Received-SPF: pass (google.com: domain of linux-kernel+bounces-20419-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.48.161 as permitted sender) client-ip=147.75.48.161;
Message-ID: <ca74243f-d1d1-4b01-95a6-58b4c85842d9@linux.intel.com>
Date: Tue, 9 Jan 2024 13:55:59 +0800
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Cc: baolu.lu@linux.intel.com, Joerg Roedel <joro@8bytes.org>,
 Will Deacon <will@kernel.org>, Robin Murphy <robin.murphy@arm.com>,
 Kevin Tian <kevin.tian@intel.com>,
 Jean-Philippe Brucker <jean-philippe@linaro.org>,
 Nicolin Chen <nicolinc@nvidia.com>, Yi Liu <yi.l.liu@intel.com>,
 Jacob Pan <jacob.jun.pan@linux.intel.com>,
 Longfang Liu <liulongfang@huawei.com>, Yan Zhao <yan.y.zhao@intel.com>,
 iommu@lists.linux.dev, kvm@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH v9 14/14] iommu: Track iopf group instead of last fault
Content-Language: en-US
To: Jason Gunthorpe <jgg@ziepe.ca>
References: <20231220012332.168188-1-baolu.lu@linux.intel.com>
 <20231220012332.168188-15-baolu.lu@linux.intel.com>
 <20240105175339.GI50608@ziepe.ca>
From: Baolu Lu <baolu.lu@linux.intel.com>
In-Reply-To: <20240105175339.GI50608@ziepe.ca>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit

On 1/6/24 1:53 AM, Jason Gunthorpe wrote:
> On Wed, Dec 20, 2023 at 09:23:32AM +0800, Lu Baolu wrote:
>>   /**
>> - * iommu_handle_iopf - IO Page Fault handler
>> - * @fault: fault event
>> - * @iopf_param: the fault parameter of the device.
>> + * iommu_report_device_fault() - Report fault event to device driver
>> + * @dev: the device
>> + * @evt: fault event data
>>    *
>> - * Add a fault to the device workqueue, to be handled by mm.
>> + * Called by IOMMU drivers when a fault is detected, typically in a threaded IRQ
>> + * handler. When this function fails and the fault is recoverable, it is the
>> + * caller's responsibility to complete the fault.
> This patch seems OK for what it does so:
> 
> Reviewed-by: Jason Gunthorpe<jgg@nvidia.com>
> 
> However, this seems like a strange design, surely this function should
> just call ops->page_response() when it can't enqueue the fault?
> 
> It is much cleaner that way, so maybe you can take this into a
> following patch (along with the driver fixes to accomodate. (and
> perhaps iommu_report_device_fault() should return void too)
> 
> Also iopf_group_response() should return void (another patch!),
> nothing can do anything with the failure. This implies that
> ops->page_response() must also return void - which is consistent with
> what the drivers do, the failure paths are all integrity validations
> of the fault and should be WARN_ON'd not return codes.

Make sense. I will integrate the code in the next version and convert
iommu_report_device_fault() to return void.

Best regards,
baolu