This is the mail archive of the libc-alpha@sources.redhat.com mailing list for the glibc project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]

Re: (usagi-core 00349) Re: getnameinfo()

To: drepper at redhat dot com
Subject: Re: (usagi-core 00349) Re: getnameinfo()
From: Hideaki YOSHIFUJI ($B5HF#1QL@(B ) <yoshfuji at ecei dot tohoku dot ac dot jp>
Date: Fri, 10 Nov 2000 09:22:47 +0900
Cc: jk at csuchico dot edu, martin at loewis dot home dot cs dot tu-berlin dot de, yoshfuji at linux-ipv6 dot org, usagi-core at linux-ipv6 dot org, libc-alpha at sources dot redhat dot com
References: <m33dh2xrlx.fsf@otr.mynet.cygnus.com><20001109150457.A9952@north.csuchico.edu><m366lwlp0i.fsf@otr.mynet.cygnus.com>

In article <m366lwlp0i.fsf@otr.mynet.cygnus.com> (at 09 Nov 2000 15:11:25 -0800), Ulrich Drepper <drepper@redhat.com> says:

> I'm using openssh 2.2.0p1 with glibc 2.1.94 and all is fine.  It's
> anyway too late for 2.2 now.

My report for getnameinfo() bugs is not for the first time.

1. bugs in scope_id support was submitted to the GNATS long time ago.
   Code I submitted should have been correct, but you broke it.
   I addressed the problem, but you ignored my reports (or me :-p).

2. did you receive these reports dated September 27th, didn't you?
   (These reports were not perfect, though.)
   If the bug reporting system had been down, you would have had to warn.

-- 
Hideaki YOSHIFUJI @ USAGI Project  <yoshfuji@linux-ipv6.org>
PGP5i FP: F731 6599 5EB2 BBA7 1515  1323 1806 A96F 5700 6B25

To: bugs at gnu dot org
Subject: [USERS 359] getnameinfo() NI_NUMERICSCOPE flag support
From: yoshfuji at linux-ipv6 dot org
Date: Wed, 27 Sep 2000 20:04:28 +0900
CC: users at jp dot ipv6 dot org
Reply-To: users at jp dot ipv6 dot org

>Submitter-Id:	net
>Originator:	YOSHIFUJI Hideaki
>Organization:
 USAGI Project / linux-ipv6.org
>Confidential:	no
>Synopsis:	getnameinfo() NI_NUMERICSCOPE support.
>Severity:	non-critical
>Priority:	medium
>Category:	libc
>Class:		support
>Release:	libc-2.x (CVS)
>Environment:
	
Host type: i386-pc-linux-gnu
System: Linux cerberus 2.2.16 #17 Fri Aug 4 11:33:10 JST 2000 i586 unknown
Architecture: i586

Addons: crypt linuxthreads nss-v1

Build CC: gcc
Compiler version: 2.95.2 20000220 (Debian GNU/Linux)
Kernel headers: UTS_RELEASE
Symbol versioning: yes
Build static: yes
Build shared: yes
Build pic-default: no
Build profile: yes
Build omitfp: no
Build bounded: no
Build static-nss: no
Stdio: libio

>Description:
	NI_NUMERICSCOPE support as defined in RFC2553bis.
>How-To-Repeat:
	
>Fix:
With this patch introduces NI_NUMERICSCOPE as defined in RFC2552bis.

--- resolv/netdb.h	Wed Aug  2 00:00:40 2000
+++ netdb.h	Wed Sep 27 17:41:27 2000
@@ -432,11 +432,12 @@
 # define NI_MAXHOST      1025
 # define NI_MAXSERV      32
 
