals = NULL;
double mean, dev, iops;
- unsigned int len;
- int i;
- struct json_object *dir_object, *tmp_object, *percentile_object, *clat_bins_object = NULL;
- char buf[120];
+ struct json_object *dir_object, *tmp_object;
double p_of_agg = 100.0;
assert(ddir_rw(ddir) || ddir_sync(ddir));
- if (ts->unified_rw_rep && ddir != DDIR_READ)
+ if ((ts->unified_rw_rep == UNIFIED_MIXED) && ddir != DDIR_READ)
return;
dir_object = json_create_object();
json_object_add_value_object(parent,
- ts->unified_rw_rep ? "mixed" : io_ddir_name(ddir), dir_object);
+ (ts->unified_rw_rep == UNIFIED_MIXED) ? "mixed" : io_ddir_name(ddir), dir_object);
if (ddir_rw(ddir)) {
bw_bytes = 0;
json_object_add_value_int(dir_object, "short_ios", ts->short_io_u[ddir]);
json_object_add_value_int(dir_object, "drop_ios", ts->drop_io_u[ddir]);
- if (!calc_lat(&ts->slat_stat[ddir], &min, &max, &mean, &dev)) {
- min = max = 0;
- mean = dev = 0.0;
- }
- tmp_object = json_create_object();
+ tmp_object = add_ddir_lat_json(ts, ts->slat_percentiles,
+ &ts->slat_stat[ddir], ts->io_u_plat[FIO_SLAT][ddir]);
json_object_add_value_object(dir_object, "slat_ns", tmp_object);
- json_object_add_value_int(tmp_object, "min", min);
- json_object_add_value_int(tmp_object, "max", max);
- json_object_add_value_float(tmp_object, "mean", mean);
- json_object_add_value_float(tmp_object, "stddev", dev);
-
- if (!calc_lat(&ts->clat_stat[ddir], &min, &max, &mean, &dev)) {
- min = max = 0;
- mean = dev = 0.0;
- }
- tmp_object = json_create_object();
+
+ tmp_object = add_ddir_lat_json(ts, ts->clat_percentiles,
+ &ts->clat_stat[ddir], ts->io_u_plat[FIO_CLAT][ddir]);
json_object_add_value_object(dir_object, "clat_ns", tmp_object);
- json_object_add_value_int(tmp_object, "min", min);
- json_object_add_value_int(tmp_object, "max", max);
- json_object_add_value_float(tmp_object, "mean", mean);
- json_object_add_value_float(tmp_object, "stddev", dev);
- } else {
- if (!calc_lat(&ts->sync_stat, &min, &max, &mean, &dev)) {
- min = max = 0;
- mean = dev = 0.0;
- }
- tmp_object = json_create_object();
+ tmp_object = add_ddir_lat_json(ts, ts->lat_percentiles,
+ &ts->lat_stat[ddir], ts->io_u_plat[FIO_LAT][ddir]);
json_object_add_value_object(dir_object, "lat_ns", tmp_object);
+ } else {
json_object_add_value_int(dir_object, "total_ios", ts->total_io_u[DDIR_SYNC]);
- json_object_add_value_int(tmp_object, "min", min);
- json_object_add_value_int(tmp_object, "max", max);
- json_object_add_value_float(tmp_object, "mean", mean);
- json_object_add_value_float(tmp_object, "stddev", dev);
+ tmp_object = add_ddir_lat_json(ts, ts->lat_percentiles | ts->clat_percentiles,
+ &ts->sync_stat, ts->io_u_sync_plat);
+ json_object_add_value_object(dir_object, "lat_ns", tmp_object);
}
- if (ts->clat_percentiles || ts->lat_percentiles) {
- if (ddir_rw(ddir)) {
- uint64_t samples;
-
- if (ts->clat_percentiles)
- samples = ts->clat_stat[ddir].samples;
- else
- samples = ts->lat_stat[ddir].samples;
-
- len = calc_clat_percentiles(ts->io_u_plat[ddir],
- samples, ts->percentile_list, &ovals,
- &maxv, &minv);
- } else {
- len = calc_clat_percentiles(ts->io_u_sync_plat,
- ts->sync_stat.samples,
- ts->percentile_list, &ovals, &maxv,
- &minv);
- }
-
- if (len > FIO_IO_U_LIST_MAX_LEN)
- len = FIO_IO_U_LIST_MAX_LEN;
- } else
- len = 0;
-
- percentile_object = json_create_object();
- if (ts->clat_percentiles)
- json_object_add_value_object(tmp_object, "percentile", percentile_object);
- for (i = 0; i < len; i++) {
- snprintf(buf, sizeof(buf), "%f", ts->percentile_list[i].u.f);
- json_object_add_value_int(percentile_object, (const char *)buf, ovals[i]);
- }
+ if (!ddir_rw(ddir))
+ return;
- if (output_format & FIO_OUTPUT_JSON_PLUS) {
- clat_bins_object = json_create_object();
- if (ts->clat_percentiles)
- json_object_add_value_object(tmp_object, "bins", clat_bins_object);
+ /* Only include per prio stats if there are >= 2 prios with samples */
+ if (get_nr_prios_with_samples(ts, ddir) >= 2) {
+ struct json_array *array = json_create_array();
+ const char *obj_name;
+ int i;
- for(i = 0; i < FIO_IO_U_PLAT_NR; i++) {
- if (ddir_rw(ddir)) {
- if (ts->io_u_plat[ddir][i]) {
- snprintf(buf, sizeof(buf), "%llu", plat_idx_to_val(i));
- json_object_add_value_int(clat_bins_object, (const char *)buf, ts->io_u_plat[ddir][i]);
- }
- } else {
- if (ts->io_u_sync_plat[i]) {
- snprintf(buf, sizeof(buf), "%llu", plat_idx_to_val(i));
- json_object_add_value_int(clat_bins_object, (const char *)buf, ts->io_u_sync_plat[i]);
- }
+ if (ts->lat_percentiles)
+ obj_name = "lat_ns";
+ else
+ obj_name = "clat_ns";
+
