2006-03-23 20:00:26 +01:00
# ifndef BLKTRACE_H
# define BLKTRACE_H
2009-01-30 20:36:52 +05:30
# include <linux/types.h>
2008-09-26 10:58:02 +02:00
# ifdef __KERNEL__
2006-03-23 20:00:26 +01:00
# include <linux/blkdev.h>
# include <linux/relay.h>
2010-07-07 16:51:26 +02:00
# include <linux/compat.h>
2008-09-26 10:58:02 +02:00
# endif
2006-03-23 20:00:26 +01:00
/*
* Trace categories
*/
enum blktrace_cat {
BLK_TC_READ = 1 < < 0 , /* reads */
BLK_TC_WRITE = 1 < < 1 , /* writes */
BLK_TC_BARRIER = 1 < < 2 , /* barrier */
2006-07-06 09:56:30 +02:00
BLK_TC_SYNC = 1 < < 3 , /* sync IO */
2009-02-16 10:25:40 +01:00
BLK_TC_SYNCIO = BLK_TC_SYNC ,
2006-03-23 20:00:26 +01:00
BLK_TC_QUEUE = 1 < < 4 , /* queueing/merging */
BLK_TC_REQUEUE = 1 < < 5 , /* requeueing */
BLK_TC_ISSUE = 1 < < 6 , /* issue */
BLK_TC_COMPLETE = 1 < < 7 , /* completions */
BLK_TC_FS = 1 < < 8 , /* fs requests */
BLK_TC_PC = 1 < < 9 , /* pc requests */
BLK_TC_NOTIFY = 1 < < 10 , /* special message */
2006-07-06 10:03:28 +02:00
BLK_TC_AHEAD = 1 < < 11 , /* readahead */
2006-07-23 02:12:01 +02:00
BLK_TC_META = 1 < < 12 , /* metadata */
2008-08-10 11:21:57 +01:00
BLK_TC_DISCARD = 1 < < 13 , /* discard requests */
2008-10-16 08:23:21 +02:00
BLK_TC_DRV_DATA = 1 < < 14 , /* binary per-driver data */
2006-03-23 20:00:26 +01:00
BLK_TC_END = 1 < < 15 , /* only 16-bits, reminder */
} ;
# define BLK_TC_SHIFT (16)
# define BLK_TC_ACT(act) ((act) << BLK_TC_SHIFT)
/*
* Basic trace actions
*/
enum blktrace_act {
__BLK_TA_QUEUE = 1 , /* queued */
__BLK_TA_BACKMERGE , /* back merged to existing rq */
__BLK_TA_FRONTMERGE , /* front merge to existing rq */
__BLK_TA_GETRQ , /* allocated new request */
__BLK_TA_SLEEPRQ , /* sleeping on rq allocation */
__BLK_TA_REQUEUE , /* request requeued */
__BLK_TA_ISSUE , /* sent to driver */
__BLK_TA_COMPLETE , /* completed by driver */
__BLK_TA_PLUG , /* queue was plugged */
__BLK_TA_UNPLUG_IO , /* queue was unplugged by io */
__BLK_TA_UNPLUG_TIMER , /* queue was unplugged by timer */
__BLK_TA_INSERT , /* insert request */
__BLK_TA_SPLIT , /* bio was split */
__BLK_TA_BOUNCE , /* bio was bounced */
__BLK_TA_REMAP , /* bio was remapped */
2008-09-13 20:31:27 +02:00
__BLK_TA_ABORT , /* request aborted */
2008-10-16 08:23:21 +02:00
__BLK_TA_DRV_DATA , /* driver-specific binary data */
2006-03-23 20:00:26 +01:00
} ;
2006-12-01 10:39:12 +01:00
/*
* Notify events .
