November 2016 - dovecot

Re: (build) crash in dcrypt when "-Bsymbolic-functions" is used on linking
by Aki Tuomi 25 Nov '16

25 Nov '16

On 26.10.2016 13:48, Christian Ehrhardt wrote: > Hi, > I was wondering about a crash when building dovecot 2.2.25 on latest > Ubuntu. > I wondered as I've had the same source building on Debian just fine. > > Some debugging led me to this weird behavior: > Using this gdb command file called autoreportissue in my case: > break dcrypt_initialize > commands > p dcrypt_vfs > p &dcrypt_vfs > watch dcrypt_vfs > c > end > break dcrypt_set_vfs > commands > p dcrypt_vfs > p &dcrypt_vfs > c > end > r > > Running test-crypto on Debian and Ubuntu reported those two behaviours: > gdb -d /root/dovecot-2.2.25/src/ -x autoreportissue ./test-crypto > > Good: > Breakpoint 1, dcrypt_initialize (backend=0x555555587c02 "openssl", > set=0x0, error_r=0x0) at dcrypt.c:15 > 15 if (dcrypt_vfs != NULL) { > $1 = (struct dcrypt_vfs *) 0x0 > $2 = (struct dcrypt_vfs **) 0x555555796370 <dcrypt_vfs> > Hardware watchpoint 3: dcrypt_vfs > Breakpoint 2, dcrypt_set_vfs (vfs=0x7ffff7835020 <dcrypt_openssl_vfs>) > at dcrypt.c:56 > 56 dcrypt_vfs = vfs; > $3 = (struct dcrypt_vfs *) 0x0 > $4 = (struct dcrypt_vfs **) 0x555555796370 <dcrypt_vfs> > Hardware watchpoint 3: dcrypt_vfs > Old value = (struct dcrypt_vfs *) 0x0 > New value = (struct dcrypt_vfs *) 0x7ffff7835020 <dcrypt_openssl_vfs> > dcrypt_set_vfs (vfs=0x7ffff7835020 <dcrypt_openssl_vfs>) at dcrypt.c:57 > 57 } > > Bad: > Breakpoint 1, dcrypt_initialize (backend=0x555555589f02 "openssl", > set=0x0, error_r=0x0) at dcrypt.c:11 > 11 { > $1 = (struct dcrypt_vfs *) 0x0 > $2 = (struct dcrypt_vfs **) 0x555555798370 <dcrypt_vfs> > Hardware watchpoint 3: dcrypt_vfs > Breakpoint 2, dcrypt_set_vfs (vfs=0x7ffff780a020 <dcrypt_openssl_vfs>) > at dcrypt.c:56 > 56 dcrypt_vfs = vfs; > $3 = (struct dcrypt_vfs *) 0x0 > $4 = (struct dcrypt_vfs **) 0x7ffff780a890 <dcrypt_vfs> > Panic: file dcrypt.c: line 34 (dcrypt_initialize): assertion failed: > (dcrypt_vfs != NULL) > Error: Raw backtrace: > /root/dovecot-2.2.25/src/lib-dcrypt/test-crypto(+0x15f7c) > [0x555555569f7c] -> > /root/dovecot-2.2.25/src/lib-dcrypt/test-crypto(default_error_handler+0) > [0x55555556a030] -> > /root/dovecot-2.2.25/src/lib-dcrypt/test-crypto(i_fatal+0) > [0x55555556a2ff] -> > /root/dovecot-2.2.25/src/lib-dcrypt/test-crypto(dcrypt_initialize+0x140) > [0x55555555f030] -> > /root/dovecot-2.2.25/src/lib-dcrypt/test-crypto(main+0x23) > [0x55555556706d] -> > /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf1) > [0x7ffff782d3f1] -> > /root/dovecot-2.2.25/src/lib-dcrypt/test-crypto(_start+0x2a) > [0x55555555edea] > Program received signal SIGABRT, Aborted. > > One can see that in the bad case the pointer of dcrypt_vfs is pointing > to something of the scope of the .libs/libdcrypt_openssl.so file and > not the dcrypt_initialize of test-crypto. > > That made me wonder even more - where would this issue of variable > scope come from. After more debugging I found that the linker flag > "-Bsymbolic-functions" is the reason. This is default on recent > Ubuntu, but not on Debian (?yet?). > > Eventually what happens is that the dcrypt_vfs becomes part of the > .libs/libdcrypt_openssl.so. So the call from there to dcrypt_set_vfs > ends up setting not the expected variable. > I was unable to come up with a reasonable fix since I'm not enough > into your sublib structure. > > For now I assume I'm gonna build the package stripping this flag in > Ubuntu. > But long term I think dovecot should fix it to work with that compiler > flag. > Therefore the report to make you aware. > > > P.S. thanks to the dovecot community for having unit tests that find > this at build time! > > -- > Christian Ehrhardt > Software Engineer, Ubuntu Server > Canonical Ltd Hi! Thank you for reporting this, we'll look into it. Aki Tuomi Dovecot oy

1 1

Could the index of dovecot be stored in database?
by eSX 25 Nov '16

25 Nov '16

Hi dovecot, I have a question about the index of dovecot. I had read the wiki, and saw that I can custom a Input streams to do read/seek/sync...by implementing a istream-*.c file. Thus maybe dovecot can read mail from a http server or any others. But I found that the index of maildir can't extend very well. There is no extern API for developer? For example, I want to put the index to database(maybe, MySQL) or any else, But I found I have no way to implement it, Unless a heavy modification? -- eSX

1 0

Patch for dovecot-antispam-plugin to support IMAP Flags indicating Junk status
by Jonas Wielicki 25 Nov '16

25 Nov '16

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Dear list, I made a patch which extends dovecot-antispam-plugin with support to take the addition and removal of a specific set of flags as trigger for learning. This allows for very tight integration with Thunderbird and KMail, which both set flags for Junk messages (Thunderbird sets "Junk", KMail sets "JUNK"). With this patch and according configuration, dovecot-antispam-plugin can trigger on the addition (learn as spam) and removal (learn as ham) of these flags. Combined with a sieve script which sets these flags for incoming mail which has been classified as spam, this allows for a tight and intuitive feedback loop between users and the server-side spam classification system. I tested that patch with dovecot 2.2 from Debian stretch and so far it works pretty well. I realise that dovecot-antispam-plugin has not been updated since 2013. It also has not been migrated to Github yet. If it is to be migrated there, I would be happy to submit a pull request there for discussion and merging. Otherwise, someone would have to indicate how to go forward with applying this patch to the official version, if that is desired. The patch is based on current tip and feedback is appreciated. I took some inspiration from dovecot-core/src/plugins/notify/notify-storage.c -- this requires the use of struct mail_private to be able to override the vfunc for mail_update_keywords, which I actually do not like. However, when I tried to make this patch base on the use of the notify plugin (via notify_register), I ran into the issue that the transaction is partially freed before the transaction_commit event from notify is emitted, which is unfortunate and breaks at least the mailtrain backend (which I use). This is in theory fixable, but I wanted to keep the patch as small as possible, as I have no idea whether this will eventually end up upstream or whether I’ll have to maintain that for myself. If upstream is not interested in this patch, please let me know! In that case, I would consider making an own plugin, which fewer backends (possibly only a mailtrain-like one) out of this, but frankly, I would consider that a waste of resources. In the hope that this is useful for others and hoping for feedback, Jonas Wielicki -----BEGIN PGP SIGNATURE----- iQIzBAEBCgAdFiEEG/EPV+Xzd5wEoQQIwGIDJZdiWIoFAlg4Fd0ACgkQwGIDJZdi WIpM3Q//TWVHDFx5LK14y0jOnxzE7kGthc6hakILnWqYgl7B/wlljD1HLrec//hg 9LEancjURnv3sPU/kEEpl6RRDX5kDMbXXKXMq7/1R7bNdmeMz5NEebAo7ktlCii7 ikk7MPL0F8UViY7IoW7y5NxfqMUldYseejn0GwKiVkEyqlyGLRbP7ijyzMWjpIWp eqOg/b5bNX6ARMC1O/hKhvHw4AhvSO2j73da60Xchu39gvDa2kC+MTU/kX4wb6po kPOEgFxGhpxG0+UMzUEBPLYwY3Q9x+R/PzlpUF6xEjYxgE7leVRVfCHiZLsMUGH1 itozwSzoAyDYV+Rbzk8OEBL6EGBJQHqYmNybULZsJkxhUenNeC4GRf7NcJ0jAisG rSfwr/+kD9+ErKNgtNJzfceLYGNWti44whV6kXxRxqGlIfJsJB7ZMdrj8+wSdn0p slXUoyHMUkRkN92tIPqZF5cb+5Ke+O+5hHE07Wohxw3eNoBSqtQWo81aigjaCraC ivcZX3HswTapyKtpgWMzjoWqp982crr+wtAI+AlEUKoCEqF54EnxM1kB1qoe0HcH nMgMdIkXI4VzPaUNUK206ffEtlnKNIqF3SkOkBX263L0ZJgXyCFRRj6W3z78VDdP He62tN/FPPBtmeVrmCj2ZDE+6gqtMQN52FsN4mpaq4BrCYlYB7g= =3+CU -----END PGP SIGNATURE-----

1 0

Updated my Dovecot certificate for the first time
by Steve Litt 25 Nov '16

25 Nov '16

Hi all, I've used Dovecot since February 2012, but because I kept reinstalling Linux with every major version, I never had a Dovecot self-signed certificate go bad on me before. Til now. I started using rolling release Void Linux about a year ago, and my Dovecot self-signed certificate just expired. The solution I used is contained in these documents: http://wiki2.dovecot.org/SSL/CertificateCreation http://wiki2.dovecot.org/SSL/CertificateClientImporting file:///etc/ssl/dovecot-openssl.cnf http://www.faqforge.com/linux/renew-the-dovecot-ssl-certificate-on-ubuntu-l… I basically moved my old /etc/ssl/certs/dovecot.pem and /etc/ssl/private/dovecot.pem, then edited /etc/ssl/dovecot-openssl.cnf specifically to give myself the common name of 192.168.100.2. I had earlier used my hostname, but that produced a conflict, so I just used the ip address. Then I ran dovecot-mkcert.sh to create the new self-signed cert, and finally, configured Claws-Mail to use /etc/ssl/certs/dovecot.pem as its cert. Obviously, if my Claws-Mail were on a different machine than my Dovecot, I would have had Claws-Mail point to a local copy. Alpine still gives me a bad cert warning, saying I should either fix it or disable checking. I haven't yet found a way to get Alpine to discriminate between a valid self-signed cert and a bad one. Anyway, all's good. SteveT Steve Litt November 2016 featured book: Quit Joblessness: Start Your Own Business http://www.troubleshooters.com/startbiz

7 9

Re: Updated my Dovecot certificate for the first time
by Joseph Tam 24 Nov '16

24 Nov '16

On Thu, 24 Nov 2016, Steve Litt wrote >> add the public part of the cert into your system's trusted CA store. > > Silly question, but how would you do that? You didn't say which OS you're running on (alpine runs on Windows as well), but I'll assume *nix. A previous poster showed you how to do it with a real certificate, and the steps are the same. However, the way I found out without too much fuss was to process trace my alpine process and see where it tied to load a cert $ strace -o trace.out alpine ... quit after connection $ grep -F cert traceout /1: open64("/etc/openssl/cert.pem", O_RDONLY) Err#2 ENOENT /1: stat("/etc/openssl/certs/cbf06781.0", 0xFFBF8E54) Err#2 ENOENT ... Your output will be different of course. The first load is the default pre-loaded root CAs (Thawte, etc.) supplied by OpenSSL, and the second, etc. are chained certificate lookups. You would replace the missing cert with your own self-signed public pem file. e.g. cp mypub.pem /etc/openssl/certs/cbf06781.0 For Windows, I don't know where it fetches it from. Joseph Tam <jtam.home(a)gmail.com>

1 0

Re: Updated my Dovecot certificate for the first time
by Joseph Tam 24 Nov '16

24 Nov '16

> Alpine still gives me a bad cert warning, saying I should either fix it > or disable checking. I haven't yet found a way to get Alpine to > discriminate between a valid self-signed cert and a bad one. Well, it can't discriminate since any certificate (except those in your trusted store) that asserts its own validity is suspect. You can either get alpine to not complain e.g. inbox-path={192.168.100.2:143/user=whatever/tls/novalidate-cert} or add the public part of the cert into your system's trusted CA store. Joseph Tam <jtam.home(a)gmail.com>

2 1

antispam plugin not working correctly
by Felix Zielcke 23 Nov '16

23 Nov '16

Hi, I've configured the antispam plugin like in the wiki to use the sa- learn-pipe.sh shell script. But it's not creating the log file in /tmp. Even if I add a "touch /tmp/x" in sa-learn-pipe.sh it's not created. So how I can check if it works correctly? dovecot version 2.2.26.0 /var/log/debug although contains: Nov 23 18:36:49 daedalus imap [fzielcke(a)z-51.de 80.187.96.201 UID MOVE]: antispam: mail copy: src spam: 1, dst spam: 0, src unsure: 0 Nov 23 18:36:49 daedalus imap [fzielcke(a)z-51.de 80.187.96.201 UID MOVE]: antispam: running mailtrain backend program /usr/bin/sa-learn-pipe.sh Nov 23 18:36:49 daedalus imap [fzielcke(a)z-51.de 80.187.96.201 UID MOVE]: antispam: running mailtrain backend program /usr/bin/sa-learn-pipe.sh Nov 23 18:36:49 daedalus imap [fzielcke(a)z-51.de 80.187.96.201 UID MOVE]: antispam: running mailtrain backend program parameter 1 --ham # doveconf -n |grep antispam antispam_backend = pipe antispam_debug_target = syslog antispam_pipe_program = /usr/bin/sa-learn-pipe.sh antispam_pipe_program_notspam_arg = --ham antispam_pipe_program_spam_arg = --spam antispam_pipe_tmpdir = /tmp antispam_spam_pattern_ignorecase = SPAM antispam_trash_pattern_ignorecase = trash;Deleted * antispam_verbose_debug = 1 mail_plugins = acl zlib quota imap_zlib imap_acl imap_quota antispam

