EPICS Home

Experimental Physics and Industrial Control System


 
1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  <20112012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024  Index 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  <20112012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 
<== Date ==> <== Thread ==>

Subject: Re: [Fwd: RE: IOC make error]
From: Andrew Johnson <[email protected]>
To: "Cid, Florencio A" <[email protected]>, "EPICS tech-talk" <[email protected]>
Date: Wed, 9 Nov 2011 12:46:14 -0600
Hi Florencio,

On 2011-11-09 Cid, Florencio A wrote:
> Thanks for the help, everyone!!  I can live with manually edit
>  configure/RELEASE as a temporary work-around.   When will the revised fix
>  to makeBaseApp.pl (mentioned below) be available?

I committed the fix to the 3.14 branch yesterday, and have just created a 
patch which is now available through the Known Problems page at
   http://www.aps.anl.gov/epics/base/R3-14/12-docs/KnownProblems.html
The fix will also be included in the 3.14.12.2 release that we're currently 
working on which is due out before Christmas.

> Also, should this thread (which didn't make it to tech-talk) be added?  Not
> sure I'm the one to make the decision.

Done with this reply.

> Thanks again, everyone!!
> 
> Flo
> 
> -----Original Message-----
> From: Andrew Johnson [mailto:[email protected]]
> Sent: Tuesday, November 08, 2011 4:35 PM
> To: Janet Anderson
> Cc: Mark Rivers; [email protected]; Cid, Florencio A
> Subject: Re: [Fwd: RE: IOC make error]
> 
> Hi Janet,
> 
> On 2011-11-08 you wrote:
> > Is the following still true?  From makeBaseApp.pl:
> >
> > -b base  Set the location of EPICS base (full path)
> >           If not specified, base path is taken from configure/RELEASE
> >           If configure does not exist, from environment
> >           If not found in environment, from makeBaseApp.pl location
> 
> That's what the code does:
> 
>     # Locate epics_base
>     my ($command) = UnixPath($0);
>     if ($opt_b) {
>         # first choice is -b base
> 	$epics_base = UnixPath($opt_b);
>     } elsif ($release{"EPICS_BASE"}) {
>         # second choice is configure/RELEASE
> 	$epics_base = UnixPath($release{"EPICS_BASE"});
> 	$epics_base =~s|^\$\(TOP\)/||;
>     } elsif ($ENV{EPICS_MBA_BASE}) {
>         # third choice is env var EPICS_MBA_BASE
>         $epics_base = UnixPath($ENV{EPICS_MBA_BASE});
>     } elsif ($command =~ m|/bin/|) {
>         # assume script was run with full path to base
> 	$epics_base = $command;
> 	$epics_base =~ s|^(.*)/bin/.*makeBaseApp.*|$1|;
>     }
>     $epics_base and -d "$epics_base" or Cleanup(1, "Can't find EPICS
>  base"); $app_epics_base = LocalPath($epics_base);
>     $app_epics_base =~ s|^\.\.|\$(TOP)/..|;
> 
> The $app_epics_base value is then used when instantiating the template file
> for configure/RELEASE.
> 
> Mark Rivers wrote:
> > The question is why did makeBaseApp use C:/EPICS/base?  I thought it
> > might have to do with how you ran makeBaseApp.pl.  I thought running it
> > by typing C:/EPICS/bin/cygwin-x86/makeBaseApp.pl would produce
> > EPICS_BASE as you saw, but running by typing
> > /cygdrive/c/EPICS/base/makeBaseApp.pl would get what you want.  But I
> > just tested this, and it does not seem to be the case, even running
> > it as /cygdrive/c/... I still get configure/RELEASE having
> > EPICS_BASE=c:/EPICS/base-3.14.11.
> 
> The LocalPath() call shown on the penultimate line above is converting the
> path string into a Windows-style path when it shouldn't be.  Hmm, we
>  removed that conversion from the tools/EPICS/Path.pm file back in May, but
>  apparently not from its duplicate in makeBaseApp.pl — this is definitely a
>  bug which I will fix (properly).
> 
> Thanks all for tracking this down.
> 
> - Andrew

- Andrew
-- 
Optimization is the process of taking something that works and
replacing it with something that almost works, but costs less.
-- Roger Needham


Navigate by Date:
Prev: RE: About the problem of areaDetector module's compiling and run Allison, Stephanie
Next: Allen Bradley VME-AB1771, 6008-SV2R Szalata, Zenon M.
Index: 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  <20112012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 
Navigate by Thread:
Prev: RE: PV in PCAS can not be archived by Channel Archiver Jeff Hill
Next: Allen Bradley VME-AB1771, 6008-SV2R Szalata, Zenon M.
Index: 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  <20112012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024