*/
enum blktrace_notify {
__BLK_TN_PROCESS = 0 , /* establish pid/name mapping */
__BLK_TN_TIMESTAMP , /* include system clock */
2008-05-27 14:54:41 +02:00
__BLK_TN_MESSAGE , /* Character string message */
2006-12-01 10:39:12 +01:00
} ;
2006-03-23 20:00:26 +01:00
/*
* Trace actions in full . Additionally , read or write is masked
*/
# define BLK_TA_QUEUE (__BLK_TA_QUEUE | BLK_TC_ACT(BLK_TC_QUEUE))
# define BLK_TA_BACKMERGE (__BLK_TA_BACKMERGE | BLK_TC_ACT(BLK_TC_QUEUE))
# define BLK_TA_FRONTMERGE (__BLK_TA_FRONTMERGE | BLK_TC_ACT(BLK_TC_QUEUE))
# define BLK_TA_GETRQ (__BLK_TA_GETRQ | BLK_TC_ACT(BLK_TC_QUEUE))
# define BLK_TA_SLEEPRQ (__BLK_TA_SLEEPRQ | BLK_TC_ACT(BLK_TC_QUEUE))
# define BLK_TA_REQUEUE (__BLK_TA_REQUEUE | BLK_TC_ACT(BLK_TC_REQUEUE))
# define BLK_TA_ISSUE (__BLK_TA_ISSUE | BLK_TC_ACT(BLK_TC_ISSUE))
# define BLK_TA_COMPLETE (__BLK_TA_COMPLETE| BLK_TC_ACT(BLK_TC_COMPLETE))
# define BLK_TA_PLUG (__BLK_TA_PLUG | BLK_TC_ACT(BLK_TC_QUEUE))
# define BLK_TA_UNPLUG_IO (__BLK_TA_UNPLUG_IO | BLK_TC_ACT(BLK_TC_QUEUE))
# define BLK_TA_UNPLUG_TIMER (__BLK_TA_UNPLUG_TIMER | BLK_TC_ACT(BLK_TC_QUEUE))
# define BLK_TA_INSERT (__BLK_TA_INSERT | BLK_TC_ACT(BLK_TC_QUEUE))
# define BLK_TA_SPLIT (__BLK_TA_SPLIT)
# define BLK_TA_BOUNCE (__BLK_TA_BOUNCE)
# define BLK_TA_REMAP (__BLK_TA_REMAP | BLK_TC_ACT(BLK_TC_QUEUE))
2008-09-13 20:31:27 +02:00
# define BLK_TA_ABORT (__BLK_TA_ABORT | BLK_TC_ACT(BLK_TC_QUEUE))
2008-10-16 08:23:21 +02:00
# define BLK_TA_DRV_DATA (__BLK_TA_DRV_DATA | BLK_TC_ACT(BLK_TC_DRV_DATA))
2006-03-23 20:00:26 +01:00
2006-12-01 10:39:12 +01:00
# define BLK_TN_PROCESS (__BLK_TN_PROCESS | BLK_TC_ACT(BLK_TC_NOTIFY))
# define BLK_TN_TIMESTAMP (__BLK_TN_TIMESTAMP | BLK_TC_ACT(BLK_TC_NOTIFY))
2008-05-27 14:54:41 +02:00
# define BLK_TN_MESSAGE (__BLK_TN_MESSAGE | BLK_TC_ACT(BLK_TC_NOTIFY))
2006-12-01 10:39:12 +01:00
2006-03-23 20:00:26 +01:00
# define BLK_IO_TRACE_MAGIC 0x65617400
# define BLK_IO_TRACE_VERSION 0x07
/*
* The trace itself
*/
struct blk_io_trace {
2008-09-26 10:58:02 +02:00
__u32 magic ; /* MAGIC << 8 | version */
__u32 sequence ; /* event number */
__u64 time ; /* in microseconds */
__u64 sector ; /* disk offset */
__u32 bytes ; /* transfer length */
__u32 action ; /* what happened */
__u32 pid ; /* who did it */
__u32 device ; /* device number */
__u32 cpu ; /* on what cpu did it happen */
__u16 error ; /* completion error */
__u16 pdu_len ; /* length of data after this trace */
2006-03-23 20:00:26 +01:00
} ;
/*
* The remap event
*/
struct blk_io_trace_remap {
2007-08-07 15:30:23 +02:00
__be32 device_from ;
2009-05-04 16:27:26 -04:00
__be32 device_to ;
__be64 sector_from ;
2006-03-23 20:00:26 +01:00
} ;
enum {
Blktrace_setup = 1 ,
Blktrace_running ,
Blktrace_stopped ,
} ;
2008-10-01 16:16:25 +02:00
# define BLKTRACE_BDEV_SIZE 32
2008-09-26 10:58:02 +02:00
