Building Ruby in IRIX64

[Joe Van Dyk]

Hi,

configure.in in Ruby 1.8.2 specifies that you need autoconf 2.5 to
build the configure script. I only have autoconf 2.13 installed.=20
Does Ruby really need autoconf 2.5 to build the script?

Thanks,
Joe

 

[nobu.nokada]

Hi,

At Thu, 19 May 2005 08:20:58 +0900,
Joe Van Dyk wrote in [ruby-talk:143107]:

configure.in in Ruby 1.8.2 specifies that you need autoconf 2.5 to
build the configure script. I only have autoconf 2.13 installed.
Does Ruby really need autoconf 2.5 to build the script?

Yes.

You can use configure in snapshot tarballs instead.

 

[Joe Van Dyk]

Hi,
=20
At Thu, 19 May 2005 08:20:58 +0900,
Joe Van Dyk wrote in [ruby-talk:143107]:

configure.in in Ruby 1.8.2 specifies that you need autoconf 2.5 to
build the configure script. I only have autoconf 2.13 installed.
Does Ruby really need autoconf 2.5 to build the script?

=20
Yes.
=20
You can use configure in snapshot tarballs instead.

Thanks. I got an updated version of m4 and autoconf, so seeing how it does=

 

[Joe Van Dyk]

Hi,

At Thu, 19 May 2005 08:20:58 +0900,
Joe Van Dyk wrote in [ruby-talk:143107]:

configure.in in Ruby 1.8.2 specifies that you need autoconf 2.5 to
build the configure script. I only have autoconf 2.13 installed.
Does Ruby really need autoconf 2.5 to build the script?

Yes.

You can use configure in snapshot tarballs instead.

=20
Thanks. I got an updated version of m4 and autoconf, so seeing how it do= es.
=20
Just failed!
=20
....
compiling dl
Generating callback.func
Generating cbtable.func
gcc -fPIC -g -O2 -fno-defer-pop -fno-omit-frame-pointer -I.
-I../.. -I../../. -I../.././ext/dl -DHAVE_DLFCN_H -DHAVE_DLOPEN
-DHAVE_DLCLOSE -DHAVE_DLSYM -DHAVE_DLERROR -I. -c dl.c
gcc -fPIC -g -O2 -fno-defer-pop -fno-omit-frame-pointer -I.
-I../.. -I../../. -I../.././ext/dl -DHAVE_DLFCN_H -DHAVE_DLOPEN
-DHAVE_DLCLOSE -DHAVE_DLSYM -DHAVE_DLERROR -I. -c ptr.c
Generating call.func
void
char
short
long
float
double
gcc -fPIC -g -O2 -fno-defer-pop -fno-omit-frame-pointer -I.
-I../.. -I../../. -I../.././ext/dl -DHAVE_DLFCN_H -DHAVE_DLOPEN
-DHAVE_DLCLOSE -DHAVE_DLSYM -DHAVE_DLERROR -I. -c sym.c
gcc -fPIC -g -O2 -fno-defer-pop -fno-omit-frame-pointer -I.
-I../.. -I../../. -I../.././ext/dl -DHAVE_DLFCN_H -DHAVE_DLOPEN
-DHAVE_DLCLOSE -DHAVE_DLSYM -DHAVE_DLERROR -I. -c handle.c
ld -shared -L'../..' -o dl.so dl.o ptr.o sym.o handle.o -ldl
-ldl -lcrypt -lm -lc
ld32: FATAL 12 : Expecting n32 objects: dl.o is o32.
*** Error code 4 (bu21)
*** Error code 1 (bu21)
=20
Any ideas?
=20
Joe
=20

It seems adding -n32 to CFLAGS works. But I shouldn't have to do that, rig=
ht?

 

[nobu.nokada]

Hi,

At Thu, 19 May 2005 08:57:34 +0900,
Joe Van Dyk wrote in [ruby-talk:143115]:

It seems adding -n32 to CFLAGS works. But I shouldn't have to do that, right?

Is -n32 necessary for ld -shaerd?