| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Sat, 26 Apr 2014 00:40:47 +0200
From: Tomasz Figa <tomasz.figa@...il.com>
To: Pankaj Dubey <pankaj.dubey@...sung.com>,
linux-samsung-soc@...r.kernel.org, linux-kernel@...r.kernel.org,
linux-arm-kernel@...ts.infradead.org
CC: kgene.kim@...sung.com, linux@....linux.org.uk, t.figa@...sung.com,
chow.kim@...sung.com, yg1004.jang@...sung.com,
vikas.sajjan@...sung.com, s.nawrocki@...sung.com,
b.zolnierkie@...sung.com
Subject: Re: [PATCH v2 10/10] ARM: EXYNOS: Add device tree based initialization
support for PMU.
Hi,
On 25.04.2014 14:32, Pankaj Dubey wrote:
> This patch adds device tree based initialization for PMU and modifies
> PMU initialization implementation in following way:
>
> 1: Let's initialize PMU based on device tree compatibility string.
> 2: Obtain PMU regmap handle using "syscon_early_regmap_lookup_by_phandle"
> so that we can reduce dependency over machine header files.
> 3: Separate each SoC's PMU initialization function and bind this initialization
> function with PMU compatibility string.
> 4 : It also removes uses of soc_is_exynosXXXX() thus making PMU implementation
> independent of "plat/cpu.h".
>
> Signed-off-by: Pankaj Dubey <pankaj.dubey@...sung.com>
> Signed-off-by: Young-Gun Jang <yg1004.jang@...sung.com>
> ---
> arch/arm/mach-exynos/pmu.c | 182 +++++++++++++++++++++++++++++++++-----------
> 1 file changed, 138 insertions(+), 44 deletions(-)
>
> diff --git a/arch/arm/mach-exynos/pmu.c b/arch/arm/mach-exynos/pmu.c
> index 67116a5..abcf753 100644
> --- a/arch/arm/mach-exynos/pmu.c
> +++ b/arch/arm/mach-exynos/pmu.c
> @@ -9,17 +9,31 @@
> * published by the Free Software Foundation.
> */
>
> -#include <linux/io.h>
> #include <linux/kernel.h>
> #include <linux/regmap.h>
> -
> -#include <plat/cpu.h>
> +#include <linux/of.h>
> +#include <linux/platform_device.h>
> +#include <linux/slab.h>
> +#include <linux/mfd/syscon.h>
>
> #include "common.h"
> #include "regs-pmu.h"
>
> +enum exynos_pmu_id {
> + PMU_EXYNOS4210,
> + PMU_EXYNOS4X12,
> + PMU_EXYNOS4412,
> + PMU_EXYNOS5250,
> +};
> +
> +struct exynos_pmu_data {
> + enum exynos_pmu_id pmu_id;
> + struct regmap *pmu_regmap;
Again, since this uses the PMU node directly and doesn't seem to access
any registers shared with other drivers (or at least not at the time
most the functions from this file are called) I don't think there is any
reason why not to use of_iomap() and access the registers directly.
What about adding more fields to this struct? For example .pmu_config,
.pmu_config_extra (for model-specific settings, like for Exynos4412, see
my comment below) and (*pmu_init)? Of course the regmap (or base
address) would have to be moved outside, as it isn't const data.
Then for each PMU variant there would be a static const struct will all
those fields already filled in and entries in OF match table would
already point to appropriate structure.
