August 2018 - dovecot

Occasional crash in db-auth.c (Valgrind: Invalid read of size 4 et al.), Dovecot 2.2.27+
by Jonas Wielicki 14 Sep '18

14 Sep '18

Dear list, We are encountering troubles with dovecot using LDAP userdbs on Debian stretch (but if I’m reading valgrind correctly, we can reproduce this with vanilla dovecot master). Minimal reproducer below. While testing an upgrade to Debian stretch (dovecot-core=1:2.2.27-3+deb9u1), auth-worker has stopped working. We are using two LDAP user databases; one which is iterable, and one which is not (for reasons; I think this is not relevant; if it is, we’re happy to elaborate). The issue seems to be present in any LDAP userdb iteration handling, but it only seems to cause a crash in certain conditions which seem to be reproducibly caused by our setup for some reason. We’re seeing one of two errors. Variant A: Jan 27 13:40:05 up2 dovecot: auth-worker: Error: *** Error in `dovecot/auth': free(): corrupted unsorted chunks: 0x000056553f0fcfa0 *** Jan 27 13:40:05 up2 dovecot: auth-worker: Error: ======= Backtrace: ========= Jan 27 13:40:05 up2 dovecot: auth-worker: Error: /lib/x86_64-linux-gnu/ libc.so.6(+0x70bcb)[0x7f78426adbcb] Jan 27 13:40:05 up2 dovecot: auth-worker: Error: /lib/x86_64-linux-gnu/ libc.so.6(+0x76f96)[0x7f78426b3f96] Jan 27 13:40:05 up2 dovecot: auth-worker: Error: /lib/x86_64-linux-gnu/ libc.so.6(+0x777de)[0x7f78426b47de] Jan 27 13:40:05 up2 dovecot: auth-worker: Error: /usr/lib/dovecot/modules/ auth/libauthdb_ldap.so(+0x5bde)[0x7f7842008bde] Jan 27 13:40:05 up2 dovecot: auth-worker: Error: /usr/lib/dovecot/ libdovecot.so.0(io_loop_call_io+0x52)[0x7f78430cfdd2] Jan 27 13:40:05 up2 dovecot: auth-worker: Error: /usr/lib/dovecot/ libdovecot.so.0(io_loop_handler_run_internal+0x109)[0x7f78430d1409] Jan 27 13:40:05 up2 dovecot: auth-worker: Error: /usr/lib/dovecot/ libdovecot.so.0(io_loop_handler_run+0x3c)[0x7f78430cfe6c] Jan 27 13:40:05 up2 dovecot: auth-worker: Error: /usr/lib/dovecot/ libdovecot.so.0(io_loop_run+0x38)[0x7f78430d0018] Jan 27 13:40:05 up2 dovecot: auth-worker: Error: /usr/lib/dovecot/ libdovecot.so.0(master_service_run+0x13)[0x7f7843057e93] Jan 27 13:40:05 up2 dovecot: auth-worker: Error: dovecot/auth(main+0x398) [0x56553da45f98] Jan 27 13:40:05 up2 dovecot: auth-worker: Error: /lib/x86_64-linux-gnu/ libc.so.6(__libc_start_main+0xf1)[0x7f784265d2b1] Jan 27 13:40:05 up2 dovecot: auth-worker: Error: dovecot/auth(_start+0x2a) [0x56553da461aa] Variant B: Jan 27 13:06:56 up2 dovecot: auth-worker(27495): Panic: file db-ldap.c: line 840 (db_ldap_result_unref): assertion failed: (res->refcount > 0) Jan 27 13:06:56 up2 dovecot: auth-worker(27495): Error: Raw backtrace: /usr/ lib/dovecot/libdovecot.so.0(+0x95272) [0x7f027a4fc272] -> /usr/lib/dovecot/ libdovecot.so.0(+0x9536d) [0x7f027a4fc36d] -> /usr/lib/dovecot/libdovecot.so. 0(i_fatal+0) [0x7f027a492951] -> /usr/lib/dovecot/modules/auth/ libauthdb_ldap.so(+0x3868) [0x7f0279447868] -> /usr/lib/dovecot/modules/auth/ libauthdb_ldap.so(+0x5d7c) [0x7f0279449d7c] -> /usr/lib/dovecot/libdovecot.so. 0(io_loop_call_io+0x52) [0x7f027a510dd2] -> /usr/lib/dovecot/libdovecot.so. 0(io_loop_handler_run_internal+0x109) [0x7f027a512409] -> /usr/lib/dovecot/ libdovecot.so.0(io_loop_handler_run+0x3c) [0x7f027a510e6c] -> /usr/lib/ dovecot/libdovecot.so.0(io_loop_run+0x38) [0x7f027a511018] -> /usr/lib/ dovecot/libdovecot.so.0(master_service_run+0x13) [0x7f027a498e93] -> dovecot/ auth(main+0x398) [0x55b981cd2f98] -> /lib/x86_64-linux-gnu/libc.so. 6(__libc_start_main+0xf1) [0x7f0279a9e2b1] -> dovecot/auth(_start+0x2a) [0x55b981cd31aa] When running auth-worker within valgrind, we get more information (valgrind extract from mail.log attached for two versions of dovecot). We have attached mail.log contents for the respective valgrind runs. This seems to affect at least 2.2.27 onwards. We can reproduce the crash with our production data 100% of the time. The valgrind error is present even with very minimal setups (see below). Unfortunately, I haven’t been able so far to create an LDAP database which produces the crash as reliable as our production database does (which we cannot share for obvious reasons). Minimal Reproducer of Valgrind error: - Install dovecot 2.2.27 from debian stretch, or compile from master. - Install valgrind - In conf.d/10-auth.conf, disable auth-system.conf.ext - In conf.d/10-auth.conf, enable auth-ldap.conf.ext - Set contents of dovecot-ldap.conf.ext to: hosts = localhost base = dc=nodomain - In conf.d/10-master.conf, in section "service auth-worker", set executable = /usr/bin/valgrind /usr/lib/dovecot/auth -w (path may differ on your system) - Install an LDAP server with a database for dc=nodomain; this is trivial to do with debian: Simply install slapd and systemctl start slapd. - Start dovecot - Run doveadm user '*' - You should find the error in the mail.log. I hope this is somehow useful to fix our crash issue. We’ll be happy to provide more information as needed. kind regards, Jonas

