https://bugs.kde.org/show_bug.cgi?id=382152
--- Comment #18 from Benoît Robin <benoit.ro...@hoshikaze.net> --- Created attachment 107245 --> https://bugs.kde.org/attachment.cgi?id=107245&action=edit Report of Profiling + SQL Hi Stéphane, I doubled up with the previous trace and the file, which is copied from the console, contains both trace and SQL. This is probably redundant. >From your diagnostic, I guess the statement that causes the issue is the following : ##executeSqliteOrder:SELECT EXISTS(SELECT 1 FROM payee WHERE EXISTS (SELECT 1 FROM payee p2 WHERE p2.id<>payee.id AND upper(p2.t_name)=upper(payee.t_name) AND p2.t_name<>payee.t_name)) TIME=27582 ms, (with fetch):27582 ms It was executed twice, which matches with my observing two CPU max-outs, one at startup and one when I updated the operation's category. Hope that could help you improve that operation. It seems a quite simple brute-force statement on all the base and I'm not SQL-wise enough to know if such a seemingly simple request can be improved ^-^ Benoît 'Mutos' Robin -- You are receiving this mail because: You are watching all bug changes.
