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;
Date:   Thu, 24 Oct 2019 12:38:03 +0300
From:   Mika Westerberg <mika.westerberg@linux.intel.com>
To:     Bjorn Helgaas <helgaas@kernel.org>
Cc:     "Rafael J. Wysocki" <rjw@rjwysocki.net>,
        Lukas Wunner <lukas@wunner.de>,
        Keith Busch <keith.busch@intel.com>,
        Andy Shevchenko <andriy.shevchenko@linux.intel.com>,
        Frederick Lawler <fred@fredlawl.com>,
        "Gustavo A . R . Silva" <gustavo@embeddedor.com>,
        Sinan Kaya <okaya@kernel.org>,
        Kai-Heng Feng <kai.heng.feng@canonical.com>,
        linux-pci@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2 2/2] PCI: pciehp: Prevent deadlock on disconnect
Message-ID: <20191024093803.GU2819@lahna.fi.intel.com>
References: <20190812143133.75319-2-mika.westerberg@linux.intel.com>
 <20191022230006.GA46791@google.com>
 <20191023075248.GO2819@lahna.fi.intel.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20191023075248.GO2819@lahna.fi.intel.com>
Organization: Intel Finland Oy - BIC 0357606-4 - Westendinkatu 7, 02160 Espoo
User-Agent: Mutt/1.12.1 (2019-06-15)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Wed, Oct 23, 2019 at 10:52:53AM +0300, Mika Westerberg wrote:
> > Shouldn't we check for slot_status being an error response (~0)
> > instead of looking for PCIBIOS_DEVICE_NOT_FOUND?  There are 7 RsvdP
> > bits in Slot Status, so ~0 is not a valid value for the register.
> > 
> > All 16 bits of Link Status are defined, but ~0 is still an invalid
> > value because the Current Link Speed and Negotiated Link Width fields
> > only define a few valid encodings.
> 
> Indeed that's a good point. I'll try that.

Just checking if I understand correctly what you are suggesting.

Currently we use pcie_capability_read_word() and check the return value.
If the device is gone it returns an error and resets *val to 0. That
only works if pci_dev_is_disconnected() is true so we would need to do
something like below.

pciehp_check_link_active():

	ret = pcie_capability_read_word(pdev, PCI_EXP_LNKSTA, &lnk_status);
	if (ret == PCIBIOS_DEVICE_NOT_FOUND || lnk_status == (u16)~0)
		return -ENODEV;

Or you mean that we check only for ~0 in which case we either need to
use pci_read_config_word() directly here or modify pcie_capability_read_word()
return ~0 instead of clearing it?