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:   Sat, 10 Aug 2019 12:12:32 +0200
From:   Lukas Wunner <lukas@wunner.de>
To:     Keith Busch <kbusch@kernel.org>
Cc:     Bjorn Helgaas <helgaas@kernel.org>, linux-pci@vger.kernel.org,
        linux-kernel@vger.kernel.org,
        Xiongfeng Wang <wangxiongfeng2@huawei.com>
Subject: Re: [PATCH] PCI: pciehp: Avoid returning prematurely from sysfs
 requests
Message-ID: <20190810101232.3fdqrbi3tdd27dwx@wunner.de>
References: <4174210466e27eb7e2243dd1d801d5f75baaffd8.1565345211.git.lukas@wunner.de>
 <20190809193216.GD28515@localhost.localdomain>
 <20190809202815.4jtpdsnnmztins34@wunner.de>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20190809202815.4jtpdsnnmztins34@wunner.de>
User-Agent: NeoMutt/20170113 (1.7.2)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Fri, Aug 09, 2019 at 10:28:15PM +0200, Lukas Wunner wrote:
> On Fri, Aug 09, 2019 at 01:32:16PM -0600, Keith Busch wrote:
> > On Fri, Aug 09, 2019 at 12:28:43PM +0200, Lukas Wunner wrote:
> > > A sysfs request to enable or disable a PCIe hotplug slot should not
> > > return before it has been carried out.  That is sought to be achieved
> > > by waiting until the controller's "pending_events" have been cleared.
> > > 
> > > However the IRQ thread pciehp_ist() clears the "pending_events" before
> > > it acts on them.  If pciehp_sysfs_enable_slot() / _disable_slot() happen
> > > to check the "pending_events" after they have been cleared but while
> > > pciehp_ist() is still running, the functions may return prematurely
> > > with an incorrect return value.
> > > 
> > > Fix by introducing an "ist_running" flag which must be false before a
> > > sysfs request is allowed to return.
> > 
> > Can you instead just call synchronize_irq(ctrl->pcie->irq) after the
> > pending events is cleared?
> 
> You mean call synchronize_irq() from pciehp_sysfs_enable_slot() /
> disable_slot()?  That's a good idea, let me think that through and
> try to make it work that way.

After a bit of thinking:

synchronize_irq() doesn't work for poll mode.

A secondary concern is that if the IRQ is shared, synchronize_irq()
waits for all the other (unrelated) IRQ threads to finish,
i.e. longer than necessary.

So I can't think of a better way to solve this.

Thanks,

Lukas