IRIX and Software

rss-0.7.6 - Page 1

Say Squeen,

How did you manage to circumvent the nagging about ImageMagick 5.5.1 when configuring? The freeware distro is 5.4.x, which configure doesn't like. And did you build it with gcc or MIPSPro?
Here's the funny thing -- I tried this build several times in the past without luck. I even have ImageMagick-5.5.7 unzipped in a directory under my build tree. I just tried building it again when I read your post and (as I now recall) it went into and infinite loop during the ./configure. Something like:

Code: Select all

Xsgi[26354]: -wrapx                 Wrap cursor in X direction
Xsgi[26354]: -wrapy                 Wrap cursor in Y direction
Xsgi[26354]: -stacked               Lay out screens vertically
Xsgi[26354]: -ptrdev <string>       Name of core pointer device
Xsgi[26354]: -xindex <int>          Index of X valuator in pointer device
Xsgi[26354]: -yindex <int>          Index of Y valuator in pointer device
Xsgi[26354]: -kbddev <string>       Name of core keyboard device
Xsgi[26354]: -keymap <string>       Name of keymap to use
Xsgi[26354]: -inputdir <string>     Specifies directory to use for input configuration files
Xsgi[26354]: -configdir <string>    Specifies directory to use for input device configuration files
Xsgi[26354]: -devdir <string>       Specifies directory to use for input device.
Xsgi[26354]: -ddxdir <string>       Specifies directory to use for DDX modules.
Xsgi[26354]: -gamma <string>        Gamma value
Xsgi[26354]: -uid <int>             X server's User ID
Xsgi[26354]: -gid <int>             X server's Group ID
Xsgi[26354]: -reportdecompress      Report names of font files being decompressed
Xsgi[26354]: -solidroot <string>    Name of solid color to initially start root window as.
Xsgi[26354]: -cursorFG <string>     Name of the Cursor foreground color.
Xsgi[26354]: -cursorBG <string>     Name of the Cursor background color.
Xsgi[26354]: -pbufferThreshold <string> Depth of visible framebuffer protected from interference with pbuffers.
Xsgi[26354]: -phantomcursors        Enable the display of phantom hardware cursors
/usr/gfx/gfxinit: Operation not permitted
gfxinit: graphics initialize failed
./config.status[579]: $Xsed:  not found
Xsgi0[26368]:
Xsgi0[26368]: Fatal server error:
Xsgi0[26368]: Couldn't create shared arena
Xsgi[26366]: use: X [:<display>] [option]
Xsgi[26366]: -a #                   mouse acceleration (pixels)
Xsgi[26366]: -ac                    disable access control restrictions
Xsgi[26366]: -audit int             set audit trail level
Xsgi[26366]: -auth file             select authorization file
Xsgi[26366]: bc                     enable bug compatibility
Xsgi[26366]: -bs                    disable any backing store support
Xsgi[26366]: -c                     turns off key-click
Xsgi[26366]: c #                    key-click volume (0-100)
Xsgi[26366]: -co file               color database file
./config.status[579]: $Xsed:  not found
/usr/gfx/gfxinit: Operation not permitted
gfxinit: graphics initialize failed
Xsgi[26366]: -core                  generate core dump on fatal error
Xsgi[26366]: -dpi int               screen resolution in dots per inch
Xsgi[26366]: dpms                   enables VESA DPMS monitor control
Xsgi[26366]: -dpms                  disables VESA DPMS monitor control
Xsgi[26366]: -deferglyphs [none|all|16] defer loading of [no|all|16-bit] glyphs
Xsgi[26366]: -f #                   bell base (0-100)
Xsgi[26366]: -fc string             cursor font
Xsgi[26366]: -fn string             default font name
Xsgi[26366]: -fp string             default font path
Xsgi[26366]: -help                  prints message with these options
Xsgi[26366]: -I                     ignore all remaining arguments
Xsgi[26366]: -logo                  enable logo in screen saver

./config.status[579]: Xno:  not found


I remember getting stuck here once before. Well I just got a fresh copy of rss-0.7.6 source and tried again from a temp build directory. THE RESULT:

Code: Select all

checking magick/api.h usability... yes
checking magick/api.h presence... no
configure: WARNING: magick/api.h: accepted by the compiler, rejected by the preprocessor!
configure: WARNING: magick/api.h: proceeding with the preprocessor's result
configure: WARNING:     ## ------------------------------------ ##
configure: WARNING:     ## Report this to [email protected]. ##
configure: WARNING:     ## ------------------------------------ ##
checking for magick/api.h... no


My settings are:

Code: Select all

setenv CC cc
setenv CFLAGS '-O2 -I/usr/local/include -I/usr/freeware/include'
setenv LDFLAGS '-L/usr/local/lib -L/usr/freeware/lib32 -L/usr/lib32'
setenv PKG_CONFIG_PATH '/usr/local/lib/pkgconfig:/usr/freeware/lib32/pkgconfig'
setenv LD_LIBRARY_PATH '/usr/local/lib:usr/freeware/lib32:/usr/lib32'

./configure --prefix=/usr/local --disable-3dnow --disable-sound

I just went ahead with the "gmake" command and ignored the error message (which I what I usually do went things go wrong).:wink:


Now for the gcc/MIPSPro question : That's a funny thing as well. If you use MIPSPro (I only have a C license, so the C++ is g++ anyways), it will chug along until it hits reallyslick/c_src/euphoria.c and then gmake will bomb with the ERROR

Code: Select all

c-1079 cc: ERROR File = euphoria.c, Line = 278
A type specifier is expected.

float ATTR_ALIGN(16) up[4], right[4], crossvec[4];
^

cc-1065 cc: ERROR File = euphoria.c, Line = 278
A semicolon is expected at this point.

float ATTR_ALIGN(16) up[4], right[4], crossvec[4];
^

cc-1020 cc: ERROR File = euphoria.c, Line = 286
The identifier "up" is undefined.

up[3] = right[3] = crossvec[3] = 0;
^

cc-1020 cc: ERROR File = euphoria.c, Line = 286
The identifier "right" is undefined.

up[3] = right[3] = crossvec[3] = 0;
^

cc-1020 cc: ERROR File = euphoria.c, Line = 286
The identifier "crossvec" is undefined.

up[3] = right[3] = crossvec[3] = 0;
^

5 errors detected in the compilation of "euphoria.c".


Alright. The first time around I went in and removed the offending ATTR_ALIGN which is a macro that expands to

Code: Select all

./src/driver.h:8:#define ATTR_ALIGN(x) __attribute__ ((aligned(x)))
./include/sse.h:22:#define ATTR_ALIGN(x) __attribute__ ((aligned(16)))

which I guess is some kind of gcc built-in function. Now here's the trippy part.... if you edit reallyslick/c_src/Makefile and change line #47 from "cc" to "gcc" and then rerun "gmake" from the command line you'll get:

Code: Select all

source='euphoria.c' object='euphoria.o' libtool=no \
depfile='.deps/euphoria.Po' tmpdepfile='.deps/euphoria.TPo' \
depmode=sgi /bin/sh ../../depcomp \
gcc -DHAVE_CONFIG_H -I. -I. -I../.. -I../../include -I../../src    -O2 -I/usr/local/include -I/usr/freeware/include  -c `test -f 'euphoria.c' || echo './'`euphoria.c
gcc: .deps/euphoria.TPo: No such file or directory
gcc: unrecognized option `-MDupdate'
gmake: *** [euphoria.o] Error 1
holmes4:~/temp/rss-glx_0.7.6/reallyslick/c_src:gmake
source='stringify.c' object='stringify.o' libtool=no \
depfile='.deps/stringify.Po' tmpdepfile='.deps/stringify.TPo' \
depmode=sgi /bin/sh ../../depcomp \
gcc -DHAVE_CONFIG_H -I. -I. -I../.. -I../../include -I../../src    -O2 -I/usr/local/include -I/usr/freeware/include  -c `test -f 'stringify.c' || echo './'`stringify.c
gcc: .deps/stringify.TPo: No such file or directory
gcc: unrecognized option `-MDupdate'
gmake: *** [stringify.o] Error 1


