Conference hydra::axp-developer

3444.0. "Iona Technologies - Point 24910" by KZIN::ASAP () Mon Apr 07 1997 05:34

    Company Name :  Iona Technologies - Point 24910
    Contact Name :  Aileen Cunningham
    Phone        :  +353 1 6625255
    Fax          :  +353 1 6625244
    Email        :  [email protected]
    Date/Time in :   7-APR-1997 09:33:53
    Entered by   :  Ian Chamberlin
    SPE center   :  REO

    Category     :  unix
    OS Version   :  3.2
    System H/W   :  


    Brief Description of Problem:
    -----------------------------

From:	RDGENG::MRGATE::"RDGMTS::PMDF::mail.dec.com::Santing"  7-APR-1997 07:55:56.83
To:	RDGENG::ASAP
CC:	
Subj:	FW: ESCALATION: POINT 24910, Iona Technologies

From:	NAME: Ben Santing <[email protected]@PMDF@INTERNET>
To:	NAME: '[email protected]' <IMCEAX400-c=US+3Ba=+20+3Bp=DIGITAL+3Bo=SBUEURMFG+3Bdda+3ASMTP=asap+40reo+2Emts+2Edec+2Ecom+3B@mail.dec.com@PMDF@INTERNET>

Hello -

POINT Log Number	 24910

Company Name 	Iona Technologies

Engineers name	Aileen Cunningham

Telephone Number 		+353 1 6625255

Fax Number		+353 1 6625244

E-mail Address	[email protected]

Operating System, Version	OSF1 V3.2 148

Platform			

Problem Statement		



I'm having problems using the gethostbyname_r system call on
OSF1 explicit V3.2 148 alpha with version 5.5 of the cxx C++ 
compiler.

The problem is demonstared by the following sample code.

The first call to gethostbyname_r works correctly but the 2nd call 
core dumps.

The program is compiled as follows

cxx -D_REENTRANT -threads -g hst.cc -ohst

and when exectued produces the following output.

./hst
res is 0 errno is 0 h_errno is 0
Segmentation fault (core dumped)

The stack trace at the moment of the crash is


 decladebug hst
Welcome to the Ladebug Debugger Version 4.0-14A
------------------
object file name: hst
Reading symbolic information ...done
(ladebug) t
Program hasn't started running
(ladebug) r
res is 0 errno is 0 h_errno is 0
Thread received signal SEGV
stopped at [???: ??? 0x400000800000000]
(ladebug) t
>0  0x400000800000000 in ???
(ladebug)


Regards,

	Aileen Cunningham

--------------------------- TEST PROGRAM 
--------------------------------------


 #include <netdb.h>
 #include <iostream.h>
 #include <errno.h>

void connect(char *name)
{
  struct hostent hptr;
  struct hostent_data hdptr;
  int res;

  memset (&hdptr, 0, sizeof(hdptr));
  res = gethostbyname_r (name, &hptr, &hdptr);
  cout << "res is " << res << " errno is " << errno << " h_errno is " 
<<
h_errno
 << endl;
}

int main()
{
  connect("explicit.dublin.iona.ie");
  connect("true.dublin.iona.ie");
  return 0;
}


Regards,


Ben Santing


	     In replying, please use [email protected]




RFC-822-headers:
Received: from reoexc1.reo.dec.com by rg71rw.reo.dec.com (PMDF V5.0-7 #15552)
 id <[email protected]> for [email protected]; Mon,
 07 Apr 1997 07:50:36 +0100
Received: by reoexc1.reo.dec.com with SMTP
 (Microsoft Exchange Server Internet Mail Connector Version 4.0.994.63)
 id <[email protected]>; Mon, 07 Apr 1997 07:53:15 +0100
X-Mailer: Microsoft Exchange Server Internet Mail Connector Version 4.0.994.63

Aileen,
	The problem you describe with gethostbyname_r is a known problem with
Ditital Unix V3.2x.

	There is a patch available; I think the Patch Id for Digital Unix V3.2c
(which you are using) is OSF350-108, but will be different for other V3.2
versions. You should be able to get this from your local Customer Support
Centre, or our www site http://www.service.digital.com, if you have a service
contract. I dont believe its available under the public patches tree on our www
site.

	The problem is fixed in Digital Unix V4.0/A/B releases, so we would
recommend you upgrade.

regards,
Ian Chamberlin,

Digital Equipment Co, Software Partner Engineering.

As of 17 April, the partner had difficulty finding the patch, either on 

http://wis.europe.digital.com:2429,

or from the Irish and UK CSC's

On checking, I found that patches had been consolidated, so OSF350-108 is now
paart of OSF350-279, but surely the CSC could have found this as well?

Haven't heard anything since 17 April, so assume partner has the patch, and
close the call.