| 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: Possible bug??? in dbReadDatabaseFP |
| From: | Michael Davidsaver via Tech-talk <tech-talk at aps.anl.gov> |
| To: | "Dudley, David" <dudleyd at frib.msu.edu>, Andrew Johnson <anj at aps.anl.gov> |
| Cc: | "tech-talk at aps.anl.gov" <tech-talk at aps.anl.gov> |
| Date: | Thu, 3 Mar 2022 08:47:59 -0800 |
On 3/3/22 08:16, Dudley, David via Tech-talk wrote:
If I do define ‘USE_TMPFILE’ (which was my original idea), it creates the temporary file, writes a proper record to it (I have checked this), and then calls ‘dbReadDatabaseFP’. In this case, the IOC shows that it loaded the database, closed the file, and then I get a ‘free(): double free detected in tcache 2’ error and the IOC aborts.
It looks like dbReadDatabaseFP() will close the provided 'FILE*' on success,
and sometimes on failure[1].
So your code should not call fclose() after dbReadDatabaseFP().
eg.
https://github.com/mdavidsaver/pyDevSup/blob/ea5f0d62412085dfc941e8fe758695abdfdab548/devsupApp/src/dbapi.c#L219-L221
It may be worth noting that pyDevSup is the only usage of dbReadDatabaseFP()
I could find, and I doubt this aspect has had much testing.
fyi. If you only need Linux support, you might look into fmemopen()
https://man7.org/linux/man-pages/man3/fmemopen.3.html
@Andrew
[1] Should the first error check should be changed to set 'status' and
'goto cleanup' instead of returning immediately?
https://github.com/epics-base/epics-base/blob/d8c537945354f13433c685b5a6c6a30bc6dc6848/modules/database/src/ioc/dbStatic/dbLexRoutines.c#L226-L227