Home | Docs | Forums | Lists | Bugs | Planet | Store | GMN | Get Gentoo!
Not eligible to see or edit group visibility for this bug.
View Bug Activity | Format For Printing | XML | Clone This Bug
Hi, You will find on attachment the complete ebuilds of the 0.6.2 and 0.7_rc1 of pam_mysql modules. I have tried to correct the lack of md5, openssl and sasl2 use. I've had some dependencies. Could it be possible for someone to test it please ? Cyrius.
Created an attachment (id=80200) [details] New ebuilds for 0.6.2 and 0.7_rc1 pam_mysql modules
Don't attach tarballs. http://www.gentoo.org/proj/en/devrel/handbook/handbook.xml?part=2&chap=3#doc_chap2 Reopen with unified diffs against current ebuilds.
(From update of attachment 80200 [details]) Doesn't take in account pam librairies
Created an attachment (id=80217) [details] New ebuilds for 0.6.2 and 0.7_rc1 pam_mysql modules Hi all, When doing ldd -r /lib/security/pam_mysql, we can see that the pam librairies are not linked with pam_mysql. I hope to have corrected it. Could someone test thoses versions please ? Cyrius
Oups, Sorry Jakub. i didn't see it was closed. But there are Two ebuilds and source, acinclude.m4, configure.in code patch in the tarball. Do you confirm me i have to create thoses attachments ? Thanks in advance Cyrius
Yeah, plaintext please...
(From update of attachment 80217 [details]) Tarball is forbiden
Created an attachment (id=80218) [details] New ebuilds for 0.6.2 pam_mysql module
Created an attachment (id=80219) [details] Source code patch for sasl2 openssl integration You must put this patch in the files directiry
Created an attachment (id=80220) [details] Source code patch for linking with pam librairies To put in the files directory
Created an attachment (id=80221) [details] New ebuild 0.7_rc1 pam_mysql module
Created an attachment (id=80222) [details] 0.7_rc1 source code patch for sasl2 openssl ipv6 integration to put in the files directory
Created an attachment (id=80223) [details] 0.7_rc1 source code patch for linking with pam librairies to put in the files directory
Hi, I've hadded thoses new ebuilds and patch wihtout the use of a tarball. Could someone test thoses ebuilds please ? Cyrius
Hello All, I'm testing the 0.7_rc1 with postfix. I'm following thoses how to : http://postfixwiki.org/index.php?title=Virtual_Users_and_Domains_with_Courier-IMAP_and_MySQL http://gentoo-wiki.com/HOWTO_Linux_Virtual_Hosting_Server#Postfix With telnet testing all seems to be ok. Mysql is accessed, data retreived and translated. I have tested only this configuration : /etc/pam.d/smtp auth required pam_nologin.so auth required pam_mysql.so user=mailsql passwd=ebonzai02152006 host=localhost db=postfixdb_config table=mailbox usercolumn=username passwdcolumn=password crypt=1 #auth required pam_unix.so auth required pam_env.so # [1] account sufficient pam_mysql.so user=mailsql passwd=xxxxxx host=localhost db=postfixdb_config table=mailbox usercolumn=username passwdcolumn=password crypt=1 account required pam_unix.so No dump, No error message. The parameter crypt put to 1 means using crypt function (see the README file) Then, perhaps you can say it's ok for p
Hello All, I'm testing the 0.7_rc1 with postfix. I'm following thoses how to : http://postfixwiki.org/index.php?title=Virtual_Users_and_Domains_with_Courier-IMAP_and_MySQL http://gentoo-wiki.com/HOWTO_Linux_Virtual_Hosting_Server#Postfix With telnet testing all seems to be ok. Mysql is accessed, data retreived and translated. I have tested only this configuration : /etc/pam.d/smtp auth required pam_nologin.so auth required pam_mysql.so user=mailsql passwd=ebonzai02152006 host=localhost db=postfixdb_config table=mailbox usercolumn=username passwdcolumn=password crypt=1 #auth required pam_unix.so auth required pam_env.so # [1] account sufficient pam_mysql.so user=mailsql passwd=xxxxxx host=localhost db=postfixdb_config table=mailbox usercolumn=username passwdcolumn=password crypt=1 account required pam_unix.so No dump, No error message. The parameter crypt put to 1 means using crypt function (see the README file) Then, perhaps you can say it's ok for pôrtage testing ? Thanks, Cyrius
