#include <stdio.h>
#include <unistd.h>
#include <string.h>
-#include <getopt.h>
#include <inttypes.h>
#include <assert.h>
+struct thread_data;
+
#include "compiler/compiler.h"
#include "flist.h"
#include "fifo.h"
#include "io_ddir.h"
#include "ioengine.h"
#include "iolog.h"
+#include "helpers.h"
+#include "options.h"
+#include "profile.h"
+#include "time.h"
+#include "lib/getopt.h"
+#include "lib/rand.h"
#ifdef FIO_HAVE_GUASI
#include <guasi.h>
unsigned long long max_bw[2], min_bw[2];
unsigned long long io_kb[2];
unsigned long long agg[2];
+ unsigned int kb_base;
};
/*
MEM_MMAPHUGE, /* memory mapped huge file */
};
+/*
+ * offset generator types
+ */
+enum {
+ RW_SEQ_SEQ = 0,
+ RW_SEQ_IDENT,
+};
+
/*
* How many depth levels to log
*/
-#define FIO_IO_U_MAP_NR 8
+#define FIO_IO_U_MAP_NR 7
#define FIO_IO_U_LAT_U_NR 10
#define FIO_IO_U_LAT_M_NR 12
+/*
+ * Aggregate clat samples to report percentile(s) of them.
+ *
+ * EXECUTIVE SUMMARY
+ *
+ * FIO_IO_U_PLAT_BITS determines the maximum statistical error on the
+ * value of resulting percentiles. The error will be approximately
+ * 1/2^(FIO_IO_U_PLAT_BITS+1) of the value.
+ *
+ * FIO_IO_U_PLAT_GROUP_NR and FIO_IO_U_PLAT_BITS determine the maximum
+ * range being tracked for latency samples. The maximum value tracked
+ * accurately will be 2^(GROUP_NR + PLAT_BITS -1) microseconds.
+ *
+ * FIO_IO_U_PLAT_GROUP_NR and FIO_IO_U_PLAT_BITS determine the memory
+ * requirement of storing those aggregate counts. The memory used will
+ * be (FIO_IO_U_PLAT_GROUP_NR * 2^FIO_IO_U_PLAT_BITS) * sizeof(int)
+ * bytes.
+ *
+ * FIO_IO_U_PLAT_NR is the total number of buckets.
+ *
+ * DETAILS
+ *
+ * Suppose the clat varies from 0 to 999 (usec), the straightforward
+ * method is to keep an array of (999 + 1) buckets, in which a counter
+ * keeps the count of samples which fall in the bucket, e.g.,
+ * {[0],[1],...,[999]}. However this consumes a huge amount of space,
+ * and can be avoided if an approximation is acceptable.
+ *
+ * One such method is to let the range of the bucket to be greater
+ * than one. This method has low accuracy when the value is small. For
+ * example, let the buckets be {[0,99],[100,199],...,[900,999]}, and
+ * the represented value of each bucket be the mean of the range. Then
+ * a value 0 has an round-off error of 49.5. To improve on this, we
+ * use buckets with non-uniform ranges, while bounding the error of
+ * each bucket within a ratio of the sample value. A simple example
+ * would be when error_bound = 0.005, buckets are {
+ * {[0],[1],...,[99]}, {[100,101],[102,103],...,[198,199]},..,
+ * {[900,909],[910,919]...} }. The total range is partitioned into
+ * groups with different ranges, then buckets with uniform ranges. An
+ * upper bound of the error is (range_of_bucket/2)/value_of_bucket
+ *
+ * For better efficiency, we implement this using base two. We group
+ * samples by their Most Significant Bit (MSB), extract the next M bit
+ * of them as an index within the group, and discard the rest of the
+ * bits.