+ json_object_add_value_array(dir_object, "prios", array);
+
+ for (i = 0; i < ts->nr_clat_prio[ddir]; i++) {
+ if (ts->clat_prio[ddir][i].clat_stat.samples > 0) {
+ struct json_object *obj = json_create_object();
+ unsigned long long class, level;
+
+ class = ts->clat_prio[ddir][i].ioprio >> 13;
+ json_object_add_value_int(obj, "prioclass", class);
+ level = ts->clat_prio[ddir][i].ioprio & 7;
+ json_object_add_value_int(obj, "prio", level);
+
+ tmp_object = add_ddir_lat_json(ts,
+ ts->clat_percentiles | ts->lat_percentiles,
+ &ts->clat_prio[ddir][i].clat_stat,
+ ts->clat_prio[ddir][i].io_u_plat);
+ json_object_add_value_object(obj, obj_name, tmp_object);
+ json_array_add_value_object(array, obj);
}
}
}
- if (!ddir_rw(ddir))
- return;
-
- if (!calc_lat(&ts->lat_stat[ddir], &min, &max, &mean, &dev)) {
- min = max = 0;
- mean = dev = 0.0;
- }
- tmp_object = json_create_object();
- json_object_add_value_object(dir_object, "lat_ns", tmp_object);
- json_object_add_value_int(tmp_object, "min", min);
- json_object_add_value_int(tmp_object, "max", max);
- json_object_add_value_float(tmp_object, "mean", mean);
- json_object_add_value_float(tmp_object, "stddev", dev);
- if (ts->lat_percentiles)
- json_object_add_value_object(tmp_object, "percentile", percentile_object);
- if (output_format & FIO_OUTPUT_JSON_PLUS && ts->lat_percentiles)
- json_object_add_value_object(tmp_object, "bins", clat_bins_object);
-
- if (ovals)
- free(ovals);
-
if (calc_lat(&ts->bw_stat[ddir], &min, &max, &mean, &dev)) {
- if (rs->agg[ddir]) {
- p_of_agg = mean * 100 / (double) (rs->agg[ddir] / 1024);
- if (p_of_agg > 100.0)
- p_of_agg = 100.0;
- }
+ p_of_agg = convert_agg_kbytes_percent(rs, ddir, mean);
} else {
min = max = 0;
p_of_agg = mean = dev = 0.0;
}
+
json_object_add_value_int(dir_object, "bw_min", min);
json_object_add_value_int(dir_object, "bw_max", max);
json_object_add_value_float(dir_object, "bw_agg", p_of_agg);
}
}
+static void add_mixed_ddir_status_json(struct thread_stat *ts,
+ struct group_run_stats *rs, struct json_object *parent)
+{
+ struct thread_stat *ts_lcl = gen_mixed_ddir_stats_from_ts(ts);
+
+ /* add the aggregated stats to json parent */
+ if (ts_lcl)
+ add_ddir_status_json(ts_lcl, rs, DDIR_READ, parent);
+
+ free_clat_prio_stats(ts_lcl);
+ free(ts_lcl);
+}
+
static void show_thread_status_terse_all(struct thread_stat *ts,
struct group_run_stats *rs, int ver,
struct buf_output *out)
log_buf(out, "%d;%s;%s;%d;%d", ver, fio_version_string,
ts->name, ts->groupid, ts->error);
- /* Log Read Status */
+ /* Log Read Status, or mixed if unified_rw_rep = 1 */
show_ddir_status_terse(ts, rs, DDIR_READ, ver, out);
- /* Log Write Status */
- show_ddir_status_terse(ts, rs, DDIR_WRITE, ver, out);
- /* Log Trim Status */
- if (ver == 2 || ver == 4 || ver == 5)
- show_ddir_status_terse(ts, rs, DDIR_TRIM, ver, out);
-
+ if (ts->unified_rw_rep != UNIFIED_MIXED) {
+ /* Log Write Status */
+ show_ddir_status_terse(ts, rs, DDIR_WRITE, ver, out);
+ /* Log Trim Status */
+ if (ver == 2 || ver == 4 || ver == 5)
+ show_ddir_status_terse(ts, rs, DDIR_TRIM, ver, out);
+ }
+ if (ts->unified_rw_rep == UNIFIED_BOTH)
+ show_mixed_ddir_status_terse(ts, rs, ver, out);
/* CPU Usage */
if (ts->total_run_time) {
double runt = (double) ts->total_run_time;
json_object_add_value_object(root, name, dir_object);
flist_for_each(entry, opt_list) {
- const char *pos = "";
-
p = flist_entry(entry, struct print_option, list);
- if (p->value)
- pos = p->value;
- json_object_add_value_string(dir_object, p->name, pos);
+ json_object_add_value_string(dir_object, p->name, p->value);
}
}
add_ddir_status_json(ts, rs, DDIR_TRIM, root);
add_ddir_status_json(ts, rs, DDIR_SYNC, root);
+ if (ts->unified_rw_rep == UNIFIED_BOTH)
+ add_mixed_ddir_status_json(ts, rs, root);
+
/* CPU Usage */
if (ts->total_run_time) {
double runt = (double) ts->total_run_time;
snprintf(buf, sizeof(buf), "%f",
ts->percentile_list[i].u.f);
json_object_add_value_int(percentile_object,
- (const char *)buf,
+ buf,
percentiles[i]);
}
* numbers. For group_reporting, we should just add those up, not make
* them the mean of everything.
*/
-static void sum_stat(struct io_stat *dst, struct io_stat *src, bool first,
- bool pure_sum)
+static void sum_stat(struct io_stat *dst, struct io_stat *src, bool pure_sum)
{
+ bool first = dst->samples == 0;
+
if (src->samples == 0)
return;
dst->sig_figs = src->sig_figs;
}
-void sum_thread_stats(struct thread_stat *dst, struct thread_stat *src,
- bool first)
+/*
+ * Free the clat_prio_stat arrays allocated by alloc_clat_prio_stat_ddir().
