Re: [RESEND BUGFIX PATCH 1/3] PCI/AER: fix pci_ops return NULL when hotplug a pci bus which was doing aer error inject

From: Yijing Wang <wangyijing@huawei.com>
To: Huang Ying <ying.huang@intel.com>
Cc: Jiang Liu <liuj97@gmail.com>, Bjorn Helgaas <bhelgaas@google.com>,
	Rusty Russell <rusty@rustcorp.com.au>,
	Mauro Carvalho Chehab <mchehab@redhat.com>,
	PCI <linux-pci@vger.kernel.org>,
	Hanjun Guo <guohanjun@huawei.com>, <linux-kernel@vger.kernel.org>
Subject: Re: [RESEND BUGFIX PATCH 1/3] PCI/AER: fix pci_ops return NULL when hotplug a pci bus which was doing aer error inject
Date: Tue, 28 Aug 2012 08:53:13 +0800	[thread overview]
Message-ID: <503C1679.4020008@huawei.com> (raw)
In-Reply-To: <1346114304.6279.341.camel@yhuang-dev>

On 2012/8/28 8:38, Huang Ying wrote:
> On Mon, 2012-08-27 at 23:05 +0800, Jiang Liu wrote:
>> Is it ok to ignore such a case? After all, aer_inject is just a test tool:)
>> It's not worth to change the core logic for such a corner case.
>> --Gerry
> 
> Why ignore?  At least you can prevent aer_inject from unload if
> something special happened.
> 

Hi Huang Ying,
   Thanks for your comments. It's my negligence. I will add some protection code when do #rmmod aer_inject(a race condition window about bus_ops),
I will correct it in the new version patch.

----------
Thanks!
Yijing

>> On 08/27/2012 09:23 AM, Huang Ying wrote:
>>> On Sat, 2012-08-25 at 17:59 +0800, Yijing Wang wrote:
>>>> When we inject aer errors to the target pci device by aer_inject module, the pci_ops of pci
>>>> bus which the target device is on will be assign to pci_ops_aer.So if the target pci device
>>>> is a bridge, once we hotplug the pci bus(child bus) which the target device bridges to, child
>>>> bus's pci_ops will be assigned to pci_ops_aer too.Now every access to the child bus's device
>>>> will result to system panic, because it return NULL pci_ops in pci_read_aer.
>>>> This patch fix this.