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

From: Rasmus Villemoes <linux@rasmusvillemoes.dk>
To: Linus Torvalds <torvalds@linux-foundation.org>,
	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: Wed, 28 Apr 2021 09:38:10 +0200	[thread overview]
Message-ID: <1de23de2-12a9-2b13-3b86-9fe4102fdc0c@rasmusvillemoes.dk> (raw)
In-Reply-To: <CAHk-=wjeUhrznxM95ni4z+ynMqhgKGsJUDU8g0vrDLc+fDtYWg@mail.gmail.com>

On 28/04/2021 09.14, Linus Torvalds wrote:
So let me just quote that first reply of mine, because you seem to not
> have seen it:
> 
>> We have '%pD' for printing a filename. It may not be perfect (by
>> default it only prints one component, you can do "%pD4" to show up to
>> four components), but it should "JustWork(tm)".
>>
>> And if it doesn't, we should fix it.
> 
> I really think %pD4 should be more than good enough. And I think maybe
> we should make plain "%pD" mean "as much of the path that is
> reasonable" rather than "as few components as possible" (ie 1).
> 
> So I don't think "%pD" (or "%pD4") is necessarily perfect, but I think
> it's even worse when people then go and do odd ad-hoc things because
> of some inconvenience in our %pD implementation.
> 
> For example, changing the default to be "show more by default" should
> be as simple as something like the attached.  I do think that would be
> the more natural behavior for %pD - don't limit it unnecessarily by
> default, but for somebody who literally just wants to see a maximum of
> 2 components, using '%pD2' makes sense.
> 
> (Similarly, changing the limit of 4  components to something slightly
> bigger would be trivial)
> 
> Hmm?
> 
> Grepping for existing users with
> 
>     git grep '%pD[^1-4]'
>
> most of them would probably like a full pathname, and the odd s390
> hmcdrv_dev.c use should just be fixed (it has a hardcoded "/dev/%pD",
> which seems very wrong).

So the patch makes sense to me. If somebody says '%pD5', it would get
capped at 4 instead of being forced down to 1. But note that while that
grep only produces ~36 hits, it also affects %pd, of which there are
~200 without a 2-4 following (including some vsprintf test cases that
would break). So I think one would first have to explicitly support '1',
switch over some users by adding that 1 in their format string
(test_vsprintf in particular), then flip the default for 'no digit
following %p[dD]'.

Rasmus