Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | backported "clean" increment of memory trim counter (not protected by mutex) | Rainer Gerhards | 2009-06-25 | 2 | -3/+6 |
| | |||||
* | named queue worker threads | Rainer Gerhards | 2009-06-25 | 2 | -1/+7 |
| | | | | ... but I don't see the name anywhere...? | ||||
* | cleanup (removed now-unused parameters) | Rainer Gerhards | 2009-06-25 | 3 | -17/+17 |
| | |||||
* | bugfix: subtle potential issue during queue shutdown | Rainer Gerhards | 2009-06-25 | 1 | -1/+1 |
| | | | | | ... this one could cause trouble, but I really don't think it caused any actual harm. | ||||
* | improvements/fixes in queue termination timeout handling | Rainer Gerhards | 2009-06-25 | 4 | -5/+14 |
| | | | | | | | | - bugfix: subtle (and usually irrelevant) issue in timout processing timeout could be one second too early if nanoseconds wrapped - set a more sensible timeout for shutdow, now 1.5 seconds to complete processing (this also removes those cases where the shutdown message was not written because the termination happened before it) | ||||
* | some memory accesses are now explicitely atomic | Rainer Gerhards | 2009-06-25 | 4 | -10/+25 |
| | | | | | ... as far as I think this mostly is to keep the thread debuggers happy | ||||
* | Merge branch 'beta' | Rainer Gerhards | 2009-06-24 | 2 | -0/+4 |
|\ | |||||
| * | Merge branch 'v4-stable' into beta | Rainer Gerhards | 2009-06-24 | 2 | -0/+4 |
| |\ | |||||
| | * | Merge branch 'v3-stable' into v4-stable | Rainer Gerhards | 2009-06-24 | 2 | -0/+4 |
| | |\ | |||||
| | | * | bugfix: invalid mutex release in msg.c | Rainer Gerhards | 2009-06-24 | 2 | -0/+4 |
| | | | | | | | | | | | | | | | | | | | | Detected under threading debugger, seems not to have any impact on actual deployments. | ||||
* | | | | bugfix: abort when using multiple mainMsgQueue worker threads | Rainer Gerhards | 2009-06-24 | 1 | -16/+9 |
| | | | | | | | | | | | | | | | | | | | | | | | | this bug was introduced by a recent change which was a bit too agressive in avoiding locking. We can probably do better than with this patch, but I think I'll move that into the v5 engine. | ||||
* | | | | Merge branch 'omfile' | Rainer Gerhards | 2009-06-24 | 116 | -2379/+5219 |
|\ \ \ \ | |/ / / |/| | | | | | | | | | | | Conflicts: ChangeLog | ||||
| * | | | fixed a race condition: generated msg string store must be mutex protected | Rainer Gerhards | 2009-06-23 | 1 | -11/+13 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | the string area that is used to build the string being passed to the output module is now part of the action structure. As such, access to it must also be guarded by the action mutex (an even more optimal solution may be to store it in thread-local storage, but there always must remain some room for improvement ;)). | ||||
| * | | | optimized TAG handling | Rainer Gerhards | 2009-06-23 | 5 | -60/+44 |
| | | | | |||||
| * | | | optimized hostname processing | Rainer Gerhards | 2009-06-23 | 6 | -49/+20 |
| | | | | |||||
| * | | | more strict parsing of the hostname in rfc3164 mode | Rainer Gerhards | 2009-06-23 | 4 | -79/+57 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | ... hopefully removes false positives (but may cause some trouble with hostname parsing). For details, see this bug tracker: http://bugzilla.adiscon.com/show_bug.cgi?id=126 This patch is not optimal for v4 - another one will follow. The spirit of this commit is to enable easier backporting if someone is interested in doing so. | ||||
| * | | | prevented unneccessary apc calls | Rainer Gerhards | 2009-06-23 | 4 | -11/+19 |
| | | | | |||||
| * | | | restored repeated message reduction processing | Rainer Gerhards | 2009-06-23 | 6 | -39/+69 |
| | | | | |||||
| * | | | fix: previous patch aborted in release mode | Rainer Gerhards | 2009-06-23 | 1 | -1/+3 |
| | | | | |||||
| * | | | some more optimizations | Rainer Gerhards | 2009-06-23 | 5 | -19/+70 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | - done malloc() instead of calloc() for msg_t, as we have large space which needs not be initialized - shrunk syslogTime structure in the hope to get better cache and write performance (non-aligned data should not hurt much here) | ||||
| * | | | optimized processing of TAG message field | Rainer Gerhards | 2009-06-22 | 4 | -38/+17 |
| | | | | |||||
| * | | | some more optimizations of the msg_t object (minor) | Rainer Gerhards | 2009-06-22 | 4 | -45/+40 |
| | | | | |||||