4 4

Status of SMTPUTF8?
by Noah Tilton 14 Sep '18

14 Sep '18

I was wondering whether there is a roadmap for adding SMTPUTF8 support to Dovecot? My delivery pattern is Postfix -> Dovecot LMTP and it is choking on utf8 messages. I might be able to volunteer some of my time as a developer. Another thread about this seemed to go unanswered: http://dovecot.org/list/dovecot/2016-September/105474.html http://unix.stackexchange.com/questions/320091/configure-postfix-and-doveco… -Noah

4 3

Letsencrypt certificate for repo.dovecot.org expired May 14th..
by B. Reino 12 Sep '18

12 Sep '18

Dear all, Just in case you've missed it, the certificate for repo.dovecot.org just expired yesterday. This causes errors in e.g. apt-get update. Thanks in advance for fixing it, -- B. Reino

2 6

Re: dying on osx
by Mike Makuch 04 Sep '18

04 Sep '18

I found and tried this work around on the man page: https://www.unix.com/man-page/all/5/ngroups_max/ but I still get the same "Too many extra groups" error even when I start dovecot with the above program to limit the # of groups. I suspect that dovecot is adding a number of groups when it starts up. I've hacked a work around to get it working for me on my laptop: diff --git a/src/lib/restrict-access.c b/src/lib/restrict-access.c @@ -224,7 +224,12 @@ static void fix_groups_list(const struct restrict_access_settings *set, - if (setgroups(gid_count, gid_list) < 0) { + if (setgroups(gid_count > NGROUPS_MAX ? 16 : gid_count, gid_list) < 0) { if (errno == EINVAL) { i_fatal("setgroups(%s) failed: Too many extra groups", set->extra_groups == NULL ? "" : and this works. I'm not sure what the right solution is for a PR. Any suggestions? Thanks Mike On 8/10/18 11:04, Aki Tuomi wrote: > Is the user member of mail group? > > > > --- > Aki Tuomi > Dovecot oy > > -------- Original message -------- > From: Mike Makuch <1mikemakuch(a)gmail.com> > Date: 10/08/2018 19:02 (GMT+02:00) > To: Aki Tuomi <aki.tuomi(a)dovecot.fi> > Cc: dovecot(a)dovecot.org > Subject: Re: dying on osx > > Maybe an old problem that has resurfaced??? > > https://bugzilla.samba.org/show_bug.cgi?id=8773 > > Mike > > > > On 8/10/18 10:54, Aki Tuomi wrote: >> I have to see if this is reproducible outside mac. >> >> >> >> --- >> Aki Tuomi >> Dovecot oy >> >> -------- Original message -------- >> From: Mike Makuch <1mikemakuch(a)gmail.com> >> Date: 10/08/2018 18:46 (GMT+02:00) >> To: Aki Tuomi <aki.tuomi(a)dovecot.fi> >> Subject: Re: dying on osx >> >> I did find that page and tried a few things there. My config has all of >> the settings there except 3: >> >> mail_access_groups = mail >> mbox_read_locks = fcntl >> mbox_write_locks = fcntl >> >> I add them to my config, restart and get the same abort >> >> 20180810-103849 imap(mkm)<79213><MkA5ihZzmdh/AAAB>: Fatal: >> setgroups(mail,505) failed: Too many extra groups >> >> I've tried numerous other settings as well. >> >> Thanks for