Experimental Physics and
Industrial Control System

On 5/6/19 10:25 AM, Johnson, Andrew N. via Tech-talk wrote:
> On 5/6/19 10:15 AM, Bruno Martins via Tech-talk wrote:
>> After removing caClientCallbackThreadId=0 in base,
> 
> I just committed that change to the Base 3.15 branch, which will appear
> in 3.15.7 and be merged into the 7.0 branch before the next 7.0.x release.

I'm just too slow in replying.  This change seems like not the right fix to not the right issue :P

epicsThreadPrivateDelete() has invalided the storage pointed to by caClientCallbackThreadId.
Wouldn't this change be replacing a NULL dereference with a use after free()?

The right issue to fix is the order of epicsAtExit() hooks, or maybe to
remove cacExitHandler() altogether.

References:

libca bug? Bruno Martins via Tech-talk

Re: libca bug? Matt Newville via Tech-talk

Re: libca bug? Michael Davidsaver via Tech-talk

Re: libca bug? Matt Newville via Tech-talk

Re: libca bug? Bruno Martins via Tech-talk

Re: libca bug? Johnson, Andrew N. via Tech-talk

Navigate by Date:

Prev: Re: libca bug? Michael Davidsaver via Tech-talk

Next: Simple test fails to find subroutine in registry Matt Rippa via Tech-talk

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 

Navigate by Thread:

Prev: Re: libca bug? Johnson, Andrew N. via Tech-talk

Next: Re: libca bug? Michael Davidsaver via Tech-talk

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