Commit Graph

376 Commits

Author SHA1 Message Date
Angus Salkeld
2b2dfd5d2c RB: use the proper struct not the typedef in the implementation.
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-02-08 22:00:50 +11:00
Angus Salkeld
dff3ff4afe RB: Fix potential mem leak
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-02-08 22:00:50 +11:00
Angus Salkeld
205336980c use random() not rand()
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-02-08 22:00:49 +11:00
Angus Salkeld
943e6b301d Remove dead code
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-02-08 22:00:49 +11:00
Angus Salkeld
58c4a80027 set umask before calling mkstemp()
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-02-08 22:00:49 +11:00
Angus Salkeld
7a6382f0b6 Use safer versions of string functions (strcpy -> strlcpy)
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-02-08 22:00:49 +11:00
The Quarterback Library Release Team
362b0dccd4 Bump the lib version to 0.10.0
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-02-07 20:30:48 +11:00
Angus Salkeld
346e4d892a LOOP: handle errors from the poll function
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-02-07 10:16:00 +11:00
Angus Salkeld
b602680e1d LOOP: make the item type applicable to jobs too.
Mainly for diagnostics

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-02-07 10:07:17 +11:00
Angus Salkeld
48e7cff259 LOOP: fix the todo calculations.
The todo system was sucky as it was calculated in different
places in the mainloop and at each level. This was exposed by
calls to qb_loop_level_item_del() which decremented the level->todo
but not the mainloop one. So now we re-calculate it each time.

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-02-07 09:41:00 +11:00
Angus Salkeld
f869f93deb LOOP: prevent jobs from consuming too much cpu.
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-02-07 09:41:00 +11:00
Angus Salkeld
42d20b5090 Get coverity to ignore this warning.
"Using uninitialized element of array"

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-02-03 14:42:25 +11:00
Angus Salkeld
ad4efc2005 LOG: pass the result of qb_log_thread_start() back to the user
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-02-03 14:37:37 +11:00
Angus Salkeld
a95618ddd3 Fix some issues found by clang
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-02-03 14:37:37 +11:00
Angus Salkeld
63bdfd1b71 Add a split timer to the stopwatch.
This is just a re-work of Steve's sample patch.

You set the number of splits and whether they overwrite using
qb_util_stopwatch_split_ctl().

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-02-01 13:46:27 +11:00
Angus Salkeld
86428907c7 IPC: merge common code into new function
This also fixes a regression caused by:
42c92fb675

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-31 11:34:50 +11:00
Angus Salkeld
7d9df9ef74 IPC: better handle a disconnect been called from within connection_created()
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-31 11:34:00 +11:00
Angus Salkeld
55e4c3048a IPC: fix scary typo
Not sure how this happened

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-27 22:28:44 +11:00
Angus Salkeld
42c92fb675 IPC: fix server error handling
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-27 22:27:08 +11:00
The Quarterback Library Release Team
e2bb6d7277 Bump the library version to 0.9.0 2012-01-26 21:52:03 +11:00
Angus Salkeld
e78820b243 PTRIE: refcount the notifier structs
This fixes: https://github.com/asalkeld/libqb/issues/29

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-25 16:09:32 +11:00
Angus Salkeld
27d2b0a882 LOG: make sure qb_log_from_external_source() takes priority into account.
Fixes: https://github.com/asalkeld/libqb/issues/28

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-24 23:08:39 +11:00
Angus Salkeld
36e77bb99d LOG: only bump messages that need it (>info)
else all the messages get their priorities distorted

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-23 16:47:30 +11:00
Angus Salkeld
29d193112a LOOP: allow a timer to be created without returning the handle
This is if the user does not plan to delete the timer.

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-23 16:04:49 +11:00
Angus Salkeld
18bcbef790 IPC: this 1 sec wait is slowing all normal dissconnects down
this works just as well with a timeout of 0

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-23 10:39:31 +11:00
Angus Salkeld
295f7e48bb IPC: check for the server liveness before disconnecting.
if the server is dead then is_connected will cause the resourses
to be properly cleaned up.

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-21 22:05:50 +11:00
Angus Salkeld
e235412497 TESTS: add tests for signal handlers
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-18 22:30:35 +11:00
Angus Salkeld
2f99276d16 IPC: add a context to the client interface
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-18 15:24:34 +11:00
Angus Salkeld
2947797897 LOG: make sure the format is checked.
In case of log clashes (same file/lineno but in different directories.