any further advice >> >> Mike >> >> >> On 8/10/18 08:45, Aki Tuomi wrote: >> > Can you try this config and report back? >> > >> > >> https://superuser.com/questions/957272/dotlock-permissions-problems-with-do… >> > >> > Aki >> > >> >> On 10 August 2018 at 16:39 Mike Makuch <1mikemakuch(a)gmail.com> wrote: >> >> >> >> >> >> OSX 10.13.6 High Sierra, dovecot 2.3.2.1 >> >> >> >> dovecot starts up and runs but dies as soon as my mail client makes a >> >> request with log and config below. >> >> >> >> And advice appreciated. >> >> >> >> Thanks >> >> >> >> Mike >> >> >> >> >> >> >> >> 20180810-083730 auth: Debug: auth client connected (pid=77432) >> >> >> >> 20180810-083730 auth: Debug: client in: AUTH 1 PLAIN >> >> service=imap secured session=xo1p2BRzZNd/AAAB lip=127.0.0.1 >> >> rip=127.0.0.1 lport=143 rport=55140 >> >> 20180810-083730 auth: Debug: client passdb out: CONT 1 >> >> 20180810-083730 auth: Debug: client in: CONT 1 AG1rbQBta20xMjM= >> >> (previous base64 data may contain sensitive data) >> >> 20180810-083730 auth: Debug: static(mkm,127.0.0.1,<xo1p2BRzZNd/AAAB>): >> >> lookup >> >> 20180810-083730 auth: Debug: client passdb out: OK 1 user=mkm >> >> host=localhost nopasswd=y >> >> 20180810-083730 auth: Debug: master in: REQUEST 4201906177 77432 >> >> 1 b8126b4b71be2959fc7716888eccc566 session_pid=77433 >> >> request_auth_token >> >> 20180810-083730 auth-worker(77426): Debug: >> >> passwd(mkm,127.0.0.1,<xo1p2BRzZNd/AAAB>): lookup >> >> 20180810-083730 auth: Debug: master userdb out: USER 4201906177 >> >> mkm system_groups_user=mkm uid=503 gid=20 home=/Users/mkm >> >> auth_token=4d2bb44168df3d63e4e1bb352e59de632bc7da49 >> >> 20180810-083730 imap-login: Info: Login: user=<mkm>, method=PLAIN, >> >> rip=127.0.0.1, lip=127.0.0.1, mpid=77433, secured, >> >> session=<xo1p2BRzZNd/AAAB> >> >> 20180810-083730 imap(mkm)<77433><xo1p2BRzZNd/AAAB>: Fatal: >> >> setgroups(505) failed: Too many extra groups >> >> >> >> # doveconf -n >> >> # 2.3.2.1 (0719df592): /usr/local/etc/dovecot/dovecot.conf >> >> # OS: Darwin 17.7.0 x86_64 >> >> # Hostname: pine >> >> auth_debug = yes >> >> auth_debug_passwords = yes >> >> default_internal_user = _dovecot >> >> default_login_user = _dovenull >> >> disable_plaintext_auth = no >> >> listen = 127.0.0.1 >> >> log_path = /var/log/dovecot >> >> log_timestamp = "%Y%m%d-%H%M%S " >> >> mail_debug = yes >> >> mail_gid = staff >> >> mail_location = mbox:~/Mail:INBOX=/var/mail/%u >> >> mail_privileged_group = mail >> >> mail_uid = _dovecot >> >> passdb { >> >> args = password=*** host=localhost nopasswd=y >> >> driver = static >> >> } >> >> passdb { >> >> args = login >> >> driver = pam >> >> } >> >> protocols = imap >> >> service auth { >> >> user = root >> >> } >> >> service imap-login { >> >> inet_listener imap { >> >> address = * >> >> port = 143 >> >> } >> >> } >> >> ssl = no >> >> userdb { >> >> driver = passwd >> >> } >> >> >> >> >> >

