Experimental Physics and
Industrial Control System

Mark Rivers via Tech-talk <tech-talk at aps.anl.gov> · Sat, 20 Jun 2020 12:01:26 +0000

> Could it be, that this commit removed the functionality that we need for  Windows ?

Yes, that is the problem.  These core-talk threads discuss it.

https://epics.anl.gov/core-talk/2020/msg01064.php
https://epics.anl.gov/core-talk/2020/msg01071.php

> Does `git revert db6e7c7a22b73f70a8b93e2` help ?

The patch I posted basically reverts that commit.  But it is not quite that simple because base 7.0.4 changes the EPICS-specific way that __declspec(dllimport)/export is specified in base.  It no longer epicsShareFunc.

Mark

________________________________
From: Torsten Bögershausen <torsten.bogershausen at ess.eu>
Sent: Saturday, June 20, 2020 2:03 AM
To: Mark Rivers; 'tech-talk'
Subject: Re: Serious problem with base 7.0.4 on Windows

Could it be, that this commit removed the functionality that we need for
Windows ?

commit db6e7c7a22b73f70a8b93e2aa4b6fa505e0218a6
Author: Michael Davidsaver <mdavidsaver at gmail.com>
Date:   Wed Feb 5 10:30:58 2020 -0800
     use one osdSockAddrReuse impl for all targets
         drop win32 specialization of osdSockAddrReuse

After reading
<https://docs.microsoft.com/en-us/windows/win32/winsock/using-so-reuseaddr-and-so-exclusiveaddruse>

it seems clear that there a different understandings what
SO_REUSEADDR should do, compared UNIX - Windows.

Does `git revert db6e7c7a22b73f70a8b93e2` help ?

On 19/06/20 22:40, Mark Rivers via Tech-talk wrote:
> Folks,
>
> I just want to give a heads-up that I have found a serious problem in
> base 7.0.4 on Windows.  Running a single IOC on a computer works fine.
> However, additional IOCs started on the same machine cannot be reached
> with Channel Access.
>
> I assume there will be a patch release soon.  Meanwhile this patch
> (actually a new file) will fix the problem.
>
> corvette:local/epics/base-7.0.4>git diff R7.0.4 modules/libcom/
>
> diff --git a/modules/libcom/src/osi/os/WIN32/osdSockAddrReuse.cpp
> b/modules/libcom/src/osi/os/WIN32/osdSockAddrReuse.cpp
>
> new file mode 100644
>
> index 0000000..cc13605
>
> --- /dev/null
>
> +++ b/modules/libcom/src/osi/os/WIN32/osdSockAddrReuse.cpp
>
> @@ -0,0 +1,44 @@
>
> +
>
> +/*************************************************************************\
>
> +* Copyright (c) 2002 The University of Chicago, as Operator of Argonne
>
> +*     National Laboratory.
>
> +* Copyright (c) 2002 The Regents of the University of California, as
>
> +*     Operator of Los Alamos National Laboratory.
>
> +* EPICS BASE Versions 3.13.7
>
> +* and higher are distributed subject to a Software License Agreement found
>
> +* in file LICENSE that is included with this distribution.
>
> +\*************************************************************************/
>
> +
>
> +/*
>
> + * Author: Jeff Hill
>
> + */
>
> +
>
> +#define epicsExportSharedSymbols
>
> +#include "osiSock.h"
>
> +#include "errlog.h"
>
> +
>
> +/*
>
> + * Note: WINSOCK appears to assign a different functionality for
>
> + * SO_REUSEADDR compared to other OS. With WINSOCK SO_REUSEADDR indicates
>
> + * that simultaneously servers can bind to the same TCP port on the
> same host!
>
> + * Also, servers are always enabled to reuse a port immediately after
>
> + * they exit ( even if SO_REUSEADDR isnt set ).
>
> + */
>
> +LIBCOM_API void epicsStdCall
>
> +    epicsSocketEnableAddressReuseDuringTimeWaitState ( SOCKET s )
>
> +{
>
> +}
>
> +
>
> +LIBCOM_API void epicsStdCall
>
> +    epicsSocketEnableAddressUseForDatagramFanout ( SOCKET s )
>
> +{
>
> +    int yes = true;
>
> +    int status;
>
> +    status = setsockopt ( s, SOL_SOCKET, SO_REUSEADDR,
>
> +        (char *) & yes, sizeof ( yes ) );
>
> +    if ( status < 0 ) {
>
> +        errlogPrintf (
>
> +            "epicsSocketEnablePortUseForDatagramFanout: "
>
> +            "unable to set SO_REUSEADDR?\n");
>
> +    }
>
> +}
>
> Mark
>

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System