Received-SPF: softfail (google.com: domain of transitioning linux-kernel-owner@vger.kernel.org does not designate 23.128.96.19 as permitted sender) client-ip=23.128.96.19;
MIME-Version: 1.0
References: <alpine.LRH.2.02.2204111023230.6206@file01.intranet.prod.int.rdu2.redhat.com>
 <CAHk-=wijDnLH2K3Rh2JJo-SmWL_ntgzQCDxPeXbJ9A-vTF3ZvA@mail.gmail.com>
 <alpine.LRH.2.02.2204111236390.31647@file01.intranet.prod.int.rdu2.redhat.com>
 <CAHk-=wgsHK4pDDoEgCyKgGyo-AMGpy1jg2QbstaCR0G-v568yg@mail.gmail.com> <YlUQ9fnc+4eP3AE5@infradead.org>
In-Reply-To: <YlUQ9fnc+4eP3AE5@infradead.org>
From:   Linus Torvalds <torvalds@linux-foundation.org>
Date:   Mon, 11 Apr 2022 19:47:44 -1000
Message-ID: <CAHk-=wjOiBc3NiMRQJC6EsVosKOWmXiskqb0up6b5MOxCKSCBQ@mail.gmail.com>
Subject: Re: [PATCH] stat: don't fail if the major number is >= 256
To:     Christoph Hellwig <hch@infradead.org>
Cc:     Mikulas Patocka <mpatocka@redhat.com>,
        Alexander Viro <viro@zeniv.linux.org.uk>,
        linux-fsdevel <linux-fsdevel@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        Matthew Wilcox <willy@infradead.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Mon, Apr 11, 2022 at 7:41 PM Christoph Hellwig <hch@infradead.org> wrote:
>
> And that is easily fixed by using a lower major for the block dynamic
> dev_t.  In theory userspace should be able to copy with any major for
> it, but I fear something might break so we could make it configurable.

We actually have a ton of major numbers free, although it's not
obvious because of the whole "they are used by character devices, not
block devices" issue.

Ie 4-6, 12, 14, 19 are all free, afaik.

But yeah, somebody might have a static /dev for some odd embedded case, I guess.

That said, it really does look like we just return -EOVERFLOW much too
eagerly, for stupid and bad reasons.

Considering that BLOCK_EXT_MAJOR has been 259 since 2008, and this is
the first time anybody has hit this, I don't think there's much reason
to change that major number when the whole error case seems to have
been largely a mistake to begin with.

                  Linus