Re: one-liner patches to sim/common

[Andrew Cagney <ac131313@cygnus.com>]

"Frank Ch. Eigler" wrote:
> 
> Hi -
> 
> I just committed a one-liner pair of patches, as suggested by
> Andrew Cagney, to get around some annoying problems related
> to the sim/gdb hanging on to executable file handles
> unnecessarily.

Er, that's not correct.

I pointed out that Insight had *hacked* around a problem with locked
files on WindowsSOMETHING by closing the BFD using the *INTERNAL*
function bfd_cache_close().

Neither GDB, sim/common or Insight should be calling functions that are
internal to BFD.  If you think the function is useful then please get
BFD to publish its interface:

cc1: warnings being treated as errors
/home/scratch/MI/src/sim/d10v/../common/sim-load.c: In function
`sim_load_file':
/home/scratch/MI/src/sim/d10v/../common/sim-load.c:169: warning:
implicit declaration of function `bfd_cache_close'
*** Error code 1

	Andrew


> 2001-04-19  Frank Ch. Eigler  <fche@redhat.com>
> 
>         * sim-utils.c (sim_analyze_program): Call bfd_cache_close after
>         we're finished with its immediate use.
>         * sim-load.c (sim_load_file): Ditto.
> 
> Index: sim-load.c
> ===================================================================
> RCS file: /cvs/cvsfiles/devo/sim/common/sim-load.c,v
> retrieving revision 1.9
> diff -u -r1.9 sim-load.c
> --- sim-load.c  2000/03/02 09:10:40     1.9
> +++ sim-load.c  2001/04/19 19:29:50
> @@ -166,6 +166,8 @@
>        report_transfer_performance (callback, data_count, start_time, end_time);
>      }
> 
> +  bfd_cache_close (result_bfd);
> +
>    return result_bfd;
>  }
> 
> Index: sim-utils.c
> ===================================================================
> RCS file: /cvs/cvsfiles/devo/sim/common/sim-utils.c,v
> retrieving revision 1.23
> diff -u -r1.23 sim-utils.c
> --- sim-utils.c 2000/03/02 09:11:13     1.23
> +++ sim-utils.c 2001/04/19 19:29:50
> @@ -287,6 +287,8 @@
>         break;
>        }
> 
> +  bfd_cache_close (prog_bfd);
> +
>    return SIM_RC_OK;
>  }
>