Sun, 31 Aug 2014 16:18:35 +0500 Disengaging imsgev is tricky. First imsgev_clear as we know there
Sunil Nimmagadda <sunil@nimmagadda.net> [Sun, 31 Aug 2014 16:18:35 +0500] rev 37
Disengaging imsgev is tricky. First imsgev_clear as we know there isn't anything left to be written. Signal imsgev termination by setting iev->terminate = 1(Mind that we are setting it from with the callback). As imsgev would again callback dispatch with IMSGEV_DONE, we cannot free imsgev yet. So iev_maildrop need to exist beyond lifetime of session until IMSGEV_DONE. Allocating it seperately from session. imsgev_close is avoided as it schedules another EV_WRITE which is not needed in our case. This fixes a crash observed consistently with a little perl script to simulate concurrent sessions. Thanks MALLOC_OPTIONS='SFG<<'
Thu, 28 Aug 2014 15:46:44 +0500 Shorten.
Sunil Nimmagadda <sunil@nimmagadda.net> [Thu, 28 Aug 2014 15:46:44 +0500] rev 36
Shorten.
Thu, 28 Aug 2014 14:17:28 +0500 space nit.
Sunil Nimmagadda <sunil@nimmagadda.net> [Thu, 28 Aug 2014 14:17:28 +0500] rev 35
space nit.
Wed, 27 Aug 2014 16:21:31 +0500 No use of iev->data, set it to NULL.
Sunil Nimmagadda <sunil@nimmagadda.net> [Wed, 27 Aug 2014 16:21:31 +0500] rev 34
No use of iev->data, set it to NULL.
Wed, 27 Aug 2014 14:25:26 +0500 Give session a chance to init its imsgev. Let session initiate the
Sunil Nimmagadda <sunil@nimmagadda.net> [Wed, 27 Aug 2014 14:25:26 +0500] rev 33
Give session a chance to init its imsgev. Let session initiate the process of MAILDROP_INIT. Maildrop now sets up itself and waits for MAILDROP_INIT from session. This would now eliminate the doubt in the earlier setup when priv process forks and the child gets cpu slice first before parent which sends MAILDROP_INIT to an uninitialized imsgev.
Wed, 27 Aug 2014 12:47:09 +0500 mtype, mpath are accessible to maildrop, simplify setup parameters.
Sunil Nimmagadda <sunil@nimmagadda.net> [Wed, 27 Aug 2014 12:47:09 +0500] rev 32
mtype, mpath are accessible to maildrop, simplify setup parameters.
Wed, 27 Aug 2014 11:50:11 +0500 Rename to differentiate setup and init operations.
Sunil Nimmagadda <sunil@nimmagadda.net> [Wed, 27 Aug 2014 11:50:11 +0500] rev 31
Rename to differentiate setup and init operations.
Wed, 27 Aug 2014 10:36:03 +0500 Bounds check array index. Fixes a crash when len == 1.
Sunil Nimmagadda <sunil@nimmagadda.net> [Wed, 27 Aug 2014 10:36:03 +0500] rev 30
Bounds check array index. Fixes a crash when len == 1.
Sat, 23 Aug 2014 19:07:45 +0500 First clear io then iobuf.
Sunil Nimmagadda <sunil@nimmagadda.net> [Sat, 23 Aug 2014 19:07:45 +0500] rev 29
First clear io then iobuf.
Sat, 23 Aug 2014 09:19:50 +0500 Latest changes from upstream with commit message...
Sunil Nimmagadda <sunil@nimmagadda.net> [Sat, 23 Aug 2014 09:19:50 +0500] rev 28
Latest changes from upstream with commit message... When the three possible return values are -1, 0, and 1, != 1 is the same as <= 0. And the latter is the normal idiom so use that. ok claudio@ henning@
Wed, 20 Aug 2014 21:46:02 +0500 Prepend session id to the log message.
Sunil Nimmagadda <sunil@nimmagadda.net> [Wed, 20 Aug 2014 21:46:02 +0500] rev 27
Prepend session id to the log message.