+ */
+void free_clat_prio_stats(struct thread_stat *ts)
+{
+ enum fio_ddir ddir;
+
+ for (ddir = 0; ddir < DDIR_RWDIR_CNT; ddir++) {
+ sfree(ts->clat_prio[ddir]);
+ ts->clat_prio[ddir] = NULL;
+ ts->nr_clat_prio[ddir] = 0;
+ }
+}
+
+/*
+ * Allocate a clat_prio_stat array. The array has to be allocated/freed using
+ * smalloc/sfree, so that it is accessible by the process/thread summing the
+ * thread_stats.
+ */
+int alloc_clat_prio_stat_ddir(struct thread_stat *ts, enum fio_ddir ddir,
+ int nr_prios)
+{
+ struct clat_prio_stat *clat_prio;
+ int i;
+
+ clat_prio = scalloc(nr_prios, sizeof(*ts->clat_prio[ddir]));
+ if (!clat_prio) {
+ log_err("fio: failed to allocate ts clat data\n");
+ return 1;
+ }
+
+ for (i = 0; i < nr_prios; i++)
+ clat_prio[i].clat_stat.min_val = ULONG_MAX;
+
+ ts->clat_prio[ddir] = clat_prio;
+ ts->nr_clat_prio[ddir] = nr_prios;
+
+ return 0;
+}
+
+static int grow_clat_prio_stat(struct thread_stat *dst, enum fio_ddir ddir)
+{
+ int curr_len = dst->nr_clat_prio[ddir];
+ void *new_arr;
+
+ new_arr = scalloc(curr_len + 1, sizeof(*dst->clat_prio[ddir]));
+ if (!new_arr) {
+ log_err("fio: failed to grow clat prio array\n");
+ return 1;
+ }
+
+ memcpy(new_arr, dst->clat_prio[ddir],
+ curr_len * sizeof(*dst->clat_prio[ddir]));
+ sfree(dst->clat_prio[ddir]);
+
+ dst->clat_prio[ddir] = new_arr;
+ dst->clat_prio[ddir][curr_len].clat_stat.min_val = ULONG_MAX;
+ dst->nr_clat_prio[ddir]++;
+
+ return 0;
+}
+
+static int find_clat_prio_index(struct thread_stat *dst, enum fio_ddir ddir,
+ uint32_t ioprio)
{
- int l, k;
+ int i, nr_prios = dst->nr_clat_prio[ddir];
+
+ for (i = 0; i < nr_prios; i++) {
+ if (dst->clat_prio[ddir][i].ioprio == ioprio)
+ return i;
+ }
+
+ return -1;
+}
+
+static int alloc_or_get_clat_prio_index(struct thread_stat *dst,
+ enum fio_ddir ddir, uint32_t ioprio,
+ int *idx)
+{
+ int index = find_clat_prio_index(dst, ddir, ioprio);
+
+ if (index == -1) {
+ index = dst->nr_clat_prio[ddir];
+
+ if (grow_clat_prio_stat(dst, ddir))
+ return 1;
+
+ dst->clat_prio[ddir][index].ioprio = ioprio;
+ }
+
+ *idx = index;
+
+ return 0;
+}
+
+static int clat_prio_stats_copy(struct thread_stat *dst, struct thread_stat *src,
+ enum fio_ddir dst_ddir, enum fio_ddir src_ddir)
+{
+ size_t sz = sizeof(*src->clat_prio[src_ddir]) *
+ src->nr_clat_prio[src_ddir];
+
+ dst->clat_prio[dst_ddir] = smalloc(sz);
+ if (!dst->clat_prio[dst_ddir]) {
+ log_err("fio: failed to alloc clat prio array\n");
+ return 1;
+ }
+
+ memcpy(dst->clat_prio[dst_ddir], src->clat_prio[src_ddir], sz);
+ dst->nr_clat_prio[dst_ddir] = src->nr_clat_prio[src_ddir];
+
+ return 0;
+}
+
+static int clat_prio_stat_add_samples(struct thread_stat *dst,
+ enum fio_ddir dst_ddir, uint32_t ioprio,
+ struct io_stat *io_stat,
+ uint64_t *io_u_plat)
+{
+ int i, dst_index;
+
+ if (!io_stat->samples)
+ return 0;
+
+ if (alloc_or_get_clat_prio_index(dst, dst_ddir, ioprio, &dst_index))
+ return 1;
+
+ sum_stat(&dst->clat_prio[dst_ddir][dst_index].clat_stat, io_stat,
+ false);
+
+ for (i = 0; i < FIO_IO_U_PLAT_NR; i++)
+ dst->clat_prio[dst_ddir][dst_index].io_u_plat[i] += io_u_plat[i];
+
+ return 0;
+}
+
+static int sum_clat_prio_stats_src_single_prio(struct thread_stat *dst,
+ struct thread_stat *src,
+ enum fio_ddir dst_ddir,
+ enum fio_ddir src_ddir)
+{
+ struct io_stat *io_stat;
+ uint64_t *io_u_plat;
+
+ /*
+ * If src ts has no clat_prio_stat array, then all I/Os were submitted
+ * using src->ioprio. Thus, the global samples in src->clat_stat (or
+ * src->lat_stat) can be used as the 'per prio' samples for src->ioprio.