/*
* User setup structure passed with BLKTRACESTART
*/
struct blk_user_trace_setup {
2008-10-01 16:16:25 +02:00
char name [ BLKTRACE_BDEV_SIZE ] ; /* output */
2008-09-26 10:58:02 +02:00
__u16 act_mask ; /* input */
__u32 buf_size ; /* input */
__u32 buf_nr ; /* input */
__u64 start_lba ;
__u64 end_lba ;
__u32 pid ;
} ;
# ifdef __KERNEL__
# if defined(CONFIG_BLK_DEV_IO_TRACE)
2009-01-26 15:00:56 -02:00
# include <linux/sysfs.h>
2006-03-23 20:00:26 +01:00
struct blk_trace {
int trace_state ;
struct rchan * rchan ;
2010-02-02 14:38:57 +09:00
unsigned long __percpu * sequence ;
unsigned char __percpu * msg_data ;
2006-03-23 20:00:26 +01:00
u16 act_mask ;
u64 start_lba ;
u64 end_lba ;
u32 pid ;
u32 dev ;
struct dentry * dir ;
struct dentry * dropped_file ;
2008-06-11 09:12:52 +02:00
struct dentry * msg_file ;
2006-03-23 20:00:26 +01:00
atomic_t dropped ;
} ;
extern int blk_trace_ioctl ( struct block_device * , unsigned , char __user * ) ;
2007-07-24 09:28:11 +02:00
extern void blk_trace_shutdown ( struct request_queue * ) ;
2009-04-14 13:58:56 +08:00
extern int do_blk_trace_setup ( struct request_queue * q , char * name ,
dev_t dev , struct block_device * bdev ,
struct blk_user_trace_setup * buts ) ;
2008-05-27 14:54:41 +02:00
extern void __trace_note_message ( struct blk_trace * , const char * fmt , . . . ) ;
2007-10-09 13:23:53 +02:00
2008-05-27 14:54:41 +02:00
/**
* blk_add_trace_msg - Add a ( simple ) message to the blktrace stream
* @ q : queue the io is for
* @ fmt : format to print message in
* args . . . Variable argument list for format
*
* Description :
* Records a ( simple ) message onto the blktrace stream .
*
* NOTE : BLK_TN_MAX_MSG characters are output at most .
* NOTE : Can not use ' static inline ' due to presence of var args . . .
*
* */
# define blk_add_trace_msg(q, fmt, ...) \
do { \
struct blk_trace * bt = ( q ) - > blk_trace ; \
if ( unlikely ( bt ) ) \
__trace_note_message ( bt , fmt , # # __VA_ARGS__ ) ; \
} while ( 0 )
2008-05-28 14:45:33 +02:00
# define BLK_TN_MAX_MSG 128
2006-03-23 20:00:26 +01:00
2008-10-30 08:34:33 +01:00
extern void blk_add_driver_data ( struct request_queue * q , struct request * rq ,
void * data , size_t len ) ;
2008-01-29 19:12:06 +01:00
extern int blk_trace_setup ( struct request_queue * q , char * name , dev_t dev ,
2009-04-14 13:58:56 +08:00
struct block_device * bdev ,
2008-01-11 10:09:43 +01:00
char __user * arg ) ;
2008-01-29 19:12:06 +01:00
extern int blk_trace_startstop ( struct request_queue * q , int start ) ;
extern int blk_trace_remove ( struct request_queue * q ) ;
2009-09-25 06:19:26 +02:00
extern void blk_trace_remove_sysfs ( struct device * dev ) ;
2009-04-14 14:00:05 +08:00
extern int blk_trace_init_sysfs ( struct device * dev ) ;
2008-01-11 10:09:43 +01:00
2009-01-26 15:00:56 -02:00
