В Птн, 04/08/2006 в 00:06 -0300, Ricardo Markiewicz пишет: > I'm trying to reproduce this bug without lucky, can anyone confirm if > this still happend? > > I saw that 'ps uax' show an increment of memory usage, but is nothing > compared to the original report : > gazer 24912 84.7 2.7 42712 21596 pts/0 Sl+ 00:03 0:06 evince > kdvi.pdf > (after 3 reloads) > gazer 24912 71.6 2.9 42336 23068 pts/0 Sl+ 00:03 1:07 evince > kdvi.pdf > > valgrind report many "XXX bytes in Y blocks are definitely lost in loss > record", mostly in Gtk+ and GnomeProgram, but I need to check that > before report those problems as bugs. > > The only thing I found is a minor memory leak in ev-window.c (patch > attached). > > Thanks Hello Ricardo First of all, nice patch. I've extended it a bit (there was a lot of places where we didn't free GValue) and committed. About memory leak, I also tried to reproduce it and failed. I think there is no memory leak but rather extensive memory usage (you know that famous bug about zoom size). I don't maximize evince window due to that, 5 pages in cache takes all my memory. Of course it's not as big as gaim (requires 150 Mb to show 6 chats) or firefox (requires 900 Mb to render 12 Mb images) but very big anyhow. Probably we are doing something wrong there, we should investigate a bit. I think that allocation/deallocation of large chunks like pages makes things work slow.
Attachment:
signature.asc
Description: =?koi8-r?Q?=FC=D4=C1?= =?koi8-r?Q?_=DE=C1=D3=D4=D8?= =?koi8-r?Q?_=D3=CF=CF=C2=DD=C5=CE=C9=D1?= =?koi8-r?Q?_=D0=CF=C4=D0=C9=D3=C1=CE=C1?= =?koi8-r?Q?_=C3=C9=C6=D2=CF=D7=CF=CA?= =?koi8-r?Q?_=D0=CF=C4=D0=C9=D3=D8=C0?=