+ */
+ assert(!src->clat_prio[src_ddir]);
+ assert(src->nr_clat_prio[src_ddir] == 0);
+
+ if (src->lat_percentiles) {
+ io_u_plat = src->io_u_plat[FIO_LAT][src_ddir];
+ io_stat = &src->lat_stat[src_ddir];
+ } else {
+ io_u_plat = src->io_u_plat[FIO_CLAT][src_ddir];
+ io_stat = &src->clat_stat[src_ddir];
+ }
+
+ return clat_prio_stat_add_samples(dst, dst_ddir, src->ioprio, io_stat,
+ io_u_plat);
+}
+
+static int sum_clat_prio_stats_src_multi_prio(struct thread_stat *dst,
+ struct thread_stat *src,
+ enum fio_ddir dst_ddir,
+ enum fio_ddir src_ddir)
+{
+ int i;
+
+ /*
+ * If src ts has a clat_prio_stat array, then there are multiple prios
+ * in use (i.e. src ts had cmdprio_percentage or cmdprio_bssplit set).
+ * The samples for the default prio will exist in the src->clat_prio
+ * array, just like the samples for any other prio.
+ */
+ assert(src->clat_prio[src_ddir]);
+ assert(src->nr_clat_prio[src_ddir]);
+
+ /* If the dst ts doesn't yet have a clat_prio array, simply memcpy. */
+ if (!dst->clat_prio[dst_ddir])
+ return clat_prio_stats_copy(dst, src, dst_ddir, src_ddir);
+
+ /* The dst ts already has a clat_prio_array, add src stats into it. */
+ for (i = 0; i < src->nr_clat_prio[src_ddir]; i++) {
+ struct io_stat *io_stat = &src->clat_prio[src_ddir][i].clat_stat;
+ uint64_t *io_u_plat = src->clat_prio[src_ddir][i].io_u_plat;
+ uint32_t ioprio = src->clat_prio[src_ddir][i].ioprio;
+
+ if (clat_prio_stat_add_samples(dst, dst_ddir, ioprio, io_stat, io_u_plat))
+ return 1;
+ }
+
+ return 0;
+}
+
+static int sum_clat_prio_stats(struct thread_stat *dst, struct thread_stat *src,
+ enum fio_ddir dst_ddir, enum fio_ddir src_ddir)
+{
+ if (dst->disable_prio_stat)
+ return 0;
+
+ if (!src->clat_prio[src_ddir])
+ return sum_clat_prio_stats_src_single_prio(dst, src, dst_ddir,
+ src_ddir);
+
+ return sum_clat_prio_stats_src_multi_prio(dst, src, dst_ddir, src_ddir);
+}
+
+void sum_thread_stats(struct thread_stat *dst, struct thread_stat *src)
+{
+ int k, l, m;
for (l = 0; l < DDIR_RWDIR_CNT; l++) {
- if (!dst->unified_rw_rep) {
- sum_stat(&dst->clat_stat[l], &src->clat_stat[l], first, false);
- sum_stat(&dst->slat_stat[l], &src->slat_stat[l], first, false);
- sum_stat(&dst->lat_stat[l], &src->lat_stat[l], first, false);
- sum_stat(&dst->bw_stat[l], &src->bw_stat[l], first, true);
- sum_stat(&dst->iops_stat[l], &src->iops_stat[l], first, true);
+ if (dst->unified_rw_rep != UNIFIED_MIXED) {
+ sum_stat(&dst->clat_stat[l], &src->clat_stat[l], false);
+ sum_stat(&dst->slat_stat[l], &src->slat_stat[l], false);
+ sum_stat(&dst->lat_stat[l], &src->lat_stat[l], false);
+ sum_stat(&dst->bw_stat[l], &src->bw_stat[l], true);
+ sum_stat(&dst->iops_stat[l], &src->iops_stat[l], true);
+ sum_clat_prio_stats(dst, src, l, l);
dst->io_bytes[l] += src->io_bytes[l];
if (dst->runtime[l] < src->runtime[l])
dst->runtime[l] = src->runtime[l];
} else {
- sum_stat(&dst->clat_stat[0], &src->clat_stat[l], first, false);
- sum_stat(&dst->slat_stat[0], &src->slat_stat[l], first, false);
- sum_stat(&dst->lat_stat[0], &src->lat_stat[l], first, false);
- sum_stat(&dst->bw_stat[0], &src->bw_stat[l], first, true);
- sum_stat(&dst->iops_stat[0], &src->iops_stat[l], first, true);
+ sum_stat(&dst->clat_stat[0], &src->clat_stat[l], false);
+ sum_stat(&dst->slat_stat[0], &src->slat_stat[l], false);
+ sum_stat(&dst->lat_stat[0], &src->lat_stat[l], false);
+ sum_stat(&dst->bw_stat[0], &src->bw_stat[l], true);
+ sum_stat(&dst->iops_stat[0], &src->iops_stat[l], true);
+ sum_clat_prio_stats(dst, src, 0, l);
dst->io_bytes[0] += src->io_bytes[l];
if (dst->runtime[0] < src->runtime[l])
dst->runtime[0] = src->runtime[l];
-
- /*
- * We're summing to the same destination, so override
- * 'first' after the first iteration of the loop
- */
- first = false;
}
}
- sum_stat(&dst->sync_stat, &src->sync_stat, first, false);
+ sum_stat(&dst->sync_stat, &src->sync_stat, false);
dst->usr_time += src->usr_time;
dst->sys_time += src->sys_time;
dst->ctx += src->ctx;
for (k = 0; k < FIO_IO_U_LAT_M_NR; k++)
dst->io_u_lat_m[k] += src->io_u_lat_m[k];
- for (k = 0; k < FIO_IO_U_PLAT_NR; k++)
- dst->io_u_sync_plat[k] += src->io_u_sync_plat[k];
-
for (k = 0; k < DDIR_RWDIR_CNT; k++) {
- if (!dst->unified_rw_rep) {
+ if (dst->unified_rw_rep != UNIFIED_MIXED) {
dst->total_io_u[k] += src->total_io_u[k];
dst->short_io_u[k] += src->short_io_u[k];
dst->drop_io_u[k] += src->drop_io_u[k];
dst->total_io_u[DDIR_SYNC] += src->total_io_u[DDIR_SYNC];
- for (k = 0; k < DDIR_RWDIR_CNT; k++) {
