EPICS on RTEMS 4.9 (for mvme2100 and mpc8313erdb)
ralf.corsepius at rtems.org
Fri Oct 3 02:16:09 UTC 2008
On Thu, 2008-10-02 at 15:39 -0400, Tom Cheng wrote:
> After a few trial, I got the EPICS 3.14.9 compiler forward a little,
> and new error came out. I guess this is related to the size_t issue.
> Here is the error message:
> /usr/bin/g++ -c -D_POSIX_C_SOURCE=199506L -D_POSIX_THREADS
> -D_XOPEN_SOURCE=500 -D_X86_ -DUNIX -D_BSD_SOURCE -Dlinux
> -D_REENTRANT -ansi -O3 -Wall -g -fPIC -I. -I..
> -I../../../include ../disconnectGovernorTimer.cpp
> /usr/include/string.h:296: error: expected identifier before numeric
> /usr/include/string.h:296: error: expected ',' or '..' before numeric
> /usr/include/string.h:296: error: declaration of C function 'void
> memset(void*, int)' conflicts with
> /usr/include/string.h:59: error: previous declaration 'void*
> memset(void*, int, size_t)' here
> ../nciu.h:124: warning: 'class privateInterfaceForIO' has virtual
> functions but non-virtual destructor
> make: *** [disconnectGovernorTimer.o] Error 1
> make: Leaving directory
> make: *** [install.linux-x86] Error 2
> make: Leaving directory `/home/ytian/EPICS/base-3.14.9/src/ca'
> make: *** [ca.install] Error 2
> make: Leaving directory `/home/ytian/EPICS/base-3.14.9/src'
> make: *** [src.install] Error 2
> The g++ complained about the string.h file. Here is the line 296
> of /usr/include/string.h:
> extern void bzero (void *__s, size_t __n) __THROW __nonnull ((1));
> Could this be caused by the size_t definition ?
Likely, something is trying to redefine bzero to memset and fails on
However, this is your native compiler trying to compile some non-RTEMS
code (EPICS's code). I don't see how this is related to RTEMS at all.
It's likely a bug in EPICS.
More information about the users