As i said, it works for me
0.6.2 version works too ! Cyrius
0.6.2 version works too ! cyrius
Hi, To see the log generated by pam_mysql, switch the parameter verbose=1. Then you will have in auth.log for an error (or message.log on very basic install)this kind of log : Feb 20 19:05:11 ebonzai saslauthd[27013]: pam_mysql - SELECT password FROM mailbox WHERE username = 'toto@linogconsulting.org' Feb 20 19:05:11 ebonzai saslauthd[27013]: pam_mysql - SELECT returned no result. Feb 20 19:05:11 ebonzai saslauthd[27013]: pam_mysql - pam_mysql_check_passwd() returning 1. Feb 20 19:05:11 ebonzai saslauthd[27013]: pam_mysql - pam_sm_authenticate() returning 10. Feb 20 19:05:11 ebonzai saslauthd[27013]: DEBUG: auth_pam: pam_authenticate failed: User not known to the underlying authentication module Feb 20 19:05:11 ebonzai saslauthd[27013]: pam_mysql - pam_mysql_release_ctx() called. And for a success : Feb 20 19:11:46 ebonzai saslauthd[27013]: pam_mysql - SELECT password FROM mailbox WHERE username = 'cyrius@linogconsulting.org' Feb 20 19:11:46 ebonzai saslauthd[27013]: pam_mysql - pam_mysql_sql_log() called. Feb 20 19:11:46 ebonzai saslauthd[27013]: pam_mysql - pam_mysql_sql_log() returning 0. Feb 20 19:11:46 ebonzai saslauthd[27013]: pam_mysql - pam_mysql_check_passwd() returning 0. Feb 20 19:11:46 ebonzai saslauthd[27013]: pam_mysql - pam_sm_authenticate() returning 0. Cyrius
Hi, For the sceptics men : Test with good user and bad password : Feb 20 19:23:04 ebonzai saslauthd[27013]: pam_mysql - SELECT password FROM mailbox WHERE username = 'cyrius@linogconsulting.org' Feb 20 19:23:04 ebonzai saslauthd[27013]: pam_mysql - pam_mysql_sql_log() called. Feb 20 19:23:04 ebonzai saslauthd[27013]: pam_mysql - pam_mysql_sql_log() returning 0. Feb 20 19:23:04 ebonzai saslauthd[27013]: pam_mysql - pam_mysql_check_passwd() returning 6. Feb 20 19:23:04 ebonzai saslauthd[27013]: pam_mysql - pam_sm_authenticate() returning 7. Feb 20 19:23:04 ebonzai saslauthd[27013]: DEBUG: auth_pam: pam_authenticate failed: Authentication failure Feb 20 19:23:04 ebonzai saslauthd[27013]: pam_mysql - pam_mysql_release_ctx() called. Feb 20 19:23:04 ebonzai saslauthd[27013]: pam_mysql - pam_mysql_destroy_ctx() called. Feb 20 19:23:04 ebonzai saslauthd[27013]: pam_mysql - pam_mysql_close_db() called. Feb 20 19:23:04 ebonzai saslauthd[27013]: do_auth : auth failure: [user=cyrius@linogconsulting.org] [service=smtp] [realm=linogconsulting.org] [mech=pam] [reason=PAM auth error] Cyrius
Hi, I don't know what is the procedure. I have put this bugg as resolved/worksforme but i don't know if it's the good way to do when i saw the three dependants buggs still open. So, i will let the gentoo team decide. regards, Cyrius
I installed using the 0.6.2 ebuild - authentication with md5 works for me. Thanks for the patches! The only problem I noticed was that the configure output indicates that sasl v1 headers are being used, v2 headers are actually being used. Any idea when this will be released, at least into ~x86?
(From update of attachment 80220 [details]) Contain absolute path
Created an attachment (id=84163) [details] 0.6.2 source code patch for sasl2 openssl integration Having suppress absolute path
Created an attachment (id=84164) [details] 0.6.2 source code patch for linking with pam librairies Having suppress absolute path
(From update of attachment 80222 [details]) It contains absolute path
Created an attachment (id=84168) [details] 0.7_rc1 source code patch for sasl2 openssl ipv6 integration Having suppress absolute path
Created an attachment (id=84170) [details] 0.7_rc1 source code patch for linking with pam librairies Having suppress absolute path
Hello Mike, Yes, thanks your apply, i saw there is a mistake into the patch part. I have let the absolute path. Then in your patch.log, you will see that, in fact, patches are not applied. I've corrected it and added the new corrected patches. It will be great if you could confirm the correction. I have no idea on the ~x86 include date. I hope soon. Regards, Cyrius
Marking dupe of the bump bug as it mostly requires a fresh support entirely. *** This bug has been marked as a duplicate of 120842 ***