Thu, 29 May 2014 16:02:35 +0500 knf and whitespace fixes. v1.0
Sunil Nimmagadda <sunil@nimmagadda.net> [Thu, 29 May 2014 16:02:35 +0500] rev 26
knf and whitespace fixes.
Wed, 28 May 2014 16:37:41 +0500 Fix a crash during concurrent sessions TIMEOUT reported by James
Sunil Nimmagadda <sunil@nimmagadda.net> [Wed, 28 May 2014 16:37:41 +0500] rev 25
Fix a crash during concurrent sessions TIMEOUT reported by James Turner (#5). imsgev_close schedules(event_add) iev and the event struct is free-ed before the event is dispatched. Include string.h after recent libressl changes.
Fri, 02 May 2014 19:17:38 +0500 Pull in latest changes for iobuf/ioev from smtpd.
Sunil Nimmagadda <sunil@sunilnimmagadda.com> [Fri, 02 May 2014 19:17:38 +0500] rev 24
Pull in latest changes for iobuf/ioev from smtpd.
Sat, 19 Apr 2014 19:27:26 +0500 This cannot truncate unless new states are added.
Sunil Nimmagadda <sunil@sunilnimmagadda.com> [Sat, 19 Apr 2014 19:27:26 +0500] rev 23
This cannot truncate unless new states are added.
Sat, 19 Apr 2014 19:10:41 +0500 Equal sized buffer copy, no truncation happens.
Sunil Nimmagadda <sunil@sunilnimmagadda.com> [Sat, 19 Apr 2014 19:10:41 +0500] rev 22
Equal sized buffer copy, no truncation happens.
Sat, 19 Apr 2014 19:07:31 +0500 args are checked for size in parse function. Cast to void as no truncation check required here.
Sunil Nimmagadda <sunil@sunilnimmagadda.com> [Sat, 19 Apr 2014 19:07:31 +0500] rev 21
args are checked for size in parse function. Cast to void as no truncation check required here.
Sat, 19 Apr 2014 17:29:17 +0500 hash is a fixed length string. Copying is always done to same length
Sunil Nimmagadda <sunil@sunilnimmagadda.com> [Sat, 19 Apr 2014 17:29:17 +0500] rev 20
hash is a fixed length string. Copying is always done to same length buffer. Explicitly cast to void for clarity.
Sat, 12 Apr 2014 07:15:21 +0500 Update README.md
Sunil Nimmagadda <sunil@poolp.org> [Sat, 12 Apr 2014 07:15:21 +0500] rev 19
Update README.md
Fri, 11 Apr 2014 10:06:47 +0500 Mark uidl correctly in list_res for the terminal. Fixes UIDL command
Sunil Nimmagadda <sunil@nimmagadda.net> [Fri, 11 Apr 2014 10:06:47 +0500] rev 18
Mark uidl correctly in list_res for the terminal. Fixes UIDL command on an empty maildrop.
Fri, 11 Apr 2014 10:01:48 +0500 Remove redundant sock member in session. io already holds the sock for each session.
Sunil Nimmagadda <sunil@nimmagadda.net> [Fri, 11 Apr 2014 10:01:48 +0500] rev 17
Remove redundant sock member in session. io already holds the sock for each session.
Thu, 10 Apr 2014 18:26:13 +0500 pull in latest master.
Sunil Nimmagadda <sunil@sunilnimmagadda.com> [Thu, 10 Apr 2014 18:26:13 +0500] rev 16
pull in latest master.
Thu, 10 Apr 2014 18:19:56 +0500 Kill pop3_main return value as we now waitpid on WAIT_ANY.
Sunil Nimmagadda <sunil@sunilnimmagadda.com> [Thu, 10 Apr 2014 18:19:56 +0500] rev 15
Kill pop3_main return value as we now waitpid on WAIT_ANY.
Thu, 10 Apr 2014 18:15:50 +0500 Since we now waitpid on WAIT_ANY, pop3_main needed return pid.
Sunil Nimmagadda <sunil@sunilnimmagadda.com> [Thu, 10 Apr 2014 18:15:50 +0500] rev 14
Since we now waitpid on WAIT_ANY, pop3_main needed return pid.
