Hey Christoph,
On Wed, Apr 06, 2022 at 09:36:55AM -0700, Christoph Hellwig wrote:
> Can you explain the split and the reason for it a little more here?
>
> > dump_file(
> > FILE *fp,
> > fs_disk_quota_t *d,
> > - uint id,
> > - uint *oid,
> > - uint type,
> > - char *dev,
> > - int flags)
> > + char *dev)
> > {
> > - if (!get_quota(d, id, oid, type, dev, flags))
> > - return 0;
>
> I think it would make more sense to move this into the previous
> patch that passes the fs_disk_quota to dump_file.
>
> And maybe this and the previous patch should be split into one for
> dump_file and one for report_mount?
I did it like this initially but it appeared to me that the diff was
messy. As there were many &d -> d and report_mount ->
get_quota/report_mount replacements, so I split it. But I'm not
against reshaping this back, should I do it?
>
> > + while ((g = getgrent()) != NULL) {
> > + get_quota(&d, g->gr_gid, NULL, type, mount->fs_name, 0);
>
> Overly long line. (and a few more below).
>
--
- Andrey