- int m;
+ for (k = 0; k < FIO_LAT_CNT; k++)
+ for (l = 0; l < DDIR_RWDIR_CNT; l++)
+ for (m = 0; m < FIO_IO_U_PLAT_NR; m++)
+ if (dst->unified_rw_rep != UNIFIED_MIXED)
+ dst->io_u_plat[k][l][m] += src->io_u_plat[k][l][m];
+ else
+ dst->io_u_plat[k][0][m] += src->io_u_plat[k][l][m];
- for (m = 0; m < FIO_IO_U_PLAT_NR; m++) {
- if (!dst->unified_rw_rep)
- dst->io_u_plat[k][m] += src->io_u_plat[k][m];
- else
- dst->io_u_plat[0][m] += src->io_u_plat[k][m];
- }
- }
+ for (k = 0; k < FIO_IO_U_PLAT_NR; k++)
+ dst->io_u_sync_plat[k] += src->io_u_sync_plat[k];
dst->total_run_time += src->total_run_time;
dst->total_submit += src->total_submit;
gs->min_bw[i] = gs->min_run[i] = ~0UL;
}
-void init_thread_stat(struct thread_stat *ts)
+void init_thread_stat_min_vals(struct thread_stat *ts)
{
- int j;
+ int i;
+
+ for (i = 0; i < DDIR_RWDIR_CNT; i++) {
+ ts->clat_stat[i].min_val = ULONG_MAX;
+ ts->slat_stat[i].min_val = ULONG_MAX;
+ ts->lat_stat[i].min_val = ULONG_MAX;
+ ts->bw_stat[i].min_val = ULONG_MAX;
+ ts->iops_stat[i].min_val = ULONG_MAX;
+ }
+ ts->sync_stat.min_val = ULONG_MAX;
+}
+void init_thread_stat(struct thread_stat *ts)
+{
memset(ts, 0, sizeof(*ts));
- for (j = 0; j < DDIR_RWDIR_CNT; j++) {
- ts->lat_stat[j].min_val = -1UL;
- ts->clat_stat[j].min_val = -1UL;
- ts->slat_stat[j].min_val = -1UL;
- ts->bw_stat[j].min_val = -1UL;
- ts->iops_stat[j].min_val = -1UL;
- }
- ts->sync_stat.min_val = -1UL;
+ init_thread_stat_min_vals(ts);
ts->groupid = -1;
}
+static void init_per_prio_stats(struct thread_stat *threadstats, int nr_ts)
+{
+ struct thread_data *td;
+ struct thread_stat *ts;
+ int i, j, last_ts, idx;
+ enum fio_ddir ddir;
+
+ j = 0;
+ last_ts = -1;
+ idx = 0;
+
+ /*
+ * Loop through all tds, if a td requires per prio stats, temporarily
+ * store a 1 in ts->disable_prio_stat, and then do an additional
+ * loop at the end where we invert the ts->disable_prio_stat values.
+ */
+ for_each_td(td, i) {
+ if (!td->o.stats)
+ continue;
+ if (idx &&
+ (!td->o.group_reporting ||
+ (td->o.group_reporting && last_ts != td->groupid))) {
+ idx = 0;
+ j++;
+ }
+
+ last_ts = td->groupid;
+ ts = &threadstats[j];
+
+ /* idx == 0 means first td in group, or td is not in a group. */
+ if (idx == 0)
+ ts->ioprio = td->ioprio;
+ else if (td->ioprio != ts->ioprio)
+ ts->disable_prio_stat = 1;
+
+ for (ddir = 0; ddir < DDIR_RWDIR_CNT; ddir++) {
+ if (td->ts.clat_prio[ddir]) {
+ ts->disable_prio_stat = 1;
+ break;
+ }
+ }
+
+ idx++;
+ }
+
+ /* Loop through all dst threadstats and fixup the values. */
+ for (i = 0; i < nr_ts; i++) {
+ ts = &threadstats[i];
+ ts->disable_prio_stat = !ts->disable_prio_stat;
+ }
+}
+
void __show_run_stats(void)
{
struct group_run_stats *runstats, *rs;
opt_lists[i] = NULL;
}
+ init_per_prio_stats(threadstats, nr_ts);
+
j = 0;
last_ts = -1;
idx = 0;
ts->clat_percentiles = td->o.clat_percentiles;
ts->lat_percentiles = td->o.lat_percentiles;
+ ts->slat_percentiles = td->o.slat_percentiles;
ts->percentile_precision = td->o.percentile_precision;
memcpy(ts->percentile_list, td->o.percentile_list, sizeof(td->o.percentile_list));
opt_lists[j] = &td->opt_list;
idx++;
- ts->members++;
if (ts->groupid == -1) {
/*
for (k = 0; k < ts->nr_block_infos; k++)
ts->block_infos[k] = td->ts.block_infos[k];
- sum_thread_stats(ts, &td->ts, idx == 1);
+ sum_thread_stats(ts, &td->ts);
+
+ ts->members++;
if (td->o.ss_dur) {
ts->ss_state = td->ss.state;
rs->kb_base = ts->kb_base;
rs->unit_base = ts->unit_base;
rs->sig_figs = ts->sig_figs;
- rs->unified_rw_rep += ts->unified_rw_rep;
+ rs->unified_rw_rep |= ts->unified_rw_rep;
for (j = 0; j < DDIR_RWDIR_CNT; j++) {
if (!ts->runtime[j])
}
for (i = 0; i < groupid + 1; i++) {
- int ddir;
+ enum fio_ddir ddir;
rs = &runstats[i];
log_info_flush();
free(runstats);
+
+ /* free arrays allocated by sum_thread_stats(), if any */
+ for (i = 0; i < nr_ts; i++) {
+ ts = &threadstats[i];
+ free_clat_prio_stats(ts);
+ }
free(threadstats);
free(opt_lists);
}
-void __show_running_run_stats(void)
+int __show_running_run_stats(void)
{
struct thread_data *td;
unsigned long long *rt;
for_each_td(td, i) {
td->update_rusage = 1;
- td->ts.io_bytes[DDIR_READ] = td->io_bytes[DDIR_READ];
- td->ts.io_bytes[DDIR_WRITE] = td->io_bytes[DDIR_WRITE];
- td->ts.io_bytes[DDIR_TRIM] = td->io_bytes[DDIR_TRIM];
+ for_each_rw_ddir(ddir) {
+ td->ts.io_bytes[ddir] = td->io_bytes[ddir];
+ }
td->ts.total_run_time = mtime_since(&td->epoch, &ts);
rt[i] = mtime_since(&td->start, &ts);
free(rt);
