[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20141001203616.GA2860@wallace>
Date: Wed, 1 Oct 2014 16:36:17 -0400
From: Eric Whitney <enwlinux@...il.com>
To: "Darrick J. Wong" <darrick.wong@...cle.com>
Cc: Eric Whitney <enwlinux@...il.com>, linux-ext4@...r.kernel.org,
tytso@....edu
Subject: Re: [PATCH] mke2fs: clarify inline data inode size error message
* Darrick J. Wong <darrick.wong@...cle.com>:
> On Sat, Sep 27, 2014 at 11:55:15AM -0400, Eric Whitney wrote:
> > The existing error message can be made more helpful by more clearly
> > implying the attempt to make a file system with undersized inodes is
> > failing and suggesting a corrective action.
> >
> > Signed-off-by: Eric Whitney <enwlinux@...il.com>
> > ---
> > misc/mke2fs.c | 3 ++-
> > 1 file changed, 2 insertions(+), 1 deletion(-)
> >
> > diff --git a/misc/mke2fs.c b/misc/mke2fs.c
> > index 2bc435b..00bb00d 100644
> > --- a/misc/mke2fs.c
> > +++ b/misc/mke2fs.c
> > @@ -2358,7 +2358,8 @@ profile_error:
> > EXT4_FEATURE_INCOMPAT_INLINE_DATA) &&
> > fs_param.s_inode_size == EXT2_GOOD_OLD_INODE_SIZE) {
> > com_err(program_name, 0,
> > - _("inode size is %d, inline data is useless"),
> > + _("%d byte inodes are too small for inline data; "
> > + "specify larger size"),
>
> How much larger? The next larger inode size is 256 bytes, so we ought to tell
> the user that. ("specify a size at least 256 bytes"?)
>
> --D
I'd be fine with that as well. My only concern is that noting the minimum
might lead to users taking that as the default when that might not best
suit their needs. I thought that the reworked message I suggested might
send them off to the mke2fs man page, where it's at least clear the legal
sizes are powers of 2 greater than 128 bytes. However, we don't seem to
have published any useful guidelines regarding inode size selection when
building a file system supporting inline data. Your inline data notes on
the wiki are about as good as we've got, unless I've missed something, and
I think we'd want a little more addressing inode size tradeoffs.
I wonder if Zheng has anything relevant we can use to improve the manpages?
Again, if an error message specifying the minimum or say, a range from minimum
to maximum suggested is preferred, I'd be happy to post another patch.
Thanks,
Eric
>
> > fs_param.s_inode_size);
> > exit(1);
> > }
> > --
> > 1.9.1
> >
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> > the body of a message to majordomo@...r.kernel.org
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists