Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Subject: Re: [PATCH v4 1/5] vfio/migration: Add debugfs to live migration
 driver
To:     Jason Gunthorpe <jgg@nvidia.com>
CC:     <alex.williamson@redhat.com>,
        <shameerali.kolothum.thodi@huawei.com>, <john.garry@huawei.com>,
        <cohuck@redhat.com>, <linux-kernel@vger.kernel.org>,
        <linuxarm@openeuler.org>
References: <20221202092625.35075-1-liulongfang@huawei.com>
 <20221202092625.35075-2-liulongfang@huawei.com> <Y4oW/ToMZEUnlYZ1@nvidia.com>
From:   liulongfang <liulongfang@huawei.com>
Message-ID: <f17dfd5d-3d65-d51e-426a-e60f7e99a3d6@huawei.com>
Date:   Mon, 5 Dec 2022 21:39:29 +0800
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:60.0) Gecko/20100101
 Thunderbird/60.8.0
MIME-Version: 1.0
In-Reply-To: <Y4oW/ToMZEUnlYZ1@nvidia.com>
Content-Type: text/plain; charset="gbk"
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 2022/12/2 23:17, Jason Gunthorpe wrote:
> On Fri, Dec 02, 2022 at 05:26:21PM +0800, Longfang Liu wrote:
> 
>> +static ssize_t vfio_pci_vf_state_read(struct file *filp, char __user *buffer,
>> +			   size_t count, loff_t *pos)
>> +{
> 
> This would be better to use debugfs_create_devm_seqfile()
> 
> Then you can simply use seq_printf()
>

The previous debugfs used this method in my patch.
Is it now recommended to use the interface with "devm" ?


>> @@ -119,6 +129,8 @@ struct vfio_migration_ops {
>>  				   enum vfio_device_mig_state *curr_state);
>>  	int (*migration_get_data_size)(struct vfio_device *device,
>>  				       unsigned long *stop_copy_length);
>> +	int (*migration_get_data)(struct vfio_device *device, char *buffer);
>> +	int (*migration_get_attr)(struct vfio_device *device, char
>>  				   *buffer);
> 
> Still no to passing through debugfs ops, create these files in the
> driver.
> 

If this part is also created and enabled in the device driver, why do we need
to put this debugfs in the vfio public framework?

Thanks,
Longfang.
> Jason
> .
>