artasoftkey / ptpv2d Goto Github PK

1. Compile ptp2d with on a gcc 4.4.3 on Linux.  Fails compile due to INT_MAC 
not being defined.  

I added this statement in constants_dep.h to fix the problem.  

#ifndef INT_MAX 
#define INT_MAX __INT_MAX__
#endif

2. A compile also causes a warning with a */ included in a comment.  

A */ should be added at the end of the 2nd line.  

/* System Dependent (AKB: Moved from constants.h to constants_dep.h */
/* Defines for MPC831X implementation with Hardware timestampinng

What steps will reproduce the problem?
1.PTP Packets are send from a user space application to a TAP interface.
2.Packet Rx count is incremented at TAP interface.
3. Now when the PTPv2 stack is bound to "tap0" using "-b" the packets are not 
received at the PTPv2 stack.

What is the expected output? What do you see instead?
Packets are exoected to be recived from teh TAP0 interface.

What version of the product are you using? On what operating system?
I am using Linux.

Please provide any additional information below.
The same with eth0 interface works fine.

I am attempting to test the Windows version as a slave using a Linux based 
master.    I currently have the Linux version working with a master and several 
slaves.

I don’ t know any reason why this configuration would not work, but maybe you 
guys know otherwise.

I have the master running on Linux using protocol version 1, with the following 
command line:

Ptpd –D –p –u192.168.1.102 

so it is operating as the master and sending PTPD v1 Synch messages to both the 
group address and the address of the Windows based slave.

The command line I am using for the Windows based slave is 

“ptpv2d  -D –g  -z 2”.   

The attachment is the log file from the Windows slave.

I can observe via Wireshark the PTP messages from the Linux PTP master

From what I can tell from the log files, it looks like ptpv2d is receiving and 
decoding the Synch Message, but stays locked in master mode.

Maybe you can point me in the right direction for getting this working under 
Windows with either a Linux or Windows master.

What steps will reproduce the problem?
1. run ptpv2d with -8 option
2. run Wireshark or other LAN analyzer
3. observe IPv4 Multicast join packets sent on network

What is the expected output? What do you see instead?

A) It would be better that if the encapsulation selected is Direct Ethernet 
that no IPv4 packets are generated.

Please use labels and text to provide additional information.

1) In the case of the -8 option, it would probably be better if only a raw 
socket was opened and the opening of IPv4 UDP sockets was skipped.  This would 
avoid the Multicast join message being generated and also would make it so that 
IPv4 packets would be ignored.
2) Downside is that I had some thoughts about allowing future versions of 
ptpv2d auto-detect the best grandmaster regardless of protocol type and then 
adapt to it.

Comments and thoughts on this are welcome.

This bug is mostly cosmetic.

What steps will reproduce the problem?
2. svn co (current version)
3. make


What is the expected output? What do you see instead?
During compilation I get tons of warnings in protocol.c and net.c:
"warning: format ‘%d’ expects type ‘int’, but argument XXX has type 
‘ssize_t’"


What version of the product are you using? On what operating system?
Using Fedora Linux 14.x86_64, GCC 4.5.1, gmake 3.82, subversion 1.6.16
svn info: Revision: 27

Please provide any additional information below.
I fixed all warning using %zd or %zu instead of %d and %u when displaying 
ssize_t and size_t.
Attached file: in ptpv2d-dev: svn di > fixed-warnings-rev27.diff