@Juzz

I am experiencing the same results on my Peg2. It would be nice to get this figured out.

_________________
Raptor BlackBird (awaiting a Miggy like OS)
Amiga 4000 3.9 / Pegasos II 4.1 F.E.

Which filesystems you have tried or does it matter? It's been a long time since this port was made, so I don't remember anything about it but archive seems to contain sources so it may be fixable.

@Mr_Capehill

I thought about that too - the first time I tried, I was using jxfs. However I moved it to another partition (I can't remember right now which filesystem it is, but I think it's sfs - however I will check that next time I boot up my AmigaOne).

Cheers
Juzz

_________________
AmigaOne Owner since 2002.
The first person to get a Linux installer running on the AmigaOne :-

@Mr_Capehill

The other partition that I have tried is SFS, I could try to make an FFS partition on one of my old USB sticks.

Cheers
Juzz

_________________
AmigaOne Owner since 2002.
The first person to get a Linux installer running on the AmigaOne :-

@Mr_Capehill

It's the same story on FFS, it creates an empty file with no attributes and then complains it can't write to the file.
Again if I add rwed then the game strips them from the file.


Cheers
Juzz

_________________
AmigaOne Owner since 2002.
The first person to get a Linux installer running on the AmigaOne :-

@Mr_Capehill

Did you get a chance to look at it?

Cheers
Juzz

_________________
AmigaOne Owner since 2002.
The first person to get a Linux installer running on the AmigaOne :-

@Juzz

Yes, briefly. I was able to reproduce issue on Ram disk and JXFS installations at least. Then I tried to rebuild it with the latest SDK. Old port used newlib. New build with newlib behaved similarly than old. Then I built a clib2 version which was able to save/load. Perhaps something has changed in newlib since...2005? :)

It could be a Moria bug but if I find the cause I will report it to newlib developers. I will test clib2 build a bit and upload it somewhere later.

If you know C you may be able to rebuild it yourself meanwhile. Moria wants an old style timerequest and some prototype (time()?) collided with clib2 declarations but otherwise it compiled with warnings.

@Mr_Capehill

FWIW I just went through the code and fixed all the warnings apart from the ones to do with deprecated OS functions, which turned out to be a lot more work than I expected .

You can get my fixed code here if you want it:
https://dl.dropboxusercontent.com/u/26599983/CWMMoria110.7z

I also fixed the makefile so that it works with cross-compiling.

I intend to do some tests later to see whether it is a bug in newlib or moria code that is causing the protection flags to get trashed.

Last edited by salass00 on 09-Jan-2014 at 07:05 AM.
Last edited by salass00 on 08-Jan-2014 at 08:04 PM.

@all
Its good to see that I am not the only one here who still likes and plays Rogue like games in 2014

Although I have to admit I am also rather enjoying playing Legend of Grimrock on my PC after picking it up in the recent GOG sale, even if it is more of a Dungeon Master/Eye of the beholder clone.

_________________
Sam440ep 667mhz, 512MB, 120GB 2.5" HD, OS4.1FE
WinUae 3.0.0, OS 3.9, BB3, Catweasel MkIV
Amiga 1200, Blizzard 040/40 (BlizzardPPC 060/200 with SCSI removed at present), mediatorSX pci, Voodoo3, PCI network card os 3.9BB2
4MB Minimig with ARM addon boar

@Mr_Capehill

Looking at the code the problem with save files is on line 481 of file save.c:

fd = open(fnam, O_RDWR | O_CREAT | O_TRUNC);

When O_CREAT is used a third parameter should be provided of type mode_t that contains the unix permission bits for the new file. Since currently one is not provided the value used is undefined.

Last edited by salass00 on 09-Jan-2014 at 07:03 AM.
Last edited by salass00 on 09-Jan-2014 at 07:02 AM.

@Mr_Capehill

I changed the line of code to:
fd = open(fnam, O_RDWR | O_CREAT | O_TRUNC, 0644);

and now the save system works.

Updated source version is here:
https://dl.dropboxusercontent.com/u/26599983/CWMMoria110.7z

@salass00

Classical failure of the way library and compiler don't technically know about each other there. Documentation says mode MUST be supplied when O_CREAT is used, but compiler does not enforce it. Astronomical pitfall in POSIX et al. Methinks it would be better to create two wrappers, one for when you can create (requiring mode) and one for when you will not create, where the compiler could easily see how many arguments to check for.

Allthough, gcc does incorporate printf style argument counting, so the same mechanism could very well be used to check that O_CREAT has a mode argument.

Anyhoo, just a morning rant.

Rogue lookalikes are always cool.

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@olegil

There is the creat() function which is more or less what you suggest:
http://pubs.opengroup.org/onlinepubs/7999959899/functions/creat.html

Also if I compile the following program using gcc on my Ubuntu system:
Quote:

#include <fcntl.h> #include <unistd.h> int main(void) { int fd = open("test.out", O_WRONLY|O_CREAT|O_TRUNC); close(fd); return 0; }

it produces the following error:
Quote:

$ gcc -O2 -o test test.c In file included from /usr/include/fcntl.h:252:0, from test.c:1: In function ‘open’, inlined from ‘main’ at test.c:5:6: /usr/include/i386-linux-gnu/bits/fcntl2.h:51:24: error: call to ‘__open_missing_mode’ declared with attribute error: open with O_CREAT in second argument needs 3 arguments

So having error checking for this is apparently not entirely impossible. Interestingly enough though it compiles without error if I remove the -O2 option.

Last edited by salass00 on 09-Jan-2014 at 09:23 AM.

