1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 <2009> 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 | Index | 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 <2009> 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 |
<== Date ==> | <== Thread ==> |
---|
Subject: | Re: File I/O from genSub Process function |
From: | Bruce Hill <[email protected]> |
To: | Andrew Johnson <[email protected]> |
Cc: | [email protected] |
Date: | Mon, 02 Nov 2009 16:38:37 -0800 |
Thanks for the suggestion. I can't risk trying anything till our next down time on Wed, but I'll give it a try then. - Bruce Andrew Johnson wrote: Hi Bruce, On Friday 30 October 2009 18:41:33 Bruce Hill wrote:I have a genSub I've setup to run test code on my IOC, and I'm getting errors when I try to read the test set data files. This is an RTEMS 4.7.1 based IOC accessing an NFS mounted filesystem. I'm able to read the same files from a genSub Init function, but not from the Process function. I can also read these files if I initiate the test from the Cexp shell on the RTEMS console. Is there something about how the Process functions are called that would preclude file I/O?No, although file I/O is blocking and as such will cause whatever scan thread is processing that record to be delayed and could affect whatever control algorithms the IOC is running at the time. We usually suggest that the best way to do file I/O inside the IOC is in a SNL program which can block whenever it likes and talk to the PVs using CA, or even global variables (provided you understand the need to lock them). I suspect the issue might be due to the fact that NFS can take up quite a lot of stack space - can you check whether the relevent thread has gone over its allotted stack? - Andrew -- Bruce Hill Member Technical Staff SLAC National Accelerator Lab 2575 Sand Hill Road M/S 94 Menlo Park, CA 94025 |