-# define NI_NUMERICHOST	1	/* Don't try to look up hostname.  */
-# define NI_NUMERICSERV 2	/* Don't convert port number to name.  */
-# define NI_NOFQDN	4	/* Only return nodename portion.  */
-# define NI_NAMEREQD	8	/* Don't return numeric addresses.  */
-# define NI_DGRAM	16	/* Look up UDP service rather than TCP.  */
+# define NI_NUMERICHOST		0x01	/* Don't try to look up hostname.  */
+# define NI_NUMERICSERV 	0x02	/* Don't convert port number to name.  */
+# define NI_NOFQDN		0x04	/* Only return nodename portion.  */
+# define NI_NAMEREQD		0x08	/* Don't return numeric addresses.  */
+# define NI_DGRAM		0x10	/* Look up UDP service rather than TCP.  */
+# define NI_NUMERICSCOPE	0x20	/* Don't try to look up scope id. */
 
 /* Translate name of a service location and/or a service name to set of
    socket addresses.  */
--- inet/getnameinfo.c	Wed Sep 27 19:39:43 2000
+++ getnameinfo.c	Wed Sep 27 19:59:38 2000
@@ -302,8 +302,9 @@
 			char scopebuf[IFNAMSIZ + 1];
 			int ni_numericscope = 0;
 
-			if (IN6_IS_ADDR_LINKLOCAL (&sin6p->sin6_addr)
-			    || IN6_IS_ADDR_MC_LINKLOCAL (&sin6p->sin6_addr))
+			if (!ni_numericscope && 
+			    (IN6_IS_ADDR_LINKLOCAL (&sin6p->sin6_addr)
+			     || IN6_IS_ADDR_MC_LINKLOCAL (&sin6p->sin6_addr)))
 			  {
 			    if (if_indextoname (scopeid, scopebuf) == NULL)
 			      ++ni_numericscope;

===---===---===---===
Get help : send mail to ppserv@jp.ipv6.org, body is "help", no subject.
unsubscribe: send mail to ppserv@jp.ipv6.org, body is "unsubscribe", no subject.

To: bugs at gnu dot org
Subject: [USERS 360] IF_NAMESIZE is missing in net/if.h
From: yoshfuji at linux-ipv6 dot org
Date: Wed, 27 Sep 2000 20:08:36 +0900
CC: users at jp dot ipv6 dot org
Reply-To: users at jp dot ipv6 dot org

>Submitter-Id:	net
>Originator:	YOSHIFUJI Hideaki
>Organization:
 USAGI Project / linux-ipv6.org
>Confidential:	no
>Synopsis:	IF_NAMESIZE defined in RFC2553 is missing in net/if.h
>Severity:	serious
>Priority:	high
>Category:	libc
>Class:		sw-bug
>Release:	libc-2.x (CVS)
>Environment:
	
Host type: i386-pc-linux-gnu
System: Linux cerberus 2.2.16 #17 Fri Aug 4 11:33:10 JST 2000 i586 unknown
Architecture: i586

Addons: crypt linuxthreads nss-v1

Build CC: gcc
Compiler version: 2.95.2 20000220 (Debian GNU/Linux)
Kernel headers: UTS_RELEASE
Symbol versioning: yes
Build static: yes
Build shared: yes
Build pic-default: no
Build profile: yes
Build omitfp: no
Build bounded: no
Build static-nss: no
Stdio: libio

>Description:
	RFC2553's IF_NAMESIZE is missing in net/if.h
>How-To-Repeat:
	
>Fix:
--- sysdeps/gnu/net/if.h	Sat Aug 19 00:01:15 2000
+++ if.h	Wed Sep 27 18:50:05 2000
@@ -112,6 +112,7 @@
   {
 #define IFHWADDRLEN	6
 #define	IFNAMSIZ	16
+#define	IF_NAMESIZE	IFNAMSIZ
     union
       {
 	char ifrn_name[IFNAMSIZ];	/* Interface name, e.g. "en0".  */


===---===---===---===
Get help : send mail to ppserv@jp.ipv6.org, body is "help", no subject.
unsubscribe: send mail to ppserv@jp.ipv6.org, body is "unsubscribe", no subject.

