Alexander Dupuy | 1 Sep 2006 22:44
Picon
Gravatar

some (minor, libopts) issues building 3.0beta11 on FC4 i686

I've registered with Trac on the tcpreplay.synfin.net site, but I guess 
you have to manually enable submission privileges for me to be able to 
submit tickets.

I got the following warning during compilation of the libopts library:
In file included from ../../libopts/libopts.c:2:
../../libopts/compat/compat.h:245:1: warning: "strchr" redefined
In file included from /usr/include/string.h:417,
                 from /usr/include/memory.h:30,
                 from ../../libopts/compat/compat.h:127,
                 from ../../libopts/libopts.c:2:
/usr/include/bits/string2.h:396:1: warning: this is the location of the 
previous definition

The problem here is that libopts is using #ifdef HAVE_STRCHR, but the 
config.h file used (generated as part of the tcpreplay/configure run) 
doesn't actually check for strchr (which is present on this system, 
really most nowadays as it is ANSI C).  Fixes would be to change the 
sense of the libopts test or add a check for strchr() in the tcpreplay 
configure.in:

--- configure.in~       2006-09-01 14:18:03.000000000 -0400
+++ configure.in        2006-09-01 16:18:18.000000000 -0400
 <at>  <at>  -119,7 +119,7  <at>  <at> 
 AC_FUNC_MEMCMP
 AC_TYPE_SIGNAL
 AC_FUNC_VPRINTF
-AC_CHECK_FUNCS([gettimeofday ctime memset regcomp strdup strerror 
strtol strncpy strtoull poll ntohll mmap snprintf vsnprintf])
+AC_CHECK_FUNCS([gettimeofday ctime memset regcomp strchr strdup 
(Continue reading)

Aaron Turner | 1 Sep 2006 23:17
Picon

Re: some (minor, libopts) issues building 3.0beta11 on FC4 i686

On 9/1/06, Alexander Dupuy <alex.dupuy <at> mac.com> wrote:
> I've registered with Trac on the tcpreplay.synfin.net site, but I guess
> you have to manually enable submission privileges for me to be able to
> submit tickets.

Hmm... that would suck, I'll have to look into that... thanks for the heads up.

[snip]

Anyways, what options did you pass to configure ??

Thanks,
Aaron

--

-- 
Aaron Turner
http://synfin.net/

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
Aaron Turner | 1 Sep 2006 23:19
Picon

Re: some (minor, libopts) issues building 3.0beta11 on FC4 i686

Actually, on second thought, can you just send me your config.log?

Thanks,
Aaron

On 9/1/06, Aaron Turner <synfinatic <at> gmail.com> wrote:
> On 9/1/06, Alexander Dupuy <alex.dupuy <at> mac.com> wrote:
> > I've registered with Trac on the tcpreplay.synfin.net site, but I guess
> > you have to manually enable submission privileges for me to be able to
> > submit tickets.
>
> Hmm... that would suck, I'll have to look into that... thanks for the heads up.
>
> [snip]
>
> Anyways, what options did you pass to configure ??
>
> Thanks,
> Aaron
>
> --
> Aaron Turner
> http://synfin.net/
>

--

-- 
Aaron Turner
http://synfin.net/

-------------------------------------------------------------------------
(Continue reading)

Alexander Dupuy | 2 Sep 2006 00:37
Picon
Gravatar

MTU length check and some other minor compilation warnings

First, a minor spelling check - the error message in tcpedit/l2_rewrite.c:

Packet length (%u) is greater then MTU (%u);

is incorrect, it should be "greater than" (with an 'a').  This error 
also occurs in the FAQ (along with a misspelling of "transmission" as 
"transmition").

Spelling errors aside, it does seem that tcpreplay 3.0 doesn't have the 
problem that tcpreplay 2.3 has with full-MTU VLAN tagged packets 
generating a spurious warning about "packet length (1518) exceeds MTU" 
so this seems to be a distinct improvement.  Thanks!

There were some compilation warnings (apart from the libopts ones I 
reported separately):

One is potentially serious, and could result in a segfault in an 
(admittedly unlikely) error condition:

tcpreplay/src/common/sendpacket.c:436: warning: ‘sp’ is used 
uninitialized in this function

This is because the SIOCGIFHWADDR error handling code precedes the 
safe_malloc allocation of sp.