fio_sem_up(stat_sem);
+
+ return 0;
}
-static bool status_interval_init;
-static struct timespec status_time;
static bool status_file_disabled;
#define FIO_STATUS_FILE "fio-dump-status"
}
if (temp_dir == NULL)
temp_dir = "/tmp";
+#ifdef __COVERITY__
+ __coverity_tainted_data_sanitize__(temp_dir);
+#endif
snprintf(fio_status_file_path, sizeof(fio_status_file_path), "%s/%s", temp_dir, FIO_STATUS_FILE);
void check_for_running_stats(void)
{
- if (status_interval) {
- if (!status_interval_init) {
- fio_gettime(&status_time, NULL);
- status_interval_init = true;
- } else if (mtime_since_now(&status_time) >= status_interval) {
- show_running_run_stats();
- fio_gettime(&status_time, NULL);
- return;
- }
- }
if (check_status_file()) {
show_running_run_stats();
return;
is->samples++;
}
+static inline void add_stat_prio_sample(struct clat_prio_stat *clat_prio,
+ unsigned short clat_prio_index,
+ unsigned long long nsec)
+{
+ if (clat_prio)
+ add_stat_sample(&clat_prio[clat_prio_index].clat_stat, nsec);
+}
+
/*
* Return a struct io_logs, which is added to the tail of the log
* list for 'iolog'.
*/
static struct io_logs *get_new_log(struct io_log *iolog)
{
- size_t new_size, new_samples;
+ size_t new_samples;
struct io_logs *cur_log;
/*
* Cap the size at MAX_LOG_ENTRIES, so we don't keep doubling
* forever
*/
- if (!iolog->cur_log_max)
- new_samples = DEF_LOG_ENTRIES;
- else {
+ if (!iolog->cur_log_max) {
+ new_samples = iolog->td->o.log_entries;
+ } else {
new_samples = iolog->cur_log_max * 2;
if (new_samples > MAX_LOG_ENTRIES)
new_samples = MAX_LOG_ENTRIES;
}
- new_size = new_samples * log_entry_sz(iolog);
-
cur_log = smalloc(sizeof(*cur_log));
if (cur_log) {
INIT_FLIST_HEAD(&cur_log->list);
- cur_log->log = malloc(new_size);
+ cur_log->log = calloc(new_samples, log_entry_sz(iolog));
if (cur_log->log) {
cur_log->nr_samples = 0;
cur_log->max_samples = new_samples;
td->flags &= ~TD_F_REGROW_LOGS;
}
+void regrow_agg_logs(void)
+{
+ enum fio_ddir ddir;
+
+ for (ddir = 0; ddir < DDIR_RWDIR_CNT; ddir++)
+ regrow_log(agg_io_log[ddir]);
+}
+
static struct io_logs *get_cur_log(struct io_log *iolog)
{
struct io_logs *cur_log;
static void __add_log_sample(struct io_log *iolog, union io_sample_data data,
enum fio_ddir ddir, unsigned long long bs,
- unsigned long t, uint64_t offset)
+ unsigned long t, uint64_t offset,
+ unsigned int priority)
{
struct io_logs *cur_log;
s = get_sample(iolog, cur_log, cur_log->nr_samples);
s->data = data;
- s->time = t + (iolog->td ? iolog->td->unix_epoch : 0);
+ s->time = t + (iolog->td ? iolog->td->alternate_epoch : 0);
io_sample_set_ddir(iolog, s, ddir);
s->bs = bs;
+ s->priority = priority;
if (iolog->log_offset) {
struct io_sample_offset *so = (void *) s;
ios->mean.u.f = ios->S.u.f = 0;
}
+static inline void reset_io_u_plat(uint64_t *io_u_plat)
+{
+ int i;
+
+ for (i = 0; i < FIO_IO_U_PLAT_NR; i++)
+ io_u_plat[i] = 0;
+}
+
+static inline void reset_clat_prio_stats(struct thread_stat *ts)
+{
+ enum fio_ddir ddir;
+ int i;
+
+ for (ddir = 0; ddir < DDIR_RWDIR_CNT; ddir++) {
+ if (!ts->clat_prio[ddir])
+ continue;
+
+ for (i = 0; i < ts->nr_clat_prio[ddir]; i++) {
+ reset_io_stat(&ts->clat_prio[ddir][i].clat_stat);
+ reset_io_u_plat(ts->clat_prio[ddir][i].io_u_plat);
+ }
+ }
+}
+
void reset_io_stats(struct thread_data *td)
{
struct thread_stat *ts = &td->ts;
ts->total_io_u[i] = 0;
ts->short_io_u[i] = 0;
ts->drop_io_u[i] = 0;
-
- for (j = 0; j < FIO_IO_U_PLAT_NR; j++) {
- ts->io_u_plat[i][j] = 0;
- if (!i)
- ts->io_u_sync_plat[j] = 0;
- }
}
+ for (i = 0; i < FIO_LAT_CNT; i++)
+ for (j = 0; j < DDIR_RWDIR_CNT; j++)
+ reset_io_u_plat(ts->io_u_plat[i][j]);
+
+ reset_clat_prio_stats(ts);
+
ts->total_io_u[DDIR_SYNC] = 0;
+ reset_io_u_plat(ts->io_u_sync_plat);
for (i = 0; i < FIO_IO_U_MAP_NR; i++) {
ts->io_u_map[i] = 0;
else
data.val = iolog->avg_window[ddir].mean.u.f + 0.50;
- __add_log_sample(iolog, data, ddir, 0, elapsed, 0);
+ __add_log_sample(iolog, data, ddir, 0, elapsed, 0, 0);
}
reset_io_stat(&iolog->avg_window[ddir]);
static void _add_stat_to_log(struct io_log *iolog, unsigned long elapsed,
bool log_max)
{
- int ddir;
+ enum fio_ddir ddir;
for (ddir = 0; ddir < DDIR_RWDIR_CNT; ddir++)
__add_stat_to_log(iolog, ddir, elapsed, log_max);
struct io_log *iolog,
union io_sample_data data,
enum fio_ddir ddir, unsigned long long bs,
- uint64_t offset)
+ uint64_t offset, unsigned int ioprio)
{
unsigned long elapsed, this_window;
* If no time averaging, just add the log sample.