extern struct attribute_group blk_trace_attr_group ;
2006-03-23 20:00:26 +01:00
# else /* !CONFIG_BLK_DEV_IO_TRACE */
2009-04-14 13:58:56 +08:00
# define blk_trace_ioctl(bdev, cmd, arg) (-ENOTTY)
# define blk_trace_shutdown(q) do { } while (0)
# define do_blk_trace_setup(q, name, dev, bdev, buts) (-ENOTTY)
# define blk_add_driver_data(q, rq, data, len) do {} while (0)
# define blk_trace_setup(q, name, dev, bdev, arg) (-ENOTTY)
# define blk_trace_startstop(q, start) (-ENOTTY)
# define blk_trace_remove(q) (-ENOTTY)
# define blk_add_trace_msg(q, fmt, ...) do { } while (0)
2009-10-01 21:16:13 +02:00
# define blk_trace_remove_sysfs(dev) do { } while (0)
2009-04-14 14:00:05 +08:00
static inline int blk_trace_init_sysfs ( struct device * dev )
{
return 0 ;
}
2006-03-23 20:00:26 +01:00
# endif /* CONFIG_BLK_DEV_IO_TRACE */
2009-04-14 13:58:56 +08:00
2010-07-09 14:24:38 +10:00
# ifdef CONFIG_COMPAT
struct compat_blk_user_trace_setup {
char name [ 32 ] ;
u16 act_mask ;
u32 buf_size ;
u32 buf_nr ;
compat_u64 start_lba ;
compat_u64 end_lba ;
u32 pid ;
} ;
# define BLKTRACESETUP32 _IOWR(0x12, 115, struct compat_blk_user_trace_setup)
# endif
2009-06-10 10:06:24 +08:00
# if defined(CONFIG_EVENT_TRACING) && defined(CONFIG_BLOCK)
tracing/events: convert block trace points to TRACE_EVENT()
TRACE_EVENT is a more generic way to define tracepoints. Doing so adds
these new capabilities to this tracepoint:
- zero-copy and per-cpu splice() tracing
- binary tracing without printf overhead
- structured logging records exposed under /debug/tracing/events
- trace events embedded in function tracer output and other plugins
- user-defined, per tracepoint filter expressions
...
Cons:
- no dev_t info for the output of plug, unplug_timer and unplug_io events.
no dev_t info for getrq and sleeprq events if bio == NULL.
no dev_t info for rq_abort,...,rq_requeue events if rq->rq_disk == NULL.
This is mainly because we can't get the deivce from a request queue.
But this may change in the future.
- A packet command is converted to a string in TP_assign, not TP_print.
While blktrace do the convertion just before output.
Since pc requests should be rather rare, this is not a big issue.
- In blktrace, an event can have 2 different print formats, but a TRACE_EVENT
has a unique format, which means we have some unused data in a trace entry.
The overhead is minimized by using __dynamic_array() instead of __array().
I've benchmarked the ioctl blktrace vs the splice based TRACE_EVENT tracing:
dd dd + ioctl blktrace dd + TRACE_EVENT (splice)
1 7.36s, 42.7 MB/s 7.50s, 42.0 MB/s 7.41s, 42.5 MB/s
2 7.43s, 42.3 MB/s 7.48s, 42.1 MB/s 7.43s, 42.4 MB/s
3 7.38s, 42.6 MB/s 7.45s, 42.2 MB/s 7.41s, 42.5 MB/s
So the overhead of tracing is very small, and no regression when using
those trace events vs blktrace.
