Received-SPF: pass (google.com: domain of linux-nfs-owner@vger.kernel.org designates 23.128.96.34 as permitted sender) client-ip=23.128.96.34;
From:   Benjamin Coddington <bcodding@redhat.com>
To:     Anna Schumaker <anna@kernel.org>
Cc:     Chuck Lever <chuck.lever@oracle.com>,
        Jeff Layton <jlayton@kernel.org>,
        trond.myklebust@hammerspace.com, linux-nfs@vger.kernel.org
Subject: Re: [PATCH 2/2] NFSv4: Allow per-mount tuning of READDIR attrs
Date:   Wed, 18 Oct 2023 15:17:22 -0400
Message-ID: <5468EC9F-A80B-4F63-90AB-1E27B9E8C227@redhat.com>
In-Reply-To: <CAFX2JfkON+5MsYuw-SsvSg04M6Fy=BY_v7RZ9aAs35P7fD15Gw@mail.gmail.com>
References: <cover.1697577945.git.bcodding@redhat.com>
 <bd900de1d19bc56e6df5b44379f373617acc894e.1697577945.git.bcodding@redhat.com>
 <ZS/V+4Cuzox7erqz@tissot.1015granger.net>
 <6157b73e380e5b625cd8ed0133ef392d0dd4bd8b.camel@kernel.org>
 <ZS/qwYzQvrgJNEv6@tissot.1015granger.net>
 <CAFX2JfkON+5MsYuw-SsvSg04M6Fy=BY_v7RZ9aAs35P7fD15Gw@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain
Precedence: bulk

On 18 Oct 2023, at 14:38, Anna Schumaker wrote:
>> It's not yet clear sysadmins will even need a switch like this, so I
>> would go further and say you should hold off on merging anything
>> like it until there is an actual reported need for it.
>>
>> Now, full control over that bitmap is still very neat thing for
>> experimentation by NFS developers. Hiding this behind a Kconfig
>> option would let you merge it but then turn it off in production
>> kernels.
>
> Definitely a neat thing to have, but I'm also in favor of hiding it
> behind a kconfig option to start.

Ah, missed replying to this part in the last message:

Ok, if my last message isn't convicing enough, are we talking about
something like CONFIG_NFS_EXPERT_SYSFS?

I'm worried that trying to make NFS nicer/safer in sysfs just means we're
gatekeeping what's already a complex and breakable set of technologies in
the one place (sysfs) where we can actually expose some complexity.

Ben