The collected works of squeen - Page 7

Can you grep for -L../../dist/bin in the Makefile.in or configure scripts and find the flag before it?

nekonoko wrote: Okay - builds now, but just cores on startup.

Doh! I hate that.

Interesting. I haven't really gotten into OpenMP much, but that good to know.

Neko, fantastic!

Second-hand Octane on ebay......................................$300
mencoder and ffmpeg from nekoware................................$0
Nekochan.net annual fees....................................................$0

Making cross platform movies from IRIX.............*priceless*.


today is a good day
thanks all

hamei wrote: ...No one can do worthwhile work with old SGI junk so just give it up and move on.

We have this on the highest authority.

Aye! Aye! Sir!

In a related note I under went an MRI scan yesterday. The OS running the system was IRIX (Onyx?)!
Just tell me what other OS you'd trust your brain to?

Cheers.

In case the last post confused anyone, unixmuseum is letting me know he works in aerospace.

MLI = multi-layer insulation or "blankets"
Kapton = kapton tape (in fact HST uses silverized teflon---kapton is the yellow stuff)

Either way, despite the progress of the past year (leaving Performer in May and writing a C base scene graph library from scratch & moving from IR4/IRIX to nVidia Quadro FX/Linux AND completely rewriting the visualization app...*whew*) there's still quite a lot missing from the model and the viewer. The immediate goal was decent frame rates with dynamic environment (cube) mapping, shadow mapping and per pixel lighting all playing nice in the scene. (5 fps...ugh!). Right now the shaders are the bottle neck. Hopefully the newer nV cards will help.

As for model improvements, stay tuned...

That said. I apologize for cluttering the nekoware thread with a tangent. What I really was psyced about was how nicely the nekoware media tools (ffmpeg and mplayer) were working for animation creation. I'm gonna try the DVD authoring tools in the near future as well.

Minor quibble on mencoder and (?) ffmpeg is that they seemed to have dropped SGI RGB image format support, so I'm stuck converting the frames to jpeg (which takes the most time).

Again thanks to the nekoware porters for putting those tools at my finger tips (before I even knew I needed them!).

@neko: I just tickles me to death that the version of Firefox I run under IRIX is more recent than all the Red Hat installations. You rock! (It also cracks me up I can't get SuperTux to run under Linux, but IRIX and Mac work just fine.)

Happy holidays!

Yes. The mp4 created by ffmpeg preserves colors better than mencoder (avi container) and has the advantage of playing nicely on IRIX, Linux (w/ mplayer) and OS X (w/ Quicktime). Neither ffmpeg nor mencoder created files played using window media player (on OS X, I don't have a Window machine).

A weirdness though on IRIX & Linux mplayer playback of the .mp4 created with ffmpeg using HDTV 1280x720 sized frames. The playback window is stretched. xwininfo report 1536x720. Why? Quicktime plays it back at the proper aspect. Weird.

Another weirdness:
neko_xscreensaver-demo now crashes and complains

screensaver-demo
Error: Cairo does not yet support the requested image format:
Depth: 32
Alpha mask: 0x00000000
Red mask: 0x000003ff
Green mask: 0x000ffc00
Blue mask: 0x3ff00000
Please file an enhacement request (quoting the above) at:
http://bugs.freedesktop.org/enter_bug.cgi?product=cairo
Assertion failed: NOT_REACHED, file cairo-image-surface.c, line 155, pid 4877
Abort (core dumped)
holmes4:~:
X error in fliptext:
X Error of failed request: BadWindow (invalid Window parameter)
Major opcode of failed request: 3 (X_GetWindowAttributes)
Resource id in failed request: 0x38000c9
Serial number of failed request: 12
Current serial number in output stream: 13

I've got neko_cairo-1.2.2 installed. Don't know why this started happening. I don't think I messed with software lately.

Ho ho! Good catch Neko. I have the desktop set to 24bpp, but...like we had in SGI freeware's gtk+ once long ago, the Octane is capable of 32bpp, so some dumb Linux apps just grap the "best" and then assume it's 24bpp. Cario must have been fallen victim to sloppy Linux programming. A patch is probably in order.

schleusel wrote:

squeen wrote: A weirdness though on IRIX & Linux mplayer playback of the .mp4 created with ffmpeg using HDTV 1280x720 sized frames. The playback window is stretched. xwininfo report 1536x720. Why? Quicktime plays it back at the proper aspect. Weird.

Hmm, do those files contain bogus aspect information for some reason? (Mplayer should print the aspect ratio it thinks the file has). You can use -noaspect to prevent mplayer from honoring such aspect information or -aspect to override it.

-noaspect fixed it thanks! It complianed some during the encoding but since the original frames (just jpegs) were corrrect movie aspect I don't know how bad aspect info got in. I'll have to learn more of the ffmpeg interface. The command I used was:

ffmpeg -r $FPS -b $bitrate -i frame%05d.jpg output.mp4

Awesome. Thanks!

