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: iocStats module on RTEMS-5 |
From: | Ben Franksen via Tech-talk <tech-talk at aps.anl.gov> |
To: | "Johnson, Andrew N." <anj at anl.gov> |
Cc: | EPICS tech-talk <tech-talk at aps.anl.gov> |
Date: | Thu, 7 Oct 2021 16:29:08 +0200 |
Hi Andrew Am 06.10.21 um 17:13 schrieb Johnson, Andrew N.:
On Oct 6, 2021, at 2:51 AM, Ben Franksen via Tech-talk <tech-talk at aps.anl.gov<mailto:tech-talk at aps.anl.gov>> wrote: I am getting lots of build errors when I try to build iocStats-3.16.1 for RTEMS-5 (using base-7.0.6). We compile for the BSPs mvme2100 and beatnik. Has anyone perhaps already been through that and would be willing to share? We have not yet, but we’re going to need it to support that OS soon, so I should be able to offer some help from the APS with code reviews and maybe more.
Okay, thanks. The main problem seems to be that the CPU load measurements rely on internal data structures that have changed from 4.10 to 5. Porting this code is going to be messy. It would be ideal if RTEMS-5 had a stable API for this stuff and perhaps it has, nowadays, but I haven't investigated this yet.
The iocStats module could use a major overhaul but it doesn’t really have a maintainer at the moment and it isn’t clear which parts are being used by which labs — nobody wants to rip out stuff that another lab might still be using. Do you have MVME2100 boards running RTEMS-5 yet?
Not yet. Still busy trying to get our support modules to build.
Any issues with them that should be fed back to Base?
Again, not yet, except one fix but that applies to all base and RTEMS versions for this board, which I made a long time ago and never came around to pushing upstream. I attached a patch file, it fixes a bad rule that can fail if run in parallel.
Anything else I will report if/when it comes up. The release notes mention that mvme2100 BSP needs patches but doesn't provide any details. So far I found one rather trivial issue: its bsp.h misses a declaration for bsp_reset() (the function itself is there).
Cheers Ben -- "This work was strictly voluntary, but any animal who absented himself from it would have his rations reduced by half." -- George Orwell, Animal Farm
Attachment:
OpenPGP_0x9E74E6F7E7B29005.asc
Description: OpenPGP public key
Attachment:
OpenPGP_signature
Description: OpenPGP digital signature