[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-Id: <20181015153936eucas1p10fc822d6161c3d95d5d0bdd98408e55f~d0mRowyUD3030630306eucas1p1K@eucas1p1.samsung.com>
Date: Mon, 15 Oct 2018 17:39:34 +0200
From: Bartlomiej Zolnierkiewicz <b.zolnierkie@...sung.com>
To: Krzysztof Kozlowski <krzk@...nel.org>
Cc: kgene@...nel.org, linux@...linux.org.uk,
linux-arm-kernel@...ts.infradead.org,
"linux-samsung-soc@...r.kernel.org"
<linux-samsung-soc@...r.kernel.org>, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] ARM: SAMSUNG: limit SAMSUNG_PM_DEBUG config option
to non-Exynos platforms
On 10/15/2018 05:22 PM, Krzysztof Kozlowski wrote:
> On Mon, 15 Oct 2018 at 15:50, Bartlomiej Zolnierkiewicz
> <b.zolnierkie@...sung.com> wrote:
>>
>> "Samsung PM Suspend debug" feature (controlled by SAMSUNG_PM_DEBUG
>> config option) is not working properly (debug messages are not
>> displayed after resume) on Exynos platforms because GPIOs restore
>> code is not implemented.
>>
>> Add PLAT_S3C24XX, ARCH_S3C64XX and ARCH_S5PV210 dependencies to
>> SAMSUNG_PM_DEBUG config option to hide it on Exynos platforms.
>> Then convert Exynos code to not require <plat/pm-common.h>
>> header (use pr_debug() directly instead of S3C_PMDBG() macro and
>> remove redundant s3c_pm_*() calls).
>>
>> Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@...sung.com>
>> ---
>> v2:
>> - convert Exynos code to not require <plat/pm-common.h> header
>>
>> Should by applied on top of "[PATCH] ARM: EXYNOS: remove no longer
>> needed s3c_pm_check_*() calls".
>>
>> arch/arm/mach-exynos/suspend.c | 14 ++++----------
>> arch/arm/plat-samsung/Kconfig | 3 ++-
>> 2 files changed, 6 insertions(+), 11 deletions(-)
>>
>> Index: b/arch/arm/mach-exynos/suspend.c
>> ===================================================================
>> --- a/arch/arm/mach-exynos/suspend.c 2018-10-15 15:11:15.296187842 +0200
>> +++ b/arch/arm/mach-exynos/suspend.c 2018-10-15 15:22:54.656205454 +0200
>> @@ -30,8 +30,6 @@
>> #include <asm/smp_scu.h>
>> #include <asm/suspend.h>
>>
>> -#include <plat/pm-common.h>
>> -
>> #include "common.h"
>>
>> #define REG_TABLE_END (-1U)
>> @@ -490,11 +488,9 @@ static int exynos_suspend_enter(suspend_
>> {
>> int ret;
>>
>> - s3c_pm_debug_init();
>> -
>> - S3C_PMDBG("%s: suspending the system...\n", __func__);
>> + pr_debug("%s: suspending the system...\n", __func__);
>>
>> - S3C_PMDBG("%s: wakeup masks: %08x,%08x\n", __func__,
>> + pr_debug("%s: wakeup masks: %08x,%08x\n", __func__,
>> exynos_irqwake_intmask, exynos_get_eint_wake_mask());
>>
>> if (exynos_irqwake_intmask == -1U
>> @@ -504,7 +500,6 @@ static int exynos_suspend_enter(suspend_
>> return -EINVAL;
>> }
>>
>> - s3c_pm_save_uarts();
>> if (pm_data->pm_prepare)
>> pm_data->pm_prepare();
>> flush_cache_all();
>> @@ -517,12 +512,11 @@ static int exynos_suspend_enter(suspend_
>>
>> if (pm_data->pm_resume_prepare)
>> pm_data->pm_resume_prepare();
>> - s3c_pm_restore_uarts();
>>
>> - S3C_PMDBG("%s: wakeup stat: %08x\n", __func__,
>> + pr_debug("%s: wakeup stat: %08x\n", __func__,
>> pmu_raw_readl(S5P_WAKEUP_STAT));
>>
>> - S3C_PMDBG("%s: resuming the system...\n", __func__);
>> + pr_debug("%s: resuming the system...\n", __func__);
>>
>> return 0;
>> }
>> Index: b/arch/arm/plat-samsung/Kconfig
>> ===================================================================
>> --- a/arch/arm/plat-samsung/Kconfig 2018-10-15 15:16:22.804195586 +0200
>> +++ b/arch/arm/plat-samsung/Kconfig 2018-10-15 15:16:22.800195586 +0200
>> @@ -238,7 +238,8 @@ comment "Power management"
>>
>> config SAMSUNG_PM_DEBUG
>> bool "Samsung PM Suspend debug"
>> - depends on PM && DEBUG_KERNEL
>> + depends on DEBUG_KERNEL
>
> I don't get this change - why moving PM to line below?
My mistake, I will correct it in v3.
[ The idea was to make SAMSUNG_PM_DEBUG consistent with other options
depending on DEBUG_KERNEL. However after I've checked them it turned
out that there is no such consistency. I intended to revert this
change but ultimately I forgot about it.. ]
Best regards,
--
Bartlomiej Zolnierkiewicz
Samsung R&D Institute Poland
Samsung Electronics
Powered by blists - more mailing lists