Another, unlikely to cause problems except for older distributions:

tcpreplay/src/common/get.c:237:24: warning: extra tokens at end of 
#ifndef directive

(Continue reading)

Alexander Dupuy | 5 Sep 2006 21:50
Picon
Gravatar

Problem building 3.0beta11 on FC5 x86_64

On an FC5 x86_64 system with libnet-devel installed, I'm getting the 
following compilation error:

In file included from pcaptools/tcpreplay/src/tcpedit/../mac.c:32:
pcaptools/tcpreplay/src/tcpedit/../defines.h:44:39: error: 
/usr/local/include/libnet.h: No such file or directory

The problem is that libnet is detected by configure, but it is installed 
in /usr (/usr/include/libnet.h and /usr/lib64/libnet.a) and not 
/usr/local; config.log fragments include:

configure:24390: checking for libnet
configure:24418: result: /usr
...
LNETINC='/usr/include/libnet.h'
LNETLIB='/usr/lib64/libnet.a'

However, src/defines.h has the following:

#ifdef HAVE_LIBNET
#include "/usr/local/include/libnet.h"
#endif

There are a few solutions to this. 

One is to modify the configure script so that it doesn't look in 
/usr/include and /usr/lib (or /usr/lib64 in this case) for libnet, and 
to only use a version in /usr/local/.  But this is pretty lame, and 
requires extra work to prevent configure from accepting a --with-libnet= 
option with any value other than /usr/local/.
(Continue reading)

Aaron Turner | 6 Sep 2006 07:08
Picon

Re: Problem building 3.0beta11 on FC5 x86_64

Since you're using Linux, I would suggest you use: ./configure
--disable-libnet.  This will force tcpreplay to use Linux's PF_PACKET
for injection which in effect removes the middleman (libnet).

Libnet IMHO sucks and in 3.0 will likely be considered deprecated
(meaning, most likely you'll have to manually set --enable-libnet if
you don't have PF_PACKET, BPF or a 0.9.x version of libpcap installed.

Anyways, regarding the defines.h error.  I can't explain it.  It's
already generated via configure & defines.h.in and uses the value of
LNETINC to include the file.   Are you sure you don't have two copies
of libnet installed?  If not, a copy of your config.log from that
machine would be helpful.

-Aaron

On 9/5/06, Alexander Dupuy <alex.dupuy <at> mac.com> wrote:
> On an FC5 x86_64 system with libnet-devel installed, I'm getting the
> following compilation error:
>
> In file included from pcaptools/tcpreplay/src/tcpedit/../mac.c:32:
> pcaptools/tcpreplay/src/tcpedit/../defines.h:44:39: error:
> /usr/local/include/libnet.h: No such file or directory
>
> The problem is that libnet is detected by configure, but it is installed
> in /usr (/usr/include/libnet.h and /usr/lib64/libnet.a) and not
> /usr/local; config.log fragments include:
>
> configure:24390: checking for libnet
> configure:24418: result: /usr
(Continue reading)

Aaron Turner | 6 Sep 2006 07:10
Picon

Re: MTU length check and some other minor compilation warnings

thanks for the bug reports.  I'll see about getting fixes in the next release.

-Aaron

On 9/1/06, Alexander Dupuy <alex.dupuy <at> mac.com> wrote:
> First, a minor spelling check - the error message in tcpedit/l2_rewrite.c:
>
> Packet length (%u) is greater then MTU (%u);
>
> is incorrect, it should be "greater than" (with an 'a').  This error
> also occurs in the FAQ (along with a misspelling of "transmission" as
> "transmition").
>
> Spelling errors aside, it does seem that tcpreplay 3.0 doesn't have the
> problem that tcpreplay 2.3 has with full-MTU VLAN tagged packets
> generating a spurious warning about "packet length (1518) exceeds MTU"
> so this seems to be a distinct improvement.  Thanks!
>
> There were some compilation warnings (apart from the libopts ones I
> reported separately):
>
> One is potentially serious, and could result in a segfault in an
> (admittedly unlikely) error condition:
>
> tcpreplay/src/common/sendpacket.c:436: warning: 'sp' is used
> uninitialized in this function
>
> This is because the SIOCGIFHWADDR error handling code precedes the
> safe_malloc allocation of sp.
>
(Continue reading)

seth | 21 Sep 2006 21:12
Picon

pcap conflicting types

Hi,

I would appreciate help on Make failure:

