On 4/18/18, Nick <haveagoodtime2...@gmail.com> wrote: > Hi, > I have one process using sqlite with “pragma mmap_size=30M”. The operations > of my process is insert-select-insert-select. So the PSS(private clean) will > increase along with the growing of the db which is treated as memory leak by > Mem-Analysor tool. > > I guess calling sqlite3_close() or pragma mmap_size=0 after querys may free > the PSS but that is not a good way for my process. > > So I am wondering is there any other way to free the PSS? As I find > unixUnmapfile() will be called when nFetchOut back to 0 but I do not know > what operation may trigger that.
I think sqlite3_close() or pragma mmap_size are the only ways to do that. Note that the PSS increase is not actually a memory leak. That it is reported as a leak is a issue in Mem-Analyzor. -- D. Richard Hipp d...@sqlite.org _______________________________________________ sqlite-users mailing list sqlite-users@mailinglists.sqlite.org http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