1 0

Active sieve script symbolic link doesn't update when changed.
by Steven Mainor 23 Nov '16

23 Nov '16

While setting up sieve on my email server I mistakenly setup a symlink (~/sieve/myfilter.svbin > ~/.dovecot.sieve) but I think that that should have been (~/sieve/myfilter.sieve > ~/.dovecot.sieve). I have fixed the mistake but dovecot/sieve still throws an error saying that the symlink still links to the wrong file. I have made sure the symlink is fixed (it links to ~/sieve/myfilter.sieve) and I tried restarting dovecot and the server but nothing works. I still get the error and my sieve filters aren't processed. Any help would be greatly appreciated. ~# grep sieve /var/log/syslog | nano temp.save Nov 23 01:32:09 1 dovecot: lda(steve(a)degga.net): Warning: sieve: file storage: Active Sieve script symlink /var/mail/vmail/degga.net/steve/.dovecot.sieve is broken: Invalid scriptname (points to myfilter.svbin). vmail@1:/var/mail/vmail/degga.net/steve/mail$ ls -l -a #redacted irrelevant stuff lrwxrwxrwx 1 vmail vmail 14 Nov 23 02:43 .dovecot.sieve -> myfilter.sieve vmail@1:/var/mail/vmail$ dovecot -n # 2.2.22 (fe789d2): /etc/dovecot/dovecot.conf # Pigeonhole version 0.4.13 (7b14904) # OS: Linux 4.4.0-47-generic x86_64 Ubuntu 16.04.1 LTS ext4 auth_mechanisms = plain login auth_verbose = yes mail_home =/var/mail/vmail/%d/%n mail_location = maildir:/var/mail/vmail/%d/%n/mail:LAYOUT=fs managesieve_notify_capability = mailto managesieve_sieve_capability = fileinto reject envelope encoded-character vacation subaddress comparator-i;ascii-numeric relational regex imap4flags copy include variables body enotify environment mailbox date index ihave duplicate mime foreverypart extracttext namespace inbox { inbox = yes location = mailbox Drafts { auto = subscribe special_use = \Drafts } mailbox Junk { auto = subscribe special_use = \Junk } mailbox Sent { auto = subscribe special_use = \Sent } mailbox "Sent Messages" { special_use = \Sent } mailbox Trash { auto = subscribe special_use = \Trash } prefix = } passdb { args = username_format=%u scheme=ssha512 /etc/dovecot/passwd.db driver = passwd-file } plugin { sieve = ~/.dovecot.sieve sieve_after = /var/mail/vmail/sieve-after sieve_before = /var/mail/vmail/sieve-before sieve_dir = ~/sieve } protocols = imap sieve service auth { unix_listener /var/spool/postfix/private/dovecot-auth { group = postfix mode = 0660 user = postfix } } # SSL information redacted here. userdb { args = uid=vmail gid=vmail home=/var/mail/vmail/%d/%n driver = static } protocol lda { deliver_log_format = msgid=%m: %$ mail_plugins = sieve postmaster_address =postmaster(a)degga.net quota_full_tempfail = yes rejection_reason = Your message to <%t> was automatically rejected:%n% r } protocol imap { imap_client_workarounds = delay-newmail tb-extra-mailbox-sep mail_max_userip_connections = 10 }

2 2

Exim still accepting emails to nonexistent users
by Marti Markov 23 Nov '16

23 Nov '16

After adding the configuration bit: deny message = invalid recipient domains = +local_domains !verify = recipient/callout=no_cache from: http://wiki2.dovecot.org/LMTP/Exim <http://wiki2.dovecot.org/LMTP/Exim> running update-exim4.conf and service exim4 restart the server is still accepting emails to recipients that do not exist in dovecot. Any ideas?

3 6

Single user to have lrw to all mailboxes
by Marti Markov 23 Nov '16

23 Nov '16

I have implemented a global acl: * owner lrwstipkae INBOX.Spam owner lrwstipeka .Trash owner lrwsti INBOX.Trash owner lrw * user=master-user(a)mydomain.com lrs * owner lrwstipkae But the master-user then loses all other permissions on his own mailbox, he is left only with: lrs Any pointers on how to fix this?

1 0