To: bugs at gnu dot org
Subject: [USERS 361] getaddrinfo() does not allow PF_INET6/SOCK_RAW/IPPROTO_ICMPV6 set
From: yoshfuji at linux-ipv6 dot org
Date: Wed, 27 Sep 2000 20:17:35 +0900
CC: users at jp dot ipv6 dot org
Reply-To: users at jp dot ipv6 dot org

>Submitter-Id:	net
>Originator:	YOSHIFUJI Hideaki
>Organization:
 USAGI Project / linux-ipv6.org
>Confidential:	no
>Synopsis:	getaddrinfo() does not allow PF_INET6/SOCK_RAW/IPPROTO_ICMPV6 set
>Severity:	serious
>Priority:	high
>Category:	libc
>Class:		sw-bug
>Release:	libc-2.x (CVS)
>Environment:
	
Host type: i386-pc-linux-gnu
System: Linux cerberus 2.2.16 #17 Fri Aug 4 11:33:10 JST 2000 i586 unknown
Architecture: i586

Addons: crypt linuxthreads nss-v1

Build CC: gcc
Compiler version: 2.95.2 20000220 (Debian GNU/Linux)
Kernel headers: UTS_RELEASE
Symbol versioning: yes
Build static: yes
Build shared: yes
Build pic-default: no
Build profile: yes
Build omitfp: no
Build bounded: no
Build static-nss: no
Stdio: libio

>Description:
	getaddrinfo() won't accept PF_INET6 / SOCK_RAW / IPPROTO_ICMPV6 set.
	In fact, for PF_INET6 / SOCK_RAW set, it accepts only IPPROTO_RAW protocol.
	It should accept other protocols such as IPPROTO_ICMPV6, too.
>How-To-Repeat:
	
>Fix:
--- sysdeps/posix/getaddrinfo.c	Fri Sep  1 00:01:09 2000
+++ getaddrinfo.c	Wed Sep 27 16:37:01 2000
@@ -99,13 +99,14 @@
 
 /* Values for `protoflag'.  */
 #define GAI_PROTO_NOSERVICE	1
+#define GAI_PROTO_PROTOANY	2
 
 static struct gaih_typeproto gaih_inet_typeproto[] =
 {
   { 0, 0, NULL, 0 },
   { SOCK_STREAM, IPPROTO_TCP, (char *) "tcp", 0 },
   { SOCK_DGRAM, IPPROTO_UDP, (char *) "udp", 0 },
-  { SOCK_RAW, IPPROTO_RAW, (char *) "raw", GAI_PROTO_NOSERVICE },
+  { SOCK_RAW, 0, (char *) "raw", GAI_PROTO_PROTOANY|GAI_PROTO_NOSERVICE },
   { 0, 0, NULL, 0 }
 };
 
@@ -150,6 +151,7 @@
 	     && ((tp->protoflag & GAI_PROTO_NOSERVICE) != 0
 		 || (req->ai_socktype != 0 && req->ai_socktype != tp->socktype)
 		 || (req->ai_protocol != 0
+		     && !(tp->protoflag & GAI_PROTO_PROTOANY)
 		     && req->ai_protocol != tp->protocol)))
 	++tp;
 
@@ -311,6 +313,7 @@
       while (tp->name != NULL
 	     && ((req->ai_socktype != 0 && req->ai_socktype != tp->socktype)
 		 || (req->ai_protocol != 0
+		     && !(tp->protoflag & GAI_PROTO_PROTOANY)
 		     && req->ai_protocol != tp->protocol)))
 	++tp;
 
	

===---===---===---===
Get help : send mail to ppserv@jp.ipv6.org, body is "help", no subject.
unsubscribe: send mail to ppserv@jp.ipv6.org, body is "unsubscribe", no subject.

To: bugs at gnu dot org
Subject: [USERS 358] getnameinfo() with scope_id
From: yoshfuji at linux-ipv6 dot org
Date: Wed, 27 Sep 2000 19:55:30 +0900
CC: users at jp dot ipv6 dot org
Reply-To: users at jp dot ipv6 dot org

>Submitter-Id:	net
>Originator:	YOSHIFUJI Hideaki
>Organization:
 USAGI Project / linux-ipv6.org
>Confidential:	no
>Synopsis:	getnameinfo() does not work correctly with scope-id.
>Severity:	serious
>Priority:	high
>Category:	libc
>Class:		sw-bug
>Release:	libc-2.x (CVS)
>Environment:
	
Host type: i386-pc-linux-gnu
System: Linux cerberus 2.2.16 #17 Fri Aug 4 11:33:10 JST 2000 i586 unknown
Architecture: i586

Addons: crypt linuxthreads nss-v1

Build CC: gcc
Compiler version: 2.95.2 20000220 (Debian GNU/Linux)
Kernel headers: UTS_RELEASE
Symbol versioning: yes
Build static: yes
Build shared: yes
Build pic-default: no
Build profile: yes
Build omitfp: no
Build bounded: no
Build static-nss: no
Stdio: libio

>Description:
getnameinfo() does not work correctly.
1. We should allow users to run old binaries without scope-id.
   note: linux-kernel allow old binaries. (1st and 2nd chunk) 
2. scope-id support was 'dead'.
   even if it were alive, it would be wrong. (3rd, 4th and 5th chunk)

>How-To-Repeat:
/*
 * Copyright (C)2000 Hideaki YOSHIFUJI / USAGI Project.
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. Neither the name of the author, project nor the names of its contributors
 *    may be used to endorse or promote products derived from this software
 *    without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR, PROJECT AND CONTRIBUTORS ``AS IS''
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR, PROJECT OR CONTRIBUTORS BE 
 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 
 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 
 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 
 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 
 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 
 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF 
 * THE POSSIBILITY OF SUCH DAMAGE.
 */

/* gnitest_nasty.c
 *
 * this test program check whether getnameinfo() work with / without scope id;
 * both version should be allowed.
 */

#include <stdio.h>
#include <sys/types.h>
#include <sys/socket.h>
#include <netinet/in.h>
#include <arpa/inet.h>
#include <net/if.h>
#include <netdb.h>

struct __sockaddr_in6_rfc2133
  {
    __SOCKADDR_COMMON (sin6_);
    uint16_t sin6_port;
    uint32_t sin6_flowinfo;
    struct in6_addr sin6_addr;
  };
#ifdef NEED_SOCKADDR_IN6_RFC2553
struct __sockaddr_in6_rfc2553
  {
    __SOCKADDR_COMMON (sin6_);
    uint16_t sin6_port;
    uint32_t sin6_flowinfo;
    struct in6_addr sin6_addr;
    uint32_t sin6_scope_id;
  };
#endif


#define IN6ADDR			"ff02::1"
#define	IFNAME			"eth0"
#define _SCOPE_DELIMITER	"%"

int main(){
	struct __sockaddr_in6_rfc2133	sin6_rfc2133;
#ifndef NEED_SOCKADDR_IN6_RFC2553
	struct sockaddr_in6		sin6_rfc2553;
#else
	struct __sockaddr_in6_rfc2553	sin6_rfc2553;
#endif

	int ifindex;
	int res;

	char hbuf[NI_MAXHOST];

	ifindex = if_nametoindex(IFNAME);

	if (!ifindex){
		fprintf(stderr, "Panic: define IFNAME correctly\n", IFNAME);
		exit(4);
	}

	/* fill RFC2133 sockaddr_in6 */
	memset(&sin6_rfc2133, 0, sizeof(sin6_rfc2133));
	sin6_rfc2133.sin6_family = AF_INET6;
#ifdef SIN6_LEN
	sin6_rfc2133.sin6_len = sizeof(sin6_rfc2133);
#endif
	res = inet_pton(AF_INET6, IN6ADDR, &sin6_rfc2133.sin6_addr);
	if (!res){
		fprintf(stderr, "Panic: inet_pton() failed.\n");
		exit(4);
	}

