Archived discussion

Test harness problems?

Other usage issuesWed, 12/29/1999 - 23:295 replies

QuestionAuthorWed, 12/29/1999 - 23:29

Has anyone else tried running the "Draw" test harness program? When I try to run the demo program as stated in the sourceREADME file, I get the following output in the text window: (used "script" command to capture")

Script started on Wed Dec 29 15:34:30 1999

[root@bobslap1 DrawResources]# ./demo

Draw[1]> Draw[2]> axo

Draw[3]> source demo.tcl

Creating a box

Draft of two lateral faces

Fillet on four lateral edges, then on the top and bottom edges

tolerance ang : 0.01

tolerance 3d : 0.0001

tolerance 2d : 1e-05

fleche : 0.001

tolblend 0.01 0.0001 1e-05 0.001

EXCEPTION Stripe compute 0x42748840 : Standard_Failure: StartSol echec

EXCEPTION Stripe compute 0x42748840 : Standard_Failure: CallPerformSurf : Echec cheminement!

Draw[4]> exit

[root@bobslap1 DrawResources]#

Script done on Wed Dec 29 15:35:52 1999

Running through the demo.tcl script manually, I notice that the graphics seem to blow up doing the first "depouille" command but I don't get any error messages at that time.

Here are some of my system specific details:

gcc version pgcc-2.91.66 19990314 (egcs-1.1.2 release)

Linux version 2.2.13-22mdk ([email removed])

(gcc version egcs-2.91.66 19990314/Linux (egcs-1.1.2

release)) #1 SMP Fri Oct 22 02:06:33 CEST 1999

Memory: 127664k/131008k available

(1172k kernel code, 424k reserved, 1668k data, 80k init, 0k bigmem)

rpm -q -a |grep Mesa

Mesa-3.0.99-4mdk

Mesa-demos-3.0.99-4mdk

Mesa-devel-3.0.99-4mdk

Everything seemed to compile OK, except for some warnings about uninitialized variables, etc. Also, the .lnk scripts apparently test the shared libs wrong, as the line that tries a compile test for each library does not specify any input files. Gcc complains about this (no input files), script says there are errors, but doesn't show them.

I have went through the entire build process twice so far, as the first time I was trying to build with Mesa 3.1 installed, and there were numerous problems in the OpenGL Visualization Module compilation.

Any clues?

Discussion

5 archived replies

Posts are displayed in their archived order.

01Commenter-1Thu, 12/30/1999 - 17:59

Bob,

I also tried this other problem of yours. Unfortunately, it worked for me again :-(

I downloaded the source and the binaries, but haven't done the recompile like I think you have. I will probably start doing this over the next couple day. If you discover a solution, can you post it? I'm afraid I may run into these problems myself...

[mark@dell1 DrawResources]$ demo Draw[1]> Draw[2]> axo Draw[3]> source demo.tcl

Creating a box

Draft of two lateral faces

? Fillet on four lateral edges, then on the top and bottom edges

tolerance ang : 0.01 tolerance 3d : 0.0001 tolerance 2d : 1e-05 fleche : 0.001 tolblend 0.01 0.0001 1e-05 0.001 tolerance ang : 0.01 tolerance 3d : 0.0001 tolerance 2d : 1e-05 fleche : 0.001 tolblend 0.01 0.0001 1e-05 0.001 Creating a profile on the top face

Creating a prism

Opening the top face

Creating a cylinder and positioning it

Demo completed

Draw[4]>

02Commenter-2Thu, 12/30/1999 - 19:42

I did not download binaries. I did not know they were available pre built. What kernel version, compiler, etc do you have?

03Commenter-2Thu, 12/30/1999 - 22:49

Well, I downloaded the binary packages and installed them and lo and behold, the test harness demo.tcl script now works. However, after removing and re-installing the SampleGeometry java application, I still can not get that to run. Now it looks like it COULD POSSIBLY be due to missing fonts? Here is the start of the stack trace dump to show this error (Whole dump is way too long and the start is always the relevant part anyway:

[root@bobslap1 java]# csh -f runme.csh

*Xw_Error_4/4*code 3/'BadWindow (invalid Window parameter)'

from Xw_error_handler routine

*Xw_Error_1/19*Bad Font Name '-adobe-TABTXT01-*-*-*-*-*-*-*-*-*-*-*-*' from Xw_def_font routine

method unknown (eip = 458f3459)

SIGSEGV 11* segmentation violation

si_signo [11]: SIGSEGV 11* segmentation violation

si_errno [0]: Success

si_code [0]: SI_USER [pid: 0, uid: 0]

stackpointer=0xbffff1e0

Full thread dump Classic VM (1.2.2-RC2-K, green threads):

Back to the Draw Test Harness, I have noticed that the compiled versions of the binaries which I compiled were only about 50% of the size of the pre-compiled ones. The shared libraries are slightly different in sizes. Must be some kind of difference in compiler versions or something as I had used the exact same *.COMP files to compile everything. Has anyone else had problems rebuilding these?

04Commenter-1Thu, 12/30/1999 - 23:10

Who needs the internet when you have a psychic net? I was posting the message to try the binaries while you were trying the binaries. Maybe it is just one of those "Great minds think alike" things ;-)

THis file mentions the TABTXT01

.../CAS3.0/src/V2d

It might give you a hint to what is going on.

05Commenter-1Thu, 12/30/1999 - 22:55

Linux Kernel 2.2.9

Xfree 3.3.3.1

JDK 1.2.2

Mesa-dev 3.0-2

pgcc 1.1.3 GLibc 2.1.1

128 Meg PIII

LD_LIBRARY_PATH=/lib:/usr/lib:/usr/X11R6/lib

It is a mandrake 6.0 installation, plus a few upgrades. I'm guessing you have a mandrake6.1 setup. Other than being slightly ahead of me on the Kernel and mesa3D, I think we match pretty closely.

The binaries are found just below the source files. I would try downloading them. This would at least tell you if it is your system or the compilation which is causing the problem.