Brombear is right. The game used OpenGL Shading Language with IRIX graphics don't support. You'd have to whack that part of the code (for the water it looks like).

See also the wiki page .

OpenSSL SSL_get_shared_ciphers Function Buffer Overflow Vulnerability

-------------------------------------------------------------------------------------

Alert Type :VULNERABILITY ALERT

Threat Type :Unintended Weakness:Buffer Overflow

IntelliShield ID :11788
Version :23

Urgency :2 - Unlikely Use

Credibility :5 - Confirmed

Severity :4 - Moderate Damage

CVSS Base Score :10.0

CVSS Temporal Score :7.4

CVSS Vector :AV:R/AC:L/Au:NR/C:C/I:C/A:C/B:N/E:U/RL:O/RC:C

First Published :Sep 28, 2006; 03:56 PM EDT
Last Published :Jan 18, 2007; 09:54 AM EST

Ports :Not Available
CVE :CVE-2006-3738

Version Summary
-------------------------------------------------------------------------------------
Oracle has released the January 2007 Critical Patch Update and updates to address the buffer overflow vulnerability in the SSL_get_shared_ciphers function of OpenSSL.

Description
-------------------------------------------------------------------------------------
OpenSSL versions 0.9.7k and prior and 0.9.8c and prior contain a vulnerability that could allow an unauthenticated, remote attacker to execute arbitrary code with elevated privileges.

This vulnerability exists due to insufficient boundary checking on user-supplied input. An unauthenticated, remote attacker could exploit this vulnerability by sending a crafted request to an application or system service depending on the OpenSSL library. An exploit could trigger a buffer overflow condition, allowing the attacker to crash the affected service or execute arbitrary code with privileges of the target application or service.

OpenSSL confirmed this vulnerability in a security advisory and released updated versions.

Impact
-------------------------------------------------------------------------------------
An unauthenticated, remote attacker could exploit this vulnerability to execute arbitrary code with privileges of the affected service or application. An exploit could allow the attacker to gain complete control of the affected system.

Warning Indicators
-------------------------------------------------------------------------------------
OpenSSL versions 0.9.7k and prior and 0.9.8c and prior are vulnerable.

Technical Information
-------------------------------------------------------------------------------------
This vulnerability exists because the SSL_get_shared_ciphers() function fails to properly limit input before use in memory operations. An attacker could exploit this vulnerability by sending a malicious set of large, crafted ciphers to an exposed application depending on the OpenSSL library and then using the affected function. An exploit could trigger a buffer overflow condition, allowing an attacker to crash the affected application or execute arbitrary code with privileges of the affected application or service.

IntelliShield Analysis
-------------------------------------------------------------------------------------
Systems that allow unfiltered user input to system applications and services depending on the vulnerable OpenSSL library are at a greater risk. An attacker can exploit this vulnerability by sending a crafted request to an exposed service linked to the OpenSSL library. Statically linked applications and services may require recompilation, while dynamically linked applications and services will simply require a service or system restart. Vulnerable services may include mail, web, and database services using OpenSSL to encrypt network communications.

Safeguards
-------------------------------------------------------------------------------------
Administrators are advised to apply the appropriate software updates.

Administrators are advised to restrict access to trusted users.

Administrators may consider employing network filtering devices to block malformed requests to effected systems.

Oh right. Sorry, my mistake as well.

Woot! Woot!

Thanks Neko.

Thanks neko.

Thanks jimmer.

One comment on the IRIX download. The tardist format is already compressed, so I imagine the size difference by making a tardist.gz in minimal.

I had a wierd problem with the DCD V12 and my Octane2. With a regular DVI cable (no dual link) and the SGI (Sony) 23" LCD monitor I could run the Octane at 1920x1200@75 Hz (xsetmon reported). However I later plugged in a new HP 23" LCD, it wouldn't support the resolution---it runs at 1920x1200@50 Hz on a Linux box. The Octane did not have a slower video format for 1920x1200, so I switched the monitors back, which is unfortunate since the HP has two DVI-I inputs and the SGI/SOny only has a DVI and an analog.

VFC is my only solution?

Thank all for the info. I'm out of the office this week, but when I get back I'll verify the full list of formats I have, and perhaps I'll take hamei's up on his offer. I am only running 6.5.29 right now.

Does anyone remember the steps to set up a dual interface IRIX box as a gateway for other systems?
Something about routed...

Thanks.

awesome thanks!

Ack.

I've got the Tzero with 2 Gig-E card working on both networks, but I can't get it to route packets from machines on the private net. I found my notes and followed the same procedure as I did on my Octane many years ago to get it work, but no dice!

My is /etc/config/routed.options

-h -Prdisc_interval=45 -q

The /etc/config/netif.options has (there are three interfaces actually, but I'm just using tg0 and tg1)

if1name=tg0
if1addr=$HOSTNAME
if2name=tg1
if2addr=gate-$HOSTNAME
if3name=tg2
if3addr=gate2-$HOSTNAME

I've chkconfiged gated on and rebuilt the kernel with /etc/autoconfig -vf.