	/* fill RFC2553 sockaddr_in6 */
	memset(&sin6_rfc2553, 0, sizeof(sin6_rfc2553));
	sin6_rfc2553.sin6_family = sin6_rfc2133.sin6_family;
#ifdef SIN6_LEN
	sin6_rfc2553.sin6_len = sin6_rfc2133.sin6_len;
#endif
	sin6_rfc2553.sin6_addr = sin6_rfc2133.sin6_addr;
	sin6_rfc2553.sin6_scope_id = ifindex;

	/* print result */
	res = getnameinfo((struct sockaddr *)&sin6_rfc2133, sizeof(sin6_rfc2133),
			  hbuf, sizeof(hbuf), NULL, 0,
			  NI_NUMERICHOST);
	if (res){
		printf("getnameinfo() for OLD sockaddr_in6{} FAILED: ret = %d\n", res);
	}
	else{
		printf("getnameinfo() for OLD sockaddr_in6{} succeeded: name=%s\n", hbuf);
		if (strcmp(hbuf, IN6ADDR) != 0){
			printf(" - but, result should be \"%s\"\n", IN6ADDR);
		}
	}
	res = getnameinfo((struct sockaddr *)&sin6_rfc2553, sizeof(sin6_rfc2553),
			  hbuf, sizeof(hbuf), NULL, 0,
			  NI_NUMERICHOST);
	if (res){
		printf("getnameinfo() for NEW sockaddr_in6{} FAILED: ret = %d\n", res);
	}
	else{
		printf("getnameinfo() for NEW sockaddr_in6{} succeeded: name=%s\n", hbuf);
		if (strcmp(hbuf, IN6ADDR _SCOPE_DELIMITER IFNAME) != 0){
			printf(" - result should be \"%s\"\n", IN6ADDR _SCOPE_DELIMITER IFNAME);
		}
	}

	exit(0);
}

>Fix:
1. We should allow users to run old binaries without scope-id.
   note: linux-kernel allow old binaries. (1st and 2nd chunk) 
2. scope-id support was 'dead'.
   even if it were alive, it would be wrong. (3rd, 4th and 5th chunk)

--- /pub/cvs/libc/inet/getnameinfo.c	Sun Apr  2 17:25:58 2000
+++ getnameinfo.c	Wed Sep 27 19:37:58 2000
@@ -63,6 +63,18 @@
 #endif /* min */
 
 
+/*
+ * RFC2133 defines sockaddr_in6 without sin6_scope_id.
+ */
+struct __sockaddr_in6_rfc2133
+  {
+    __SOCKADDR_COMMON (sin6_);
+    uint16_t sin6_port;		/* Transport layer port # */
+    uint32_t sin6_flowinfo;	/* IPv6 flow information */
+    struct in6_addr sin6_addr;	/* IPv6 address */
+  };
+
+
 static char *
 internal_function
 nrl_domainname (void)
@@ -189,7 +201,11 @@
 	return -1;
       break;
     case AF_INET6:
-      if (addrlen < sizeof (struct sockaddr_in6))
+      /*
+       * glibc 2.1.x and linux-2.2.x are based on old spec (RFC2133).
+       * We should allow users to run old binaries.
+       */
+      if (addrlen < sizeof (struct __sockaddr_in6_rfc2133))
 	return -1;
       break;
     default:
@@ -295,18 +311,25 @@
 
 		    c = inet_ntop (AF_INET6,
 				   (void *) &sin6p->sin6_addr, host, hostlen);
-		    if (addrlen > sizeof (struct sockaddr_in6)
+		    if (addrlen >= sizeof (struct sockaddr_in6)
 			&& (scopeid = sin6p->sin6_scope_id))
 		      {
 			/* Buffer is >= IFNAMSIZ+1.  */
 			char scopebuf[IFNAMSIZ + 1];
 			int ni_numericscope = 0;
+			size_t real_hostlen = __strnlen(host, hostlen);
+			size_t scopelen = 0;
 