2 1

Does domain quota work and how?
by Kai Schaetzl 01 Sep '18

01 Sep '18

Hello, I'm in the process of moving users from an old dovecot setup with system users to a modern virtual user setup. I can't find good documentation on domain quota. It is mentioned in the docs as quota2 = dict:domain:%d:proxy::quota_domain but not much more. Searching the mailing list or Google doesn't reveal much more info. Searching the German mailing list at jpberlin finds mostly information where they say that it doesn't work. So, does it or does it not? And if it works, does it work like I'm used to? e.g. with system users and aquota the filesystem quota kicks in whenever either the user or the domain quota is reached. That makes sense. It prohibits that single users can "fill" the whole domain (group) quota. I found one particular example where they say the dict quota works different, e.g. it's an either/or. They give this example for the user_query (the relevant part): CONCAT('*:bytes=', \ IF(mailbox.quota = 0, domain.maxquota*1024000, mailbox.quota)) \ as quota_rule So, this works as a default domain quota that is overridden on a user by user basis. But this means that a user quota that is set cannot honor the domain quota. So, in theory all users together can reach a maximum quota that is way over the domain quota. Not desirable. Concluding from the example in 90-quota.conf I want to set up this user_query (relevant part): concat('*:bytes=', m.quota) AS quota_rule, concat('*:bytes=', (s.quota*1024*1024)) AS quota2_rule (with m and s being the two tables for user and domain quota and s holding the domain quota in MB) Does this work together with this quota plugin setup and would provide both a user and domain quota? quota = dict:user::proxy::quota quota2 = dict:domain:%d:proxy::quota_domain quota_rule = *:storage=100M # default quota2_rule = *:storage=1000M # default I would like to know if this is supposed to work (and how) before I start testing in vain. One thing that I can confirm from several statements on the mailing lists is that the recalc option of doveadm doesn't work correctly. A doveadm quota recalc -A sets the used domain quota to the user used quota that it handles last for the domain (and a recalc -u sets the whole domain used quota to the one of that user). Which sets the domain quota completely wrong. I tried iterating thru the various users but got the same result as with -A. So, this is clearly broken, at least from my point of view. I'm using dovecot 2.3 from this repo: deb https://repo.dovecot.org/ce-2.3-latest/ubuntu/xenial xenial main I installed it because of the native blowfish encryption support. So far, it seems to be running great. Except for this particular aspect. Thanks! Kai

2 5

Re: Dovecot User Listing Error - getpwent() failed: Invalid Argument
by Michael Marley 31 Aug '18