I even get from netstat -s -p ip |grep forward

86 packets forwarded (forwarding enabled)
0 packets not forwardable

But the machines behind the firewall just sit and spin. The ugly part is I configured a Linux box as the gateway previously and all the firewalled machines could get web access OK (but were slow). I figured--based on my experiences with the Octane--that the Tzero would make a much better gateway.

Help!

From the Tzero with the IP address modified to protect the innocent

From the Linux box behind the Tzero


networking has never been one of my strengths, so your help is greatly appreciated

Boy. Lately, nothing seems to be working for me.

I ran vncserver on the IRIX box and x2vnc on a Linux box, but even though the x2vnc client prompts me for the desktop password and says I've connected, and the pointer disappears when I go off the edge of the screen. There is no pointer motion on the Tzero.

Here's the Xvnc log on the Tzero:

22/06/07 16:43:28 Got connection from client 192.168.10.101
22/06/07 16:43:28 Protocol version 3.3
22/06/07 16:43:31 Full-control authentication passed by 192.168.10.101
22/06/07 16:43:31 Pixel format for client 192.168.10.101:
22/06/07 16:43:31 32 bpp, depth 24, big endian
22/06/07 16:43:31 true colour: max r 255 g 255 b 255, shift r 16 g 8 b 0
22/06/07 16:43:31 no translation needed
22/06/07 16:43:31 Using hextile encoding for client 192.168.10.101

See anything fishy? Man I'm been batting zero lately with this Tzero!

Still trying to ditch one keyboard & mouse.

Here my latest puzzler:

I can run the vncserver on the Tzero and vnviewer into it from the Linux box...
-OR-
I can run the vncserver and vncviewer into it from the IRIX box....
-BUT-
I can't get x2vnc to take over the mouse or keyboard (accept input) in either direction, even though the the client logs in fine, the server log shows a connection, and the pointer disappears when it goes off the east (right) edge of the screen on the client.

Worse still, I've gotten this working many times before with my Octane. Yeesh!

Any ideas? Does the resolution of the two screen have to match?

I just tried the x2vnc software between Linux and IRIX again from my Octane--everything connects but I still get zero mouse/keyboard activity on either system. Definitely not the way I remember it working.

However I got a small ray of sunshine. I installed x2x on the Linux box and did the following via SSH (with X11 forwarding enabled on the Linux ssh deamon)

That's it. It working fine the other way as well if you build x2x on the IRIX machine. The USB mouse (w/ wheel) works in IRIX too.

I did an

at some point as well, but I don't think it was needed.


Peace at last!

Ack! I really need to resolve this one. Help me in my ignorance. I'll beg!

Here's the instructions in IRIX Admin: Networking and Mail Chapter 3

Configuring a Router With Two Interfaces

The /etc/init.d/network script is designed to automatically detect and configure a router with two interfaces if the default naming scheme for the interfaces is used. By default, the Internet addresses of the primary and secondary interfaces are derived from the /etc/sys_id file. The primary interface uses the name in the sys_id file. The secondary interface prefixes gate- to the name specified in the sys_id file.

To set up a router with two interfaces using the default naming scheme, follow this procedure:

1. Log in as root.

2. Assign valid Internet names and addresses to both interfaces in the /etc/hosts file. For example, the /etc/hosts file entries for the primary and secondary interfaces on the station biway might look like this:

198.70.75.2 biway.salad.com biway
198.70.80.3 gate-biway.salad.com gate-biway

3. Ensure that the router has the appropriate name in its /etc/sys_id file. Following this example, the /etc/sys_id file should look like this:

biway

4. Reconfigure the kernel and reboot the station to initialize your changes and interfaces. Some systems prompt you for permission, as in the following example. Others simply return a shell prompt. In either case, enter the reboot command when the kernel has been reconfigured:

/etc/autoconfig

Automatically reconfigure the operating system? (y/n)y

reboot

Which I did. My /etc/hosts looks like


The contents of /etc/config/netif.options are


I also have a static-route set up in /etc/config/static-route.options


Right now the Tezro see both networks and can ping, SSH, HTML etc. just fine. I have

routed on
gated off

and told the network (through the System Manager GUI) to forward IP packets .

From the private network (Linux and OS X boxes) I can ssh into heimdall (the gateway) just fine. They are set up to use a default route of 192.168.10.1 (again, the Tezro gateway).

If I traceroute from the Tezro (gateway) to our Onyx server (eno) in another building I get:



however, from the linux box behind the gateway I get

so I goes through the gateway to the gateway's default route and then stops.

I had a Linux box gateway working using the commands


Help!

@foetz: No the didn't remove the localhost entry, but the netstat output is exactly as I reported.


Also, I tried


but all I get is