Parallel sockets or?

bruce.labitt at autoliv.com bruce.labitt at autoliv.com
Fri Oct 9 10:09:16 EDT 2009


Somewhat beating a dead horse...

"Shawn O'Shea" <shawn at eth0.net> wrote on 10/09/2009 01:46:28 AM:
> 
> On Thu, Oct 8, 2009 at 5:06 PM, <bruce.labitt at autoliv.com> wrote:
> "Shawn O'Shea" <shawn at eth0.net> wrote on 10/08/2009 01:27:57 PM:
> >
> > The originators of psock were at cesnet.cz, there is a link to the
> library
> > at http://www.cesnet.cz/project/qosip
> >
> > The make doesn't finish.  Several compile errors.  I played about a 
bit,
> > but did not make much headway...
> 
> > I was curious and downloaded this. I'm working on CentOS 5.3 (32-bit).
> >
<snip>

<bruce said>
> It really completes without error?  Not for me... :(
> 
> With a 64 bit system, with gcc-3.4 (Ubuntu 8.10 on x86_64) I had to edit
> atomicptr.h  from "__i86_64__" to "__x86_64__", since the system reports
> itself as x86_64 rather than i86_64
> This causes the make to die unless the change is made.

<shawn said> 
> So again out of sheer curiosity, I installed an Ubuntu 8.10 amd64 
> desktop default install in VMware Fusion. I installed gcc 3.4 (apt-get
> install gcc-3.4), untarred, configured (CC=gcc-3.4 ./configure) and 
> make. I get this:
>  gcc-3.4 -DHAVE_CONFIG_H -I. -I. -I.. -g -O2 -MT file.lo -MD -MP -MF .
> deps/file.Tpo -c file.c  -fPIC -DPIC -o .libs/file.o
> In file included from ../psock/file.h:30,
>                  from file.c:35:
> ../psock/atomicptr.h:38:2: warning: #warning Running architecture is 
> unapproved, please refer psock/atomicptr.h
> 
> followed by a slew of warnings and errors. It would seem by the nature
> of that message that the architecture (amd64 aka x86_64) is not 
supported.
> 

<bruce says>
The original author apparently did not test his code with a 64 bit 
machine, remember package dates from ~2005.  The return machine type for 
an intel Pentium+ or AMD is "x86_64" NOT "i86_64".  If atomicptr.h is 
editted to make that change, as I indicated above, x86_64, rather than 
i86_64, the compile completes if the wrap_##f change is done and a bunch 
of undone functions are commented out.

<shawn said>
> Compiling it 32-bit works fine on the same VM (apt-get install libc6-
> dev-i386 lib32gcc1  to get the necessary 32-bit build pieces and tell 
> confogure and gcc to build 32-bit with CC=gcc-3.4 CFLAGS='-m32' .
> /configure). Make completes fine.
> 
> -Shawn
>  

<bruce really wants to know this...>
Hmm, did you have to do the wrap_##f change or not?  Any other edits?


<snip some of bruce's stuff>

<bruce, imploring now...>
If I could really impose on you...  If you cd to the psock-testperf dir 
and run

./psock-testperf --xmit --in mem --len 10000 --connect host

and on another window
and run
./psock-testperf --recv --out mem --listen

This works across a network for me.  (One cannot just copy the executable 
however.  One needs to build it on the other node as well.)

However, if len is 1,000,000 it does not.  20,000 does not work either.

Does it do the same for you?


> This library certainly looks far from complete to me...  Or I am missing
> something big?
> 
> Thanks!
> -Bruce


******************************
Neither the footer nor anything else in this E-mail is intended to or constitutes an <br>electronic signature and/or legally binding agreement in the absence of an <br>express statement or Autoliv policy and/or procedure to the contrary.<br>This E-mail and any attachments hereto are Autoliv property and may contain legally <br>privileged, confidential and/or proprietary information.<br>The recipient of this E-mail is prohibited from distributing, copying, forwarding or in any way <br>disseminating any material contained within this E-mail without prior written <br>permission from the author. If you receive this E-mail in error, please <br>immediately notify the author and delete this E-mail.  Autoliv disclaims all <br>responsibility and liability for the consequences of any person who fails to <br>abide by the terms herein. <br>
******************************



More information about the gnhlug-discuss mailing list