+ *
+ * E.g., assume a sample 'x' whose MSB is bit n (starting from bit 0),
+ * and use M bit for indexing
+ *
+ * | n | M bits | bit (n-M-1) ... bit 0 |
+ *
+ * Because x is at least 2^n, and bit 0 to bit (n-M-1) is at most
+ * (2^(n-M) - 1), discarding bit 0 to (n-M-1) makes the round-off
+ * error
+ *
+ * 2^(n-M)-1 2^(n-M) 1
+ * e <= --------- <= ------- = ---
+ * 2^n 2^n 2^M
+ *
+ * Furthermore, we use "mean" of the range to represent the bucket,
+ * the error e can be lowered by half to 1 / 2^(M+1). By using M bits
+ * as the index, each group must contains 2^M buckets.
+ *
+ * E.g. Let M (FIO_IO_U_PLAT_BITS) be 6
+ * Error bound is 1/2^(6+1) = 0.0078125 (< 1%)
+ *
+ * Group MSB #discarded range of #buckets
+ * error_bits value
+ * ----------------------------------------------------------------
+ * 0* 0~5 0 [0,63] 64
+ * 1* 6 0 [64,127] 64
+ * 2 7 1 [128,255] 64
+ * 3 8 2 [256,511] 64
+ * 4 9 3 [512,1023] 64
+ * ... ... ... [...,...] ...
+ * 18 23 17 [8838608,+inf]** 64
+ *
+ * * Special cases: when n < (M-1) or when n == (M-1), in both cases,
+ * the value cannot be rounded off. Use all bits of the sample as
+ * index.
+ *
+ * ** If a sample's MSB is greater than 23, it will be counted as 23.
+ */
+
+#define FIO_IO_U_PLAT_BITS 6
+#define FIO_IO_U_PLAT_VAL (1 << FIO_IO_U_PLAT_BITS)
+#define FIO_IO_U_PLAT_GROUP_NR 19
+#define FIO_IO_U_PLAT_NR (FIO_IO_U_PLAT_GROUP_NR * FIO_IO_U_PLAT_VAL)
+#define FIO_IO_U_LIST_MAX_LEN 20 /* The size of the default and user-specified
+ list of percentiles */
+
+#define MAX_PATTERN_SIZE 512
+
struct thread_stat {
char *name;
char *verror;
struct io_log *slat_log;
struct io_log *clat_log;
+ struct io_log *lat_log;
struct io_log *bw_log;
/*
*/
struct io_stat clat_stat[2]; /* completion latency */
struct io_stat slat_stat[2]; /* submission latency */
+ struct io_stat lat_stat[2]; /* total latency */
struct io_stat bw_stat[2]; /* bandwidth stats */
unsigned long long stat_io_bytes[2];
/*
* IO depth and latency stats
*/
+ unsigned int clat_percentiles;
+ double* percentile_list;
+
unsigned int io_u_map[FIO_IO_U_MAP_NR];
unsigned int io_u_submit[FIO_IO_U_MAP_NR];
unsigned int io_u_complete[FIO_IO_U_MAP_NR];
unsigned int io_u_lat_u[FIO_IO_U_LAT_U_NR];
unsigned int io_u_lat_m[FIO_IO_U_LAT_M_NR];
- unsigned long total_io_u[2];
- unsigned long short_io_u[2];
+ unsigned int io_u_plat[2][FIO_IO_U_PLAT_NR];
+ unsigned long total_io_u[3];
+ unsigned long short_io_u[3];
unsigned long total_submit;
unsigned long total_complete;
unsigned long long io_bytes[2];
- unsigned long runtime[2];
+ unsigned long long runtime[2];
unsigned long total_run_time;
+
+ /*
+ * IO Error related stats
+ */
+ unsigned continue_on_error;
+ unsigned long total_err_count;
+ int first_error;
+
+ unsigned int kb_base;
};
struct bssplit {
char *opendir;
char *ioengine;
enum td_ddir td_ddir;
- unsigned int ddir_nr;
+ unsigned int rw_seq;
+ unsigned int kb_base;
+ unsigned int ddir_seq_nr;
unsigned int iodepth;
unsigned int iodepth_low;
unsigned int iodepth_batch;
unsigned int iodepth_batch_complete;
unsigned long long size;
+ unsigned int size_percent;
unsigned int fill_device;
unsigned long long file_size_low;
unsigned long long file_size_high;
unsigned int verifysort;
unsigned int verify_interval;
unsigned int verify_offset;
