| Re: Problems with spawn on Linux [message #18294] |
Thu, 16 December 1999 00:00 |
ugly_alaskan_at_work
Messages: 1
Registered: December 1999
|
Junior Member |
|
|
Peter --
That was my first though too, that I was somehow running into some sort
of resource issue. I inserted a huge 200 second wait before the spawn
and it still does the same thing. That should be more than enough time
for the os to clean up after the process and free the pipes.
It turns out it must be a problem or something different with IDL 5.3 -
I tried with IDL 5.2 and everything works fine.
Thanks,
-Jay
> Hi Jay,
>
> Sorry, I cannot reproduce this error on my computer (SuSE Linux 6.1
> running IDL 5.2). All pipes are read w/o errors.
>
> Are you sure that you are not hitting any shell-level limits,
> e.g. number of open pipes or processes etc.?
>
> Peter
>
> --
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~
> Dr. Peter "Pit" Suetterlin http://www.astro.uu.nl/~suetter
> Sterrenkundig Instituut Utrecht
> Tel.: +31 (0)30 253 5225 P.Suetterlin@astro.uu.nl
> ____________________________________________________________ ______________
>
Sent via Deja.com http://www.deja.com/
Before you buy.
|
|
|
|
| Re: Problems with spawn on Linux [message #18296 is a reply to message #18294] |
Thu, 16 December 1999 00:00  |
Craig Markwardt
Messages: 1869
Registered: November 1996
|
Senior Member |
|
|
ftjec@uaf.edu writes:
> Hello,
> I am having problems with the spawn command on Linux. I am using the
> unit parameter to spawn to obtain a pipe to the child process and use
> this pipe to moniter the child's progress. The code I have works fine
> on the sun I coded it for originally -- but on linux I get very
> unpredictable results. Sometimes the process is spawned, yet no output
> is generated, sometimes it works fine, and sometimes it crashes idl
> completely with a "bus error" message.
>
> Does anyone know of a solution to this problem?
>
> Also any advice on the general state of idl on Linux would be greatly
> apprecated.
I can't reproduce your problem either:
RedHat 6.1
IDL> print, !version
{ x86 linux unix 5.2.1 Jun 4 1999}
What you describe is very strange.
However, I did discover that the ON_IOERROR procedure interferes with
reads from SPAWNed pipes. In particular, ON_IOERROR occasionally
causes the input buffer to be lost (along with your data of course).
Usually this can be solved by moving ON_IOERROR outside of your read
loop, which in your case you have done already. Perhaps you could try
CATCH just to see if anything different happens.
Craig
--
------------------------------------------------------------ --------------
Craig B. Markwardt, Ph.D. EMAIL: craigmnet@cow.physics.wisc.edu
Astrophysics, IDL, Finance, Derivatives | Remove "net" for better response
------------------------------------------------------------ --------------
|
|
|
|
| Re: Problems with spawn on Linux [message #18299 is a reply to message #18294] |
Thu, 16 December 1999 00:00 |
pit
Messages: 92
Registered: January 1996
|
Member |
|
|
In article <8396av$p6o$1@nnrp1.deja.com>,
ftjec@uaf.edu writes:
Hi Jay,
Sorry, I cannot reproduce this error on my computer (SuSE Linux 6.1
running IDL 5.2). All pipes are read w/o errors.
Are you sure that you are not hitting any shell-level limits,
e.g. number of open pipes or processes etc.?
Peter
--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~
Dr. Peter "Pit" Suetterlin http://www.astro.uu.nl/~suetter
Sterrenkundig Instituut Utrecht
Tel.: +31 (0)30 253 5225 P.Suetterlin@astro.uu.nl
____________________________________________________________ ______________
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