In file included from tcpreplay.c:68:
fakepcap.h:40: error: conflicting types for 'pcap_datalink_val_to_description'/usr/local/include/pcap.h:245: error: previous declaration of 'pcap_datalink_val_to_description' was here
fakepcap.h:40: error: conflicting types for 'pcap_datalink_val_to_description'/usr/local/include/pcap.h:245: error: previous declaration of 'pcap_datalink_val_to_description' was here

The conflicting types are:
fakepcap.h,68: char *pcap_datalink_val_to_description(int dlt);
pcap.h, 245:     const char *pcap_datalink_val_to_description(int);

Sys info:

tcpreplay-2.3.5
pcap.h,v 1.52.2.5 2005/07/07
libpcap-0.9.4-sol10-x86-local
tcpdump-3.9.4-sol10-x86-local

Solaris 10x86 6-06
Laptop IBM T43

Thank you

Seth



-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Tcpreplay-users mailing list
Tcpreplay-users <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/tcpreplay-users
Aaron Turner | 21 Sep 2006 21:32
Picon

Re: pcap conflicting types

Hi Seth,

Looks like ./configure is confused about finding your installed
version of libpcap.  Can you send me your config.log?

-Aaron

-- 
Aaron Turner
http://synfin.net/

On 9/21/06, seth <s1redh <at> gmail.com> wrote:
> Hi,
>
>  I would appreciate help on Make failure:
>
>  In file included from tcpreplay.c:68:
>  fakepcap.h:40: error: conflicting types for
> 'pcap_datalink_val_to_description'/usr/local/include/pcap.h:245:
> error: previous declaration of
> 'pcap_datalink_val_to_description' was here
>  fakepcap.h:40: error: conflicting types for
> 'pcap_datalink_val_to_description'/usr/local/include/pcap.h:245:
> error: previous declaration of
> 'pcap_datalink_val_to_description' was here
>
>  The conflicting types are:
>  fakepcap.h,68: char *pcap_datalink_val_to_description(int
> dlt);
>  pcap.h, 245:     const char
> *pcap_datalink_val_to_description(int);
>
>  Sys info:
>
> tcpreplay-2.3.5
>  pcap.h,v 1.52.2.5 2005/07/07
>  libpcap-0.9.4-sol10-x86-local
>  tcpdump-3.9.4-sol10-x86-local
>
>  Solaris 10x86 6-06
>  Laptop IBM T43

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
seth | 21 Sep 2006 21:44
Picon

Re: pcap conflicting types

Hi Aaron,

Here it is.

Thanks,

Seth

On 9/21/06, Aaron Turner <synfinatic <at> gmail.com> wrote:
Hi Seth,

Looks like ./configure is confused about finding your installed
version of libpcap.  Can you send me your config.log?

-Aaron

--
Aaron Turner
http://synfin.net/


On 9/21/06, seth <s1redh <at> gmail.com> wrote:
> Hi,
>
>  I would appreciate help on Make failure:
>
>  In file included from tcpreplay.c :68:
>  fakepcap.h:40: error: conflicting types for
> 'pcap_datalink_val_to_description'/usr/local/include/pcap.h:245:
> error: previous declaration of
> 'pcap_datalink_val_to_description' was here
>  fakepcap.h:40: error: conflicting types for
> 'pcap_datalink_val_to_description'/usr/local/include/pcap.h:245:
> error: previous declaration of
> 'pcap_datalink_val_to_description' was here
>
>  The conflicting types are:
>  fakepcap.h,68: char *pcap_datalink_val_to_description(int
> dlt);
>  pcap.h, 245:     const char
> *pcap_datalink_val_to_description(int);
>
>  Sys info:
>
> tcpreplay-2.3.5
>  pcap.h,v 1.52.2.5 2005/07/07
>  libpcap-0.9.4-sol10-x86-local
>  tcpdump-3.9.4-sol10-x86-local
>
>  Solaris 10x86 6-06
>  Laptop IBM T43

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Tcpreplay-users mailing list
Tcpreplay-users <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/tcpreplay-users

Attachment (config.log): text/x-log, 56 KiB
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Tcpreplay-users mailing list
Tcpreplay-users <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/tcpreplay-users

Gmane