@salass00

Nice work with all cleanups and fixes, thank you.

BTW: there has also been some other efforts:

http://free-moria.sourceforge.net/
http://sourceforge.net/projects/umoria/

Moria, a game that wouldn't die :)

@Mr_Capehill

Quote:

Falcon's Eye: http://www.hut.fi/~jtpelto2/nhfeatures.html#SCREENSHOTS

This looks awesome! Do we have this port for os4? I'm kind of confused on the thread as I dont know about these games, but I really like the look of the "Falcons eye" it reminds me when we were kids and used to play dungeons and dragons with floormaps and pewter miniature characters.

@Mr_Capehill

Quote:

Mr_Capehill wrote: @salass00 Nice work with all cleanups and fixes, thank you. BTW: there has also been some other efforts: http://free-moria.sourceforge.net/ http://sourceforge.net/projects/umoria/ Moria, a game that wouldn't die :)

Ahhh... Yes... But those are the NON-graphical projects. I like that the CWM was based on the graphical Umoria.

@ALL: So any builds yet?

_________________
AmigaOne Owner since 2002.
The first person to get a Linux installer running on the AmigaOne :-

@Juzz

Yes! Try salass00 new build above. Post #91. It works great!
edit: oops, you probably meant the others that you quoted

@salass00

Thank you so much for your fix.

Last edited by TheKorn on 10-Jan-2014 at 02:05 AM.
Last edited by TheKorn on 10-Jan-2014 at 02:04 AM.

_________________
Raptor BlackBird (awaiting a Miggy like OS)
Amiga 4000 3.9 / Pegasos II 4.1 F.E.

@TheKorn

Nopes, I meant of the CWM Moria - but all that ever was mentioned as being updated was the sources...

_________________
AmigaOne Owner since 2002.
The first person to get a Linux installer running on the AmigaOne :-

@salass00

Extremely interesting, I've tried it myself and any sort of -O parameter will give an error, so it should be turned on by something in the basic -O. Using each individual flag that makes up -O (it's a bit of a list) didn't give anything, though. So there's some magic involved.

Quote:

olegil@olegil-laptop:~$ gcc -o testcreat -fauto-inc-dec -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -fcompare-elim -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -fcprop-registers -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -fdce -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -fdefer-pop -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -fdelayed-branch -c testcreat.c testcreat.c:1:0: warning: this target machine does not have delayed branches [enabled by default] olegil@olegil-laptop:~$ gcc -o testcreat -fdse -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -fguess-branch-probability -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -fif-conversion2 -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -fif-conversion -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -fipa-pure-const -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -fipa-profile -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -fipa-reference -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -fmerge-constants -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -fsplit-wide-types -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -ftree-bit-ccp -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -ftree-builtin-call-dce -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -ftree-ccp -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -ftree-ch -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -ftree-copyrename -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -ftree-dce -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -ftree-dominator-opts -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -ftree-dse -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -ftree-forwprop -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -ftree-fre -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -ftree-phiprop -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -ftree-slsr -c testcreat.c cc1: error: unrecognized command line option ‘-ftree-slsr’ olegil@olegil-laptop:~$ gcc -o testcreat -ftree-sra -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -ftree-pta -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -ftree-ter -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -funit-at-a-time -c testcreat.c olegil@olegil-laptop:~$ gcc -o testcreat -Os -c testcreat.c In file included from /usr/include/fcntl.h:252:0, from testcreat.c:1: In function ‘open’, inlined from ‘main’ at testcreat.c:5:5: /usr/include/i386-linux-gnu/bits/fcntl2.h:51:24: error: call to ‘__open_missing_mode’ declared with attribute error: open with O_CREAT in second argument needs 3 arguments olegil@olegil-laptop:~$ gcc -o testcreat -Ofast -c testcreat.c In file included from /usr/include/fcntl.h:252:0, from testcreat.c:1: In function ‘open’, inlined from ‘main’ at testcreat.c:5:5: /usr/include/i386-linux-gnu/bits/fcntl2.h:51:24: error: call to ‘__open_missing_mode’ declared with attribute error: open with O_CREAT in second argument needs 3 arguments olegil@olegil-laptop:~$ gcc -o testcreat -O -c testcreat.c In file included from /usr/include/fcntl.h:252:0, from testcreat.c:1: In function ‘open’, inlined from ‘main’ at testcreat.c:5:5: /usr/include/i386-linux-gnu/bits/fcntl2.h:51:24: error: call to ‘__open_missing_mode’ declared with attribute error: open with O_CREAT in second argument needs 3 arguments olegil@olegil-laptop:~$ gcc -o testcreat -O2 -c testcreat.c In file included from /usr/include/fcntl.h:252:0, from testcreat.c:1: In function ‘open’, inlined from ‘main’ at testcreat.c:5:5: /usr/include/i386-linux-gnu/bits/fcntl2.h:51:24: error: call to ‘__open_missing_mode’ declared with attribute error: open with O_CREAT in second argument needs 3 arguments olegil@olegil-laptop:~$ gcc -o testcreat -O3 -c testcreat.c In file included from /usr/include/fcntl.h:252:0, from testcreat.c:1: In function ‘open’, inlined from ‘main’ at testcreat.c:5:5: /usr/include/i386-linux-gnu/bits/fcntl2.h:51:24: error: call to ‘__open_missing_mode’ declared with attribute error: open with O_CREAT in second argument needs 3 arguments

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@Juzz

Quote:

Nopes, I meant of the CWM Moria - but all that ever was mentioned as being updated was the sources...

There's an updated exe too in the same archive.