Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Subject: Re: [PATCH 09/10] timer-riscv: Fix undefined riscv_time_val
To:     Kefeng Wang <wangkefeng.wang@huawei.com>,
        Palmer Dabbelt <palmer@dabbelt.com>, tglx@linutronix.de
Cc:     Paul Walmsley <paul.walmsley@sifive.com>, aou@eecs.berkeley.edu,
        linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org,
        hulkci@huawei.com
References: <mhng-0c491e9e-41fd-4f76-a048-55c03d9359f0@palmerdabbelt-glaptop1>
 <66121f9a-48f3-d3a5-7c96-d71397e12aed@linaro.org>
 <0bc3eb36-7b9d-7c86-130c-68b566e85c10@huawei.com>
 <e1916079-c209-bfad-6b0c-ccfb2e136ca4@linaro.org>
 <29dc112e-d8c2-2749-7f5d-7c0c19aa9092@huawei.com>
From:   Daniel Lezcano <daniel.lezcano@linaro.org>
Message-ID: <8c5ecbd3-c23a-ccd4-b5d8-2e7d2bd10699@linaro.org>
Date:   Tue, 19 May 2020 15:51:08 +0200
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101
 Thunderbird/68.7.0
MIME-Version: 1.0
In-Reply-To: <29dc112e-d8c2-2749-7f5d-7c0c19aa9092@huawei.com>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 8bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 19/05/2020 14:39, Kefeng Wang wrote:
> 
> On 2020/5/19 4:23, Daniel Lezcano wrote:
>> Hi Kefeng,
>>
>> On 18/05/2020 17:40, Kefeng Wang wrote:
>>> On 2020/5/18 22:09, Daniel Lezcano wrote:
>>>> On 13/05/2020 23:14, Palmer Dabbelt wrote:
>>>>> On Sun, 10 May 2020 19:20:00 PDT (-0700), wangkefeng.wang@huawei.com
>>>>> wrote:
>>>>>> ERROR: modpost: "riscv_time_val" [crypto/tcrypt.ko] undefined!
>>>>>>
>>>>>> Reported-by: Hulk Robot <hulkci@huawei.com>
>>>>>> Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
>>>>>> ---
>>>>>>    drivers/clocksource/timer-riscv.c | 1 +
>>>>>>    1 file changed, 1 insertion(+)
>>>>>>
>>>>>> diff --git a/drivers/clocksource/timer-riscv.c
>>>>>> b/drivers/clocksource/timer-riscv.c
>>>>>> index c4f15c4068c0..071b8c144027 100644
>>>>>> --- a/drivers/clocksource/timer-riscv.c
>>>>>> +++ b/drivers/clocksource/timer-riscv.c
>>>>>> @@ -19,6 +19,7 @@
>>>>>>
>>>>>>    u64 __iomem *riscv_time_cmp;
>>>>>>    u64 __iomem *riscv_time_val;
>>>>>> +EXPORT_SYMBOL(riscv_time_val);
>>>>>>
>>>>>>    static inline void mmio_set_timer(u64 val)
>>>>>>    {
>>>>> Reviewed-by: Palmer Dabbelt <palmerdabbelt@google.com>
>>>>> Acked-by: Palmer Dabbelt <palmerdabbelt@google.com>
>>>>>
>>>>> Adding the clocksource maintainers.  Let me know if you want this
>>>>> through my
>>>>> tree, I'm assuming you want it through your tree.
>>>> How can we end up by an export symbol here ?!
>>> Hi Danile,
>> s/Danile/Daniel/
> Sorry for typing error.
>>
>>> Found this build error when CONFIG_RISCV_M_MODE=y and CONFIG_RISCV_SBI
>>> is not,
>>>
>>> see patch "4f9bbcefa142 riscv: add support for MMIO access to the timer
>>> registers"
>> Thanks for the pointer.
>>
>> The question still remains, how do we end up with this EXPORT_SYMBOL?
>>
>> There is something wrong if the fix is an EXPORT_SYMBOL for a global
>> variable.
> 
> Not very clear, there are some global variable( eg, acpi_disabled,
> memstart_addr in arm64,) is exported by EXPORT_SYMBOL,  do you mean that
> export riscv_time_val is wrong way?

I do not maintain acpi neither arm64.mm.

AFAICT, riscv_time_val is globally declared in
drivers/clocksource/timer-riscv.c

The driver does not use this variable at all. Then there is a readl on
it in the header file arch/riscv/include/asm/timex.h

And finally it is initialized in arch/riscv/kernel/clint.c

Same thing for riscv_time_cmp.

The correct fix is to initialize the variables in the place where they
belong to (drivers/clocksource/timer-riscv.c), create a function to read
their content and export-symbol-gpl the function.


-- 
<http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs

Follow Linaro:  <http://www.facebook.com/pages/Linaro> Facebook |
<http://twitter.com/#!/linaroorg> Twitter |
<http://www.linaro.org/linaro-blog/> Blog