delayacct: improve the average delay precision of getdelay tool to microsecond
authorWang Yong <wang.yong12@zte.com.cn>
Mon, 13 Feb 2023 06:08:08 +0000 (14:08 +0800)
committerAndrew Morton <akpm@linux-foundation.org>
Sat, 8 Apr 2023 20:45:36 +0000 (13:45 -0700)
commiteca7de7cdc382eb6e0d344c07b1449ed75f5b435
tree3ba1c86dc613b4a5647c82a00d2c73403c494850
parentaa318c48808c0aa73216bd94c54c4553d3663add
delayacct: improve the average delay precision of getdelay tool to microsecond

Improve the average delay precision of getdelay tool to microsecond.  When
using the getdelay tool, it is sometimes found that the average delay
except CPU is not 0, but display is 0, because the precison is too low.
For example, see delay average of SWAP below when using ZRAM.

print delayacct stats ON
PID 32915
CPU             count     real total  virtual total    delay total  delay average
               339202     2793871936     9233585504        7951112          0.000ms
IO              count    delay total  delay average
                   41      419296904             10ms
SWAP            count    delay total  delay average
               242589     1045792384              0ms

This wrong display is misleading, so improve the millisecond precision of
the average delay to microsecond just like CPU.  Then user would get more
accurate information of delay time.

Link: https://lkml.kernel.org/r/202302131408087983857@zte.com.cn
Signed-off-by: Wang Yong <wang.yong12@zte.com.cn>
Reviewed-by: Yang Yang <yang.yang29@zte.com.cn>
Cc: David Hildenbrand <david@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Documentation/accounting/delay-accounting.rst
Documentation/translations/zh_CN/accounting/delay-accounting.rst
tools/accounting/getdelays.c