Michael Davidsaver
32b3eddb94
epicsThreadMustJoin() warn only on double self-join
2019-06-23 19:06:22 -07:00
Michael Davidsaver
86a942872a
epicsThreadTest check explicit join
...
Also remove all the sleeps to improve chances
of catching some kind of race.
2019-06-23 08:12:58 -07:00
Michael Davidsaver
9e5c63fb91
epicsThreadMustJoin() clear joinable flag
...
Clear so that repeated calls will error correctly.
Only well defined for self join.
2019-06-23 07:53:23 -07:00
Michael Davidsaver
5efce9ecc0
epicsThreadJoin() -> epicsThreadMustJoin()
2019-06-23 07:30:48 -07:00
Michael Davidsaver
a64cb4a877
update PVA modules
2019-06-17 17:48:41 -07:00
Andrew Johnson
e8db975e7f
Initial implementation for VxWorks 6.9 and later
...
This almost seems too simple...
2019-06-17 14:33:28 -05:00
Michael Davidsaver
048975ccc7
asLib more string size...
2019-06-05 16:33:55 +02:00
Michael Davidsaver
a83a85af7c
asLib one short
2019-06-04 16:06:07 +02:00
Michael Davidsaver
932e9f3b21
asLib: asUseIP name lookup soft-fail
2019-06-04 15:56:51 +02:00
Michael Davidsaver
72985505ef
further no need to search for perl modules under bin
2019-06-04 11:24:35 +02:00
Michael Davidsaver
52b9c8b947
iocsh further on error
...
Split Scope into iocshContext and iocshScope
to separate per-thread and per-call state.
Error handling is per-call.
2019-05-27 21:24:58 -07:00
Bruce Hill
76506991da
Fixed logic errors and added some prompts.
2019-05-27 20:12:48 -07:00
Michael Davidsaver
655c5bed92
osiClockTime Darwin doesn't use.
...
Darwin/osdTime.cpp unconditionally provides osdTimeGetCurrent()
and never calls ClockTime_Init().
2019-05-20 15:32:57 -07:00
Michael Davidsaver
2c99c36ef9
Merge remote-tracking branch 'lp-Com/timeopt' into 7.0
...
* lp-Com/timeopt:
!gtMoreThanDefault -> useOsdGetCurrent
generalTime short circuit current time
2019-05-20 13:49:10 -07:00
251304e280
use dynamic binding of vxWorks BSP functions because some BSPs don't provide them
2019-05-13 09:43:44 +02:00
0bb80e373b
add simple calculations to macros
2019-05-10 14:44:17 +02:00
4acdd83984
also build pcas
2019-05-10 11:13:23 +02:00
Andrew Johnson
71d4d3ecbd
epicsErrlogTest.c improvements
...
Instrument testEqInt() macro with caller's source line-number.
Added another epicsEvent to wait for the listeners to run.
Not guaranteed to fix all issues, but should be more reliable.
2019-05-09 15:03:56 -05:00
Ralph Lange
8c993405b4
doc/ca: clarify variable size array subscription
...
(cherry-picked from branch 3.15)
2019-05-09 17:32:05 +02:00
Michael Davidsaver
bf8ff4e5f1
!gtMoreThanDefault -> useOsdGetCurrent
2019-05-08 19:05:58 -07:00
Michael Davidsaver
73cdea5517
as-hostname address review comments
2019-05-08 19:03:25 -07:00
Michael Davidsaver
4d5a677239
use iocshSetError()
2019-05-02 20:38:45 -07:00
Michael Davidsaver
eba8a13a2c
iocsh allow setting of error code
2019-05-02 20:38:45 -07:00
Michael Davidsaver
89c269e2d5
iocsh control error behavior
2019-05-02 20:38:45 -07:00
Michael Davidsaver
2e80a97da9
iocsh catch exceptions
2019-05-02 20:29:13 -07:00
Andrew Johnson
a9dbd4dd39
Update libcom and base version numbers, no -DEV
2019-04-23 16:07:43 -05:00
Michael Davidsaver
ccf34ae3d0
update PVD and PVA
2019-04-17 14:17:37 -07:00
Andrew Johnson
f559c36a6f
Update tagged modules: pvDatabase 4.4.2, pvaClient 4.5.0
2019-04-15 11:54:13 -05:00
Andrew Johnson
5a55aa17d9
Update pvData and pvDatabase modules
...
