[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#354178: MIT-KERBEROS-5



Package: xorg-x11
Version: 6.9.0.dfsg.1-4

Struggling to build X.Org with MIT Kerberos -


[...]
gcc -m32 -c -ansi -Wall -Wpointer-arith -Wstrict-prototypes                     
 -Wmissing-prototypes -Wmissing-declarations                      -Wredundant-de
cls -Wnested-externs -Wundef    -I../.. -I../../exports/include   -Dlinux -D__i3
86__ -D_POSIX_C_SOURCE=199309L                          -D_POSIX_SOURCE -D_XOPEN
_SOURCE                                 -D_BSD_SOURCE -D_SVID_SOURCE            
                     -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64                 
-D_GNU_SOURCE                             -DFUNCPROTO=15 -DNARROWPROTO -DXTHREAD
S  -D_REENTRANT -DXUSE_MTSAFE_API       -g -O2 -fno-strict-aliasing -g  -fPIC k5
encode.c -o unshared/k5encode.o
k5encode.c:75: warning: function declaration isn't a prototype
k5encode.c:81:34: error: macro "krb5_princ_realm" requires 2 arguments, but only
 1 given
k5encode.c: In function 'XauKrb5Encode':
k5encode.c:81: error: 'krb5_princ_realm' undeclared (first use in this function)
k5encode.c:81: error: (Each undeclared identifier is reported only once
k5encode.c:81: error: for each function it appears in.)
k5encode.c:82:37: error: macro "krb5_princ_size" requires 2 arguments, but only 
1 given
k5encode.c:82: error: 'krb5_princ_size' undeclared (first use in this function)
[...]


I used xorg-x11 6.9.0.dfsg.1-4 & libkrb5-dev 1.4.3-6

I used this patch to build-depend on libkrb5-dev & set HasKrb5,
Krb5Includes & Krb5Libraries -
http://cgi.sfu.ca/~jdbates/tmp/xorg/200602230/patch

There is probably a better place for HasKrb5, Krb5Includes &
Krb5Libraries - but I couldn't find where?

Many thanks - Jack

>On Tue, Apr 13, 2004 at 03:41:20AM -0700, ms419 <at> freezone.co.uk
>wrote:
>> I gather that XFree86 for Debian is not compiled with Kerberos
>> support?
>[...]
>
>Historical reasons and inertia.  Way back in the day, crypto wasn't in
>main, and having XFree86 depend on libraries outside of non-us to build
>and work would have meant putting XFree86 itself into non-us.
>
>Since crypto has gone into main, the primary reason is lack of anyone
>with the right combination of clues, will, and working patches.
>
>> Before I go galloping off to recompile it, I should probably ask if 
>> this is appropriate? Why isn't  XFree86 available w/ Kerberos?
>
>No good reason; please do engage in this experiment and report your
>findings back to this list if you'd like.
>
>I personally have no objection to including Kerberos support in XFree86,
>and I doubt anyone else does either.  If they do, now's the time to
>speak up.  :)
>
>> I'm not the X guru I'd like to be ... I want to authorize xclients to 
>> connect to a remote xserver, provided they are run by the same person 
>> who's running the server. I have a working Kerberos setup and read 
>> about the "MIT-KERBEROS-5" security mechanism in the man pages. It's 
>> not like I've used this mechanism before ... I thought I'd give it a 
>> go. Not much help from Google. Can you spot any problems already?
>
>I tried to educate myself on this a few years ago, but failed for the
>same reasons you apparently did.  There's just not many people who both
>use anything more than MIT-MAGIC-COOKIE-1 or XDM-AUTHORIZATION-1, and
>have documented their experiences publicly enough for Google to snag
>them.
>
>-- 
>G. Branden Robinson                |        People with power understand
>Debian GNU/Linux                   |        exactly one thing: violence.
>branden <at> debian.org                 |        -- Noam Chomsky
>http://people.debian.org/~branden/ |

Attachment: signature.asc
Description: Digital signature


Reply to: