Build on HP 11.31

vir vir vitrou2004 at yahoo.com
Mon May 17 16:35:17 EDT 2010


no warning for Linux 64
and I am chaging in those 3 functions
OM_uint32 to   unsigned int  I got a new errors
 
"pkinit_srv.c", line 59: error #2757: variable "krb5_error_code" is not a type name
  static krb5_error_code
         ^
"pkinit_srv.c", line 65: error #2757: variable "krb5_error_code" is not a type name
                      krb5_error_code err_code,
                      ^
"pkinit_srv.c", line 68: error #2065: expected a ";"
      krb5_error_code retval = KRB5KRB_ERR_GENERIC;
                      ^
"pkinit_srv.c", line 74: error #2020: identifier "retval" is undefined
          retval = pkinit_create_td_trusted_certifiers(context,
          ^
"pkinit_srv.c", line 95: error #2020: identifier "retval" is undefined
      return retval;
             ^
"pkinit_srv.c", line 98: error #2757: variable "krb5_error_code" is not a type name
  static krb5_error_code
         ^
"pkinit_srv.c", line 107: error #2065: expected a ";"
      krb5_error_code retval = 0;



--- On Mon, 5/17/10, Douglas E. Engert <deengert at anl.gov> wrote:


From: Douglas E. Engert <deengert at anl.gov>
Subject: Re: Build on HP 11.31
To: "vir vir" <vitrou2004 at yahoo.com>
Cc: kerberos at mit.edu
Received: Monday, May 17, 2010, 1:09 PM




vir vir wrote:
> Hi Douglas,
> thank you for your response.
>  I saw that too.
> Yes, I  compile for 64 bit
> It did work for Linux 64 .

Any warning on Linux 64? The idea is OM_uint32 should be
an unsigned integer that is at least 32 bits. Maybe on the Linux 64bit,
it is actually 64 bits, where as on HPUX 64 is it 32.

You may want to ask on the IETF kitten mail list too as this came up in 2004:

http://www.mail-archive.com/ietf-cat-wg@lists.stanford.edu/msg00086.html

Can you track down the definition of OM_uint32 on your system?

IMHO the issue is the spnego_mech.c should not be using OM_uint32, for these 3 routines
but unsigned int, as it redefines a prototype which uses unsigned int.

>  Instead of OM_uint32 I define in the function  unsigned int but got a new compilation errors
>   
> 
> --- On Mon, 5/17/10, Douglas E. Engert <deengert at anl.gov> wrote:
> 
> 
> From: Douglas E. Engert <deengert at anl.gov>
> Subject: Re: Build on HP 11.31
> To: "vir vir" <vitrou2004 at yahoo.com>
> Cc: kerberos at mit.edu
> Received: Monday, May 17, 2010, 11:30 AM
> 
> 
> 
> 
> vir vir wrote:
>> Hi,
>>   I am trying to build on HP 11.31 I ran ./configure and then ran make.
>> and got compile error.
>>   
>>> From the MIT website http://web.mit.edu/Kerberos/krb5-1.8/krb5-1.8.1/doc/krb5-install.html#HPUX
>> it has been tested only on HP 10.20
>> does it mean that MIT Kereberos doesn't support 11.31?
>>   Please find att. the output  make file ( with the error compilations)
> 
> I don't have any HP systems any more, but it looks like in spnego_mech.c
> 
>     84  /* der routines defined in libgss */
>     85  extern unsigned int gssint_der_length_size(OM_uint32);
>     86  extern int gssint_get_der_length(unsigned char **, OM_uint32, unsigned int*);
>     87  extern int gssint_put_der_length(OM_uint32, unsigned char **, unsigned int);
>     88
> 
> The OM_uint32 should all be unsigned int. For some reason the
> HP compiler does not thing so. Are you compiling for 64 bit?
> 
> This might be more then a HP problem...
> 
>>   Thanks,
>> Alex
>>   
>> 
>> ------------------------------------------------------------------------
>> 
>> ________________________________________________
>> Kerberos mailing list           Kerberos at mit.edu
>> https://mailman.mit.edu/mailman/listinfo/kerberos
> 

-- 
Douglas E. Engert  <DEEngert at anl.gov>
Argonne National Laboratory
9700 South Cass Avenue
Argonne, Illinois  60439
(630) 252-5444





More information about the Kerberos mailing list