On Wed, 2024-11-20 at 12:49 +0100, AngeloGioacchino Del Regno wrote:
> External email : Please do not click links or open attachments until
> you have verified the sender or the content.
>
>
> Il 20/11/24 07:36, Friday Yang ha scritto:
> > In order to avoid handling glitch signal when MTCMOS on/off, SMI
> > need
> > clamp and reset operation. Parse power reset settings for LARBs
> > which
> > need to reset. Register genpd callback for SMI LARBs and apply
> > reset
> > operations in the callback.
> >
> > Signed-off-by: Friday Yang <friday.yang@xxxxxxxxxxxx>
> > ---
> > drivers/memory/mtk-smi.c | 175
> > ++++++++++++++++++++++++++++++++++++++-
> > 1 file changed, 171 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/memory/mtk-smi.c b/drivers/memory/mtk-smi.c
> > index 2bc034dff691..c7119f655350 100644
> > --- a/drivers/memory/mtk-smi.c
> > +++ b/drivers/memory/mtk-smi.c
> > @@ -10,15 +10,21 @@
> > #include <linux/err.h>
> > #include <linux/io.h>
> > #include <linux/iopoll.h>
> > +#include <linux/mfd/syscon.h>
> > #include <linux/module.h>
> > #include <linux/of.h>
> > #include <linux/of_platform.h>
> > #include <linux/platform_device.h>
> > +#include <linux/pm_domain.h>
> > #include <linux/pm_runtime.h>
> > +#include <linux/regmap.h>
> > +#include <linux/reset.h>
> > +#include <linux/reset-controller.h>
> > #include <linux/soc/mediatek/mtk_sip_svc.h>
> > #include <soc/mediatek/smi.h>
> > #include <dt-bindings/memory/mt2701-larb-port.h>
> > #include <dt-bindings/memory/mtk-memory-port.h>
> > +#include <dt-bindings/reset/mt8188-resets.h>
> >
> > /* SMI COMMON */
> > #define SMI_L1LEN 0x100
> > @@ -36,6 +42,13 @@
> > #define SMI_DCM 0x300
> > #define SMI_DUMMY 0x444
> >
> > +#define SMI_COMMON_CLAMP_EN 0x3c0
> > +#define SMI_COMMON_CLAMP_EN_SET 0x3c4
> > +#define SMI_COMMON_CLAMP_EN_CLR 0x3c8
> > +#define SMI_COMMON_CLAMP_MASK(inport) BIT(inport)
> > +
> > +#define SMI_SUB_COMM_INPORT_NR (8)
> > +
> > /* SMI LARB */
> > #define SMI_LARB_SLP_CON 0xc
> > #define SLP_PROT_EN BIT(0)
> > @@ -134,6 +147,7 @@ struct mtk_smi_larb_gen {
> > unsigned int larb_direct_to_common_mask;
> > unsigned int flags_general;
> > const
> > u8 (*ostd)[SMI_LARB_PORT_NR_MAX];
> > + const u8 *clamp_port;
> > };
> >
> > struct mtk_smi {
> > @@ -150,6 +164,7 @@ struct mtk_smi {
> > };
> >
> > struct mtk_smi_larb { /* larb: local arbiter */
> > + struct device *dev;
> > struct mtk_smi smi;
> > void __iomem *base;
> > struct device *smi_common_dev; /* common or
> > sub-common dev */
> > @@ -157,6 +172,10 @@ struct mtk_smi_larb { /* larb: local arbiter
> > */
> > int larbid;
> > u32 *mmu;
> > unsigned char *bank;
> > + struct regmap *sub_comm_syscon;
> > + u8 sub_comm_inport;
> > + struct notifier_block nb;
> > + struct reset_control *rst_con;
> > };
> >
> > static int
> > @@ -377,6 +396,19 @@ static const u8
> > mtk_smi_larb_mt8195_ostd[][SMI_LARB_PORT_NR_MAX] = {
> > [28] = {0x1a, 0x0e, 0x0a, 0x0a, 0x0c, 0x0e, 0x10,},
> > };
> >
> > +static const u8 mtk_smi_larb_clamp_port_mt8188[] = {
>
> You can just set these to BIT(x) directly, like:
>
> [MT8188_SMI_RST_LARB10] = BIT(0),
>
> so that you can check if there's actually a supported/valid clamp
> port in function
> mtk_smi_larb_parse_clamp_info() - check below for comments in that
> function.
>
> Note that you are declaring SMI_SUB_COMM_INPORT_NR = 8 and this means
> that there
> are a maximum of 8 ports, with each port having its own BIT, starting
> from BIT 0
> and ending at BIT 7 (so [7:0]).
