Re: [PATCH 2/5] mkfs: add initial ini format config file parsing support

[Dave Chinner <david@fromorbit.com>]

On Wed, Oct 14, 2020 at 10:46:33PM -0700, Darrick J. Wong wrote:
> On Thu, Oct 15, 2020 at 02:29:22PM +1100, Dave Chinner wrote:
> > From: Dave Chinner <dchinner@redhat.com>
> > 
> > Add the framework that will allow the config file to be supplied on
> > the CLI and passed to the library that will parse it. This does not
> > yet do any option parsing from the config file.
> > 
> > Signed-off-by: Dave Chinner <dchinner@redhat.com>
> > ---
> >  mkfs/Makefile   |   2 +-
> >  mkfs/xfs_mkfs.c | 115 +++++++++++++++++++++++++++++++++++++++++++++++-
> >  2 files changed, 115 insertions(+), 2 deletions(-)
> > 
> > diff --git a/mkfs/Makefile b/mkfs/Makefile
> > index 31482b08d559..b8805f7e1ea1 100644
> > --- a/mkfs/Makefile
> > +++ b/mkfs/Makefile
> > @@ -11,7 +11,7 @@ HFILES =
> >  CFILES = proto.c xfs_mkfs.c
> >  
> >  LLDLIBS += $(LIBXFS) $(LIBXCMD) $(LIBFROG) $(LIBRT) $(LIBPTHREAD) $(LIBBLKID) \
> > -	$(LIBUUID)
> > +	$(LIBUUID) $(LIBINIH)
> >  LTDEPENDENCIES += $(LIBXFS) $(LIBXCMD) $(LIBFROG)
> >  LLDFLAGS = -static-libtool-libs
> >  
> > diff --git a/mkfs/xfs_mkfs.c b/mkfs/xfs_mkfs.c
> > index 8fe149d74b0a..e84be74fb100 100644
> > --- a/mkfs/xfs_mkfs.c
> > +++ b/mkfs/xfs_mkfs.c
> > @@ -11,6 +11,7 @@
> >  #include "libfrog/fsgeom.h"
> >  #include "libfrog/topology.h"
> >  #include "libfrog/convert.h"
> > +#include <ini.h>
> >  
> >  #define TERABYTES(count, blog)	((uint64_t)(count) << (40 - (blog)))
> >  #define GIGABYTES(count, blog)	((uint64_t)(count) << (30 - (blog)))
> > @@ -44,6 +45,11 @@ enum {
> >  	B_MAX_OPTS,
> >  };
> >  
> > +enum {
> > +	C_OPTFILE = 0,
> > +	C_MAX_OPTS,
> > +};
> > +
> >  enum {
> >  	D_AGCOUNT = 0,
> >  	D_FILE,
> > @@ -237,6 +243,28 @@ static struct opt_params bopts = {
> >  	},
> >  };
> >  
> > +/*
> > + * Config file specification. Usage is:
> > + *
> > + * mkfs.xfs -c file=<name>
> 
> I thought it was -c options=/dev/random ?

I fixed that, refreshed the patch and updated the change log! How
the hell did I lose that change?

> >  	{ 'd', &dopts, data_opts_parser },
> >  	{ 'i', &iopts, inode_opts_parser },
> >  	{ 'l', &lopts, log_opts_parser },
> > @@ -3562,6 +3611,61 @@ check_root_ino(
> >  	}
> >  }
> >  
> > +/*
> > + * INI file format option parser.
> > + *
> > + * This is called by the file parser library for every valid option it finds in
> > + * the config file. The option is already broken down into a
> > + * {section,name,value} tuple, so all we need to do is feed it to the correct
> 
> XFS, SAX style.
> 
> > + * suboption parser function and translate the return value.
> > + *
> > + * Returns 0 on failure, 1 for success.
> > + */
> > +static int
> > +cfgfile_parse_ini(
> > +	void			*user,
> > +	const char		*section,
> > +	const char		*name,
> > +	const char		*value)
> > +{
> > +	struct cli_params	*cli = user;
> > +
> > +	fprintf(stderr, "Ini debug: file %s, section %s, name %s, value %s\n",
> > +		cli->cfgfile, section, name, value);
> > +
> > +	return 1;
> > +}
> > +
> > +void
> > +cfgfile_parse(
> > +	struct cli_params	*cli)
> > +{
> > +	int			error;
> > +
> > +	if (!cli->cfgfile)
> > +		return;
> > +
> > +	error = ini_parse(cli->cfgfile, cfgfile_parse_ini, cli);
> > +	if (error) {
> > +		if (error > 0) {
> > +			fprintf(stderr,
> > +		_("%s: Unrecognised input on line %d. Aborting.\n"),
> > +				cli->cfgfile, error);
> > +		} else if (error == -2) {
> > +			fprintf(stderr,
> > +		_("Memory allocation failure parsing %s. Aborting.\n"),
> > +				cli->cfgfile);
> > +		} else {
> > +			fprintf(stderr,
> > +		_("Unable to open config file %s. Aborting.\n"),
> > +				cli->cfgfile);
> 
> I worry about libinih someday defining more negative error codes.  -1 is
> "open failed", -2 is OOM, and positive is the line number of a parsing
> error, at least according to the documentation.
> 
> Maybe we should handle -1 specifically and use the else as a catchall
> for unrecognized error codes?

Makes sense.

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com