And the binary output of TRACE_EVENT is much smaller than blktrace:
# ls -l -h
-rw-r--r-- 1 root root 8.8M 06-09 13:24 sda.blktrace.0
-rw-r--r-- 1 root root 195K 06-09 13:24 sda.blktrace.1
-rw-r--r-- 1 root root 2.7M 06-09 13:25 trace_splice.out
Following are some comparisons between TRACE_EVENT and blktrace:
plug:
kjournald-480 [000] 303.084981: block_plug: [kjournald]
kjournald-480 [000] 303.084981: 8,0 P N [kjournald]
unplug_io:
kblockd/0-118 [000] 300.052973: block_unplug_io: [kblockd/0] 1
kblockd/0-118 [000] 300.052974: 8,0 U N [kblockd/0] 1
remap:
kjournald-480 [000] 303.085042: block_remap: 8,0 W 102736992 + 8 <- (8,8) 33384
kjournald-480 [000] 303.085043: 8,0 A W 102736992 + 8 <- (8,8) 33384
bio_backmerge:
kjournald-480 [000] 303.085086: block_bio_backmerge: 8,0 W 102737032 + 8 [kjournald]
kjournald-480 [000] 303.085086: 8,0 M W 102737032 + 8 [kjournald]
getrq:
kjournald-480 [000] 303.084974: block_getrq: 8,0 W 102736984 + 8 [kjournald]
kjournald-480 [000] 303.084975: 8,0 G W 102736984 + 8 [kjournald]
bash-2066 [001] 1072.953770: 8,0 G N [bash]
bash-2066 [001] 1072.953773: block_getrq: 0,0 N 0 + 0 [bash]
rq_complete:
konsole-2065 [001] 300.053184: block_rq_complete: 8,0 W () 103669040 + 16 [0]
konsole-2065 [001] 300.053191: 8,0 C W 103669040 + 16 [0]
ksoftirqd/1-7 [001] 1072.953811: 8,0 C N (5a 00 08 00 00 00 00 00 24 00) [0]
ksoftirqd/1-7 [001] 1072.953813: block_rq_complete: 0,0 N (5a 00 08 00 00 00 00 00 24 00) 0 + 0 [0]
rq_insert:
kjournald-480 [000] 303.084985: block_rq_insert: 8,0 W 0 () 102736984 + 8 [kjournald]
kjournald-480 [000] 303.084986: 8,0 I W 102736984 + 8 [kjournald]
Changelog from v2 -> v3:
- use the newly introduced __dynamic_array().
Changelog from v1 -> v2:
- use __string() instead of __array() to minimize the memory required
to store hex dump of rq->cmd().
- support large pc requests.
- add missing blk_fill_rwbs_rq() in block_rq_requeue TRACE_EVENT.
- some cleanups.
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
LKML-Reference: <4A2DF669.5070905@cn.fujitsu.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2009-06-09 13:43:05 +08:00
static inline int blk_cmd_buf_len ( struct request * rq )
{
2010-08-07 18:17:56 +02:00
return ( rq - > cmd_type = = REQ_TYPE_BLOCK_PC ) ? rq - > cmd_len * 3 : 1 ;
tracing/events: convert block trace points to TRACE_EVENT()
TRACE_EVENT is a more generic way to define tracepoints. Doing so adds
these new capabilities to this tracepoint:
- zero-copy and per-cpu splice() tracing
- binary tracing without printf overhead
- structured logging records exposed under /debug/tracing/events
- trace events embedded in function tracer output and other plugins
- user-defined, per tracepoint filter expressions
...
Cons:
- no dev_t info for the output of plug, unplug_timer and unplug_io events.
no dev_t info for getrq and sleeprq events if bio == NULL.
no dev_t info for rq_abort,...,rq_requeue events if rq->rq_disk == NULL.
This is mainly because we can't get the deivce from a request queue.
But this may change in the future.
- A packet command is converted to a string in TP_assign, not TP_print.
While blktrace do the convertion just before output.
Since pc requests should be rather rare, this is not a big issue.
- In blktrace, an event can have 2 different print formats, but a TRACE_EVENT
has a unique format, which means we have some unused data in a trace entry.
The overhead is minimized by using __dynamic_array() instead of __array().
I've benchmarked the ioctl blktrace vs the splice based TRACE_EVENT tracing:
dd dd + ioctl blktrace dd + TRACE_EVENT (splice)
1 7.36s, 42.7 MB/s 7.50s, 42.0 MB/s 7.41s, 42.5 MB/s
2 7.43s, 42.3 MB/s 7.48s, 42.1 MB/s 7.43s, 42.4 MB/s
3 7.38s, 42.6 MB/s 7.45s, 42.2 MB/s 7.41s, 42.5 MB/s
So the overhead of tracing is very small, and no regression when using
those trace events vs blktrace.