- unsigned int verify_pattern;
+ char verify_pattern[MAX_PATTERN_SIZE];
unsigned int verify_pattern_bytes;
unsigned int verify_fatal;
+ unsigned int verify_dump;
+ unsigned int verify_async;
+ unsigned long long verify_backlog;
+ unsigned int verify_batch;
unsigned int use_thread;
unsigned int unlink;
unsigned int do_disk_util;
unsigned int override_sync;
unsigned int rand_repeatable;
+ unsigned int use_os_rand;
unsigned int write_lat_log;
unsigned int write_bw_log;
unsigned int norandommap;
unsigned int thinktime_spin;
unsigned int thinktime_blocks;
unsigned int fsync_blocks;
- unsigned int start_delay;
+ unsigned int fdatasync_blocks;
+ unsigned int barrier_blocks;
+ unsigned long long start_delay;
unsigned long long timeout;
unsigned long long ramp_time;
unsigned int overwrite;
unsigned long long zone_size;
unsigned long long zone_skip;
enum fio_memtype mem_type;
+ unsigned int mem_align;
unsigned int stonewall;
unsigned int new_group;
unsigned int numjobs;
os_cpu_mask_t cpumask;
unsigned int cpumask_set;
+ os_cpu_mask_t verify_cpumask;
+ unsigned int verify_cpumask_set;
unsigned int iolog;
unsigned int rwmixcycle;
unsigned int rwmix[2];
unsigned int file_service_type;
unsigned int group_reporting;
unsigned int fadvise_hint;
+ enum fio_fallocate_mode fallocate_mode;
unsigned int zero_buffers;
unsigned int refill_buffers;
unsigned int time_based;
+ unsigned int disable_lat;
unsigned int disable_clat;
unsigned int disable_slat;
unsigned int disable_bw;
unsigned int gtod_reduce;
unsigned int gtod_cpu;
unsigned int gtod_offload;
+ enum fio_cs clocksource;
+ unsigned int no_stall;
+ unsigned int trim_percentage;
+ unsigned int trim_batch;
+ unsigned int trim_zero;
+ unsigned long long trim_backlog;
+ unsigned int clat_percentiles;
+ unsigned int overwrite_plist;
+ double percentile_list[FIO_IO_U_LIST_MAX_LEN];
char *read_iolog_file;
char *write_iolog_file;
char *bw_log_file;
char *lat_log_file;
+ char *replay_redirect;
/*
* Pre-run and post-run shell
*/
unsigned int cpuload;
unsigned int cpucycle;
+
+ /*
+ * I/O Error handling
+ */
+ unsigned int continue_on_error;
+
+ /*
+ * Benchmark profile type
+ */
+ char *profile;
+
+ /*
+ * blkio cgroup support
+ */
+ char *cgroup;
+ unsigned int cgroup_weight;
+ unsigned int cgroup_nodelete;
+
+ unsigned int uid;
+ unsigned int gid;
+
+ unsigned int sync_file_range;
};
#define FIO_VERROR_SIZE 128
union {
unsigned int next_file;
os_random_state_t next_file_state;
+ struct frand_state __next_file_state;
};
int error;
int done;
unsigned int ioprio;
unsigned int ioprio_set;
unsigned int last_was_sync;
+ enum fio_ddir last_ddir;
char *mmapfile;
int mmapfd;
char *sysfs_root;
- unsigned long rand_seeds[6];
+ unsigned long rand_seeds[7];
+
+ union {
+ os_random_state_t bsrange_state;
+ struct frand_state __bsrange_state;
+ };
+ union {
+ os_random_state_t verify_state;
+ struct frand_state __verify_state;
+ };
+ union {
+ os_random_state_t trim_state;
+ struct frand_state __trim_state;
+ };
- os_random_state_t bsrange_state;
- os_random_state_t verify_state;
+ unsigned int verify_batch;
+ unsigned int trim_batch;
int shm_id;
struct flist_head io_u_freelist;
struct flist_head io_u_busylist;
struct flist_head io_u_requeues;
+ pthread_mutex_t io_u_lock;
+ pthread_cond_t free_cond;
+
+ /*
+ * async verify offload
+ */
