New Samba/VMS Release : build 20040908

John E. Malmberg wb8tyw at qsl.net
Thu Sep 9 13:22:14 GMT 2004


In article <CF0913E9C3D53D4E94DE47FDE25C633D05048920 at hermes.cofiroute.com>,
COLLOT Jean-Yves <jean-yves.collot at cofiroute.fr> writes:
>
> Now, the news.

Good, I will pull down the new kit immediately.

I am trying to find out two things.

1. I can not get guest access to work.

2. When I turn up the debugging level, the server takes too long and the
client times out.  I am also not seeing anything in the debug messages
from the server about it trying to access the guest account.

> 1. A new "VMS Record Format" share parameter has been added. It defines the
> RMS record type of the files created by Samba, i.e. the way VMS will see
> those files. The default is "stream", but it can be forced to "stream_lf".

I hope that it's decision is based on a file extension.  SAMBA 2.0.6 had
this feature, but it was global by file extension, not by share type.

The main reason for this is that SAMBA 2.0.6 was trying to make sure that text
files looked correctly on either the client or the server, regardless of the
original format.  I never fully achieved that goal.

2.0.6 used the format of a previous generation of a file to determine
the record type on creation, but if no previous generation existed, it
would look in the "template" directory for a file with that extension, and
use that instead.

> 7. The code has been checked in order to remove the /STANDARD=VAXC compiler
> option in COMPILE.COM, and to remove irrelevant warnings and informations.
> However, depending upon the version of the DECC compiler you use, you may
> encounter a lot of informational messages if you choose to recompile
> Samba/VMS from the sources. You may prefer to use the old procedure, which
> is now included under the name COMPILE_STD_VAXC.COM

When you did this check, did you fix the bug in blocking.c that the
compiler flags if you have the /warn=enable=(level4,questcode) set?

In the old code, since blr->expire_time is of unsigned type time_t, this
expression will never be true, so locks can never expire.

This appears to be a bug in the UNIX distribution for any compiler were
time_t is unsigned.

As of yet, I do not know the impact of this bug on the VMS platform.

$ diff [-.smbd]blocking.c/ignore=trail
************
File PROJECT_ROOT:[SAMBA-2_2_8-SRC.SOURCE.SMBD]BLOCKING.C;3
  644       if((blr->expire_time != (time_t)-1) && (blr->expire_time <= t)) {
  645         /*
******
File PROJECT_ROOT:[SAMBA-2_2_8-SRC.SOURCE.SMBD]BLOCKING.C;2
  644       if((blr->expire_time != -1) && (blr->expire_time <= t)) {
  645         /*
************

Adding the following to config.h should remove most of the noise for
a current C compiler.

 /* Disable noise about compiler extensions */
/*=========================================*/
#pragma message disable pragma
#pragma message disable hexoctunsign
#pragma message disable valuepres

#pragma message disable intconcastsgn /* Commonly assumed correct */
#pragma message disable questcompare2 /* Commonly assumed correct */

I have not yet inspected all the questcompare2 warnings to verify that
they are indeed nothing to be concerned about.

I have still not completed a build with that warning level as I am adjusting
the code to use either a 64 bit or 48 bit ino_t, and fixing all the minor
informationals except those messages above.

Also most of the #ifdef __ALPHA need to be change to #ifndef __VAX,
otherwise the IA64 binaries will be built wrong.

There is also a problem in the includes.h where two needed include files
are not included before proto.h.  In this case the C compiler will do
the correct code even though diagnostics are generated.

#include "nsswitch/sys_nss.h" /* needed by proto.h */

#include "nsswitch/winbindd_nss.h" /* needed by proto.h */


Also, with 8.2, a number of new features and fixes are expected to be in
the CRTL, including a standards compliant stat()/fstat() and statvfs(),
including handling the dev_t type the way SAMBA expects.

The [.vms]grp.h, pwd.h, and utime.h supplied should only be needed
prior to 7.3.

Regards,
-John
wb8tyw at qsl.net
Personal Opinion Only



More information about the samba-vms mailing list