 			if (IN6_IS_ADDR_LINKLOCAL (&sin6p->sin6_addr)
 			    || IN6_IS_ADDR_MC_LINKLOCAL (&sin6p->sin6_addr))
 			  {
-			    if (if_indextoname (scopeid, scopebuf) == NULL)
+			    if (if_indextoname (scopeid, &scopebuf[1]) == NULL)
 			      ++ni_numericscope;
+			    else
+			      {
+				scopebuf[0] = SCOPE_DELIMITER;
+				scopelen = __strnlen(scopebuf, sizeof(scopebuf));
+			      }
 			  }
 			else
 			  ++ni_numericscope;
@@ -314,8 +337,6 @@
 			if (ni_numericscope)
 			  {
 			    char *scopeptr = &scopebuf[1];
-			    size_t real_hostlen;
-			    size_t scopelen;
 
 			    scopebuf[0] = SCOPE_DELIMITER;
 			    scopelen = 1 + snprintf (scopeptr,
@@ -323,12 +344,10 @@
 						      + sizeof scopebuf
 						      - scopeptr),
 						     "%u", scopeid);
-
-			    real_hostlen = __strnlen (host, hostlen);
-			    if (real_hostlen + scopelen + 1 > hostlen)
-			      return -1;
-			    memcpy (host + real_hostlen, scopebuf, scopelen);
 			  }
+			if (real_hostlen + scopelen + 1 > hostlen)
+			  return -1;
+			memcpy (host + real_hostlen, scopebuf, scopelen);
 		      }
 		  }
 		else



===---===---===---===
Get help : send mail to ppserv@jp.ipv6.org, body is "help", no subject.
unsubscribe: send mail to ppserv@jp.ipv6.org, body is "unsubscribe", no subject.

To: bugs at gnu dot org
Subject: [USERS 362] CMSG_NXTHDR(msgh,NULL) support
From: yoshfuji at linux-ipv6 dot org
Date: Wed, 27 Sep 2000 20:23:37 +0900
CC: users at jp dot ipv6 dot org
Reply-To: users at jp dot ipv6 dot org

>Submitter-Id:	net
>Originator:	YOSHIFUJI Hideaki
>Organization:
 USAGI Project / linux-ipv6.org
>Confidential:	no
>Synopsis:	CMSG_NXTHDR(msgh,NULL) support
>Severity:	non-critical
>Priority:	medium
>Category:	libc
>Class:		support
>Release:	libc-2.x (CVS)
>Environment:
	
Host type: i386-pc-linux-gnu
System: Linux cerberus 2.2.16 #17 Fri Aug 4 11:33:10 JST 2000 i586 unknown
Architecture: i586

Addons: crypt linuxthreads nss-v1

Build CC: gcc
Compiler version: 2.95.2 20000220 (Debian GNU/Linux)
Kernel headers: UTS_RELEASE
Symbol versioning: yes
Build static: yes
Build shared: yes
Build pic-default: no
Build profile: yes
Build omitfp: no
Build bounded: no
Build static-nss: no
Stdio: libio

>Description:
	CMSG_NXTHDR(msgh,NULL) should return CMSG_FIRSTHDR(msgh)
	(RFC2553).
>How-To-Repeat:
	
>Fix:
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
This patch introduce RFC2292 / RFC2292bis extension:
if msgh == NULL, do CMSG_FIRSTHDR().

--- sysdeps/unix/sysv/linux/cmsg_nxthdr.c	Wed Nov 18 22:21:15 1998
+++ cmsg_nxthdr.c	Wed Sep 27 16:22:12 2000
@@ -23,6 +23,9 @@
 struct cmsghdr *
 __cmsg_nxthdr (struct msghdr *mhdr, struct cmsghdr *cmsg)
 {
+  if (cmsg == NULL)
+    return CMSG_FIRSTHDR(mhdr);
+
   if ((size_t) cmsg->cmsg_len < sizeof (struct cmsghdr))
     /* The kernel header does this so there may be a reason.  */
     return NULL;



===---===---===---===
Get help : send mail to ppserv@jp.ipv6.org, body is "help", no subject.
unsubscribe: send mail to ppserv@jp.ipv6.org, body is "unsubscribe", no subject.

