Message-id: <564B332E.6010608@samsung.com>
Date: Tue, 17 Nov 2015 15:01:18 +0100
From: Sylwester Nawrocki <s.nawrocki@samsung.com>
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101
 Thunderbird/31.8.0
MIME-version: 1.0
To: Krzysztof Kozlowski <k.kozlowski@samsung.com>
Cc: "pankaj.dubey" <pankaj.dubey@samsung.com>,
        Tomasz Figa <tomasz.figa@gmail.com>,
        Michael Turquette <mturquette@baylibre.com>,
        Stephen Boyd <sboyd@codeaurora.org>, linux-samsung-soc@vger.kernel.org,
        linux-clk@vger.kernel.org, linux-kernel@vger.kernel.org,
        Catalin Marinas <catalin.marinas@arm.com>,
        Will Deacon <will.deacon@arm.com>, Kukjin Kim <kgene@kernel.org>,
        Olof Johansson <olof@lixom.net>, Arnd Bergmann <arnd@arndb.de>,
        linux-arm-kernel@lists.infradead.org,
        Kevin Hilman <khilman@linaro.org>
Subject: Re: [PATCH 1/2] clk: samsung: Don't build ARMv8 clock drivers on ARMv7
References: <1447637775-9887-1-git-send-email-k.kozlowski@samsung.com>
 <1447637775-9887-2-git-send-email-k.kozlowski@samsung.com>
 <564AADB2.3030400@samsung.com> <564AAF70.9080507@samsung.com>
In-reply-to: <564AAF70.9080507@samsung.com>
Content-type: text/plain; charset=windows-1252
Content-transfer-encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 3600
Lines: 94

On 17/11/15 05:39, Krzysztof Kozlowski wrote:
> On 17.11.2015 13:31, pankaj.dubey wrote:
>> On Monday 16 November 2015 07:06 AM, Krzysztof Kozlowski wrote:
>>> Currently the Exynos5433 (ARMv8 SoC) clock driver depends on ARCH_EXYNOS
>>> so it is built also on ARMv7. This does not bring any kind of benefit.
>>> There won't be a single kernel image for ARMv7 and ARMv8 SoCs (like
>>> multi_v7 for ARMv7).
>>>
>>> Instead build clock drivers only for respective SoC's architecture.
>>>
>>> Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
>>> ---
>>>  drivers/clk/samsung/Kconfig  | 13 +++++++++++++
>>>  drivers/clk/samsung/Makefile |  4 ++--
>>>  2 files changed, 15 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/drivers/clk/samsung/Kconfig b/drivers/clk/samsung/Kconfig
>>> index 84196ecdaa12..5f138fc4d84d 100644
>>> --- a/drivers/clk/samsung/Kconfig
>>> +++ b/drivers/clk/samsung/Kconfig
>>> @@ -2,6 +2,7 @@ config COMMON_CLK_SAMSUNG
>>>  	bool
>>>  	select COMMON_CLK
>>>  
>>> +# ARMv7 SoCs:
>>>  config S3C2410_COMMON_CLK
>>>  	bool
>>>  	select COMMON_CLK_SAMSUNG
>>> @@ -24,3 +25,15 @@ config S3C2443_COMMON_CLK
>>>  	bool
>>>  	select COMMON_CLK_SAMSUNG
>>>  
>>> +# ARMv8 SoCs:
>>> +config EXYNOS5433_COMMON_CLK
>>> +	bool
>>> +	depends on ARM64 || COMPILE_TEST
>>> +	default ARCH_EXYNOS
>>> +	select COMMON_CLK_SAMSUNG
>>> +
>>> +config EXYNOS7_COMMON_CLK
>>> +	bool
>>> +	depends on ARM64 || COMPILE_TEST
>>> +	default ARCH_EXYNOS
>>> +	select COMMON_CLK_SAMSUNG
>>> diff --git a/drivers/clk/samsung/Makefile b/drivers/clk/samsung/Makefile
>>> index 5f6833ea355d..a31332a24ef4 100644
>>> --- a/drivers/clk/samsung/Makefile
>>> +++ b/drivers/clk/samsung/Makefile
>>> @@ -10,11 +10,11 @@ obj-$(CONFIG_SOC_EXYNOS5250)	+= clk-exynos5250.o
>>>  obj-$(CONFIG_SOC_EXYNOS5260)	+= clk-exynos5260.o
>>>  obj-$(CONFIG_SOC_EXYNOS5410)	+= clk-exynos5410.o
>>>  obj-$(CONFIG_SOC_EXYNOS5420)	+= clk-exynos5420.o
>>> -obj-$(CONFIG_ARCH_EXYNOS)	+= clk-exynos5433.o
>>> +obj-$(CONFIG_EXYNOS5433_COMMON_CLK)	+= clk-exynos5433.o
>>>  obj-$(CONFIG_SOC_EXYNOS5440)	+= clk-exynos5440.o
>>>  obj-$(CONFIG_ARCH_EXYNOS)	+= clk-exynos-audss.o
>>>  obj-$(CONFIG_ARCH_EXYNOS)	+= clk-exynos-clkout.o
>>> -obj-$(CONFIG_ARCH_EXYNOS7)	+= clk-exynos7.o
>>> +obj-$(CONFIG_EXYNOS7_COMMON_CLK)	+= clk-exynos7.o
>>>  obj-$(CONFIG_S3C2410_COMMON_CLK)+= clk-s3c2410.o
>>>  obj-$(CONFIG_S3C2410_COMMON_DCLK)+= clk-s3c2410-dclk.o
>>>  obj-$(CONFIG_S3C2412_COMMON_CLK)+= clk-s3c2412.o
>>>
>>
>> So in this approach we need to add separate config for clock support of
>> each ARM64 Exynos64 SoC. Is this fine?
>>
>> Can we club compilation of each ARM64 Exynos SoC clock file under
>> EXYNOS7_COMMON_CLK? As for all ARM64 SoC there is single defconfig and
>> binary.
> 
> Yes, it can be one config symbol for all clocks of ARMv8 Exynos SoCs.
> From my point of view both has some advantages and disadvantages (kernel
> size, granularity, number of Kconfig symbols etc.) and I don't mind
> choosing different than I selected before.
> 
> Any opinion from Samsung clock maintainers? Which do you prefer?

It would have been a bit unfortunate to not be able to exclude
the unneeded clk drivers from build.  From my side both patches
look like a step in right direction.

For the $subject patch:

Acked-by: Sylwester Nawrocki <s.nawrocki@samsung.com>

-- 
Thanks
Sylwester
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/