On Wed, Sep 8, 2010 at 8:48 PM, <sm@xxxxxxxxxxxx> wrote: >> Not sure what you did to the text, but quoting is quite borked here, >> hard to read... > I'm using Squirrelmail-1.4.21 and the reply link. When I go to reference > a previous mail, I save to draft folder and come back to the draft and > continue, nothing else. > >> >> On Wed, Sep 8, 2010 at 7:04 PM, <sm@xxxxxxxxxxxx> wrote: >>>>>>> think 'step-5' applied to our situation or issue; I thought we >>>>>>> tested >>> 'change_passwd' at least once. However, I will try 'step-5' now. >>>>>> It says if the binary isn't working, recompilation is probably called >>> for... your situation definitely qualifies for that. >>>>>>> RESULTS: >>>>>>> gcc -lcrypt -O -o chpasswd chpasswd.c >>>>>>> chpasswd.c:33:19: error: crypt.h: No such file or directory >>>>>>> A little googling leads us to believe that "the crypt() function is >>> defined in unistd.h, not crypt.h". >>>>>> Either one might work. Not sure. What does "man crypt" tell you? >>>>>> Did >>> you search for those header file and actually look in them for crypt? > No I have not. I thought you were meaning the chpasswd.c was a header file. > >>>>>>> I tried replacing #include <crypt.h> >>>>>>> with #include <unistd.h>, remember we're not programmers; chpasswd >>> continues to not function on FreeBSD-7.3. >>>>>> You have to give more details. >>>>> find / -name 'crypt.h' >>>>> /usr/src/lib/libcrypt/crypt.h >>>>>>What is the compilation output? >>>>> # gcc -lcrypt -O -o chpasswd chpasswd.c >>>>> chpasswd.c:33:19: error: crypt.h: No such file or directory >>>> Looks like you have crypt on your system. You just need to find the >>> right combination of compilation arguments to get it to work. Try >>> adding >>> "-L/usr/src/lib/libcrypt" for starters. Or just adding >>>> "-lcrypt" might work. You'll have to play around and/or do some >>> searching. > None of the above is working. Please allow me to point out that I have at > least one other port that has used "-lcrypt" recently; here's a snipp... > > cflags (mysql_config) = -I/usr/local/include/mysql > -fno-strict-aliasing -pipe > embedded (mysql_config) = > libs (mysql_config) = -L/usr/local/lib/mysql -lmysqlclient -lz > -lcrypt -lm > > It compiled without any issues. I don't have a BSD test machine, so I can't really help you any more than you can help yourself. Google and whatnot. The best solution would be to convert the program to use automake, but I have not the time to fiddle with that right now. If you are willing to wait around and test it when I do, I can try to put that on my list in the next few weeks. Otherwise, as I said, have a try at figuring out the right compilation arguments and let us know the correct ones when you have it compiling. > Please report back if you get it to compile. >>>>>>What happens when you run the binary from the command line? >>>>> # ./chpasswd <user> <correct.current.passwd> <new.passwd> >>>>> RESULT: Current password is incorrect >>>> How is this when you could not compile it? I just looked and unistd.h >>> is > already included, so if you replaced crypt.h with unistd.h, then >>> you >>> have > it twice. >>> >>> It was removed once we saw it had no affect. >>> - >>> The results I posted here, starting with the preceding email was with a >>> fresh install and with the removal of 'all' the '^M'. That being said, >>> you are correct gcc/cc compiles continue to fail with the results we >>> previously posted; here are the compile failure results again: >>> >>> # gcc -lcrypt -O -o chpasswd chpasswd.c >>> chpasswd.c:33:19: error: crypt.h: No such file or directory >>> >>> # gcc -Wall -lcrypt -O -o chpasswd chpasswd.c >>> chpasswd.c:33:19: error: crypt.h: No such file or directory >>> chpasswd.c: In function 'main': >>> chpasswd.c:103: warning: pointer targets in passing argument 1 of >>> 'fixpwd' >>> differ in signedness >>> chpasswd.c:104: warning: pointer targets in passing argument 1 of >>> 'fixpwd' >>> differ in signedness >>> chpasswd.c:105: warning: pointer targets in passing argument 1 of >>> 'fixpwd' >>> differ in signedness >>> >>> Once again this is a fresh install using: cp config_example.php >>> config.php >>> afterwards removing 'all' of the '^M'. >>> - >>> Using ./chpassword from the cli: >>> change_passwd/chpasswd 'userID' 'old_pass' 'new_pass' 2>&1 >>> RESULT: [3] 18811 >>> 1: Command not found. >>> [2] - Exit 9 ./chpasswd userid old_pass new_pass >>> > 2 - >>> Above nothing truncated or