04ff9eb8a1
rec: mbbiDirect and mbboDirect refactoring
...
This includes changes to functionality, documented.
2012-12-11 18:25:01 -06:00
475fc8e122
rec: Code format cleanup to mbbi, mbbo records
...
No intentional changes to functionality.
2012-12-11 18:23:17 -06:00
b7018dd16d
ca: Fix bug in "Firewalls" section of Channel Access reference manual
2012-12-11 23:38:42 +01:00
5a498e26a1
Added Undefined Severity field UDFS
...
With this, records can be configured to have a lower
alarm severity when they are undefined.
Sets initial severity at iocInit too.
2012-12-10 09:25:58 -06:00
85ed3b9f4e
rec: Delete ununsed fields from mbbxDirect record types
...
LALM was being initialized but is not otherwise used.
2012-12-07 22:15:48 -06:00
319ed9b394
rec: Delete ununsed fields from mbbxDirect record types
...
SDEF removed from mbbiDirect
LALM removed from both mbbiDirect and mbboDirect
2012-12-07 17:20:38 -06:00
79bc3f2265
Make blockingSockTest use a free port number
...
Instead of using the CA server port 5064 which is often
in use, have the listening socket pick a free port.
2012-12-06 14:31:50 -06:00
ef5688f8de
libCom: Accept hex literals in CALC expressions
...
Code back-ported from the 3.15 branch.
2012-12-03 12:58:26 -06:00
d622bbad05
Fix compiler warnings on linux-x86
2012-12-03 12:48:11 -06:00
cbde3ea9fd
cas: Added propertyEventMask support
...
Allows server tools to send DBE_PROPERTY events to clients.
New functionality not tested, but fairly trivial and doesn't
break existing server tools.
2012-11-30 17:42:47 -06:00
fc12c72809
Device support for new long string out record type
2012-11-30 16:51:37 -06:00
b64bc41d6b
Add new long string output record type.
2012-11-30 16:50:31 -06:00
080e362743
Cleanup.
2012-11-30 16:49:31 -06:00
01819b0b02
Use new helper.
2012-11-30 16:48:58 -06:00
f2716f88f3
stdRecords.dbd depends on rec/Makefile
2012-11-30 16:47:11 -06:00
d741c10635
Another helper, dbLoadLinkLS
...
Replaces recGblInitConstantLink(..., DBF_STRING, ...)
2012-11-30 16:43:16 -06:00
6f9c2db8a8
libCom: Revert 12374 which breaks MSVC++ builds
...
Both the Microsoft 2008 and 2010 compilers fail with this change.
2012-11-30 13:32:25 -06:00
e01a428059
Added new Long String Input record type "lsi"
...
With Soft Channel device support.
2012-11-29 13:45:13 -06:00
77c837bcbe
printf record: Fix bug in minimum buffer size logic.
2012-11-29 13:41:46 -06:00
02348e770b
Use new helper function.
2012-11-29 13:40:42 -06:00
d27747a7f3
Added helper functions dbGetLinkLS() and dbPutLinkLS()
2012-11-29 13:21:11 -06:00
2e4f6443e0
dbNameToAddr() changes for long string support
...
Modified dbNameToAddr() to match how dbChannelOpen() calls
prset->cvt_dbaddr() before handling the '$' field modifier.
This lets DB links and the dbgf/dbpf commands work with '$'.
Elsewhere changed array handling to check fldDes->special
for SPC_DBADDR instead of paddr->special to allow long string
records cvt_dbaddr() routines to change paddr->special
without breaking the array handling stuff.
2012-11-29 13:20:02 -06:00
493d6014b9
Add new global menuPost
...
To reduce duplication.
2012-11-29 12:53:21 -06:00
01276d952f
ca: Build caRepeater etc. by default
...
Exceptions for vxWorks, RTEMS and iOS which don't support
executable programs like the workstation OSs.
2012-11-26 15:37:54 -06:00
d69f3904d7
Fix for win32 builds.
2012-11-20 17:27:38 -06:00
7778c0c3ff
Solaris & vxWorks 5.5.x fixes
...
Remove C++-isms.
Make internal routines static when duplicate names appear
in the epicsMutexTest.cpp code; on vxWorks and RTEMS all
test programs get linked into a single binary.
2012-11-20 15:55:53 -06:00
f320cbcecd
Fix vxWorks implementation.
2012-11-20 15:04:28 -06:00
6fd9fb3066
libCom: Removed epicsSingleton workaround
...
This undoes commit 11565, "workaround for archaic T202 g++ bug"
which doesn't actually seem to be necessary for T202 compilation.
It also replaces 'typename' with 'class', for another compiler.
Modification provided by Dirk Zimoch.
2012-11-16 16:14:13 -06:00
36be0aa783
rec: Fix minor string in + out issue.
...
Undo strncmp+strncpy to strcmp+strcpy change in r12048.
The 'n' versions are needed because device support was
never explicitly required to Nil-terminate the string.
The dbAccess routines do enforce termination though.
2012-11-16 12:40:57 -06:00
694ac0a650
epicsTime::addNanoSec function size is large now so it shouldnt be inline
2012-11-14 13:10:44 -07:00
5ce291facf
fixed spelling in comment
2012-11-14 13:03:30 -07:00
b86c4107c1
After qptimizing the epicsTime::addNanoSec function we now have a larger code which uses only unsigned integer arithmetic. This is for the benefit of embedded cpu's lacking a hardware floating point coprocessor at the expense of some additional code to maintain. I measered the cpu load improving from 40% free to 80% free after making this change, on a nios2 soft core processor.
2012-11-14 12:51:30 -07:00
e3f59fd5c2
Support variable array size
...
Added LEN field and get_array_info() routine.
Removed len arg from device support write_string() routine.
Modified device support to match.
2012-11-12 13:45:23 -06:00
b8eec29e43
libCom: Added errlogSetConsole() functionality.
2012-11-09 17:06:32 -06:00
fc61cc34ce
iocLogServer fix for unsigned char arch's
...
Reported by Guy Jennings
2012-10-31 13:56:12 -05:00
3f18327d2d
libCom/osi: add tests for epicsSpin (based on epicsMutex tests)
2012-10-29 20:26:11 +01:00
a977edfaa6
libCom/osi: add epicsSpin with default, posix, and vxWorks implementations
...
- posix uses pthread_spin_ interface when supported, pthread_mutex_ otherwise
- default uses epicsMutex
- vxWorks (single core) uses intLock()
2012-10-29 20:25:23 +01:00
c06f3cd9a0
db: Fix dbGet so attributes and value can be fetched simultaneously
...
Only the Async Soft Channel support did this, so that's where it
was noticed that this didn't work.
2012-10-29 01:20:56 -05:00
35a2793c23
rec: Cleanup of compress & histogram record source code
...
No functional changes.
2012-10-19 15:06:37 -05:00
c9da766e9e
rec: Fix compress & histogram attribute bugs
...
Problems introduced during Dirk's non-val attribute changes.
2012-10-19 15:01:43 -05:00
7f21c8eec1
Merged changes from 3.14 branch
...
Merged up to revno 12369 on 2012-10-17
2012-10-18 00:10:12 -05:00
e889336458
db: Undefined alarm limits should return NaN
...
Where the record doesn't provide get_alarm_double() or for fields
where the default applies, we were still returning 0.
2012-10-17 18:08:39 -05:00
2613c62b40
rec/aSub: Add cleanup subroutine
2012-10-15 12:35:04 -05:00
7cac267a4b
db: Fix possible error handling in dbPut
2012-10-11 18:10:00 -05:00
71079ede4b
rec/aSub: Fixes from Michael Davidsaver
...
* Changing the number of elements in a VALx array triggers monitors
* Post monitors on the NEVx fields, following EFLG
* Remove artificial limit on array sizes
2012-10-11 17:51:00 -05:00
ab7d73c537
cas: Fix syntax error in last commit
...
errMessage() is not printf-like, use errPrintf() instead.
2012-10-11 11:28:28 -05:00
7d37c31cc5
added the PV name to the unable to read application types "enums" message
2012-10-10 19:08:18 -06:00
26173c1e0d
devLib: add compatibility typedef for VME table type
2012-10-10 15:52:34 -05:00
9c1b8ba952
rec: Fix problem with NAN in MLST/ALST fields
...
If MLST or ALST became NAN the associated monitor would never trigger.
Now a NAN will cause it to always trigger.
2012-10-06 21:48:01 -05:00
68bde36d75
Merged changes from 3.14 branch.
...
Merged up to commit 12364 (2012-09-11) inclusive.
2012-10-01 00:54:10 -05:00
Andrew Johnson