Thu, 10 Apr 2014 18:13:08 +0500 Merge pull request #3 from jturner/killzombies
Sunil Nimmagadda <sunil@poolp.org> [Thu, 10 Apr 2014 18:13:08 +0500] rev 13
Merge pull request #3 from jturner/killzombies Make sure zombie processes are not left over
Thu, 10 Apr 2014 17:59:07 +0500 Simplify maildrop path handling. Regardless of the order of options path is now correctly handled.
Sunil Nimmagadda <sunil@sunilnimmagadda.com> [Thu, 10 Apr 2014 17:59:07 +0500] rev 12
Simplify maildrop path handling. Regardless of the order of options path is now correctly handled.
Wed, 09 Apr 2014 21:36:45 -0400 Make sure zombie processes are left over
James Turner <james@calminferno.net> [Wed, 09 Apr 2014 21:36:45 -0400] rev 11
Make sure zombie processes are left over
Mon, 07 Apr 2014 12:28:10 +0500 Fix clang warnings. offset is always long.
Sunil Nimmagadda <sunil@nimmagadda.net> [Mon, 07 Apr 2014 12:28:10 +0500] rev 10
Fix clang warnings. offset is always long.
Wed, 02 Apr 2014 14:44:04 +0500 Return correct nlines for RETR in maildir handling.
Sunil Nimmagadda <sunil@nimmagadda.net> [Wed, 02 Apr 2014 14:44:04 +0500] rev 9
Return correct nlines for RETR in maildir handling.
Wed, 02 Apr 2014 13:14:04 +0500 Use correct format of listing.
Sunil Nimmagadda <sunil@nimmagadda.net> [Wed, 02 Apr 2014 13:14:04 +0500] rev 8
Use correct format of listing.
Wed, 02 Apr 2014 11:40:41 +0500 Move it closer to list.
Sunil Nimmagadda <sunil@nimmagadda.net> [Wed, 02 Apr 2014 11:40:41 +0500] rev 7
Move it closer to list.
Tue, 01 Apr 2014 20:24:49 +0500 Merge branch 'refactor'
Sunil Nimmagadda <sunil@sunilnimmagadda.com> [Tue, 01 Apr 2014 20:24:49 +0500] rev 6
Merge branch 'refactor'
Tue, 01 Apr 2014 20:22:19 +0500 merge from cvs.
Sunil Nimmagadda <sunil@sunilnimmagadda.com> [Tue, 01 Apr 2014 20:22:19 +0500] rev 5
merge from cvs.
Tue, 01 Apr 2014 20:19:43 +0500 Import.
Sunil Nimmagadda <sunil@sunilnimmagadda.com> [Tue, 01 Apr 2014 20:19:43 +0500] rev 4
Import.
Tue, 01 Apr 2014 19:34:51 +0500 Properly NUL terminate expanded path by memsetting it.
Sunil Nimmagadda <sunil@sunilnimmagadda.com> [Tue, 01 Apr 2014 19:34:51 +0500] rev 3
Properly NUL terminate expanded path by memsetting it.
Tue, 01 Apr 2014 16:15:19 +0500 Simplify list_all handling. Send an imsg for each msg instead of
Sunil Nimmagadda <sunil@nimmagadda.net> [Tue, 01 Apr 2014 16:15:19 +0500] rev 2
Simplify list_all handling. Send an imsg for each msg instead of consolidating into one imsg. We are no longer limited by MAX_IMSGSIZE. End of list is signalled by sending an imsg with hash as "" or sz as 0.
Thu, 27 Mar 2014 09:57:27 +0500 Create README.md
Sunil Nimmagadda <sunil@poolp.org> [Thu, 27 Mar 2014 09:57:27 +0500] rev 1
Create README.md
Thu, 27 Mar 2014 09:53:52 +0500 Import pop3d.
Sunil Nimmagadda <sunil@nimmagadda.net> [Thu, 27 Mar 2014 09:53:52 +0500] rev 0
Import pop3d.
(0) tip