![](https://secure.gravatar.com/avatar/36fe2e1a70ee9905ec5b6d8e53503684.jpg?s=120&d=mm&r=g)
On Debian testing/unstable with texlive-metapost/testing uptodate 2005.dfsg.2-2 texlive-base/testing uptodate 2005.dfsg.2-3 texlive-common/testing uptodate 2005.dfsg.2-3 texlive-base-bin/testing uptodate 2005.dfsg.2-2 and the new context package from www.tug.org/texlive/Debian version 2006.08.08 the following text file processes without error but has no graphic \starttext \startuseMPgraphic{test} draw (0,0)..(1,0)..(1,1)..(1,0)..cycle scaled 5cm withcolor black ; \stopuseMPgraphic \useMPgraphic{test} \stoptext Further investigation revealed that johans@smoke:~$ mpost -progname=metafun -mem=metafun mptest-mpgraph.mp This is MetaPost, Version 0.901 (Web2C 7.5.5) (mptest-mpgraph.mpSegmentation fault Mpost segfaulted, which I have never seen and have no idea how to pursue. More details: johans@smoke:~$ mpost --version MetaPost 0.901 (Web2C 7.5.5) kpathsea version 3.5.5 Copyright 2005 AT&T Bell Laboratories. Kpathsea is copyright 2005 Karl Berry and Olaf Weber. There is NO warranty. Redistribution of this software is covered by the terms of both the MetaPost copyright and the Lesser GNU General Public License. For more information about these matters, see the file named COPYING and the MetaPost source. Primary author of MetaPost: John Hobby. Kpathsea written by Karl Berry, Olaf Weber, and others. johans@smoke:~$ locate metafun.mem /etc/texmf/web2c/metafun.mem /var/lib/texmf/web2c/metafun.mem -- Johan Sandblom N8, MRC, Karolinska sjh t +46851776108 17176 Stockholm m +46735521477 Sweden "What is wanted is not the will to believe, but the will to find out, which is the exact opposite" - Bertrand Russell
![](https://secure.gravatar.com/avatar/1aaa8171e46f2c29d4441fa99a7fc634.jpg?s=120&d=mm&r=g)
On Mit, 08 Nov 2006, Johan Sandblom wrote:
johans@smoke:~$ locate metafun.mem /etc/texmf/web2c/metafun.mem /var/lib/texmf/web2c/metafun.mem
The mem file in /etc/texmf/web2c seems VERY FISHY!!! Please remove it.
Probably metapost loads a wrong metafun.mem file.
Best wishes
Norbert
-------------------------------------------------------------------------------
Dr. Norbert Preining
![](https://secure.gravatar.com/avatar/36fe2e1a70ee9905ec5b6d8e53503684.jpg?s=120&d=mm&r=g)
I did. No change.
Regards, Johan
2006/11/8, Norbert Preining
On Mit, 08 Nov 2006, Johan Sandblom wrote:
johans@smoke:~$ locate metafun.mem /etc/texmf/web2c/metafun.mem /var/lib/texmf/web2c/metafun.mem
The mem file in /etc/texmf/web2c seems VERY FISHY!!! Please remove it. Probably metapost loads a wrong metafun.mem file.
Best wishes
Norbert
------------------------------------------------------------------------------- Dr. Norbert Preining
Università di Siena Debian Developer Debian TeX Group gpg DSA: 0x09C5B094 fp: 14DF 2E6C 0307 BE6D AD76 A9C0 D2BF 4AA3 09C5 B094 ------------------------------------------------------------------------------- AMERSHAM (n.) The sneeze which tickles but never comes. (Thought to derive from the Metropolitan Line tube station of the same name where the rails always rattle but the train never arrives.) --- Douglas Adams, The Meaning of Liff
-- Johan Sandblom N8, MRC, Karolinska sjh t +46851776108 17176 Stockholm m +46735521477 Sweden "What is wanted is not the will to believe, but the will to find out, which is the exact opposite" - Bertrand Russell
![](https://secure.gravatar.com/avatar/a76811e924ec756bf9f8fc09e8337d30.jpg?s=120&d=mm&r=g)
On Wed, Nov 08, 2006 at 07:01:00PM +0100, Johan Sandblom wrote:
I did. No change.
I suspect this is 'the' mismatched array bug in web2c. If you have both main_memory.mpost and|or main_memory.metafun in your texmf.cnf, and they are not the same value(s), then you can get segfaults (as well as various other bugs). It is hard to be more verbose right now because my connection is awful, but there should be some more info on this in the mailinglist archive. Greetings, Taco
![](https://secure.gravatar.com/avatar/36fe2e1a70ee9905ec5b6d8e53503684.jpg?s=120&d=mm&r=g)
2006/11/8, taco@glenfiddich.elvenkind.com
On Wed, Nov 08, 2006 at 07:01:00PM +0100, Johan Sandblom wrote:
I did. No change.
I suspect this is 'the' mismatched array bug in web2c.
If you have both main_memory.mpost and|or main_memory.metafun in your texmf.cnf, and they are not the same value(s), then you can get segfaults (as well as various other bugs).
No, that does not appear to be it: root@smoke:/etc/texmf# grep main_memory `kpsewhich texmf.cnf` % main_memory.hugetex = 20000000 % main_memory is relevant only to initex, extra_mem_* only to non-ini. % Thus, have to redump the .fmt file after changing main_memory; to add main_memory = 1000000 % words of inimemory available; also applies to inimf&mp main_memory.context = 1500000 main_memory.mpost = 1500000 main_memory.lambda = 2200000
It is hard to be more verbose right now because my connection is awful, but there should be some more info on this in the mailinglist archive.
Greetings, Taco
-- Johan Sandblom N8, MRC, Karolinska sjh t +46851776108 17176 Stockholm m +46735521477 Sweden "What is wanted is not the will to believe, but the will to find out, which is the exact opposite" - Bertrand Russell
![](https://secure.gravatar.com/avatar/36fe2e1a70ee9905ec5b6d8e53503684.jpg?s=120&d=mm&r=g)
It was the crucial clue, however, since I found
/usr/share/texmf-texlive/web2c/texmf.cnf
and removed it, which solved the problem!
Thank you all, Johan Sandblom
2006/11/8, Johan Sandblom
2006/11/8, taco@glenfiddich.elvenkind.com
: On Wed, Nov 08, 2006 at 07:01:00PM +0100, Johan Sandblom wrote:
I did. No change.
I suspect this is 'the' mismatched array bug in web2c.
If you have both main_memory.mpost and|or main_memory.metafun in your texmf.cnf, and they are not the same value(s), then you can get segfaults (as well as various other bugs).
No, that does not appear to be it:
root@smoke:/etc/texmf# grep main_memory `kpsewhich texmf.cnf` % main_memory.hugetex = 20000000 % main_memory is relevant only to initex, extra_mem_* only to non-ini. % Thus, have to redump the .fmt file after changing main_memory; to add main_memory = 1000000 % words of inimemory available; also applies to inimf&mp main_memory.context = 1500000 main_memory.mpost = 1500000 main_memory.lambda = 2200000
It is hard to be more verbose right now because my connection is awful, but there should be some more info on this in the mailinglist archive.
Greetings, Taco
-- Johan Sandblom N8, MRC, Karolinska sjh t +46851776108 17176 Stockholm m +46735521477 Sweden "What is wanted is not the will to believe, but the will to find out, which is the exact opposite" - Bertrand Russell
-- Johan Sandblom N8, MRC, Karolinska sjh t +46851776108 17176 Stockholm m +46735521477 Sweden "What is wanted is not the will to believe, but the will to find out, which is the exact opposite" - Bertrand Russell
![](https://secure.gravatar.com/avatar/565476157eb3895a9c928ffce26d988d.jpg?s=120&d=mm&r=g)
main_memory = 1000000 % words of inimemory available; also applies to inimf&mp main_memory.context = 1500000 main_memory.mpost = 1500000
On Wed, Nov 08, 2006 at 08:05:48PM +0100, Johan Sandblom wrote: please try deleting this line and see what happens. If that is not, then probably your executable is the problem (i cannot reproduce the crash otherwise, on linux nor windows) Good luck, Taco
![](https://secure.gravatar.com/avatar/4f90ced6bd47c80395b33c0fde81a0d0.jpg?s=120&d=mm&r=g)
Hi, Norbert Preining wrote:
On Mit, 08 Nov 2006, Johan Sandblom wrote:
johans@smoke:~$ locate metafun.mem /etc/texmf/web2c/metafun.mem /var/lib/texmf/web2c/metafun.mem
The mem file in /etc/texmf/web2c seems VERY FISHY!!! Please remove it. Probably metapost loads a wrong metafun.mem file.
Not necessarily. One might actually be a symbolic link to the other. Try the following to check this ls -l /etc/texmf/web2c/metafun.mem /var/lib/texmf/web2c/metafun.mem and just to make sure also: ls -ld /etc/texmf/web2c /var/lib/texmf/web2c The file used by TeX can actually found using: kpsewhich metafun.mem and gives here: /var/lib/texmf/web2c/metafun.mem Tobias
![](https://secure.gravatar.com/avatar/36fe2e1a70ee9905ec5b6d8e53503684.jpg?s=120&d=mm&r=g)
Hi,
2006/11/8, Tobias Burnus
Hi,
Norbert Preining wrote:
On Mit, 08 Nov 2006, Johan Sandblom wrote:
johans@smoke:~$ locate metafun.mem /etc/texmf/web2c/metafun.mem /var/lib/texmf/web2c/metafun.mem
The mem file in /etc/texmf/web2c seems VERY FISHY!!! Please remove it. Probably metapost loads a wrong metafun.mem file.
