Re: [PATCH 3/3] mkfs.ubifs: use libubi to format UBI volume

[Artem Bityutskiy <dedekind@infradead.org>]

On Wed, 2009-05-13 at 10:38 +0200, Corentin Chary wrote:
> > On Sat, 2009-05-09 at 11:41 +0200, Corentin Chary wrote:
> >> libubi is now used to format directly UBI volume.
> >> Typing mkfs.ubifs /dev/ubi0_0 is now possible.
> >> dtypes should be ok as they are taken from UBIFS code.
> >
> > Did you test that this works fine on UBI volume and there
> > are no regressions when you output to a file?
> 
> I did some test with and without --root, it worked fine.
> But I'll do some more test with the corrected patch.

Ok, thanks.

> Because I wanted mkfs.ubifs to work without this parameter, like other mkfs.*.
> I know mounting an empty volume will works, but we should also be able
> to use mkfs.ubifs to format
> empty volumes.

Ok.

> >>       ino->uid        = cpu_to_le32(st->st_uid);
> >>       ino->gid        = cpu_to_le32(st->st_gid);
> >> -     ino->uid        = cpu_to_le32(st->st_uid);
> >> -     ino->gid        = cpu_to_le32(st->st_gid);
> >>       ino->mode       = cpu_to_le32(st->st_mode);
> >>       ino->flags      = cpu_to_le32(use_flags);
> >>       ino->data_len   = cpu_to_le32(data_len);
> >
> > Why the 2 lines are removed?
> 
> Theses should have been removed in another patch, but look, they are
> here twice =).

Ah, OK.

> > Yeah, I think you should either ask the user, or just refuse
> > writing to it. Users may just 'ubiupdatevol -t' to erase the
> > volume before invocating mkfs.ubifs.
> >
> > Probably it is easier for now to just remove whole
> > 'erase_volume()' ?
> >
> mkfs.ext{2/3/4} and mkfs.vfat don't ask for confirmation
> I believe mkfs.reiserfs does.
> But I won't hurt to ask.
> Maybe first we can remove erase_volume(), and then, in another patch,
> add erase volume with confirmation.

Well, OK, let's be consistent with other mkfs'es then.

> New patch is comming at the end of the week ...

Thanks!

-- 
Best regards,
Artem Bityutskiy (Битюцкий Артём)