segmentation fault in sockets library of sicstus prolog

Message has been deleted

pragye...@gmail.com

unread,

Jul 23, 2008, 9:42:09 AM7/23/08

to

Hi all,

I am using the linda library in sicstus prolog 4.0.4 for
commmunication between processes.

The linda server is calling the predicate socket_select/7 which
sometimes gives segmentation fault and I could trace the error to this
particular line in socket_select/7:

prolog:'$select'(ListenSocketsRaw,LReadyRaw,
ConnectSocketsRaw,CReadyRaw,
ReadStreamsRaw,RReadyRaw,
WriteStreamsRaw,WReadyRaw,
TimeoutParam, ECODE)

I could further find out that ultimately this call is making a system
call select but I could not find out what happens in between.

Could anyone of you tell how to fix this or where does the control go
after you make a call to prolog:'$select'?

Thanks,
Pragyesh

A.L.

unread,

Jul 23, 2008, 9:44:02 AM7/23/08

to

On Wed, 23 Jul 2008 06:39:05 -0700 (PDT), pragye...@gmail.com
wrote:

>Hi all,
>
>I am using the linda library in sicstus prolog for commmunication

>between processes.
>
>The linda server is calling the predicate socket_select/7 which
>sometimes gives segmentation fault and I could trace the error to this
>particular line in socket_select/7:
>
>
>prolog:'$select'(ListenSocketsRaw,LReadyRaw,
> ConnectSocketsRaw,CReadyRaw,
> ReadStreamsRaw,RReadyRaw,
> WriteStreamsRaw,WReadyRaw,
> TimeoutParam, ECODE)
>
>
>I could further find out that ultimately this call is making a system
>call select but I could not find out what happens in between.
>
>Could anyone of you tell how to fix this or where does the control go
>after you make a call to prolog:'$select'?
>

I am looking into my crystal ball, and asking the question: what
version of SICStus?... What operating system?... What hardware?...

Unfortunately, even my crystal ball doesn't know the answers...

A.L.

pragye...@gmail.com

unread,

Jul 23, 2008, 9:57:11 AM7/23/08

to

On Jul 23, 3:44 pm, A.L. <alewa...@zanoza.com> wrote:
> On Wed, 23 Jul 2008 06:39:05 -0700 (PDT), pragyesh.i...@gmail.com

Sicstus version 4.0.4 on Debian Linux 4.0 and I am runnning it on a
cluster with AMD Opteron(tm) Processor 252.

A.L.

unread,

Jul 23, 2008, 10:15:56 AM7/23/08

to

On Wed, 23 Jul 2008 06:57:11 -0700 (PDT), pragye...@gmail.com
wrote:

>
>Sicstus version 4.0.4 on Debian Linux 4.0 and I am runnning it on a
>cluster with AMD Opteron(tm) Processor 252.

Sorry, no help from me. I am not using Linux. I was testing Linda
extensively on Windows XP/INTEL, with SICStus 4.04 and 3.12.8 without
problems.

A.L.

pragye...@gmail.com

unread,

Jul 23, 2008, 11:21:18 AM7/23/08

to

On Jul 23, 4:15 pm, A.L. <alewa...@zanoza.com> wrote:
> On Wed, 23 Jul 2008 06:57:11 -0700 (PDT), pragyesh.i...@gmail.com

> wrote:
>
>
>
> >Sicstus version 4.0.4 on Debian Linux 4.0 and I am runnning it on a
> >cluster with AMD Opteron(tm) Processor 252.
>
> Sorry, no help from me. I am not using Linux. I was testing Linda
> extensively on Windows XP/INTEL, with SICStus 4.04 and 3.12.8 without
> problems.
>
> A.L.

Do you have any idea what does a call like prolog:'$select' transfer
control to?

A.L.

unread,

Jul 23, 2008, 11:42:31 AM7/23/08

to

On Wed, 23 Jul 2008 08:21:18 -0700 (PDT), pragye...@gmail.com
wrote:

Sorry - since I had no problems with Linda, I have never studied its
internals...

A.L.

Per.M...@sics.se

unread,

Aug 19, 2008, 11:18:17 AM8/19/08

to

On Jul 23, 3:42 pm, pragyesh.i...@gmail.com wrote:

> The linda server is calling the predicate socket_select/7 which
> sometimes gives segmentation fault and I could trace the error to this

If SICStus Prolog gives a segmentation fault then it is a bug. The
best way to resolve SICStus bugs is to report them at the SICStus
Prolog home page (www.sics.se/sicstus/ under "Support").

Regards,