Re: [PATCH] irqchip/mbigen: stop printing kernel addresses

[Marc Zyngier]

On 18/06/2019 09:35, Kefeng Wang wrote:
> 
> 
> On 2019/6/18 15:48, Marc Zyngier wrote:
>> Hi Kefeng,
>>
>> On Tue, 18 Jun 2019 04:22:02 +0100,
>> Kefeng Wang <wangkefeng.wang@xxxxxxxxxx> wrote:
>>>
>>> After commit ad67b74d2469d9b8 ("printk: hash addresses printed with %p"),
>>> it will print "____ptrval____" instead of actual addresses when mbigen
>>> create domain fails,
>>>
>>>   Hisilicon MBIGEN-V2 HISI0152:00: Failed to create mbi-gen@(____ptrval____) irqdomain
>>>   Hisilicon MBIGEN-V2: probe of HISI0152:00 failed with error -12
>>>
>>> Instead of changing the print to "%px", and leaking kernel addresses,
>>> just remove the print completely.
>>>
>>> Signed-off-by: Kefeng Wang <wangkefeng.wang@xxxxxxxxxx>
>>> ---
>>>  drivers/irqchip/irq-mbigen.c | 3 +--
>>>  1 file changed, 1 insertion(+), 2 deletions(-)
>>>
>>> diff --git a/drivers/irqchip/irq-mbigen.c b/drivers/irqchip/irq-mbigen.c
>>> index 98b6e1d4b1a6..d0cf596c801b 100644
>>> --- a/drivers/irqchip/irq-mbigen.c
>>> +++ b/drivers/irqchip/irq-mbigen.c
>>> @@ -355,8 +355,7 @@ static int mbigen_device_probe(struct platform_device *pdev)
>>>  		err = -EINVAL;
>>>  
>>>  	if (err) {
>>> -		dev_err(&pdev->dev, "Failed to create mbi-gen@%p irqdomain",
>>> -			mgn_chip->base);
>>> +		dev_err(&pdev->dev, "Failed to create mbi-gen irqdomain");
>>
>> The alternative would be to print res as a resource, which would still
>> help identifying the offending device by printing its physical
>> layout, and still not reveal much.
> 
> It's better to print res to show the physical layout, and add missing "\n",
> will resend v2.

As Hanjun mentioned in a separate email, pdev->dev seems to be enough to
identify which MBIGEN has failed to probe. So maybe all you need to do
is to add the missing '\n', and tidy up the commit message to reflect that.

Thanks,

	M.
-- 
Jazz is not dead. It just smells funny...