added. Given that we aren't able to compile >>> chpasswd, watch this: >>> >>> # ./chpasswd user1 >>> Missing current password >>> # ./chpasswd user1 12345 >>> Missing new password >>> # ./chpasswd user1 12345 67890 >>> User does not exist: user1 >>> - >>> Correct user1 does not exist, however it appears 'chpasswd' runs in some >>> way from the cli but it doesn't actually change any password when you >>> have >>> a valid userid in place. chpasswd also 'appears' to run via SM-1.4.21 > >>> options > change password. Currently, using the latter method tells the >>> user, in red; "Your password was changed successfully" <-- verbatim; on >>> the change password form/page. >> >> You originally reported that running the binary gave you "ELF binary >> type "0" not known" >> >> Now you say it is (half) working. I can't follow what you are doing, > > The output that include ELF was shown on the change password page/form > after we attempted to change a valid users password while both: $debug = > 1; and $seeOutput = 1; <--config.php. > >> what version of which binary you are using, > If the answer to your question is not the version number in the subject > line, I have not a clue, nor to I know How to find the answer to that > question. I'd be more than happy to follow your instructions. > >> but the most relevant >> thing to say is that all this testing is useless if you couldn't >> compile the binary in the first place. I'm not sure what you expect >> out of it otherwise. I already gave you some tips to help find the >> right compilation arguments. > > At this point the focus is on why chpasswd is not compiling and/or finding > crypt.h when in fact crypt is on the machine and other ports can find and > use crypt.h when they need to. Nothing really customized on this machine > other than the kernel, as in nothing exotic in /etc/make.conf. > > Yes the reply quotes are annoying, this is the default setting for SM eith > the exception of a theme change. > >> >>> The indented lines directly below this line still hold true. >>> >>>>> There is no doubt about the <correct.current.password>; I use that >>>>> passwd >>>>> everyday when logging into Squirrelmail. As you can see (above) >>> running the ./chpasswd from the cli shows (incorrect), however running >>> the >>> binary >>>>> via SM > Options > change password, the results are printed in red >>>>> that >>> the password was successfully changed. >>> >>>> You said it gives an error (...contact administrator...). Which is it? >>> Can't remember exactly what I was doing to get that msg; and I'm >>> currently >>> unable to reproduce it. >>> >>>>Please try to keep your output reports consistent with what version of >>> the >binary/binaries you are using. >>> >>> Everything in this email I have repeatedly reproduced, before sending >>> the >>> mail a minimum of 10-times, with a fresh install of the plugin. I hope >>> that falls under the 'consistent' definition. >>> >>> Going to read 'man crypt' Enigma(1) FreeBSD-7.3. Maybe I'll learn >>> something. Let me know if I left any detail out that you need. >>> >>> >>>>> Any attempts to reload or go to >>>>> another page results in SM directing you to login. At this point >>> Attempting to login with what you think is the new.passwd; fails. >>>>>>> Given this crypt.h issue, >>>>>> How do you know that is the case? Supposedly you fixed that. >>>>> I tried to fix it; didn't work. What you see above is >>>>> native/original; >>> nothing edited, with the exception of removing the '^M' from the >>> aforementioned file(s). >>>> The native/original supposedly won't even run at all on your system, >>>> and >>> recompiling it fails with an error, so I'm not sure what you've done >>> now. >> >> -- Paul Lesniewski SquirrelMail Team Please support Open Source Software by donating to SquirrelMail! http://squirrelmail.org/donate_paul_lesniewski.php ------------------------------------------------------------------------------ This SF.net Dev2Dev email is sponsored by: Show off your parallel programming skills. Enter the Intel(R) Threading Challenge 2010. http://p.sf.net/sfu/intel-thread-sfd ----- squirrelmail-users mailing list Posting guidelines: http://squirrelmail.org/postingguidelines List address: squirrelmail-users@xxxxxxxxxxxxxxxxxxxxx List archives: http://news.gmane.org/gmane.mail.squirrelmail.user List info (subscribe/unsubscribe/change options): https://lists.sourceforge.net/lists/listinfo/squirrelmail-users
