c2b17feaf1
libCom: Missing unlock in epicsThreadShow for RTEMS
...
fixes lp:1628323
2016-09-27 21:36:28 -04:00
9de65e05d4
libCom/osi: fix timespec definition for MinGW
...
(cherry-picked from 3.15)
2016-09-27 15:59:53 +02:00
661058e43e
ca: Fix client bug (race condition) in ipAddrToAsciiEngine
2016-09-02 15:12:50 -05:00
80eaad4118
GeneralTime updates to improve time provider synchronization
2016-08-29 13:04:48 -05:00
064cc94f59
Allow up to 100 network interfaces on Windows.
...
Fixes lp:1606848
2016-07-27 10:14:35 -05:00
6dda3decf9
Merged remove-keywords-3.14 branch
2016-07-07 16:43:29 -05:00
b9aa877727
Fix RTEMS build
2016-07-06 14:06:21 -05:00
8da6c172d1
Removed epicsExceptionTest.cpp
...
Too many build failures on newer C++ compilers. Since this
is just testing that the compiler runtime follows the C++
standard it isn't really necessary any more (it was in the
early days of C++ before standardization).
2016-07-06 11:36:09 -05:00
bf85143381
Fix for g++ 6.x on Fedora 24, cleanup
2016-06-22 12:25:17 -05:00
a5b8b0b890
Use fabs() in calcPerform
2016-05-24 11:12:25 +02:00
f4a16a1e1e
Remaining Revision-Id keywords in source
2016-05-22 05:43:09 +02:00
c03d3eb69a
Removed commit date from CONFIG_BASE_VERSION
2016-05-22 05:41:50 +02:00
3b8fd13152
libCom: NTP Time Provider adjusts to OS tick rate changes
...
Allow the NTP Time provider (used on VxWorks and RTEMS only)
to adapt to changes in the OS clock tick rate after the provider
has been initialized.
Changing the tick rate after iocInit() is not advisable,
other software might still misbehave if initialized before
an OS tick rate change.
Back-ported from the 3.15 branch (commit 12468).
2016-05-17 16:03:27 -05:00
55895e6d66
ellSort: comments and redundant asserts
2016-04-27 07:58:39 -04:00
77d6609bbd
fixup comments
2016-04-26 20:19:19 -04:00
054a234c70
Parentheses in macro
2016-04-28 13:27:47 -05:00
24ddcd2524
Fix epicsTimeZoneTest.cpp => .c and remove conditional
2016-04-27 12:09:10 -05:00
2226f3acb9
libCom/test: don't include epicsTimeZoneTest for vxWorks
2016-04-27 12:45:18 -04:00
f0f89b4b1c
Fix file paths in comments
2016-04-27 10:28:57 -05:00
fdda079b8f
BSD's finite expects a double; make glibc 2.23 happy
2016-04-27 10:28:01 -05:00
b035bd9963
test ellSortStable
2016-04-26 13:57:22 -04:00
40d110fd7d
ellSort
2016-04-26 13:56:57 -04:00
8d6875fe15
Merged changes from the 3.14 branch to revno 12664
2016-10-03 15:21:48 -05:00
4686a062c8
Merged generalTime update from 3.14 branch, revno 12659
2016-09-02 12:47:23 -05:00
25b3eae1c7
Merged changes from the 3.14 branch to revno 12658
2016-09-02 11:54:23 -05:00
e9abac25bc
Merged changes from 3.14 branch to revo 12652
2016-07-21 23:37:54 -05:00
f169781de0
Merged remove-keywords-3.15 branch
2016-07-21 14:12:51 -05:00
4223b31f9a
bswap16/32 as functions
...
Turn bswap16() and bswap32() into functions
to avoid evaluating the argument more than once.
2016-07-08 16:20:35 -04:00
22d6ebe7e1
Merge changes from 3.14 branch up to rev 12642
2016-05-26 13:18:38 +02:00
23fd4e202b
ca: Fix client bug (race condition) in ipAddrToAsciiEngine
2016-05-25 12:26:32 +02:00
b9f0ef5158
Merge 3.14 version and add 3.15 edits
2016-05-22 14:38:18 +02:00
04a9fdb4e3
Clean up warnings from gcc 5.2
2016-05-16 17:02:34 -05:00
0f6c997288
osdNetIntf: ignore 0.0.0.0 as a broadcast address
...
It seems in some situations Linux (at least)
sets IFF_BROADCAST bug leaves the bcast address
set to 0.0.0.0.
Fixes lp:1581505
2016-05-13 09:31:26 -04:00
704c748fbd
Revert "libCom/osi: osiSockDiscoverBroadcastAddresses() finds 127.255.255.255"
...
This reverts commit e805abe971
2016-05-13 09:14:20 -04:00
b8b259de6f
Merge Andrew's libs-prereq branch
2016-04-29 10:42:03 +02:00
7ce8e5ea01
Fix RTEMS Makefile issue
2016-04-28 23:59:39 -05:00
2acde8bac7
Merged changes from 3.14 branch to revno 12630
2016-04-28 21:39:55 -05:00
08fd987c60
Perl script clean-up
2016-04-28 19:00:37 -05:00
5a605fa0c3
Set various EPICS_VERSION_* environment variables
...
These are set at the same time as the ARCH variable.
Had to add EPICS_VERSION_SHORT and EPICS_VERSION_FULL to epicsVersion.h.
This uses more uniform names, should we change epicsVersion.h names?
We would continue to support the old names in parallel for some time.
2016-04-28 18:58:07 -05:00
0e5dc2a21c
More build parameters.
...
Renamed EPICS_TARGET_ARCH to EPICS_BUILD_TARGET_ARCH
Added EPICS_BUILD_OS_CLASS and EPICS_BUILD_COMPILER_CLASS
2016-04-28 18:56:03 -05:00
396ff3c484
Add EPICS_TARGET_ARCH parameter
...
Set its default value automatically in the generated envData.c
2016-04-28 18:49:10 -05:00
c933d77963
Add rule to install libraries before linking executables
...
The rule must be disabled for building libCom and gdd though.
2016-04-26 15:18:31 -05:00
93597e20d5
add epicsTimeZoneTest to libCom test harness
2016-04-26 09:59:22 -04:00
f0453faf36
libCom/test: add test for HST time zone
...
Test US/Hawaii zone which does not observe daylight saving time.
2016-04-26 09:28:09 -04:00
8358580190
Merged commit 12623 from 3.14 branch.
2016-04-15 17:38:41 -05:00
177090e35e
Merged fixes for lp:1558206 and lp:1563191 from 3.14 branch, revno 12622
2016-03-30 11:26:53 -05:00
3179e65791
Clean-up after lp:1558206 fix
2016-03-16 18:15:14 -05:00
6b9bfb09a5
pWaitRelease in wrong place
...
waitRelease==false indicates that pThread
has not be delete'd
2016-03-16 17:43:52 -04:00
4e312b9f64
libCom: exitWait() from thread exit handler corrupts stack
...
The epicsThreadCallEntryPoint() function stores a pointer
to a local variable in epicsThread::pWaitReleaseFlag.
Calling epicsAtThreadExit::exitWait() from that thread's
epicsAtThreadExit() handler writes to this pointer
after epicsThreadCallEntryPoint() has returned.
Thus corrupting the stack.
Set pWaitReleaseFlag=NULL before return to prevent this.
fixes lp:1558206
2016-03-16 14:40:37 -04:00
f73f3332fa
Merge RTEMS NTP fix from 3.14 branch, to revno 12615
2016-02-29 17:49:08 -06:00
Jeff Hill