31 Aug '18

I think this might be caused by glibc 2.28. I saw a similar error with the "doveadm purge" command after upgrading my system to that version. Michael Marley On 2018-08-31 13:10, Aki Tuomi wrote: > I'll see if this is reproducible > > --- > Aki Tuomi > Dovecot oy > > -------- Original message -------- > From: Reuben Farrelly <reuben-dovecot(a)reub.net> > Date: 31/08/2018 17:41 (GMT+02:00) > To: Aki Tuomi <aki.tuomi(a)dovecot.fi>, Dovecot Mailing List <dovecot(a)dovecot.org> > Subject: Re: Dovecot User Listing Error - getpwent() failed: Invalid Argument > Sure: > > https://www.reub.net/files/dovecot/lightning-dovecot.conf > https://www.reub.net/files/dovecot/thunderstorm-dovecot.conf > > Updated nightly. > > Reuben > > On 1/09/2018 12:26 am, Aki Tuomi wrote: >> Can you provide doveconf -n? >> >> >> >> --- >> Aki Tuomi >> Dovecot oy >> >> -------- Original message -------- >> From: Reuben Farrelly <reuben-dovecot(a)reub.net> >> Date: 31/08/2018 17:12 (GMT+02:00) >> To: Aki Tuomi <aki.tuomi(a)dovecot.fi>, Dovecot Mailing List >> <dovecot(a)dovecot.org> >> Subject: Re: Dovecot User Listing Error - getpwent() failed: Invalid >> Argument >> >> No. Neither of those are installed on either system. >> >> Reuben >> >> On 1/09/2018 12:09 am, Aki Tuomi wrote: >>> Could apparmor or selinux be causing this! >>> >>> >>> >>> --- >>> Aki Tuomi >>> Dovecot oy >>> >>> -------- Original message -------- >>> From: Reuben Farrelly <reuben-dovecot(a)reub.net> >>> Date: 31/08/2018 16:50 (GMT+02:00) >>> To: Dovecot Mailing List <dovecot(a)dovecot.org> >>> Subject: Dovecot User Listing Error - getpwent() failed: Invalid Argument >>> >>> Hi, >>> >>> I'm running dovecot-2.3 git and seeing on 2 of my Dovecot installations, >>> the following message logged quite frequently: >>> >>> Aug 31 16:55:53 lightning.reub.net dovecot[7698]: auth-worker(7707): >>> Error: getpwent() failed: Invalid argument >>> Aug 31 16:55:53 lightning.reub.net dovecot[7698]: replicator: Error: >>> User listing returned failure >>> Aug 31 16:55:53 lightning.reub.net dovecot[7698]: replicator: Error: >>> listing users failed, can't replicate existing data >>> >>> It appears that this could be having an effect on replication, but it's >>> unclear because the secondary/replica doesn't see much traffic. >>> >>> Authentication is via PAM and the system is Gentoo Linux x86_64. passb >>> driver = pam, userdb driver = static . >>> >>> doveadm user -u '*' shows the system users listed twice like this (is >>> this a problem?) and then prints: >>> >>> ---- >>> reuben >>> liam >>> reuben >>> liam >>> <some other users> >>> Error: User listing returned failure >>> Fatal: user listing failed >>> ---- >>> >>> To me that doesn't look right. >>> >>> Can anyone suggest what could be causing this? I don't recall seeing >>> this message some time ago so I suspect it's a recent change in either >>> dovecot, or pam or glibc etc.. >>> >>> Thanks, >>> Reuben

2 1

Re: Dovecot User Listing Error - getpwent() failed: Invalid Argument
by Reuben Farrelly 31 Aug '18

31 Aug '18

No. Neither of those are installed on either system. Reuben On 1/09/2018 12:09 am, Aki Tuomi wrote: > Could apparmor or selinux be causing this! > > > > --- > Aki Tuomi > Dovecot oy > > -------- Original message -------- > From: Reuben Farrelly <reuben-dovecot(a)reub.net> > Date: 31/08/2018 16:50 (GMT+02:00) > To: Dovecot Mailing List <dovecot(a)dovecot.org> > Subject: Dovecot User Listing Error - getpwent() failed: Invalid Argument > > Hi, > > I'm running dovecot-2.3 git and seeing on 2 of my Dovecot installations, > the following message logged quite frequently: > > Aug 31 16:55:53 lightning.reub.net dovecot[7698]: auth-worker(7707): > Error: getpwent() failed: Invalid argument > Aug 31 16:55:53 lightning.reub.net dovecot[7698]: replicator: Error: > User listing returned failure > Aug 31 16:55:53 lightning.reub.net dovecot[7698]: replicator: Error: > listing users failed, can't replicate existing data > > It appears that this could be having an effect on replication, but it's > unclear because the secondary/replica doesn't see much traffic. > > Authentication is via PAM and the system is Gentoo Linux x86_64. passb > driver = pam, userdb driver = static . > > doveadm user -u '*' shows the system users listed twice like this (is > this a problem?) and then prints: > > ---- > reuben > liam > reuben > liam > <some other users> > Error: User listing returned failure > Fatal: user listing failed > ---- > > To me that doesn't look right. > > Can anyone suggest what could be causing this? I don't recall seeing > this message some time ago so I suspect it's a recent change in either > dovecot, or pam or glibc etc.. > > Thanks, > Reuben

