[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20191113063519.ivv2ehejxonkfufe@wunner.de>
Date: Wed, 13 Nov 2019 07:35:19 +0100
From: Lukas Wunner <lukas@...ner.de>
To: Bjorn Helgaas <helgaas@...nel.org>
Cc: Mika Westerberg <mika.westerberg@...ux.intel.com>,
"Rafael J. Wysocki" <rjw@...ysocki.net>,
Keith Busch <keith.busch@...el.com>,
Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
Frederick Lawler <fred@...dlawl.com>,
"Gustavo A . R . Silva" <gustavo@...eddedor.com>,
Sinan Kaya <okaya@...nel.org>,
Kai-Heng Feng <kai.heng.feng@...onical.com>,
linux-pci@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3 2/2] PCI: pciehp: Prevent deadlock on disconnect
On Tue, Nov 12, 2019 at 09:17:52PM -0600, Bjorn Helgaas wrote:
> On Tue, Oct 29, 2019 at 08:00:22PM +0300, Mika Westerberg wrote:
> > parent PCIe downstream port, who got the hot-remove event, starts
> > removing devices below it taking pci_lock_rescan_remove() lock. When the
> > child PCIe port is runtime resumed it calls pciehp_check_presence()
> > which ends up calling pciehp_card_present() and pciehp_check_link_active().
>
> Oh, I see, pciehp_resume() calls pciehp_check_presence(), which
> schedules the IRQ thread via pciehp_request(). So does this deadlock
> only happen if the port(s) have been runtime suspended?
No, there is a multitude of situations when the deadlock may occur and
this is just one of them. See my comment on v1 of this patch:
https://patchwork.ozlabs.org/patch/1117870/#2230798
Thanks,
Lukas
Powered by blists - more mailing lists