Re: [PATCH v9 06/10] usb: dwc3: qcom: Add support to read IRQ's related to multiport

[Krishna Kurapati PSSNV <quic_kriskura@quicinc.com>]

On 7/14/2023 2:35 PM, Johan Hovold wrote:
> On Wed, Jul 12, 2023 at 11:56:33PM +0530, Krishna Kurapati PSSNV wrote:
>> On 7/12/2023 5:42 PM, Johan Hovold wrote:
>>> On Wed, Jun 21, 2023 at 10:06:24AM +0530, Krishna Kurapati wrote:
>>>> Add support to read Multiport IRQ's related to quad port controller
>>>> of SA8295 Device.
>>>>
>>>> Signed-off-by: Krishna Kurapati <quic_kriskura@quicinc.com>
>>>> ---
>>>>    drivers/usb/dwc3/dwc3-qcom.c | 108 +++++++++++++++++++++++++++++------
>>>>    1 file changed, 91 insertions(+), 17 deletions(-)
>>>
>>>> +static int dwc3_qcom_setup_mp_irq(struct platform_device *pdev)
>>>> +{
>>>> +	struct dwc3_qcom *qcom = platform_get_drvdata(pdev);
>>>> +	char irq_name[15];
>>>
>>> The interrupt device-name string can not be allocated on the stack or
>>> reused as it is stored directly in each irqaction structure.
>>>
>>> This can otherwise lead to random crashes when accessing
>>> /proc/interrupts:
>>>
>>> 	https://lore.kernel.org/lkml/ZK6IV_jJPICX5r53@hovoldconsulting.com/
> 
>>     Sure, will create a static array of names if possible in global
>> section of file and use it to read interrupts.
> 
> Or just use devm_kasprintf(), which should allow for a cleaner
> implementation.
> 
> I've fixed it up like this for my X13s wip branches:
> 
> 	https://github.com/jhovold/linux/commit/0898b54456bc2f4bd4d420480db98e6758771ace
>   
>>     Are you fine with seperating out setup_irq and setup_mp_irq functions
>> ? Can you please review comments and suggestion on [1].
> 
> I haven't had time to look at your latest replies yet, but as I already
> said when reviewing v9, it seems you should be using a common helper for
> non-mp and mp.
> 
Hi Johan,

  The gist of my mail was to see if I can defer qcom probe when dwc3 
probe fails/or doesn't happen on of_plat_pop (which is logical) so that 
we can move setup_irq to after dwc3_register_core so that we know 
whether we are MP capable or not. This would help us move all IRQ 
reading into one function.

Regards,
Krishna,