It's crazy, but then just toggle the Makefile line back to 'cc' and run 'gmake' again. You will next bump into a identical problem with lattice.c . Just toggle the Makefile CC varaible back and forth again and it will build too. Too weird , but both executables run fine, and the remainder builds without incident. So to answer your question -- it's both! :)

As for the ImageMagick issue, here's a clue: there used to be a note on the RSS-GLX web site that MatrixView used a function only found in the newer verion of libpng, which caused me some grief on a previous attempt. I even went and got the new libpng and libz, which built fine as I noted in the news blog -- but was later surprised went nothing seems to link to them. My guess, the version 0.7.6 of the source is really a 0.7.6a modified version that removed some of the extra library dependencies.

Sorry, for the long winded answer, but as you can tell it really was just dumb luck this time around and not genius on my part. I am, and probably always be just a hack as this sort of thing, not a pro or even a real serious student. Good thing it's just computers and not rocket science. :wink:
Sheesh, the horrors MIPSPro compiler developers put us though to get our apps compiled... :)

Thanks for the answer, i'll attempt an optimized 0.7.6 with MIPSPro and put back some of the notexture stuff from Lisa back in it. I also try a static build with libbz2 libpng and who knows libimagemagick as well.

Just being masochistic today. Wonder why i'm not out there igniting fireworks :)
Nekochan´s folks is never enough tired to let a compilation session for tomorrow, hhhhuuuuuu???

Really Cool!!!, this bunch is really so much crazy like me!!!

Just let MIPSPRO two seconds!!!, heheheheheh!!!! :lol:

I really appreciate your dedication, guys!!! Nekonoko´s friends are the best!!!

Happy New Year For All !!! :P /
You know dexter1, you were the first person I thought of when I skimpt on the rss build. Sorry about that, and good luck on the complete build!
heh.
Seems like more people than just me are being masochistic today.

I've been trying to find out a way to do commandline file downloads from http/ftp with a base irix install. After some hours I have decided to just make a static compiled CURL and put that along with IRIX-Ports. Hm, got to ask those dudes if it's ok if I do that...

holidays are best for nerding :)
Shall I describe it to you? Or do you want me to get you a box?
Any luck yet dexter1?
As a matter of fact, yes.

I have built ImageMagick 5.5.7-15 statically without perl and managed to get rss-glx0.7.6 compiled with MIPSPro 7.3.1.3m and applied the patch from lisa's 0.7.4. There were only three things, which needed to be fixed in order for it to compile on MIPSPro:

Code: Select all

--- oglc_src/FirePart.h.save    Fri Jan  2 14:22:27 2004
+++ oglc_src/FirePart.h Fri Jan  2 14:23:38 2004
@@ -122,7 +122,7 @@
Particle *p = TblP;     //+1;

n = 0;
-       float da = pow (FIREDA, dt);
+       float da = pow ((float)FIREDA, dt);
//float ds = pow (FIREDS, dt);
//SVector3D v;

@@ -162,7 +162,7 @@
p->s.x = size;
p->s.y = size;
p->s.z = size;
-                               p->a = alpha * pow (FIREDA, (t - LastPartTime));
+                               p->a = alpha * pow ((float)FIREDA, (t - LastPartTime));
p->s *= 0.5f + nrnd (0.5);
}
} else
--- reallyslick/cpp_src/skyrocket_smoke.cpp.save        Fri Jan  2 14:31:13 2004
+++ reallyslick/cpp_src/skyrocket_smoke.cpp     Fri Jan  2 14:31:44 2004
@@ -17,6 +17,7 @@
*/

#include <stdlib.h>
+#include <stdio.h>
#include <GL/gl.h>
#include <GL/glu.h>

--- reallyslick/cpp_src/skyrocket_world.cpp.save        Fri Jan  2 14:31:30 2004
+++ reallyslick/cpp_src/skyrocket_world.cpp     Fri Jan  2 14:31:58 2004
@@ -16,6 +16,7 @@
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
*/