To: bugs at gnu dot org
Subject: [USERS 363] last ancillary data may / may not have trailing pad
From: yoshfuji at linux-ipv6 dot org
Date: Wed, 27 Sep 2000 20:31:30 +0900
CC: users at jp dot ipv6 dot org
Reply-To: users at jp dot ipv6 dot org

>Submitter-Id:	net
>Originator:	YOSHIFUJI Hideaki
>Organization:
 USAGI Project / linux-ipv6.org
>Confidential:	no
>Synopsis:	CMSG_NXTHDR does not allow ancillary data object without pad.
>Severity:	critical
>Priority:	high
>Category:	libc
>Class:		sw-bug
>Release:	libc-2.x (CVS)
>Environment:
	
Host type: i386-pc-linux-gnu
System: Linux cerberus 2.2.16 #17 Fri Aug 4 11:33:10 JST 2000 i586 unknown
Architecture: i586

Addons: crypt linuxthreads nss-v1

Build CC: gcc
Compiler version: 2.95.2 20000220 (Debian GNU/Linux)
Kernel headers: UTS_RELEASE
Symbol versioning: yes
Build static: yes
Build shared: yes
Build pic-default: no
Build profile: yes
Build omitfp: no
Build bounded: no
Build static-nss: no
Stdio: libio

>Description:
	Linux kernel do not send the last pad.
	The lates spec allow this behavior; the last acillary 
	data element may / may not have trailing "pad".

RFC2292bis:
   |<--------------------------- msg_controllen -------------------------->|
   |                                  OR                                   |
   |<--------------------------- msg_controllen ----------------------->|
   |                                                                       |
   |<----- ancillary data object ----->|<----- ancillary data object ----->|
   |<------ min CMSG_SPACE() --------->|<------ min CMSG_SPACE() --------->|
   |                                   |                                   |
   |<---------- cmsg_len ---------->|  |<--------- cmsg_len ----------->|  |
   |<--------- CMSG_LEN() --------->|  |<-------- CMSG_LEN() ---------->|  |
   |                                |  |                                |  |
   +-----+-----+-----+--+-----------+--+-----+-----+-----+--+-----------+--+
   |cmsg_|cmsg_|cmsg_|XX|           |XX|cmsg_|cmsg_|cmsg_|XX|           |XX|
   |len  |level|type |XX|cmsg_data[]|XX|len  |level|type |XX|cmsg_data[]|XX|
   +-----+-----+-----+--+-----------+--+-----+-----+-----+--+-----------+--+
   ^
   |
   msg_control
   points here

>How-To-Repeat:
	
>Fix:
--- sysdeps/unix/sysv/linux/cmsg_nxthdr.c	Wed Nov 18 22:21:15 1998
+++ cmsg_nxthdr.c	Wed Sep 27 16:21:07 2000
@@ -31,7 +31,7 @@
 			     + CMSG_ALIGN (cmsg->cmsg_len));
   if ((unsigned char *) (cmsg + 1) >= ((unsigned char *) mhdr->msg_control
 				       + mhdr->msg_controllen)
-      || ((unsigned char *) cmsg + CMSG_ALIGN (cmsg->cmsg_len)
+      || ((unsigned char *) cmsg + cmsg->cmsg_len
 	  >= ((unsigned char *) mhdr->msg_control + mhdr->msg_controllen)))
     /* No more entries.  */
     return NULL;


===---===---===---===
Get help : send mail to ppserv@jp.ipv6.org, body is "help", no subject.
unsubscribe: send mail to ppserv@jp.ipv6.org, body is "unsubscribe", no subject.

Follow-Ups:
- Re: (usagi-core 00349) Re: getnameinfo()
  - From: Ulrich Drepper

References:
- Re: (usagi-core 00277) Re: glibc-2.1.97 fixes from USAGI Project
  - From: Ulrich Drepper
- getnameinfo()
  - From: John Kennedy
- Re: getnameinfo()
  - From: Ulrich Drepper

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]