Michael Hamlin <hamlin_at_pa.msu.edu> writes:
>
> Here's a follow up with more information to anyone who can help.
> Again, please copy any replies to hamlin_at_pa.msu.edu.
>
> pavo:~/bin/fvwm-2.2.5>uname -a
> SunOS pavo 5.7 Generic_106541-11 sun4u sparc SUNW,Ultra-1
> pavo:~/bin/fvwm-2.2.5>gcc -v
> Reading specs from
> /usr/local/lib/gcc-lib/sparc-sun-solaris2.5.1/2.7.2.2/specs
> gcc version 2.7.2.2
>
> I wouldn't know what is interesting enough to post to the list; except
> perhaps for one thing:
> fvwm-2.2.5 managed to finish the configure, but fvwm-2.3.32 did NOT.
>
> configure: failed program was:
> #line 5031 "configure"
> #include "confdefs.h"
> #ifdef HAVE_SYS_TIME_H
> #include <sys/time.h>
> #endif
> #ifdef HAVE_SYS_TYPES_H
> #include <sys/types.h>
> #endif
> #ifdef HAVE_UNISTD_H
> #include <unistd.h>
> #endif
> #ifdef HAVE_SYS_SELECT_H
> #include <sys/select.h>
> #endif
> #ifdef HAVE_SYS_SOCKET_H
> #include <sys/socket.h>
> #endif
> int main() {
> extern select (int,
> fd_set *, fd_set *, fd_set *,
> struct timeval *);
> ; return 0; }
>
> I went and looked in the header files to find the source of the errors.
>
> /usr/include/sys/siginfo.h:74: parse error before `pthread_attr_t'
> /usr/include/sys/siginfo.h:74: warning: no semicolon at end of struct or
> union
> /usr/include/sys/siginfo.h:76: parse error before `}'
>
> 063: /*
> 064: * The sigevent structure is also defined in <time.h> as per X/Open
> 065: * and POSIX requirements.
> 066: */
> 067: #ifndef _SIGEVENT
> 068: #define _SIGEVENT
> 069: struct sigevent {
> 070: int sigev_notify; /* notification mode */
> 071: int sigev_signo; /* signal number */
> 072: union sigval sigev_value; /* signal value */
> 073: void (*sigev_notify_function)(union sigval);
> 074: pthread_attr_t *sigev_notify_attributes;
> 075: int __sigev_pad2;
> 076: };
> 077: #endif /* _SIGEVENT */
>
>
> /usr/include/sys/socket.h:47: parse error before `sa_family_t'
> /usr/include/sys/socket.h:47: warning: data definition has no type or
> storage class
>
> 045: #ifndef _SA_FAMILY_T
> 046: #define _SA_FAMILY_T
> 047: typedef uint16_t sa_family_t;
> 048: #endif
>
>
> Appears as a syntax issue in the first one (semicolon to end the union
> definition BEFORE the close parenthesis?). The second looks like a
> missing definition of "uint16_t". The next error in socket.h is caused
> by similar missing symbol "uint32_t", and later errors appear to be
> problems with these data typedefs that failed in the first two errors.
>
> There follows many pages of the same "parse errors" in siginfo.h
> and socket.h, every time configure tries to compile something.
>
> I went back and configured fvwm-2.2.5 from scratch and examined the
> log, and found the -same- errors.. so why was 2.2.5 able to build,
> but not 2.3.32 ?
>From what I can tell, the configure tests have been upgraded
between 2.2.5 and 2.3.32 with some fixes from the
autoconf maintainers. It looks to me like they
are working better in the 2.3.32 series.
I tried compiling the code you show above on one of our Solaris 7
machines and it compiles without a problem. I think it should
on your machine also.
I made it look like this:
#include <sys/time.h>
#include <sys/types.h>
#include <unistd.h>
#include <sys/select.h>
#include <sys/socket.h>
extern select(int, fd_set *, fd_set *, fd_set *, struct timeval *);
int main() {
; return 0; }
Maybe your feature tests aren't returning all the things they
should be, I looked in my config.h to see which of the flags were
defined.
Right now, I'm out of ideas. Maybe someone else on the list
has an idea?
--
Dan Espen
444 Hoes Lane Room RRC 1C-214 E-mail: dane_at_mk.telcordia.com
Piscataway, NJ 08854 Phone: (732) 699-5570
--
Visit the official FVWM web page at <URL: http://www.fvwm.org/>.
To unsubscribe from the list, send "unsubscribe fvwm" in the body of a
message to majordomo_at_fvwm.org.
To report problems, send mail to fvwm-owner_at_fvwm.org.
Received on Fri Jun 01 2001 - 19:14:03 BST