Re: [PATCH 1/2] btrfs: open code log helpers in device_list_add()

[Anand Jain <anand.jain@oracle.com>]

On 14/1/20 12:25 AM, David Sterba wrote:
> On Sat, Jan 11, 2020 at 07:41:51AM +0800, Anand Jain wrote:
>>>>    			if (device->bdev != path_bdev) {
>>>>    				bdput(path_bdev);
>>>>    				mutex_unlock(&fs_devices->device_list_mutex);
>>>> -				btrfs_warn_in_rcu(device->fs_info,
>>>> -			"duplicate device fsid:devid for %pU:%llu old:%s new:%s",
>>>> +				rcu_read_lock();
>>>> +				printk_ratelimited(
>>>
>>> Avoiding fs_info here is correct but we don't want to use raw printk or
>>> printk_ratelimited anywhere.
>>>
>>
>>    I think I discussed this a long time back, that we should rather pass
>>    fs_devices in btrfs_warn_in_rcu().
>>
>>    I am ok to make such a change, are you ok?
> 
> No, this does not sound right at all. Why should be btrfs_warn_in_rcu
> special from the other message callbacks? We need to fix one context, so
> let's find something less hacky.
> 
>>    Or I wonder if there is
>>    any other way?
> 
> We could add a fs_info stub that will get recognized in btrfs_printk.
> Eg.
> 
> #define	NO_FS_INFO		((void*)0x1)
> 
> btrfs_printk() {
> 
> 	if (fs_info == NULL)
> 		devname = "<unknown>";
> 	else if (fs_info == NO_FS_INFO)
> 		devname = "...";
> 	else
> 		devname = fs_info->sb->sb_id;
> 

Yeah it makes sense to me. Patches sent.

Thanks, Anand