[KPhotoAlbum] KPhotoAlbum crashes while generating thumbnails during start

[Miika Turkia]

On Wed, May 1, 2013 at 12:26 PM, Manfred Usselmann <
usselmann.m at icg-online.de> wrote:

> **
>
> Am 01.05.2013 06:31, schrieb Miika Turkia:
>
>  On Tue, Apr 30, 2013 at 11:41 PM, Manfred Usselmann <
> usselmann.m at icg-online.de> wrote:
>
>>    * KPhotoAlbum is insanely searching for icon "file-broken" (png, svg
>> and possibly some other versions of it). At least on my system this occurs
>> and renders KPA useless. I did not find any package providing such an icon,
>> but a quick hack to just create an empty file with such name gets KPA
>> functioning properly (sudo touch
>> /usr/share/icons/oxygen/8x8/emblems/file-broken.png). Of course I cannot
>> test with Manfred's images, but at least the empty index.xml file loads.
>>   - KPA should be fixed to handle the case when broken image icon is not
>> found
>>
>>   For now I'm using the workaround to create file-broken.png, but it
>> seems there is already a fix in the development version. Thanks...
>> But something is still wrong even when /usr/share/icons/oxygen/8x8/emblems/file-broken.png
>> exists, KPA is still very slow opening the detail view and switching
>> between pictures and also going back to Home. Maybe it's looking for
>> different icons besides file-broken.png? Or I need to use a different
>> folder than the one mentioned above?
>>
>
> You could run strace to see what is so slow. Run the KPA from a terminal
> with the command strace in front:
> strace kphotoalbum
>
> It will print out a lot of information about system calls that are called
> and slows everything down somewhat. When you do action that takes too long,
> see what is being written on the terminal. If there is e.g. plenty of
> attempts to open some image file from different directories, that is
> probably causing the slow down.
>
>
>   Thanks for mentioning strace, I didn't know this utility. During the
> delays the following is displayed:
>
> poll([{fd=6, events=POLLIN|POLLOUT}], 1, 4294967295) = 1 ([{fd=6,
> revents=POLLOUT}])
> writev(6, [{"(\0\4\0\217\0\0\0005\0\340\7\222\0\264\4", 16}, {NULL, 0},
> {"", 0}], 3) = 16
> poll([{fd=6, events=POLLIN}], 1, 4294967295) = 1 ([{fd=6, revents=POLLIN}])
> recvfrom(6,
> "\1\1\254\r\0\0\0\0\0\0\0\0s\0#\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0", 4096,
> 0, NULL, NULL) = 32
> recvfrom(6, 0x15a1614, 4096, 0, 0, 0) = -1 EAGAIN (Resource temporarily
> unavailable)
>
The recvfrom sounds suspicious and there are quite many of those in full
strace log. The fd 6 is used to communicate with X server when it returns
those Resource unavailables:

socket(PF_FILE, SOCK_STREAM|SOCK_CLOEXEC, 0) = 6
connect(6, {sa_family=AF_FILE, path=@"/tmp/.X11-unix/X0"}, 20) = 0
getpeername(6, {sa_family=AF_FILE, path=@"/tmp/.X11-unix/X0"...}, [20]) = 0

I just have no idea whether it is normal or not and what to do with it and
if this is causing the slowness. Anyone: any ideas?

miika
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kphotoalbum/attachments/20130502/c26c45cf/attachment.htm>