On Sun, Aug 21, 2016 at 11:46:14PM -0500, Eric Sandeen wrote:
> On 8/21/16 10:34 PM, Eric Sandeen wrote:
> > On 8/21/16 9:06 PM, Dave Chinner wrote:
> >> On Tue, Aug 16, 2016 at 09:16:38AM -0500, Bill O'Donnell wrote:
>
> ...
>
> >> static int
> >> init_check_command(
> >> const cmdinfo_t *ct)
> >> {
> >> if (!fspath)
> >> return 1;
> >>
> >> /* Always run commands that we are told to skip here */
> >> if (ct->flags & CMD_SKIP_CHECK)
> >> return 1;
> >>
> >> /* if it's an XFS filesystem, always run the command */
> >> if (!(fs_path->fs_flags & FS_FOREIGN))
> >> return 1;
> >
> > Sorry for the late review; thanks for getting on it, Dave - but,
> > isn't "foreign ok" exactly == "skip check?"
> >
> > The only check that gets skipped is the foreign check, so just
> > setting FOREIGN_OK seems to accomplish the same thing without
> > more flag complexity, no?
>
> Oh, the subliminal brain reminded me that we want to be able to
> issue help or quit whether or not we had the "-f" flag, regardless
> of the filesystem, and that "foreign" isn't ok unless the -f flag
> is set, so we do need a class of "always works" commands.
Right, but that was something that was done in patch 1/3. I pointed
out that no mention of it was made in the cmmit message there....
> I guess that was the point of the patch, but I suppose some clarity
> in comments or commitlog would help slow people like me. ;)
Right. better explanations needed all round :P
Cheers,
Dave.
--
Dave Chinner
david@xxxxxxxxxxxxx
|