On Fri, Nov 28, 2014 at 5:50 PM, Yaroslav Halchenko <deb...@onerussian.com> wrote: > Package: libsqlite3-0 > Version: 3.7.13-1+deb7u1 > Severity: serious > > Originally detected I believe on sid installation, but forgot to capture the > version, I will try to replicate/report later. But very consistent with > wheezy > (from which I am reporting now). May you give me some details how it happened in Sid?
> Triggered by the backport fail2ban 0.9.1-1~nd70+1 (available from > http://neuro.debian.net/debian-devel/ wheezy/main amd64 Packages apt repo) it > gets to [...] The problem is, I don't see the segfault in the mentioned gdb output. > unfortunately we haven't logged the sql queries so not sure on exact one, but > I > think it was this one, which if executed from the shell seems to not cause the > segfault... > > n {1..100}; do sqlite3 -list fail2ban.sqlite3 'SELECT ip, timeofban, data > FROM bans WHERE 1 AND jail="sshd" AND ip="111.74.239.35" ORDER BY ip, > timeofban' >/dev/null && echo success; done Then how often do you get segfault? Do you have any additional information if it happens in a given daytime, when there are several bots try to get into your system or anything else? > Please help me to troubleshoot this one if more information is necessary > to point the issue I'm the SQLite3 maintainer and not the fail2ban one. But please note the changelog the version you use[1]: - 0.9 series is quite a big leap in development, especially since 0.8.6 which made it to previous Debian stable wheezy. Please consult upstream ChangeLog about changes Did you check it, reviewed your configuration? Does a segfault happen in other applications that link to SQLite3? Regards, Laszlo/GCS [1] https://packages.qa.debian.org/f/fail2ban/news/20141028T041852Z.html -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org