+#include <stdio.h>
#include <math.h>
#include <GL/gl.h>
#include <GL/glu.h>


The code was built on my O2 at work, and haven't had a chance to package it and give it a spin. Will do that as soon as i get to work next monday. I'm attempting to compile it on my crimson though. So hold on...
Fantastic! BTW, No "inifite loop" configure on ImageMagick for you? I would be interested in any and all results. :) Have a good weekend!
Well, all looks good, except for cyclone ofcourse. Lisa's patch was for lattice to run without textures, so everything looks cool, except for sound. I'll bench my stuff along with squeens and see if it really makes a difference. And because everybody has an Octane, i'll make it a -mips4 -Ofast=ip30 -IPA build along with a build -mips3 and more general optimisations.
Fantastic and sorry I haven't had the time to follow through on this. When you finish, please go ahead and replace the tardist on the downloads page.
dexter1 wrote: Lisa's patch was for lattice to run without texture.


Will we still be able to run with textures if desired with the upcoming tardist, or should we keep a TRAM enabled version around as well?
Twitter: @neko_no_ko
IRIX Release 4.0.5 IP12 Version 06151813 System V
Copyright 1987-1992 Silicon Graphics, Inc.
All Rights Reserved.
It seems that for the lattice, Lisa has added an option -T if you want textures. I'm not sure if that means you get testures only with -T or you can specify a texture. I'm gonna try this on an XZ Indy to see if that makes a difference. Otherwise we have to hack a bit more or build an unpatched rss build next to a patched one.
Great; just wanted to be certain as losing texture support completely wouldn't be ideal :)
Twitter: @neko_no_ko
IRIX Release 4.0.5 IP12 Version 06151813 System V
Copyright 1987-1992 Silicon Graphics, Inc.
All Rights Reserved.
Neko, i've upped two tardist for the MIPSPro build of RSS GLX 0.7.6 with lisa's patch and my compile patch in incoming.
The mips3 is a general -Ofast -mips3 -IPA build for all machines
The mips4 is a -Ofast=ip30 -mips4 -IPA build optimized for Octanes, but should run on every R5K and up.
Prereqs is only libbz2, which the package checks for.

Let me know if there are any probs. Oh and Cyclone is still broken :(
hello,

neko is right. 1.7a needs newer gettext.
all versions that include libintl.so.4 are fine.

i know it might not be ideal but gettext from freeware is too old and a bunch of other apps need newer versions to work. my gettext and that one from freeware can exist together but be sure to notice that only one of them must link to libintl.so or set paths strictly when building stuff on your own.

happy surfin :-D

_________________
r-a-c.de
Could you post detailed compiling instructions for Mozilla? To use Mozilla at work I have to compile from the mozilla.org source or use a downloaded binary from mozilla.org or sgi.com. We use MIPSPro C 7.3 and MIPSPro C++ 7.3. I could also get GCC 3.3 from sgi.com. I've tried your compile at home and at a friend's house on SGI O2, see comment on front page of site article. It works ok but it a little slow.
SpiffyIRIX wrote:
Could you post detailed compiling instructions for Mozilla? To use Mozilla at work I have to compile from the mozilla.org source or use a downloaded binary from mozilla.org or sgi.com. We use MIPSPro C 7.3 and MIPSPro C++ 7.3. I could also get GCC 3.3 from sgi.com. I've tried your compile at home and at a friend's house on SGI O2, see comment on front page of site article. It works ok but it a little slow.


hello,

i read your comment and i guess you want firefox.
firefox is mozilla but without all the exrta stuff.
look at mozilla.org for more information.

all the mozilla based apps i provide are mipspro only! (-O3)

runs fine on octanes with r12k and up. o2 should be a little slower but i guess you know you can't compare just the cpu without looking at the whole machine.

gcc builds suck in general :D

_________________
r-a-c.de
GCC at least provides an option for those without MIPSpro.

_________________
- Jim
:Indigo: :Indigo: :Indy: :Indy: <- signed by The Screensavers :) :Indigo2IMP: (230L) (230L) :540: :1600SW: :1600SW: <- touchscreen :PI: :Octane2:
directedition wrote:
GCC at least provides an option for those without MIPSpro.


sure, but he has both.
this advice was just for him.

_________________
r-a-c.de