epics/epics-base
0
0
Fork 0
Code Issues Pull Requests Projects Wiki Activity
10,529 Commits 8 Branches 203 Tags
e7f78a7d69d8b4ee8488cec2141c08f1535c5e69
Commit Graph

8613 Commits

Author SHA1 Message Date
Andrew Johnson
e7f78a7d69 Fixed problem with previously generated scan.c which was not case-independent. 
Regenerated scan.c, and added instructions for later maintainers.
 2006-03-28 00:11:51 +00:00
W. Eric Norum
51018ac066 Add support for RTEMS tasks with string names. 2006-03-27 22:21:18 +00:00
W. Eric Norum
f648ab5f09 Ensure that RTEMS-style names are properly terminated. 2006-03-27 20:47:05 +00:00
W. Eric Norum
fc2e1fd651 Add infrastructure for spy command. 2006-03-27 16:30:27 +00:00
Jeff Hill
cf231b5af3 fixed repeated sentance 2006-03-27 15:47:49 +00:00
Andrew Johnson
4ee5d5130e Oops, EPICS_CA_BEACON_PORT was wrong... 2006-03-24 20:53:25 +00:00
Andrew Johnson
8e9e846c34 Build the S99 scripts on Linux. 
Modified the scripts and build process to work on Linux.
 2006-03-24 20:29:27 +00:00
Jeff Hill
f4068d6a49 upgraded tests for mantis 245 2006-03-21 00:59:25 +00:00
Jeff Hill
a49961afe6 I fixed the following issues: 
1) The epicsTime constructor does not throw an exception when an invalid epicsTimeStamp is passed to it.
2) The "epicsTime::operator epicsTimeStamp ()" conversion operator should throw an exception instead of assert failing when it discovers that the internal state of the time stamp is invalid.
3) William Lupton's epicsTime::strftime doesnt add .OVF if the nano-seconds field is invalid. Kay's original issue.
4) William Lupton's epicsTime::strftime has a buffer overrun vulnerability.
5) Certain C interfaces to epicsTime were not bracketed by try catch blocks.
 2006-03-21 00:58:43 +00:00
Jeff Hill
f83e11167d accomodate visual c++ 6 2006-03-20 15:53:25 +00:00
Jeff Hill
99d1b3dd04 improved string hash table distribution 2006-03-15 19:02:15 +00:00
Andrew Johnson
df07bfc9e9 Lower-case operands a thru l are never used, the infix buffer was already 
converted to upper-case.
 2006-03-01 19:55:13 +00:00
Janet B. Anderson
c542b4f4d1 Initial version. 2006-03-01 19:44:55 +00:00
Andrew Johnson
14ad2c5055 Modified scan.l.DISTRIB to accept DOS line-endings. Used itself to convert 
this into the scan.c file.  The initscan.c file is not used at all.
 2006-02-28 23:35:22 +00:00
Andrew Johnson
a57f10fb90 Build devLib on all archs now there's a default devLibOSD.c file. 2006-02-17 23:00:30 +00:00
Jeff Hill
0f4acafeae fixed mantis 239 2006-02-17 22:54:49 +00:00
Andrew Johnson
7b383a18b4 Make devLib available from other OSs iff they provide the right table in an 
external support application.
 2006-02-17 22:51:26 +00:00
Jeff Hill
b55729875d fixed mantis 239 2006-02-17 22:33:54 +00:00
Andrew Johnson
a5b4cb0766 PACT must be set during FLNK. 2006-02-15 22:25:24 +00:00
Jeff Hill
dd63c3177d improved test 2006-02-13 16:33:59 +00:00
Jeff Hill
ceb11f1fb4 fix for a potential race in ca_context_destroy 2006-02-13 16:33:46 +00:00
Ralph Lange
85eacf584c Integrating change by Malcolm Waters: fix for "-c -w 0" hanging forever 2006-02-06 16:08:04 +00:00
Ralph Lange
d6b6955903 Integrating change by Malcolm Waters: += put_callback option - heavily based on caget 2006-02-06 16:04:55 +00:00
W. Eric Norum
4074b45a30 Add epicsThreadResume to IOCSH. 2006-02-03 19:36:09 +00:00
Andrew Johnson
ba9499c786 Better solution to zero-sized alloc requests. 2006-01-31 20:44:55 +00:00
