RE: CVS Source Code

[Jones, Larry]

There is no validation, people porting CVS to a particular OS are expected to 
set those configuration variables sanely (or fix the code). That said, the code 
is still suspect and may even be wrong; there are comments in the most recent 
version (for both 1.11 and 1.12) that you may find interesting 
(http://cvs.savannah.nongnu.org/viewvc/?root=cvs).
 
-Larry

-----Original Message-----
From: bug-cvs-bounces+lawrence.jones=siemens.com@nongnu.org 
[mailto:bug-cvs-bounces+lawrence.jones=siemens.com@nongnu.org] On Behalf Of 
Flávio Medeiros
Sent: Thursday, June 06, 2013 10:08 PM
To: bug-cvs@gnu.org
Subject: CVS Source Code

Hello,

I'm studying the source code of CVS. I found an interesting things and I'm
really excited to get an answer from a real developer. I found the
following code (file buffer.c):

#ifdef (SHUTDOWN_SERVER)
if (current_parsed_root->method != server_method)
#endif
#ifndef (NO_SOCKET_TO_FD)
{
if (S_ISSOCK (s.st_mode))
shutdown (fileno (bc->fp), 0);
}
#endif
#ifdef (START_RSH_WITH_POPEN_RW)
  else if (pclose (bc->fp) == EOF){
error (1, errno, "closing connection to
%s",current_parsed_root->hostname); closefp
= 0;
}
#endif

It is clear to me that we can generate invalid source codes depending on
the configuration used. For example, if we have SHUTDOWN_SERVER not defined
and START_RSH_WITH_POPEN_RW defined. Is it a valid configuration? Is there
any place on the source code that you ensure that (START_RSH_WITH_POPEN_RW)
requires SHUTDOWN_SERVER and NO_SOCKET_TO_FD. Is it an error?


Thanks in advance. I'm looking forward to hear from you.

Best regards,
Flávio
_______________________________________________
Bug-cvs mailing list
Bug-cvs@nongnu.org
https://lists.nongnu.org/mailman/listinfo/bug-cvs