Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Tue, 14 Sep 2021 06:26:28 -0500
From:   Bjorn Helgaas <helgaas@kernel.org>
To:     Heiner Kallweit <hkallweit1@gmail.com>
Cc:     Dave Jones <davej@codemonkey.org.uk>,
        Linus Torvalds <torvalds@linux-foundation.org>,
        linux-kernel@vger.kernel.org, linux-pci@vger.kernel.org
Subject: Re: Linux 5.15-rc1
Message-ID: <20210914112628.GA1412445@bjorn-Precision-5520>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <870c3332-db60-9cf5-0439-247f91ce7808@gmail.com>
Precedence: bulk

On Tue, Sep 14, 2021 at 08:21:46AM +0200, Heiner Kallweit wrote:
> On 14.09.2021 01:46, Bjorn Helgaas wrote:

> > +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_ATTANSIC, PCI_ANY_ID, quirk_blacklist_vpd);
> >  /*
> 
> Leaving the quirks in FIXUP_HEADER stage would have the advantage that for
> blacklisted devices the vpd sysfs attribute isn't visibale. The needed
> changes to the patch are minimal.

What do you have in mind?  The only thing I can think of would be to
add a "pci_dev.no_vpd" bit.  "vpd.cap == 0" means the device has no
VPD, and "vpd.len == 0" means we haven't determined the size yet.  All
devices start off with vpd.cap == 0 and vpd.len == 0, so a
FIXUP_HEADER quirk would have to set a sentinel value or some other
bit.

Bjorn