*/
if (!iolog->avg_msec) {
- __add_log_sample(iolog, data, ddir, bs, elapsed, offset);
+ __add_log_sample(iolog, data, ddir, bs, elapsed, offset,
+ ioprio);
return 0;
}
__add_stat_to_log(iolog, ddir, elapsed, td->o.log_max != 0);
- iolog->avg_last[ddir] = elapsed - (this_window - iolog->avg_msec);
+ iolog->avg_last[ddir] = elapsed - (elapsed % iolog->avg_msec);
+
return iolog->avg_msec;
}
_add_stat_to_log(td->iops_log, elapsed, td->o.log_max != 0);
}
-void add_agg_sample(union io_sample_data data, enum fio_ddir ddir, unsigned long long bs)
+void add_agg_sample(union io_sample_data data, enum fio_ddir ddir,
+ unsigned long long bs)
{
struct io_log *iolog;
return;
iolog = agg_io_log[ddir];
- __add_log_sample(iolog, data, ddir, bs, mtime_since_genesis(), 0);
+ __add_log_sample(iolog, data, ddir, bs, mtime_since_genesis(), 0, 0);
}
void add_sync_clat_sample(struct thread_stat *ts, unsigned long long nsec)
add_stat_sample(&ts->sync_stat, nsec);
}
-static void add_clat_percentile_sample(struct thread_stat *ts,
- unsigned long long nsec, enum fio_ddir ddir)
+static inline void add_lat_percentile_sample(struct thread_stat *ts,
+ unsigned long long nsec,
+ enum fio_ddir ddir,
+ enum fio_lat lat)
{
unsigned int idx = plat_val_to_idx(nsec);
assert(idx < FIO_IO_U_PLAT_NR);
- ts->io_u_plat[ddir][idx]++;
+ ts->io_u_plat[lat][ddir][idx]++;
+}
+
+static inline void
+add_lat_percentile_prio_sample(struct thread_stat *ts, unsigned long long nsec,
+ enum fio_ddir ddir,
+ unsigned short clat_prio_index)
+{
+ unsigned int idx = plat_val_to_idx(nsec);
+
+ if (ts->clat_prio[ddir])
+ ts->clat_prio[ddir][clat_prio_index].io_u_plat[idx]++;
}
void add_clat_sample(struct thread_data *td, enum fio_ddir ddir,
unsigned long long nsec, unsigned long long bs,
- uint64_t offset)
+ uint64_t offset, unsigned int ioprio,
+ unsigned short clat_prio_index)
{
const bool needs_lock = td_async_processing(td);
unsigned long elapsed, this_window;
add_stat_sample(&ts->clat_stat[ddir], nsec);
+ /*
+ * When lat_percentiles=1 (default 0), the reported per priority
+ * percentiles and stats are used for describing total latency values,
+ * even though the variable names themselves start with clat_.
+ *
+ * Because of the above definition, add a prio stat sample only when
+ * lat_percentiles=0. add_lat_sample() will add the prio stat sample
+ * when lat_percentiles=1.
+ */
+ if (!ts->lat_percentiles)
+ add_stat_prio_sample(ts->clat_prio[ddir], clat_prio_index,
+ nsec);
+
if (td->clat_log)
add_log_sample(td, td->clat_log, sample_val(nsec), ddir, bs,
- offset);
+ offset, ioprio);
- if (ts->clat_percentiles)
- add_clat_percentile_sample(ts, nsec, ddir);
+ if (ts->clat_percentiles) {
+ /*
+ * Because of the above definition, add a prio lat percentile
+ * sample only when lat_percentiles=0. add_lat_sample() will add
+ * the prio lat percentile sample when lat_percentiles=1.
+ */
+ add_lat_percentile_sample(ts, nsec, ddir, FIO_CLAT);
+ if (!ts->lat_percentiles)
+ add_lat_percentile_prio_sample(ts, nsec, ddir,
+ clat_prio_index);
+ }
if (iolog && iolog->hist_msec) {
struct io_hist *hw = &iolog->hist_window[ddir];
if (!hw->hist_last)
hw->hist_last = elapsed;
this_window = elapsed - hw->hist_last;
-
+
if (this_window >= iolog->hist_msec) {
uint64_t *io_u_plat;
struct io_u_plat_entry *dst;
* located in iolog.c after printing this sample to the
* log file.
