Re: [PATCH 01/11] brcmfmac: expose device memory to devcoredump subsystem

[Kalle Valo <kvalo@codeaurora.org>]

Arend van Spriel <arend@broadcom.com> writes:

> On 10/14/2015 01:01 PM, Kalle Valo wrote:
>> Arend van Spriel <arend@broadcom.com> writes:
>>
>>> On 10/12/2015 10:59 AM, Kalle Valo wrote:
>>>
>>>> Arend, are the kbuild errors valid? I would guess they are as
>>>> wireless-drivers-next should be in linux-next.
>>>
>>> Actually they are not. I am not sure about the build infrastructure,
>>> but the errors above can only occur when CONFIG_BRCMDBG is set, but
>>> DEBUG is not. However, that can not happen as we have following line
>>> in drivers/net/wireless/brcm80211/Makefile:
>>>
>>> subdir-ccflags-$(CONFIG_BRCMDBG)	+= -DDEBUG
>>
>> I didn't check the details but I'm still not convinced :) In my opinion
>> the code should always compile, with and without -DDEBUG. Anyways, I
>> pushed this script to the pending branch to get kbuild test this patch
>> once again.
>
> Ok, Kalle
>
> Let me try to be a bit more convincing ;-)
>
> I checked out linux-next/master, taken the provided .config and ran:
>
> $ make prepare
> $ make M=scripts
> $ make M=drivers/net/wireless/brcm80211

Thanks, I'm convinced now :) But the kbuild found this warning (I guess
you also got an email about that):

drivers/net/wireless/brcm80211/brcmfmac/sdio.c:3562:2: warning: format '%zu' expects argument of type 'size_t', but argument 5 has type 'unsigned int' [-Wformat=]

Can you fix that and respin, please?

-- 
Kalle Valo