Not updating pva2pva, the next 7.0.2.2 release should
keep the previous tagged version
2019-04-10 16:45:35 -05:00
Andrew Johnson
f0f7a1ef06
Move rules from modules/Makefile to new RULES_MODULES file
2019-04-10 16:16:12 -05:00
Andrew Johnson
e92a9ae426
Make submodule builds more generic
...
Adds PARENT_MODULE to detect when building a submodule.
RULES_TOP uses that to disable uninstall rules and abort instead of
deleting the parent's configure/ directory.
2019-04-10 12:13:33 -05:00
de378eec34
also build pcas
2019-04-03 15:14:34 +02:00
Michael Davidsaver
597393a8ee
libCom: drop CLOCK_MONOTONIC_RAW
...
Turns out this is ~10x slower to query than CLOCK_MONOTONIC
2019-03-28 09:59:24 -07:00
Andrew Johnson
a5b3157ec1
Update version numbers for libcom and database
2019-03-20 15:18:05 -05:00
Andrew Johnson
9a062cd6a1
Update submodules
2019-03-20 15:16:55 -05:00
Michael Davidsaver
cde682f7ba
nonEpicsThreadPriorityTest is Linux only
2019-03-18 16:08:20 -07:00
Michael Davidsaver
1d2637a04e
update submodules
2019-03-18 13:26:47 -07:00
Michael Davidsaver
a2123db9fb
Merge branch 'fix_1816841_only' into 7.0
...
* fix_1816841_only:
Another hack to deal with RTEMS which is POSIX but still different
stripped-down fix for 1816841 only
2019-03-17 18:43:34 -07:00
till straumann
7f55bb0386
Another hack to deal with RTEMS which is POSIX but still different
2019-03-14 11:47:52 -07:00
till straumann
511a541f31
stripped-down fix for 1816841 only
2019-03-14 10:36:55 -07:00
Michael Davidsaver
632d1f45c8
Merge branch 'putf-pact' into 7.0
...
* putf-pact:
RPRO/PUTF log instead of assert()
RPRO/PUTF test all three link types
RPRO/PUTF rename
RPRO/PUTF also handle self link case
fix RPRO/PUTF regression
test for RPRO/PUTF regression
dbRec2Pvt()
Remove Warning: 'blah.PUTF' found true with PACT false
dbNotify set PUTF
2019-03-13 18:04:18 -07:00
Michael Davidsaver
0f75e0aa7f
RPRO/PUTF log instead of assert()
2019-03-13 17:57:41 -07:00
Michael Davidsaver
c93ec231a2
update PVD and PVA
2019-03-11 19:45:37 -07:00
Michael Davidsaver
d1149a0ba9
iocsh epicsMutexShowAll accepts 2 arguments
2019-03-11 17:53:51 -07:00
Michael Davidsaver
95cb81c286
generalTime short circuit current time
...
epicsTimeGetCurrent() is called frequently.
Unless more than the default provider is registered,
locking and going through the timeProvider list each
time is a waste.
Instead, short circuit to directly call the default
time current time provider unless a another has
been registered.
2019-03-10 16:36:54 -07:00
Andrew Johnson
44510f2fb2
Merge 3.15 branch into 7.0
2019-03-07 12:56:57 -06:00
7889dda6f8
merged with upstream
2019-02-20 10:51:02 +01:00
Andrew Johnson
933733465e
Improve timeout diagnostics from netget.plt
2019-02-13 16:02:16 -06:00
Michael Davidsaver
688f32cff0
RPRO/PUTF test all three link types
...
Shouldn't be any difference, but check IN_LINK
and FWD_LINK to be sure.
2019-02-06 16:07:40 -08:00
Michael Davidsaver
d3feb1e2f9
RPRO/PUTF rename
2019-02-04 16:37:20 -08:00