RE: Bus errors accessing VME with base 7.0.6.1 and latest synApps modules

[Mark Rivers via Tech-talk <tech-talk at aps.anl.gov>]

Don't get me wrong, I was not being critical at all.  I just mentioned the comment about the "complexity" to suggest that perhaps it is not surprising that this PR has introduced a very subtle bug.

Mark


-----Original Message-----
From: Michael Davidsaver <mdavidsaver at gmail.com> 
Sent: Wednesday, June 1, 2022 2:03 PM
To: Mark Rivers <rivers at cars.uchicago.edu>; Torsten Bögershausen <Torsten.Bogershausen at ess.eu>; Benjamin Franksen <benjamin.franksen at helmholtz-berlin.de>
Cc: tech-talk at aps.anl.gov
Subject: Re: Bus errors accessing VME with base 7.0.6.1 and latest synApps modules

On 6/1/22 10:54, Mark Rivers wrote:
> Both of those commits were part of this PR 
> https://github.com/epics-base/epics-base/pull/99 
> <https://github.com/epics-base/epics-base/pull/99>
> 
> One of the comments on that PR (January 27, 2021) was:
> 
> We expect to be nervous about this change because of the complex area it's working in, so we will merge it /after/ the next release which should be before the Codeathon in March, but we don't see anything obviously wrong with it.

This PR was recognized as having a high risk change of regression.
The code involved is some of the most complex in Base.  (Try to follow all of the possible paths through dbGet() )

This change was judged to be worthwhile since it actually reduces that complexity.  Sadly, we don't get many PRs with a net reduction in source line count.

Merging after the 7.0.5 release, and before the 2021 codeathon, was an attempt to mitigate this risk by maximizing the amount of testing before 7.0.6.

Unfortunately, this seems not to have been enough.