W. Eric Norum
dbac42746c [cm]allocMustSucceed() now accept values of 0 for their size and count arguments and can return a NULL in these cases. 2006-01-31 14:54:19 +00:00
Andrew Johnson
10c03d3e19 We took epicsShareAPI out of Win32 already, which was the only arch that 
might have needed it!
 2006-01-05 21:12:32 +00:00
Andrew Johnson
7aa6a2ca98 Better fix to the epicsStrtod problem: Don't compile our function if epicsStrtod 
is a macro; that way you can't set a breakpoint on it which would never fire.
 2006-01-05 21:08:02 +00:00
W. Eric Norum
8d3dc9eb33 Avoid infinite recursion on systems with 'working' strtod() functions. 2005-12-26 18:32:43 +00:00
Janet B. Anderson
b01b09240c Added deprecated EPICS_UPDATE_LEVEL.. 2005-12-20 14:56:46 +00:00
W. Eric Norum
e0887d197e Get rid of a bunch of conditional compile statements. 2005-12-19 22:27:47 +00:00
Janet B. Anderson
147e21b71f Changed EPICS_UPDATE_LEVEL to EPICS_PATCH_LEVEL. 2005-12-19 21:42:40 +00:00
Janet B. Anderson
c57beaad68 Win32 changes. 2005-12-19 21:10:30 +00:00
Janet B. Anderson
45685c52d2 Changed EPICS_UPDATE_LEVEL to EPICS_PATCH_LEVEL. 2005-12-19 21:09:44 +00:00
Andrew Johnson
d9184abd62 Version numbering system change. 2005-12-16 16:47:52 +00:00
W. Eric Norum
a65943481d Improved epicsThreadGetName() so that it works for all threads, not just those 
created by epicsThreadCreate().
 2005-12-15 16:49:42 +00:00
Marty Kraimer
de797d11e2 fix win32 compiler warnings 2005-12-13 12:57:19 +00:00
Janet B. Anderson
a0e61daaac Removed epicsShareAPI from epicsStrtod. 2005-12-12 21:27:07 +00:00
Janet B. Anderson
5506587d06 Removed epicsShareAPI where not needed. 2005-12-12 21:26:01 +00:00
Andrew Johnson
225f816f0e epicsShare macro fixes from Mark Rivers (copied from the main trunk). 2005-12-12 21:25:17 +00:00
Marty Kraimer
cdda915248 o and x format should have unsigned args 2005-12-12 20:09:02 +00:00
Marty Kraimer
b283cb1fcd clean up epicsStdlib 2005-12-12 19:55:11 +00:00
Marty Kraimer
3499fbe51d get rid of epicsShareAPI; epicsStrtok_r is new; clean up epicsStdlib 2005-12-12 19:54:48 +00:00
Jeff Hill
bd14b17c94 fixed send thread blocks trying to destroy the IIU because the 
callback lock was taken without using a callback lock manager
 2005-12-08 16:23:15 +00:00
Janet B. Anderson
e93f9fd86e Updated path rewriting rules for various OSs. 2005-11-30 21:46:13 +00:00
Janet B. Anderson
f8527e3303 WIN32 directory specification changes. 2005-11-29 22:46:57 +00:00
Jeff Hill
5d31c721ad fixed msvc++ 8 assert failure. 
I had a closer look at the MS and the POSIX doc. It appears that
there is confusion between the two related to how the arguments
are interpreted when the user buffer pointer is null. In the POSIX
doc it clearly states that the count only applies if the user pointer
isn't nill. In the MS doc it clearly states that the count is also the
size of the internally allocated buffer when the user pointer is nill.

It seems that a reasonable compromise will be to specify a buffer
size of BUFSIZ (defined by POSIX and fortunately also by windows).
 2005-11-29 17:05:00 +00:00
Janet B. Anderson
f4a06313db Removed TD options in call to _fdopen for VISUAL C++ 2005. 2005-11-22 21:52:29 +00:00
Marty Kraimer
2ad879e1d1 simplify output 2005-11-22 13:45:28 +00:00
Jeff Hill
036243c9a2 made win32 stdcall consistent 2005-11-22 00:28:17 +00:00