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-17 21:37:32 +11:00
Angus Salkeld
db29539544 make the pc file auto detect -lrt
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-16 22:33:21 +11:00
Angus Salkeld
477fac4e01 IPC: fix resource cleanup if the server dies
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-12 15:09:58 +11:00
Angus Salkeld
7eaa3dc09a LOG: add %P (pid) as a format option.
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-09 16:03:15 +11:00
Angus Salkeld
b711f73ce7 LOG: add %N (log name) as a format option
This will log the name passed into qb_log_init()

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-09 16:02:53 +11:00
Angus Salkeld
ef64fd1186 UTIL: fix qb_timespec_add_ms()
It was overwriting the timespec value instead of
add to it.

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-09 10:24:08 +11:00
Angus Salkeld
46098d1cbe LOG: make it possible to pass in a NULL filename/function into qb_log_from_external_source()
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-06 10:39:10 +11:00
The Quarterback Library Release Team
e0d55fe8f7 Bump the library version.
Signed-off-by: The Quarterback Library Release Team <quarterback-devel@fedorahosted.org>
2012-01-05 17:46:38 +11:00
Angus Salkeld
e15a48254a ipc: make coverity happy.
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-05 16:46:59 +11:00
Angus Salkeld
f7a08eae5c trie: correct the free'ing of node mem
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-05 16:12:19 +11:00
Angus Salkeld
7ff28b99bc map: free unused leaf nodes
So if a node has no children and no value or notifier
then it is freed.

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-05 13:37:24 +11:00
Angus Salkeld
5d8c96314e Allow the array to automatically grow.
this makes it a bit friendlier to use, as
the user will not have to call qb_array_grow().

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-04 21:40:03 +11:00
Angus Salkeld
fa70ec98e8 IPC: remove fd from poll loop in the disconnect
Until now we have been relying on getting a POLLHUP, but
under heavy load we seem to get an old poll event
that cause a double deref of the connection object.

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-03 13:49:18 +11:00
Angus Salkeld
c3eb1be7ed IPC: add a new state to the connection state
This to handle disconnecting in a failure state (before
the connection has been established.

Another aspect to this is we don't want to call
connection_destroyed() if we haven't called
connection_created().

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2012-01-03 13:49:15 +11:00
Yunkai Zhang
c8e97a1c2e Fixed bug: incorrect array length definition
When MAX_BINS is larger than MAX_BIN_ELEMENTS, this bug will cause
boudary overflow in qb_array_create function.

Signed-off-by: Yunkai Zhang <qiushu.zyk@taobao.com>
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2011-12-29 10:12:07 +11:00
Fabio M. Di Nitto
d823a480dc Fix a compile warning on sparc (epoll_create1)
The problem is that sys/epoll.h was broken on sparc and alpha for a
bunch of glibc releases.

glibc has the symbol correctly exported, that's why ./configure finds
it, but the header is not always correct. Recent versions of glibc have
the correct header, so we simply workaround the broken ones.

This fix is only necessary to build with --enable-fatal-warnings,
otherwise you simply see a build warning, but there are no runtime
issues at any stage.

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2011-12-22 22:55:27 +11:00
Angus Salkeld
ca534c4df8 re-fix qb_strerror_r() in case caller frees buf.
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2011-12-22 14:03:18 +11:00
Fabio M. Di Nitto
9f5d0aebbb Fix some sparc test failures.
I (Angus) have changed some of Fabio's changes.

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2011-12-22 13:56:20 +11:00
Angus Salkeld
ab3dc60f7f Merge some portability changes from the mingw branch
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2011-12-21 14:23:43 +11:00
Andrew Beekhof
61d166abbc Support compilation on Mac OSX
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2011-12-19 22:16:30 +11:00
Angus Salkeld
e6f008cc09 fix strerror on non-linux
Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2011-12-19 22:12:08 +11:00
Angus Salkeld
f0a6ef28ac trie: don't create children array on all nodes
(save mem on leaf nodes)

Signed-off-by: Angus Salkeld <asalkeld@redhat.com>
2011-12-18 22:52:29 +11:00