summaryrefslogtreecommitdiffstats
path: root/usr.sbin/iscsid (follow)
Commit message (Collapse)AuthorAgeFilesLines
* Honour DEBUGkn2019-11-251-2/+1
| | | | OK claudio dlg
* snprintf/vsnprintf return < 0 on error, rather than -1.deraadt2019-07-031-2/+2
|
* fix format string found by clang -Wformat-securitygsoares2017-04-061-2/+2
| | | | OK deraadt millert
* From a syslog perspective it does not make sense to log fatal andbluhm2017-03-211-6/+6
| | | | | | warn with the same severity. Switch log_warn() to LOG_ERR and keep fatal() at LOG_CRIT. OK reyk@ florian@
* copy log.c/h from bgpd.benno2017-01-233-10/+24
| | | | ok claudio
* fix an fd leak in an error pathjsg2016-09-101-1/+2
|
* work on making log.c similar in all daemons:benno2016-09-025-100/+127
| | | | | | | move daemon-local functions into new logmsg.c, and reduce the (mostly whitespace) differences so that log.c's can be diffed easily. ok claudio@
* stop including sys/param.h for nitems. define locally as needed.tedu2016-08-164-7/+8
| | | | ok natano reyk
* add format attributes to log functionsbenno2016-07-181-8/+16
| | | | ok claudio@
* Remove the iterator for control connections unfinished and unused.yasuoka2016-04-051-11/+1
| | | | ok claudio
* Currently we have about a 50/50 split over fcntl(n, F_GETFL [,0])krw2016-03-201-2/+2
| | | | | | | | | | idioms. Adopt the more concise fcntl(n, F_GETFL) over fcntl(n, F_GETFL, 0) where it is obvious further investigation will not yield and even better way. Obviousness evaluation and ok guenther@
* strings.h -> string.h as necessary to prevent implicit declarationmmcc2015-12-053-6/+6
| | | | warnings
* Don't check for NULL before free()mmcc2015-12-051-5/+3
|
* typo in log entrymmcc2015-12-021-2/+2
|
* use file system path (.Pa) semantic markup macros where appropriate.sobrado2015-07-271-5/+5
| | | | ok jmc@
* move to <limits.h> where possible, annotate <sys/param.h> otherwisederaadt2015-01-164-11/+9
|
* TargetPortalGroupTag is a binary 16-bit value so 0 is allowed.claudio2014-11-231-2/+3
| | | | | Still not perfect since binary values can also be base64 encoded but one step closer.
* Change the error handling a bit to ensure we do not double free the taskclaudio2014-11-231-4/+7
| | | | | | by calling conn_task_cleanup() twice (once direct and once via conn_fail()). The error handling needs some more work but at least this fixes the crash found by jasper@
* Clear the work in progress PDU before calling the task handler.claudio2014-11-231-2/+2
| | | | This ensures that we don't have multiple pointers pointing to the same PDU.
* actually use macro argument instead of shadowed variable nametedu2014-11-031-5/+5
|
* Set KERN_PROC_NOBROADCASTKILL so that iscsid is not killed by initclaudio2014-07-131-2/+8
| | | | | when terminating multi-user so that we have a chance to sync the file systems mounted via iscsi.
* Extend the connection and session FSMs so that connection failure isclaudio2014-05-105-86/+197
| | | | | handled more gracefully. Losing the TCP connection no longer results in an unrecoverable stop requiring a restart of iscsid.
* Move the task cleanup from the task fail callback to taskq_cleanup.claudio2014-05-102-11/+6
| | | | This makes the one failure callback a lot simpler.
* Bind to localAddr if specified and add some XXX comments about stuffclaudio2014-04-212-3/+14
| | | | that is not prefect yet.
* Start passing the initiator and target name in the show command.claudio2014-04-211-4/+23
|
* Redo most of the control message handling. Switch it back to useclaudio2014-04-213-42/+68
| | | | | SOCK_SEQPACKET and provide functions to build the control messages that can also be used by iscsictl.