*/
- io_u_plat = (uint64_t *) td->ts.io_u_plat[ddir];
+ io_u_plat = (uint64_t *) td->ts.io_u_plat[FIO_CLAT][ddir];
dst = malloc(sizeof(struct io_u_plat_entry));
memcpy(&(dst->io_u_plat), io_u_plat,
FIO_IO_U_PLAT_NR * sizeof(uint64_t));
flist_add(&dst->list, &hw->list);
__add_log_sample(iolog, sample_plat(dst), ddir, bs,
- elapsed, offset);
+ elapsed, offset, ioprio);
/*
* Update the last time we recorded as being now, minus
}
void add_slat_sample(struct thread_data *td, enum fio_ddir ddir,
- unsigned long usec, unsigned long long bs, uint64_t offset)
+ unsigned long long nsec, unsigned long long bs,
+ uint64_t offset, unsigned int ioprio)
{
const bool needs_lock = td_async_processing(td);
struct thread_stat *ts = &td->ts;
if (needs_lock)
__td_io_u_lock(td);
- add_stat_sample(&ts->slat_stat[ddir], usec);
+ add_stat_sample(&ts->slat_stat[ddir], nsec);
if (td->slat_log)
- add_log_sample(td, td->slat_log, sample_val(usec), ddir, bs, offset);
+ add_log_sample(td, td->slat_log, sample_val(nsec), ddir, bs,
+ offset, ioprio);
+
+ if (ts->slat_percentiles)
+ add_lat_percentile_sample(ts, nsec, ddir, FIO_SLAT);
if (needs_lock)
__td_io_u_unlock(td);
void add_lat_sample(struct thread_data *td, enum fio_ddir ddir,
unsigned long long nsec, unsigned long long bs,
- uint64_t offset)
+ uint64_t offset, unsigned int ioprio,
+ unsigned short clat_prio_index)
{
const bool needs_lock = td_async_processing(td);
struct thread_stat *ts = &td->ts;
if (td->lat_log)
add_log_sample(td, td->lat_log, sample_val(nsec), ddir, bs,
- offset);
-
- if (ts->lat_percentiles)
- add_clat_percentile_sample(ts, nsec, ddir);
+ offset, ioprio);
+ /*
+ * When lat_percentiles=1 (default 0), the reported per priority
+ * percentiles and stats are used for describing total latency values,
+ * even though the variable names themselves start with clat_.
+ *
+ * Because of the above definition, add a prio stat and prio lat
+ * percentile sample only when lat_percentiles=1. add_clat_sample() will
+ * add the prio stat and prio lat percentile sample when
+ * lat_percentiles=0.
+ */
+ if (ts->lat_percentiles) {
+ add_lat_percentile_sample(ts, nsec, ddir, FIO_LAT);
+ add_lat_percentile_prio_sample(ts, nsec, ddir, clat_prio_index);
+ add_stat_prio_sample(ts->clat_prio[ddir], clat_prio_index,
+ nsec);
+ }
if (needs_lock)
__td_io_u_unlock(td);
}
if (td->bw_log)
add_log_sample(td, td->bw_log, sample_val(rate), io_u->ddir,
- bytes, io_u->offset);
+ bytes, io_u->offset, io_u->ioprio);
td->stat_io_bytes[io_u->ddir] = td->this_io_bytes[io_u->ddir];
next_log = avg_time;
spent = mtime_since(parent_tv, t);
- if (spent < avg_time && avg_time - spent >= LOG_MSEC_SLACK)
+ if (spent < avg_time && avg_time - spent > LOG_MSEC_SLACK)
return avg_time - spent;
if (needs_lock)
if (td->o.min_bs[ddir] == td->o.max_bs[ddir])
bs = td->o.min_bs[ddir];
- next = add_log_sample(td, log, sample_val(rate), ddir, bs, 0);
+ next = add_log_sample(td, log, sample_val(rate), ddir,
+ bs, 0, 0);
next_log = min(next_log, next);
}
stat_io_bytes[ddir] = this_io_bytes[ddir];
}
- timespec_add_msec(parent_tv, avg_time);
+ *parent_tv = *t;
if (needs_lock)
__td_io_u_unlock(td);
if (td->iops_log)
add_log_sample(td, td->iops_log, sample_val(1), io_u->ddir,
- bytes, io_u->offset);
+ bytes, io_u->offset, io_u->ioprio);
td->stat_io_blocks[io_u->ddir] = td->this_io_blocks[io_u->ddir];
int calc_log_samples(void)
{
struct thread_data *td;
- unsigned int next = ~0U, tmp;
+ unsigned int next = ~0U, tmp = 0, next_mod = 0, log_avg_msec_min = -1U;
struct timespec now;
int i;
+ long elapsed_time = 0;
fio_gettime(&now, NULL);
for_each_td(td, i) {
+ elapsed_time = mtime_since_now(&td->epoch);
+
if (!td->o.stats)
continue;
if (in_ramp_time(td) ||
if (!td->bw_log ||
(td->bw_log && !per_unit_log(td->bw_log))) {
tmp = add_bw_samples(td, &now);
- if (tmp < next)
- next = tmp;
+
+ if (td->bw_log)
+ log_avg_msec_min = min(log_avg_msec_min, (unsigned int)td->bw_log->avg_msec);
}
if (!td->iops_log ||
(td->iops_log && !per_unit_log(td->iops_log))) {
tmp = add_iops_samples(td, &now);
- if (tmp < next)
- next = tmp;
+
+ if (td->iops_log)
+ log_avg_msec_min = min(log_avg_msec_min, (unsigned int)td->iops_log->avg_msec);
}
+
+ if (tmp < next)
+ next = tmp;
}
+ /* if log_avg_msec_min has not been changed, set it to 0 */
+ if (log_avg_msec_min == -1U)
+ log_avg_msec_min = 0;
+
+ if (log_avg_msec_min == 0)
+ next_mod = elapsed_time;
+ else
+ next_mod = elapsed_time % log_avg_msec_min;
+
+ /* correction to keep the time on the log avg msec boundary */
+ next = min(next, (log_avg_msec_min - next_mod));
+
return next == ~0U ? 0 : next;
}
assert(idx < td->ts.nr_block_infos);
return info;
}
+