And the binary output of TRACE_EVENT is much smaller than blktrace:
# ls -l -h
-rw-r--r-- 1 root root 8.8M 06-09 13:24 sda.blktrace.0
-rw-r--r-- 1 root root 195K 06-09 13:24 sda.blktrace.1
-rw-r--r-- 1 root root 2.7M 06-09 13:25 trace_splice.out
Following are some comparisons between TRACE_EVENT and blktrace:
plug:
kjournald-480 [000] 303.084981: block_plug: [kjournald]
kjournald-480 [000] 303.084981: 8,0 P N [kjournald]
unplug_io:
kblockd/0-118 [000] 300.052973: block_unplug_io: [kblockd/0] 1
kblockd/0-118 [000] 300.052974: 8,0 U N [kblockd/0] 1
remap:
kjournald-480 [000] 303.085042: block_remap: 8,0 W 102736992 + 8 <- (8,8) 33384
kjournald-480 [000] 303.085043: 8,0 A W 102736992 + 8 <- (8,8) 33384
bio_backmerge:
kjournald-480 [000] 303.085086: block_bio_backmerge: 8,0 W 102737032 + 8 [kjournald]
kjournald-480 [000] 303.085086: 8,0 M W 102737032 + 8 [kjournald]
getrq:
kjournald-480 [000] 303.084974: block_getrq: 8,0 W 102736984 + 8 [kjournald]
kjournald-480 [000] 303.084975: 8,0 G W 102736984 + 8 [kjournald]
bash-2066 [001] 1072.953770: 8,0 G N [bash]
bash-2066 [001] 1072.953773: block_getrq: 0,0 N 0 + 0 [bash]
rq_complete:
konsole-2065 [001] 300.053184: block_rq_complete: 8,0 W () 103669040 + 16 [0]
konsole-2065 [001] 300.053191: 8,0 C W 103669040 + 16 [0]
ksoftirqd/1-7 [001] 1072.953811: 8,0 C N (5a 00 08 00 00 00 00 00 24 00) [0]
ksoftirqd/1-7 [001] 1072.953813: block_rq_complete: 0,0 N (5a 00 08 00 00 00 00 00 24 00) 0 + 0 [0]
rq_insert:
kjournald-480 [000] 303.084985: block_rq_insert: 8,0 W 0 () 102736984 + 8 [kjournald]
kjournald-480 [000] 303.084986: 8,0 I W 102736984 + 8 [kjournald]
Changelog from v2 -> v3:
- use the newly introduced __dynamic_array().
Changelog from v1 -> v2:
- use __string() instead of __array() to minimize the memory required
to store hex dump of rq->cmd().
- support large pc requests.
- add missing blk_fill_rwbs_rq() in block_rq_requeue TRACE_EVENT.
- some cleanups.
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
LKML-Reference: <4A2DF669.5070905@cn.fujitsu.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2009-06-09 13:43:05 +08:00
}
extern void blk_dump_cmd ( char * buf , struct request * rq ) ;
extern void blk_fill_rwbs ( char * rwbs , u32 rw , int bytes ) ;
extern void blk_fill_rwbs_rq ( char * rwbs , struct request * rq ) ;
2009-06-10 10:06:24 +08:00
# endif /* CONFIG_EVENT_TRACING && CONFIG_BLOCK */
tracing/events: convert block trace points to TRACE_EVENT()
TRACE_EVENT is a more generic way to define tracepoints. Doing so adds
these new capabilities to this tracepoint:
- zero-copy and per-cpu splice() tracing
- binary tracing without printf overhead
- structured logging records exposed under /debug/tracing/events
- trace events embedded in function tracer output and other plugins
- user-defined, per tracepoint filter expressions
...
Cons:
- no dev_t info for the output of plug, unplug_timer and unplug_io events.
no dev_t info for getrq and sleeprq events if bio == NULL.
no dev_t info for rq_abort,...,rq_requeue events if rq->rq_disk == NULL.