> +};
> +
> +struct exynos_pmu_data *pmu_data;
> static const struct exynos_pmu_conf *exynos_pmu_config;
> -static struct regmap *pmu_regmap;
> +typedef void (*exynos_pmu_init_t)(void);
>
> static const struct exynos_pmu_conf exynos4210_pmu_config[] = {
> /* { .reg = address, .val = { AFTR, LPA, SLEEP } */
> @@ -348,28 +362,31 @@ static void exynos5_init_pmu(void)
> * Enable both SC_FEEDBACK and SC_COUNTER
> */
> for (i = 0 ; i < ARRAY_SIZE(exynos5_list_both_cnt_feed) ; i++) {
> - regmap_read(pmu_regmap, exynos5_list_both_cnt_feed[i], &tmp);
> + regmap_read(pmu_data->pmu_regmap,
> + exynos5_list_both_cnt_feed[i], &tmp);
> tmp |= (EXYNOS5_USE_SC_FEEDBACK |
> EXYNOS5_USE_SC_COUNTER);
> - regmap_write(pmu_regmap, exynos5_list_both_cnt_feed[i], tmp);
> + regmap_write(pmu_data->pmu_regmap,
> + exynos5_list_both_cnt_feed[i], tmp);
> }
>
> /*
> * SKIP_DEACTIVATE_ACEACP_IN_PWDN_BITFIELD Enable
> */
> - regmap_read(pmu_regmap, EXYNOS5_ARM_COMMON_OPTION, &tmp);
> + regmap_read(pmu_data->pmu_regmap, EXYNOS5_ARM_COMMON_OPTION, &tmp);
> tmp |= EXYNOS5_SKIP_DEACTIVATE_ACEACP_IN_PWDN;
> - regmap_write(pmu_regmap, EXYNOS5_ARM_COMMON_OPTION, tmp);
> + regmap_write(pmu_data->pmu_regmap, EXYNOS5_ARM_COMMON_OPTION, tmp);
>
> /*
> * Disable WFI/WFE on XXX_OPTION
> */
> for (i = 0 ; i < ARRAY_SIZE(exynos5_list_diable_wfi_wfe) ; i++) {
> - tmp = regmap_read(pmu_regmap, exynos5_list_diable_wfi_wfe[i],
> - &tmp);
> + tmp = regmap_read(pmu_data->pmu_regmap,
> + exynos5_list_diable_wfi_wfe[i], &tmp);
> tmp &= ~(EXYNOS5_OPTION_USE_STANDBYWFE |
> EXYNOS5_OPTION_USE_STANDBYWFI);
> - regmap_write(pmu_regmap, exynos5_list_diable_wfi_wfe[i], tmp);
> + regmap_write(pmu_data->pmu_regmap,
> + exynos5_list_diable_wfi_wfe[i], tmp);
> }
> }
>
> @@ -377,52 +394,129 @@ void exynos_sys_powerdown_conf(enum sys_powerdown mode)
> {
> unsigned int i;
>
> - if (soc_is_exynos5250())
> + if (pmu_data->pmu_id == PMU_EXYNOS5250)
> exynos5_init_pmu();
Next field could be added to exynos_pmu_data struct, called
(*powerdown_conf) and then the code above replaced with:
if (pmu_data->powerdown_conf)
pmu_data->powerdown_conf(mode);
>
> for (i = 0; (exynos_pmu_config[i].offset != PMU_TABLE_END) ; i++)
> - regmap_write(pmu_regmap, exynos_pmu_config[i].offset,
> + regmap_write(pmu_data->pmu_regmap, exynos_pmu_config[i].offset,
> exynos_pmu_config[i].val[mode]);
>
> - if (soc_is_exynos4412()) {
> + if (pmu_data->pmu_id == PMU_EXYNOS4412) {
> for (i = 0; exynos4412_pmu_config[i].offset != PMU_TABLE_END; i++)
> - regmap_write(pmu_regmap, exynos4412_pmu_config[i].offset,
> + regmap_write(pmu_data->pmu_regmap,
> + exynos4412_pmu_config[i].offset,
> exynos4412_pmu_config[i].val[mode]);
> }
As I mentioned above, the pmu_data struct could have a field called
pmu_config_extra that would be non-NULL for Exynos4412 and then the code
above could be replaced with:
if (pmu_data->pmu_config_extra) {
// iterate over pmu_config_extra
// ...
}
This way you could completely remove the pmu_id field and checks for
particular SoCs from the code.
> }
>
> -static int __init exynos_pmu_init(void)
> +static void exynos4210_pmu_init(void)
> {
> - unsigned int value;
> -
> exynos_pmu_config = exynos4210_pmu_config;
> - pmu_regmap = get_exynos_pmuregmap();
> -
> - if (soc_is_exynos4210()) {
> - exynos_pmu_config = exynos4210_pmu_config;
> - pr_info("EXYNOS4210 PMU Initialize\n");
> - } else if (soc_is_exynos4212() || soc_is_exynos4412()) {
> - exynos_pmu_config = exynos4x12_pmu_config;
> - pr_info("EXYNOS4x12 PMU Initialize\n");
> - } else if (soc_is_exynos5250()) {
> - /*
> - * When SYS_WDTRESET is set, watchdog timer reset request
> - * is ignored by power management unit.
> - */
> - regmap_read(pmu_regmap, EXYNOS5_AUTO_WDTRESET_DISABLE, &value);
> - value &= ~EXYNOS5_SYS_WDTRESET;
> - regmap_write(pmu_regmap, EXYNOS5_AUTO_WDTRESET_DISABLE, value);
> -
> - regmap_read(pmu_regmap, EXYNOS5_MASK_WDTRESET_REQUEST, &value);
> - value &= ~EXYNOS5_SYS_WDTRESET;
> - regmap_write(pmu_regmap, EXYNOS5_MASK_WDTRESET_REQUEST, value);
> -
> - exynos_pmu_config = exynos5250_pmu_config;
> - pr_info("EXYNOS5250 PMU Initialize\n");
> - } else {
> - pr_info("EXYNOS: PMU not supported\n");
> + pmu_data->pmu_id = PMU_EXYNOS4210;
> +
> + pr_info("EXYNOS4210 PMU Initialize\n");
> +}
> +
> +static void exynos4x12_pmu_init(void)
> +{
> + exynos_pmu_config = exynos4x12_pmu_config;
> + pmu_data->pmu_id = PMU_EXYNOS4X12;
> +
> + pr_info("EXYNOS4x12 PMU Initialize\n");
> +}
> +
> +static void exynos4412_pmu_init(void)
> +{
> + exynos_pmu_config = exynos4x12_pmu_config;
> + pmu_data->pmu_id = PMU_EXYNOS4412;
> +
> + pr_info("EXYNOS4412 PMU Initialize\n");
> +}
> +
> +static void exynos5250_pmu_init(void)
> +{
> + unsigned int tmp;
> +
> + /*
> + * When SYS_WDTRESET is set, watchdog timer reset request
> + * is ignored by power management unit.
> + */
> + regmap_read(pmu_data->pmu_regmap, EXYNOS5_AUTO_WDTRESET_DISABLE, &tmp);
> + tmp &= ~EXYNOS5_SYS_WDTRESET;
> + regmap_write(pmu_data->pmu_regmap, EXYNOS5_AUTO_WDTRESET_DISABLE, tmp);
> +
> + regmap_read(pmu_data->pmu_regmap, EXYNOS5_MASK_WDTRESET_REQUEST, &tmp);
> + tmp &= ~EXYNOS5_SYS_WDTRESET;
> + regmap_write(pmu_data->pmu_regmap, EXYNOS5_MASK_WDTRESET_REQUEST, tmp);
> +
> + exynos_pmu_config = exynos5250_pmu_config;
> + pmu_data->pmu_id = PMU_EXYNOS5250;
> +
> + pr_info("EXYNOS5250 PMU Initialize\n");
> +}
This part is very nice. Finally there is no huge if with else if clause
for every SoC. Thanks.
> +
> +/*
> + * PMU platform driver and devicetree bindings.
> + */
> +static struct of_device_id exynos_pmu_of_device_ids[] = {
> + {
> + .compatible = "samsung,exynos4210-pmu",
> + .data = (void *)exynos4210_pmu_init
> + },
> + {
> + .compatible = "samsung,exynos4212-pmu",
> + .data = (void *)exynos4x12_pmu_init
> + },
> + {
> + .compatible = "samsung,exynos4412-pmu",
> + .data = (void *)exynos4412_pmu_init
> + },
> + {
> + .compatible = "samsung,exynos5250-pmu",
> + .data = (void *)exynos5250_pmu_init
> + },
> + {},
> +};
> +
> +static int exynos_pmu_probe(struct device_node *np)
> +{
> + const struct of_device_id *match;
> + exynos_pmu_init_t exynos_pmu_init;
> +
> + pmu_data = kzalloc(sizeof(struct exynos_pmu_data), GFP_KERNEL);
> + if (!pmu_data) {
> + pr_err("exynos_pmu driver probe failed\n");
> + return -ENOMEM;
> }
>
> + match = of_match_node(exynos_pmu_of_device_ids, np);
> + if (!match) {
> + pr_err("fail to get matching of_match struct\n");
> + return -EINVAL;
> + }
> +
> + exynos_pmu_init = match->data;
> +
> + pmu_data->pmu_regmap = syscon_early_regmap_lookup_by_phandle(np,
> + "samsung,syscon-phandle");
> + if (IS_ERR(pmu_data->pmu_regmap)) {
> + pr_err("failed to find exynos_pmu_regmap\n");
> + return PTR_ERR(pmu_data->pmu_regmap);
> + }
> +
> + exynos_pmu_init();
> +
> return 0;
> +};
> +
> +static int __init exynos_pmu_of_init(void)
> +{
> + int ret = 0;
> + struct device_node *np;
> +
> + for_each_matching_node_and_match(np, exynos_pmu_of_device_ids, NULL)
> + ret = exynos_pmu_probe(np);
I wouldn't expect more than one PMU in the system, so probably a simple
for_each_matching_node_and_match(np, exynos_pmu_of_device_ids...
return exynos_pmu_probe(np);
return -ENODEV;
would be enough here.
Best regards,
Tomasz
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists