[Digikam-devel] [Bug 185265] digikam constantly using some CPU in a poll loop due a timeout

[Roman Fietze]

https://bugs.kde.org/show_bug.cgi?id=185265





--- Comment #2 from Roman Fietze <rfkd fietze-home de>  2009-02-22 20:34:02 ---
(In reply to comment #1)

> Can you identify which process use this cpu ?

Seems to be a thread of the originally started process. Strace shows e.g the
following lines:

15685 <first line in strace output, the original process>
...
15685 20:17:22.664063 clone(child_stack=0xade46474,
flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|CLONE_SYSVSEM|CLONE_SETTLS|CLONE_PARENT
_SETTID|CLONE_CHILD_CLEARTID, parent_tidptr=0xade46bd8, {entry_number:6,
base_addr:0xade46b90, limit:1048575, seg_32bit:1, contents:0, read_exec_only:0,
limit_in_pages:1, seg_not_present:0, useable:1}, child_tidptr=0xade46bd8) =
15695
...
15685 20:25:58.280403 clock_gettime(CLOCK_MONOTONIC,  <unfinished ...>
15695 20:25:58.280535 <... futex resumed> ) = 0
15685 20:25:58.280649 <... clock_gettime resumed> {23161, 638303372}) = 0
15695 20:25:58.280870 clock_gettime(CLOCK_MONOTONIC,  <unfinished ...>
15685 20:25:58.280994 clock_gettime(CLOCK_MONOTONIC,  <unfinished ...>
15695 20:25:58.281125 <... clock_gettime resumed> {23161, 638786876}) = 0
15685 20:25:58.281264 <... clock_gettime resumed> {23161, 638893226}) = 0
15695 20:25:58.281464 clock_gettime(CLOCK_MONOTONIC,  <unfinished ...>
15685 20:25:58.281589 clock_gettime(CLOCK_MONOTONIC,  <unfinished ...>
15695 20:25:58.281720 <... clock_gettime resumed> {23161, 639382182}) = 0
15685 20:25:58.281857 <... clock_gettime resumed> {23161, 639488045}) = 0
15695 20:25:58.282046 clock_gettime(CLOCK_MONOTONIC,  <unfinished ...>
15685 20:25:58.282169 clock_gettime(CLOCK_MONOTONIC,  <unfinished ...>
15695 20:25:58.282299 <... clock_gettime resumed> {23161, 639963573}) = 0
15685 20:25:58.282437 <... clock_gettime resumed> {23161, 640067675}) = 0
15695 20:25:58.282616 gettimeofday( <unfinished ...>
15685 20:25:58.282738 read(7,  <unfinished ...>
15695 20:25:58.282873 <... gettimeofday resumed> {1235330758, 282723}, NULL) =
0
15685 20:25:58.283012 <... read resumed> 0x839c124, 4096) = -1 EAGAIN (Resource
temporari
ly unavailable)
15695 20:25:58.283182 clock_gettime(CLOCK_REALTIME,  <unfinished ...>
15685 20:25:58.283337 poll([{fd=3, events=POLLIN}, {fd=8, events=POLLIN},
{fd=7, events=P
OLLIN}, {fd=9, events=POLLIN}, {fd=10, events=POLLIN}, {fd=11, events=POLLIN},
{fd=35, ev
ents=POLLIN}, {fd=33, events=POLLIN}, {fd=34, events=POLLIN}, {fd=32,
events=POLLIN}, {fd
=39, events=POLLIN}], 11, 191 <unfinished ...>
15695 20:25:58.283692 <... clock_gettime resumed> {1235330758, 283317687}) = 0
...
... <forever>

PID 15685 is the originally started process with the executable digikam. Top
also show PID 15685 as the process using the CPU, I do not (yet) know which one
of the two threads use that CPU, 15685  or 15695.

> Also, we need to know in which conditions/action in digiKam the cpu is
> consumed.

The system is completely idle, esp. I waited quite some while after I accessed
any file, even if it was just to read it (a "sitecopy --update ..."). Digikam
is showing a completely empty directory (my "incoming" directory AAA-Temp)
while the strace was taken.

-- 
Configure bugmail: https://bugs.kde.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.