| * | | | reduced memory footprint / "memory leak" | Rainer Gerhards | 2009-06-22 | 2 | -1/+19 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | Testing has shown that at least the glibc malloc() subsystem returns memory to the OS far too late in our case. So we need to help it a bit, by calling malloc_trim(), which will tell the alloc subsystem to consolidate and return to the OS. | ||||
| * | | | optimized template string generation | Rainer Gerhards | 2009-06-19 | 7 | -91/+125 |
| | | | | |||||
| * | | | switched default to keep in line with traditional syslogd behaviour | Rainer Gerhards | 2009-06-19 | 2 | -4/+7 |
| | | | | |||||
| * | | | optimized action.c a bit | Rainer Gerhards | 2009-06-19 | 2 | -24/+21 |
| | | | | |||||
| * | | | removed pthread_testcancel() as it is no longer necessary | Rainer Gerhards | 2009-06-19 | 1 | -22/+0 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | we usually stay long enough inside the actions, so there should be no problem with reaching a cancellation point. Actually, if we really need to cancel, the thread is in an output action (otherwise it would have willingly terminated). | ||||
| * | | | removed uniprocessor optimization | Rainer Gerhards | 2009-06-19 | 7 | -42/+8 |
| | | | | | | | | | | | | | | | | | | | | | | | | ... as it was not even optimal on uniprocessors any longer ;) I keep the config directive in, maybe we can utilize it again at some later point in time (questionable). | ||||
| * | | | typo fix | Rainer Gerhards | 2009-06-19 | 1 | -1/+1 |
| | | | | |||||
| * | | | some cleanup | Rainer Gerhards | 2009-06-19 | 2 | -11/+14 |
| | | | | |||||
| * | | | fully optimized datetime module and enhanced test suite | Rainer Gerhards | 2009-06-19 | 31 | -110/+298 |
| | | | | | | | | | | | | | | | | tests for the various timestamp formats have been added | ||||
| * | | | optimized rfc3339 timestamp string generation | Rainer Gerhards | 2009-06-19 | 1 | -30/+73 |
| | | | | |||||
| * | | | fixed abort condition with oversize tags | Rainer Gerhards | 2009-06-18 | 5 | -23/+16 |
| | | | | | | | | | | | | | | | | this was a regression I introduced this afternoon | ||||
| * | | | optimized handling of MSG part in msg object | Rainer Gerhards | 2009-06-18 | 6 | -16/+33 |
| | | | | | | | | | | | | | | | | | | | | WARNING: currently, message repeation processing is disabled, must be reenabled (but prefer to do some other tests first) | ||||
| * | | | slight optimization of template generation | Rainer Gerhards | 2009-06-18 | 4 | -7/+15 |
| | | | | |||||
| * | | | some optimization in the property replacer | Rainer Gerhards | 2009-06-18 | 1 | -40/+48 |
| | | | | |||||
| * | | | optimized TAG handling in msg object | Rainer Gerhards | 2009-06-18 | 5 | -52/+85 |
| | | | | |||||
| * | | | cleaned up/optimized raw message handling in msg object | Rainer Gerhards | 2009-06-18 | 12 | -84/+77 |
| | | | | |||||
| * | | | Merge branch 'master' into newmsg | Rainer Gerhards | 2009-06-18 | 3 | -5/+22 |
| |\ \ \ | | | | | | | | | | | | | | | | | | | | | Conflicts: tools/omfile.c | ||||
| * | | | | going forward in moving string-handling functions to new interface... | Rainer Gerhards | 2009-06-17 | 13 | -105/+126 |
| | | | | | |||||
| * | | | | some more optimization in msg object string handling | Rainer Gerhards | 2009-06-17 | 2 | -56/+35 |
| | | | | | |||||
| * | | | | further optimized message object | Rainer Gerhards | 2009-06-17 | 6 | -137/+234 |
| | | | | | | | | | | | | | | | | | | | | pri, facility and severity string generation simplified | ||||
| * | | | | added basic plumbing to support message properties separate from message | Rainer Gerhards | 2009-06-16 | 4 | -0/+175 |
| | | | | | | | | | | | | | | | | | | | | | | | | | will fill this with live somewhat later, noticed I need to do some stage work first (at least this is useful). | ||||
| * | | | | some cleanup | Rainer Gerhards | 2009-06-16 | 1 | -6/+4 |
| | | | | | |||||
| * | | | | implemented first version of multi-enqueue support, queue side | Rainer Gerhards | 2009-06-16 | 3 | -2/+128 |
| | | | | | |||||
| * | | | | milestone: input-side multiSubmit capability | Rainer Gerhards | 2009-06-16 | 4 | -11/+64 |
| | | | | | | | | | | | | | | | | | | | | ... commit before I try to touch the queue side ;) | ||||
| * | | | | fixing problems that occurred during the last merge | Rainer Gerhards | 2009-06-16 | 2 | -293/+0 |
| | | | | | |||||
| * | | | | minor cleanup | Rainer Gerhards | 2009-06-16 | 2 | -3/+0 |
| | | | | | |||||
| * | | | | Merge branch 'master' into omfile | Rainer Gerhards | 2009-06-16 | 29 | -502/+699 |
| |\ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Conflicts: ChangeLog runtime/msg.h tcps_sess.c tcpsrv.c tools/syslogd.c | ||||
| * | | | | | cleaned up small nit | Rainer Gerhards | 2009-06-16 | 1 | -1/+1 |
| | | | | | |