+ struct flist_head verify_list;
+ pthread_t *verify_threads;
+ unsigned int nr_verify_threads;
+ pthread_cond_t verify_cond;
+ int verify_thread_exit;
/*
* Rate state
*/
- unsigned long rate_usec_cycle[2];
+ unsigned long rate_nsec_cycle[2];
long rate_pending_usleep[2];
unsigned long rate_bytes[2];
unsigned long rate_blocks[2];
struct timeval lastrate[2];
unsigned long long total_io_size;
+ unsigned long long fill_device_size;
unsigned long io_issues[2];
unsigned long long io_blocks[2];
/*
* State for random io, a bitmap of blocks done vs not done
*/
- os_random_state_t random_state;
+ union {
+ os_random_state_t random_state;
+ struct frand_state __random_state;
+ };
struct timeval start; /* start of this loop */
struct timeval epoch; /* time job was started */
/*
* read/write mixed workload state
*/
- os_random_state_t rwmix_state;
+ union {
+ os_random_state_t rwmix_state;
+ struct frand_state __rwmix_state;
+ };
unsigned long rwmix_issues;
enum fio_ddir rwmix_ddir;
- unsigned int ddir_nr;
+ unsigned int ddir_seq_nr;
/*
* IO history logs for verification. We use a tree for sorting,
*/
struct rb_root io_hist_tree;
struct flist_head io_hist_list;
+ unsigned long io_hist_len;
/*
* For IO replaying
*/
struct flist_head io_log_list;
+ /*
+ * For tracking/handling discards
+ */
+ struct flist_head trim_list;
+ unsigned long trim_entries;
+
/*
* for fileservice, how often to switch to a new file
*/
unsigned int file_service_left;
struct fio_file *file_service_file;
+ unsigned int sync_file_range_nr;
+
/*
* For generating file sizes
*/
- os_random_state_t file_size_state;
+ union {
+ os_random_state_t file_size_state;
+ struct frand_state __file_size_state;
+ };
+
+ /*
+ * Error counts
+ */
+ unsigned int total_err_count;
+ int first_error;
+
+ /*
+ * Can be overloaded by profiles
+ */
+ struct prof_io_ops prof_io_ops;
+ void *prof_data;
};
/*
break; \
int e = (err); \
(td)->error = e; \
- snprintf(td->verror, sizeof(td->verror) - 1, "file:%s:%d, func=%s, error=%s", __FILE__, __LINE__, (func), (msg)); \
+ if (!(td)->first_error) \
+ snprintf(td->verror, sizeof(td->verror) - 1, "file:%s:%d, func=%s, error=%s", __FILE__, __LINE__, (func), (msg)); \
} while (0)
+#define td_clear_error(td) \
+ (td)->error = 0;
#define td_verror(td, err, func) \
__td_verror((td), (err), strerror((err)), (func))
#define td_vmsg(td, err, msg, func) \
extern char *job_section;
extern int fio_gtod_offload;
extern int fio_gtod_cpu;
+extern enum fio_cs fio_clock_source;
+extern int warnings_fatal;
extern struct thread_data *threads;
assert(!(io_u->ddir == DDIR_WRITE && !td_write(td)));
}
-#define BLOCKS_PER_MAP (8 * sizeof(int))
+#define BLOCKS_PER_MAP (8 * sizeof(unsigned long))
#define TO_MAP_BLOCK(f, b) (b)
#define RAND_MAP_IDX(f, b) (TO_MAP_BLOCK(f, b) / BLOCKS_PER_MAP)
#define RAND_MAP_BIT(f, b) (TO_MAP_BLOCK(f, b) & (BLOCKS_PER_MAP - 1))
-#define MAX_JOBS (1024)
+#define REAL_MAX_JOBS 2048
+
+#define td_non_fatal_error(e) ((e) == EIO || (e) == EILSEQ)
+
+static inline void update_error_count(struct thread_data *td, int err)
+{
+ td->total_err_count++;
+ if (td->total_err_count == 1)
+ td->first_error = err;
+}
static inline int should_fsync(struct thread_data *td)
{
return 0;
}
-/*
- * Time functions
- */
-extern unsigned long long utime_since(struct timeval *, struct timeval *);
-extern unsigned long long utime_since_now(struct timeval *);
-extern unsigned long mtime_since(struct timeval *, struct timeval *);
-extern unsigned long mtime_since_now(struct timeval *);
-extern unsigned long time_since_now(struct timeval *);
-extern unsigned long mtime_since_genesis(void);
-extern void usec_spin(unsigned int);
-extern void usec_sleep(struct thread_data *, unsigned long);
-extern long rate_throttle(struct thread_data *, unsigned long, unsigned long, enum fio_ddir);
-extern void fill_start_time(struct timeval *);
-extern void fio_gettime(struct timeval *, void *);
-extern void fio_gtod_init(void);
-extern void fio_gtod_update(void);
-extern void set_genesis_time(void);
-extern int ramp_time_over(struct thread_data *);
-extern int in_ramp_time(struct thread_data *);
-
/*
* Init/option functions
*/
extern int __must_check parse_options(int, char **);
extern int fio_options_parse(struct thread_data *, char **, int);
+extern void fio_keywords_init(void);
extern int fio_cmd_option_parse(struct thread_data *, const char *, char *);
extern void fio_fill_default_options(struct thread_data *);
extern int fio_show_option_help(const char *);
extern void options_mem_dupe(struct thread_data *);
extern void options_mem_free(struct thread_data *);
extern void td_fill_rand_seeds(struct thread_data *);
+extern void add_job_opts(const char **);
+extern char *num2str(unsigned long, int, int, int);
+
#define FIO_GETOPT_JOB 0x89988998
-#define FIO_NR_OPTIONS 128
+#define FIO_NR_OPTIONS (FIO_MAX_OPTS + 128)
/*
* ETA/status stuff
if (!(cond)) { \
int *__foo = NULL; \
fprintf(stderr, "file:%s:%d, assert %s failed\n", __FILE__, __LINE__, #cond); \
- (td)->runstate = TD_EXITED; \
+ td_set_runstate((td), TD_EXITED); \
(td)->error = EFAULT; \
*__foo = 0; \
} \
return 0;
}
-/*
- * Cheesy number->string conversion, complete with carry rounding error.
- */
-static inline char *num2str(unsigned long num, int maxlen, int base, int pow2)
-{
- char postfix[] = { ' ', 'K', 'M', 'G', 'P', 'E' };
- unsigned int thousand;
- char *buf;
- int i;
-
- if (pow2)
- thousand = 1024;
- else
- thousand = 1000;
-
- buf = malloc(128);
-
- for (i = 0; base > 1; i++)
- base /= thousand;
-
- do {
- int len, carry = 0;
-
- len = sprintf(buf, "%'lu", num);
- if (len <= maxlen) {
- if (i >= 1) {
- buf[len] = postfix[i];
- buf[len + 1] = '\0';
- }
- return buf;
- }
-
- if ((num % thousand) >= (thousand / 2))
- carry = 1;
-
- num /= thousand;
- num += carry;
- i++;
- } while (i <= 5);
-
- return buf;
-}
-
static inline int __should_check_rate(struct thread_data *td,
enum fio_ddir ddir)
{
return ret;
}
+static inline int is_power_of_2(unsigned int val)
+{
+ return (val != 0 && ((val & (val - 1)) == 0));
+}
+
+/*
+ * We currently only need to do locking if we have verifier threads
+ * accessing our internal structures too
+ */
+static inline void td_io_u_lock(struct thread_data *td)
+{
+ if (td->o.verify_async)
+ pthread_mutex_lock(&td->io_u_lock);
+}
+
+static inline void td_io_u_unlock(struct thread_data *td)
+{
+ if (td->o.verify_async)
+ pthread_mutex_unlock(&td->io_u_lock);
+}
+
+static inline void td_io_u_free_notify(struct thread_data *td)
+{
+ if (td->o.verify_async)
+ pthread_cond_signal(&td->free_cond);
+}
+
#endif
projects / fio.git / blobdiff