This is mainly because we can't get the deivce from a request queue.
But this may change in the future.
- A packet command is converted to a string in TP_assign, not TP_print.
While blktrace do the convertion just before output.
Since pc requests should be rather rare, this is not a big issue.
- In blktrace, an event can have 2 different print formats, but a TRACE_EVENT
has a unique format, which means we have some unused data in a trace entry.
The overhead is minimized by using __dynamic_array() instead of __array().
I've benchmarked the ioctl blktrace vs the splice based TRACE_EVENT tracing:
dd dd + ioctl blktrace dd + TRACE_EVENT (splice)
1 7.36s, 42.7 MB/s 7.50s, 42.0 MB/s 7.41s, 42.5 MB/s
2 7.43s, 42.3 MB/s 7.48s, 42.1 MB/s 7.43s, 42.4 MB/s
3 7.38s, 42.6 MB/s 7.45s, 42.2 MB/s 7.41s, 42.5 MB/s
So the overhead of tracing is very small, and no regression when using
those trace events vs blktrace.
And the binary output of TRACE_EVENT is much smaller than blktrace:
# ls -l -h
-rw-r--r-- 1 root root 8.8M 06-09 13:24 sda.blktrace.0
-rw-r--r-- 1 root root 195K 06-09 13:24 sda.blktrace.1
-rw-r--r-- 1 root root 2.7M 06-09 13:25 trace_splice.out
Following are some comparisons between TRACE_EVENT and blktrace:
plug:
kjournald-480 [000] 303.084981: block_plug: [kjournald]
kjournald-480 [000] 303.084981: 8,0 P N [kjournald]
unplug_io:
kblockd/0-118 [000] 300.052973: block_unplug_io: [kblockd/0] 1
kblockd/0-118 [000] 300.052974: 8,0 U N [kblockd/0] 1
remap:
kjournald-480 [000] 303.085042: block_remap: 8,0 W 102736992 + 8 <- (8,8) 33384
kjournald-480 [000] 303.085043: 8,0 A W 102736992 + 8 <- (8,8) 33384
bio_backmerge:
kjournald-480 [000] 303.085086: block_bio_backmerge: 8,0 W 102737032 + 8 [kjournald]
kjournald-480 [000] 303.085086: 8,0 M W 102737032 + 8 [kjournald]
getrq:
kjournald-480 [000] 303.084974: block_getrq: 8,0 W 102736984 + 8 [kjournald]
kjournald-480 [000] 303.084975: 8,0 G W 102736984 + 8 [kjournald]
bash-2066 [001] 1072.953770: 8,0 G N [bash]
bash-2066 [001] 1072.953773: block_getrq: 0,0 N 0 + 0 [bash]
rq_complete:
konsole-2065 [001] 300.053184: block_rq_complete: 8,0 W () 103669040 + 16 [0]
konsole-2065 [001] 300.053191: 8,0 C W 103669040 + 16 [0]
ksoftirqd/1-7 [001] 1072.953811: 8,0 C N (5a 00 08 00 00 00 00 00 24 00) [0]
ksoftirqd/1-7 [001] 1072.953813: block_rq_complete: 0,0 N (5a 00 08 00 00 00 00 00 24 00) 0 + 0 [0]
rq_insert:
kjournald-480 [000] 303.084985: block_rq_insert: 8,0 W 0 () 102736984 + 8 [kjournald]
kjournald-480 [000] 303.084986: 8,0 I W 102736984 + 8 [kjournald]
Changelog from v2 -> v3:
- use the newly introduced __dynamic_array().
Changelog from v1 -> v2:
- use __string() instead of __array() to minimize the memory required
to store hex dump of rq->cmd().
- support large pc requests.
- add missing blk_fill_rwbs_rq() in block_rq_requeue TRACE_EVENT.
- some cleanups.
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
LKML-Reference: <4A2DF669.5070905@cn.fujitsu.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2009-06-09 13:43:05 +08:00
2007-10-09 13:23:53 +02:00
# endif /* __KERNEL__ */
2006-03-23 20:00:26 +01:00
# endif