2 1

Dovecot User Listing Error - getpwent() failed: Invalid Argument
by Reuben Farrelly 31 Aug '18

31 Aug '18

Hi, I'm running dovecot-2.3 git and seeing on 2 of my Dovecot installations, the following message logged quite frequently: Aug 31 16:55:53 lightning.reub.net dovecot[7698]: auth-worker(7707): Error: getpwent() failed: Invalid argument Aug 31 16:55:53 lightning.reub.net dovecot[7698]: replicator: Error: User listing returned failure Aug 31 16:55:53 lightning.reub.net dovecot[7698]: replicator: Error: listing users failed, can't replicate existing data It appears that this could be having an effect on replication, but it's unclear because the secondary/replica doesn't see much traffic. Authentication is via PAM and the system is Gentoo Linux x86_64. passb driver = pam, userdb driver = static . doveadm user -u '*' shows the system users listed twice like this (is this a problem?) and then prints: ---- reuben liam reuben liam <some other users> Error: User listing returned failure Fatal: user listing failed ---- To me that doesn't look right. Can anyone suggest what could be causing this? I don't recall seeing this message some time ago so I suspect it's a recent change in either dovecot, or pam or glibc etc.. Thanks, Reuben

2 1

Shared mailbox name
by Kacper 31 Aug '18

31 Aug '18

Hello, Is it possible to have a custom name on a shared inbox? Currently prefix = shared/%%u/ makes the shared inbox have the name of the user whos sharing the inbox. I would like to set a custom name on the shared inbox such as "My Recipes" (spaces included). Background: I've implemented shared inboxes by using groups in Active Directory. Each group is a user with a mailbox and has it's mailbox shared using a namespace. A post login script fetches all groups a user is member of and includes them into the ACL_GROUPS environment variable. namespace { type = shared separator = / prefix = Shared/%%u/ location = maildir:/var/mail/maildir/%%u:INDEXPVT=/var/mail/maildir/shared/%u/%%u subscriptions = no list = children } I've tried using groups which contain spaces and mixing lower/upper case letters. "doveadm acl set -u "My Recipes" inbox group="My Recipes" create expunge insert lookup read write write-deleted write-seen" works and creates "/var/mail/maildir/My Recepies", but for some reason when refreshing the imap subscriptions in Thunderbird dovecot creates "/var/mail/maildir/my recepies" and "/var/mail/maildir/myrecepies". Thanks, Kacper

1 0

SNI Dovecot
by Nicolas 31 Aug '18

31 Aug '18

Hi all, I'm testing the SNI configuration from dovecot's wiki page, to have multiple domains. I'm using letsencrypt certificates. On the 10-ssl.conf, when I only use one domain, like this, it works : ssl_ca = </etc/letsencrypt/live/mail.mydomain.fr/chain.pem ssl_cert = </etc/letsencrypt/live/mail.mydomain.fr/cert.pem ssl_key = </etc/letsencrypt/live/mail.mydomain.fr/privkey.pem I got a warning of course when using my second domain, mydomain2.fr. If I do the config : local_name mail.mydomain.fr { ssl_ca = </etc/letsencrypt/live/mail.mydomain.fr/chain.pem ssl_cert = </etc/letsencrypt/live/mail.mydomain.fr/cert.pem ssl_key = </etc/letsencrypt/live/mail.mydomain.fr/privkey.pem } local_name mail.mydomain2.fr { ssl_ca = </etc/letsencrypt/live/mail.mydomain2.fr/chain.pem ssl_cert = </etc/letsencrypt/live/mail.mydomain2.fr/cert.pem ssl_key = </etc/letsencrypt/live/mail.mydomain2.fr/privkey.pem } I got this on dovecot's start : dovecot[930]: master: Error: service(imap-login): command startup failed, throttling for 8 secs dovecot[932]: imap-login: Fatal: Couldn't parse private ssl_key: error:0906D06C:PEM routines:PEM_read_bio:no start line: Expecting: ANY PRIVATE KEY It's working without local_name, so why it can be a certificate issue? Any idea? I'm using dovecot 2.2.27-3+deb9u2 from debian. Thanks, Nicolas

3 3