Re: [GIT PULL] iomap: new code for 5.13-rc1

From: Linus Torvalds <torvalds@linux-foundation.org>
To: Christoph Hellwig <hch@lst.de>
Cc: "Darrick J. Wong" <djwong@kernel.org>, Jia He <justin.he@arm.com>,
	Al Viro <viro@zeniv.linux.org.uk>,
	linux-fsdevel <linux-fsdevel@vger.kernel.org>,
	linux-xfs <linux-xfs@vger.kernel.org>,
	Dave Chinner <david@fromorbit.com>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	Eric Sandeen <sandeen@sandeen.net>
Subject: Re: [GIT PULL] iomap: new code for 5.13-rc1
Date: Tue, 27 Apr 2021 23:38:24 -0700	[thread overview]
Message-ID: <CAHk-=whWnFu4wztnOtySjFVYXmBR4Mb2wxrp6OayZqnpKeQw0g@mail.gmail.com> (raw)
In-Reply-To: <20210428061706.GC5084@lst.de>

On Tue, Apr 27, 2021 at 11:17 PM Christoph Hellwig <hch@lst.de> wrote:
>
> On Tue, Apr 27, 2021 at 01:05:13PM -0700, Linus Torvalds wrote:
> > So how many _would_ be enough? IOW, what would make %pD work better
> > for this case?
>
> Preferably all.

WHY?

You guys are making no sense at all. You're stating silly things,
backing it up with absolutely nothing.

> Nothing needs to be secure.  It just needs to not scare users because
> they can see that the first usually two components clearly identify
> this is the test file system.

This is inane blathering.

What "scary message"? It will never happen in any normal circumstance,
and the trivial thing to do for any xfs test is to make the last
component name be something really obvious for the tester - who is the
only one who will ever see it.

And if it ever *does* happen in real life, the full path really isn't
necessary either. We're talking swap files. They aren't going to be in
random places.

The "I need the whole path" thing is just crazy, and you seem to be in
denial about it. There is absolutely zero reason for it.

I don't particularly care about this code sequence, but I do care when
people start making completely pointless arguyments to make excuses
for stupid code. You have extra silly code for "oh, the temporary
allocation that we did for no good reason can fail, so now we print
"<unknown>" for that case.

So it's all kinds of odd extra code for something that never used to
even bother with a pathname at all before, and now it's suddenly
"scary" and "really important to have all the components" instead of
just being simple and straightforward.

It's a purely informational message, and you guys made it pointlessly
overcomplicated for absolutely zero reason, and now you're too
embarrassed to just admit how pointless it was.

                   Linus