Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Fri, 9 Jun 2023 14:46:21 -0500
From:   Bjorn Helgaas <helgaas@kernel.org>
To:     Sui Jingfeng <suijingfeng@loongson.cn>
Cc:     Li Yi <liyi@loongson.cn>, etnaviv@lists.freedesktop.org,
        dri-devel@lists.freedesktop.org, linux-kernel@vger.kernel.org,
        Russell King <linux+etnaviv@armlinux.org.uk>,
        Bjorn Helgaas <bhelgaas@google.com>,
        Sui Jingfeng <15330273260@189.cn>
Subject: Re: [PATCH v8 6/8] drm/etnaviv: add driver support for the PCI
 devices
Message-ID: <20230609194621.GA1257228@bhelgaas>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <79e07134-4f89-22dd-5a9c-3c8dfac50bf2@loongson.cn>
Precedence: bulk

On Sat, Jun 10, 2023 at 02:07:58AM +0800, Sui Jingfeng wrote:
> On 2023/6/10 01:52, Bjorn Helgaas wrote:
> > On Fri, Jun 09, 2023 at 09:37:02AM +0800, Sui Jingfeng wrote:
> > > On 2023/6/9 01:32, Bjorn Helgaas wrote:
> > > > On Wed, Jun 07, 2023 at 06:55:49PM +0800, Sui Jingfeng wrote:
> > > > > From: Sui Jingfeng <suijingfeng@loongson.cn>
> > > > > 
> > > > > This patch adds PCI driver support on top of what we already have. Take
> > > > > the GC1000 in LS7A1000/LS2K1000 as the first instance of the PCI device
> > > > > driver. There is only one GPU core for the GC1000 in the LS7A1000 and
> > > > > LS2K1000. Therefore, component frameworks can be avoided.
> > > > > +	{PCI_VENDOR_ID_LOONGSON, 0x7a15, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0},
> > > > > +	{PCI_VENDOR_ID_LOONGSON, 0x7a05, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0},
> > > > PCI_VDEVICE()
> > > This make it impossible to hook device-specific data in the future.
> > > 
> > > But currently there no device specific data associated with the
> > > 0x7a05 and 0x7a15,
> > > 
> > > so it's acceptable for now. Thanks.
> > Haha, ISTR having this conversation before, sorry for repeating it.
> > 
> > Indeed, it's fine as-is.  But PCI_VDEVICE() actually *does* allow for
> > vendor-specific data because it doesn't include the data element,
> > which defaults to zero if you don't specify it.
> > 
> > So for example, drivers/net/ethernet/realtek/r8169_main.c has this:
> > 
> >    { PCI_VDEVICE(REALTEK, 0x8129) },
> >    { PCI_VDEVICE(REALTEK, 0x8136), RTL_CFG_NO_GBIT },
> > 
> > where 0x8129 has no driver_data (it defaults to zero), but 0x8136
> > does.
> 
> PCI_VDEVICE macro end with two zero. (I thought it was three)

No worries, I thought the same thing the first five times I read it :)