Not necessarily. One might actually be a symbolic link to the other. Try the following to check this ls -l /etc/texmf/web2c/metafun.mem /var/lib/texmf/web2c/metafun.mem and just to make sure also: ls -ld /etc/texmf/web2c /var/lib/texmf/web2c
They were not links. No idea where they came from. Removing them did not solve anything.
The file used by TeX can actually found using: kpsewhich metafun.mem and gives here: /var/lib/texmf/web2c/metafun.mem
johans@smoke:~$ kpsewhich metafun.mem /var/lib/texmf/web2c/metafun.mem Regards, Johan
Tobias
-- Johan Sandblom N8, MRC, Karolinska sjh t +46851776108 17176 Stockholm m +46735521477 Sweden "What is wanted is not the will to believe, but the will to find out, which is the exact opposite" - Bertrand Russell
![](https://secure.gravatar.com/avatar/1aaa8171e46f2c29d4441fa99a7fc634.jpg?s=120&d=mm&r=g)
On Mit, 08 Nov 2006, Johan Sandblom wrote:
texlive-metapost/testing uptodate 2005.dfsg.2-2 texlive-base/testing uptodate 2005.dfsg.2-3 texlive-common/testing uptodate 2005.dfsg.2-3 texlive-base-bin/testing uptodate 2005.dfsg.2-2
and the new context package from www.tug.org/texlive/Debian version 2006.08.08 the following text file processes without error but has no graphic
Strange enough, on my computer I don't see this problem. The file contains the graphics.
johans@smoke:~$ mpost -progname=metafun -mem=metafun mptest-mpgraph.mp This is MetaPost, Version 0.901 (Web2C 7.5.5) (mptest-mpgraph.mpSegmentation fault
johans@smoke:~$ locate metafun.mem /etc/texmf/web2c/metafun.mem /var/lib/texmf/web2c/metafun.mem
Maybe some other files in /etc/texmf/web2c are interfering. metapost
pool files would be an idea. Can you send me the list of files in
/etc/texmf/web2c?
I have thee ONLY mktex.cnf!!!
Best wishes
Norbert
-------------------------------------------------------------------------------
Dr. Norbert Preining
![](https://secure.gravatar.com/avatar/36fe2e1a70ee9905ec5b6d8e53503684.jpg?s=120&d=mm&r=g)
Hi,
2006/11/8, Norbert Preining
On Mit, 08 Nov 2006, Johan Sandblom wrote:
texlive-metapost/testing uptodate 2005.dfsg.2-2 texlive-base/testing uptodate 2005.dfsg.2-3 texlive-common/testing uptodate 2005.dfsg.2-3 texlive-base-bin/testing uptodate 2005.dfsg.2-2
and the new context package from www.tug.org/texlive/Debian version 2006.08.08 the following text file processes without error but has no graphic
Strange enough, on my computer I don't see this problem. The file contains the graphics.
johans@smoke:~$ mpost -progname=metafun -mem=metafun mptest-mpgraph.mp This is MetaPost, Version 0.901 (Web2C 7.5.5) (mptest-mpgraph.mpSegmentation fault
johans@smoke:~$ locate metafun.mem /etc/texmf/web2c/metafun.mem /var/lib/texmf/web2c/metafun.mem
Maybe some other files in /etc/texmf/web2c are interfering. metapost pool files would be an idea. Can you send me the list of files in /etc/texmf/web2c?
I have thee ONLY mktex.cnf!!!
There were several other files, among them a highly suspicious plain.mem. However, removing all of them, and reinstalling the context debian package (and thereby rebuilding the formats) did not solve the problem. No new files ended up in /etc/texmf/web2c.
Best wishes
Norbert
------------------------------------------------------------------------------- Dr. Norbert Preining
Università di Siena Debian Developer Debian TeX Group gpg DSA: 0x09C5B094 fp: 14DF 2E6C 0307 BE6D AD76 A9C0 D2BF 4AA3 09C5 B094 ------------------------------------------------------------------------------- DUNGENESS (n.) The uneasy feeling that the plastic handles of the overloaded supermarket carrier bag you are carrying are getting steadily longer. --- Douglas Adams, The Meaning of Liff
-- Johan Sandblom N8, MRC, Karolinska sjh t +46851776108 17176 Stockholm m +46735521477 Sweden "What is wanted is not the will to believe, but the will to find out, which is the exact opposite" - Bertrand Russell
participants (5)
-
Johan Sandblom
-
Norbert Preining
-
taco@elvenkind.com
-
taco@glenfiddich.elvenkind.com
-
Tobias Burnus