* Handle EAGAIN, ENOBUFS and EINTR a bit better. Ignore them one layer aboveclaudio2014-04-212-11/+8
| | | | and do not fail and tear down the world when they happen.
* It is possible that we can't burst all of the data in the immediate dataclaudio2014-04-211-12/+19
| | | | | | | | part of the write request. In that case R2T requests with be sent for the rest of the data. The amount of data we can send is defined by MaxRecvDataSegmentLength which is per connection and so a bit hackish to read out. In the long run the min() of all connection MaxRecvDataSegmentLength should be stored in the session struct.
* Now that we properly negotiate the session params we can enable sendingclaudio2014-04-213-14/+27
| | | | | of immediate data for write commands. This gives a nice 20% performance boost on writes compared to the solicited data transaction we did before.
* Introduce some basic stats for the vscsi layer. Just counting the commandsclaudio2014-04-203-5/+53
| | | | and the bytes we push through vscsi(4).
* After doing all of the loging dance to get the target's params it makesclaudio2014-04-202-5/+9
| | | | | sense to merge them into the active set. At least that way we run with the right params and finally do reads and writes that are > 8k.
* Use only one style for "return foo;"claudio2014-04-203-31/+35
|
* Move session params initialization to when we start a session and notclaudio2014-04-201-5/+10
| | | | | when it is created. That way the config params from the config file have a chance to stick.
* Fix conn_gen_kvp and its caller to fill the kvp array properly (includingclaudio2014-04-202-7/+8
| | | | | the NULL terminator at the end). Now iscsid does proper LoginOperational negotiation (which will bump the MaxRecvDataSegmentLength to 64k)
* Cleanup a bit, switch on bcopy to memcpy and move a function a bit upclaudio2014-04-201-19/+18
|
* COnveret the bcopy() to memcpy()claudio2014-04-196-19/+19
|
* Replace a magic number.claudio2014-04-191-2/+2
|
* To win the startup race on the control socket we need to listen to it asclaudio2014-04-073-16/+14
| | | | | | well only then the connect() call from iscsictl will not fail. Move listen() into the init function and rename control_listen() to control_event_init() since it is now only doing that.
* Call control_init() before daemon() so that iscsid && iscsictl reload worksclaudio2014-02-171-3/+4
|
* no longer any need to quote macro lines with >9 args;jmc2013-08-141-3/+3
| | | | From: Jan Stary
* use .Mt for email addresses; from Jan Stary <hans at stare dot cz>; ok jmc@schwarze2013-07-161-3/+3
|
* handle ECONNABORTED errors from accept(). In many code blocks they can bederaadt2013-03-111-2/+3
| | | | | ignored silently and without aborting, much like EINTR and EWOULDBLOCK are. ok's from various maintainers of these directories...
* last stage of rfc changes, using consistent Rs/Re blocks, and moving thejmc2012-09-261-6/+18
| | | | references into a STANDARDS section;
* tweak previous;jmc2012-06-261-3/+3
|
* There is a iscsi.conf man page now. Reminded by jmc@claudio2012-06-261-2/+7
|
* accept() pacing on the control socket if E*FILE errors are returned.claudio2012-04-111-6/+35
| | | | Based on work by Theo for ospfd and friends. OK deraadt@
* Add support for -v (enable verbose logging on the command line).sthen2011-08-203-9/+16
| | | | Previously only available via iscsictl. ok claudio@
* Massive diff to handle logins more correctly. iscsid will now doclaudio2011-05-047-198/+567
| | | | | better operational parameter negotiation but more is needed. Tested by todd@ and myself.
* Rework the logout code and use this to do a proper logout whenclaudio2011-05-025-69/+255
| | | | | | | | exiting. This works well for idle sessions but still has some issues on busy session. It seems more task scheduler changes are needed to make this work. This also includes some mem-leak fixes in error pathes found by Igor Zinovik. go for it dlg@
* Try to schedule a new task for the connection in conn_task_cleanup()claudio2011-04-282-3/+4
| | | | this removes the session_schedule() in the task_pdu_cb().
Copyright © 1996 – 2023 Jason A. Donenfeld. All Rights Reverse Engineered.