Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Message-ID: <1549445274.2544.57.camel@pengutronix.de>
Subject: Re: [PATCH RESEND] PCI: Check for USB xHCI class for HAPS platform
From:   Lucas Stach <l.stach@pengutronix.de>
To:     Bjorn Helgaas <helgaas@kernel.org>,
        Thinh Nguyen <thinh.nguyen@synopsys.com>
Cc:     linux-kernel@vger.kernel.org, linux-pci@vger.kernel.org,
        "Lukas F. Hartmann" <lukas@mntmn.com>,
        Trent Piepho <tpiepho@impinj.com>,
        John Youn <john.youn@synopsys.com>,
        Richard Zhu <hongxing.zhu@nxp.com>
Date:   Wed, 06 Feb 2019 10:27:54 +0100
In-Reply-To: <20190205233159.GC7268@google.com>
References: <a28cfcee61f0842efe114c97b566bb3d967ba3d3.1549326366.git.thinhn@synopsys.com>
         <20190205233159.GC7268@google.com>
Content-Type: text/plain; charset="UTF-8"
Mime-Version: 1.0
Content-Transfer-Encoding: 8bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Hi Bjorn,

Am Dienstag, den 05.02.2019, 17:31 -0600 schrieb Bjorn Helgaas:
> [+cc Richard, Lucas]
> 
> On Tue, Feb 05, 2019 at 01:04:28PM -0800, Thinh Nguyen wrote:
> > The Synopsys HAPS USB controller has a VID PID (16c3,abcd) that
> > matches
> > to an existing PCIe controller. This quirk is intended for USB HAPS
> > devices only. To fix this, check for the PCI class USB xHCI to
> > prevent
> > matching the PCIe controllers.
> 
> So there are at least three different parts with the same Vendor &
> Device ID ([16c3:abcd]):
> 
>   1) Synopsys HAPS USB3 controller
>   2) Synopsys PCIe IP in the NXP i.MX6QP (reported by Lukas)

The first incorporation of this IP in a Freescale SoC (i.MX6Q) also
uses the device ID 0xabcd.

So this seems to be a major miscommunication between NXP/Freescale and
Synopsys.

Regards,
Lucas

>   3) Synopsys PCIe IP in the NXP i.MX7D (reported by Trent)
> 
> I don't know if Synopsys is to blame for 2 & 3, or if NXP was
> expected
> to change the Vendor ID when incorporating the Synopsys IP into the
> i.MX designs.  But even leaving the default Device ID of the PCIe IP
> the same as another Synopsys device was probably a bad idea.
> 
> dwc3-haps claims by Vendor & Device ID; it doesn't look at the Class
> Code.  What happens when it tries to claim the PCIe IP in i.MX?
> 
> lspci also doesn't look at the Class Code when it looks up the text
> name of devices, so it will print the same name for both.  Of course,
> the Linux PCI ID database at https://pci-ids.ucw.cz/read/PC/16c3
> shows
> basically no information for the 0x16c3 Vendor ID, so it won't print
> anything useful no matter what.  You could add some things there :)
> 
> > Fixes: 03e6742584af ("PCI: Override Synopsys USB 3.x HAPS device
> > class")
> > Signed-off-by: Thinh Nguyen <thinhn@synopsys.com>
> > ---
> >  drivers/pci/quirks.c | 5 +++--
> >  1 file changed, 3 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
> > index b0a413f3f7ca..e2a879e93d86 100644
> > --- a/drivers/pci/quirks.c
> > +++ b/drivers/pci/quirks.c
> > @@ -639,8 +639,9 @@ static void quirk_synopsys_haps(struct pci_dev
> > *pdev)
> >  		break;
> >  	}
> >  }
> > -DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_SYNOPSYS, PCI_ANY_ID,
> > -			 quirk_synopsys_haps);
> > +DECLARE_PCI_FIXUP_CLASS_HEADER(PCI_VENDOR_ID_SYNOPSYS, PCI_ANY_ID,
> > +			       PCI_CLASS_SERIAL_USB_XHCI, 0,
> > +			       quirk_synopsys_haps);
> >  
> >  /*
> >   * Let's make the southbridge information explicit instead of
> > having to
> > -- 
> > 2.11.0
> >