>
> This means that we have a maximum of 8 bits here, so even if you
> assign BIT(x) it
> all still fits in a u8! :-)
Thanks for your comments, I will modify the definition for
mtk_smi_larb_clamp_port_mt8188, and macro 'SMI_COMMON_CLAMP_MASK'
could be removed.
>
> > + [MT8188_SMI_RST_LARB10] = 1,
> > + [MT8188_SMI_RST_LARB11A] = 2,
> > + [MT8188_SMI_RST_LARB11C] = 3,
> > + [MT8188_SMI_RST_LARB12] = 0,
> > + [MT8188_SMI_RST_LARB11B] = 2,
> > + [MT8188_SMI_RST_LARB15] = 1,
> > + [MT8188_SMI_RST_LARB16B] = 2,
> > + [MT8188_SMI_RST_LARB17B] = 3,
> > + [MT8188_SMI_RST_LARB16A] = 2,
> > + [MT8188_SMI_RST_LARB17A] = 3,
> > +};
> > +
> > static const struct mtk_smi_larb_gen mtk_smi_larb_mt2701 = {
> > .port_in_larb = {
> > LARB0_PORT_OFFSET, LARB1_PORT_OFFSET,
> > @@ -423,6 +455,7 @@ static const struct mtk_smi_larb_gen
> > mtk_smi_larb_mt8188 = {
> > .flags_general = MTK_SMI_FLAG_THRT_UPDATE |
> > MTK_SMI_FLAG_SW_FLAG |
> > MTK_SMI_FLAG_SLEEP_CTL |
> > MTK_SMI_FLAG_CFG_PORT_SEC_CTL,
> > .ostd = mtk_smi_larb_mt8188_ostd,
> > + .clamp_port = mtk_smi_larb_clamp_port_mt8188,
> > };
> >
> > static const struct mtk_smi_larb_gen mtk_smi_larb_mt8192 = {
> > @@ -472,6 +505,60 @@ static void
> > mtk_smi_larb_sleep_ctrl_disable(struct mtk_smi_larb *larb)
> > writel_relaxed(0, larb->base + SMI_LARB_SLP_CON);
> > }
> >
> > +static int mtk_smi_larb_clamp_protect_enable(struct device *dev)
> > +{
> > + struct mtk_smi_larb *larb = dev_get_drvdata(dev);
> > + int ret;
> > +
> > + /* sub_comm_syscon could be NULL if larb directly linked to
> > SMI common */
> > + if (!larb->sub_comm_syscon)
> > + return -EINVAL;
> > +
> > + ret = regmap_write(larb->sub_comm_syscon,
> > SMI_COMMON_CLAMP_EN_SET,
> > + SMI_COMMON_CLAMP_MASK(larb-
> > >sub_comm_inport));
> > + if (ret)
> > + dev_err(dev, "Unable to enable clamp, inport %d, ret
> > %d\n",
> > + larb->sub_comm_inport, ret);
> > +
> > + return ret;
> > +}
> > +
> > +static int mtk_smi_larb_clamp_protect_disble(struct device *dev)
>
> typo: mtk_smi_larb_clamp_protect_disable
>
> > +{
> > + struct mtk_smi_larb *larb = dev_get_drvdata(dev);
> > + int ret;
> > +
> > + /* sub_comm_syscon could be NULL if larb directly linked to
> > SMI common */
> > + if (!larb->sub_comm_syscon)
> > + return -EINVAL;
> > +
> > + ret = regmap_write(larb->sub_comm_syscon,
> > SMI_COMMON_CLAMP_EN_CLR,
> > + SMI_COMMON_CLAMP_MASK(larb-
> > >sub_comm_inport));
> > + if (ret)
> > + dev_err(dev, "Unable to disable clamp, inport %d, ret
> > %d\n",
> > + larb->sub_comm_inport, ret);
> > +
> > + return ret;
> > +}
>
> ...but anyway, I would rather do it like that:
>
> static int mtk_smi_larb_clamp_protect_enable(struct device *dev, bool
> enable)
> {
> struct mtk_smi_larb *larb = dev_get_drvdata(dev);
> u32 reg;
> int ret;
>
> /* sub_comm_syscon could be NULL if larb directly linked to
> SMI common */
> if (!larb->sub_comm_syscon)
> return -EINVAL;
>
> reg = enable ? SMI_COMMON_CLAMP_EN_SET :
> SMI_COMMON_CLAMP_EN_CLR;
>
> ret = regmap_write(larb->sub_comm_syscon, reg,
> SMI_COMMON_CLAMP_MASK(larb-
> >sub_comm_inport));
> if (ret) {
> dev_err(dev, "Unable to %s clamp for input port %d:
> %d\n",
> enable ? "enable" : "disable",
> larb->sub_comm_inport, ret);
> return ret;
> }
>
> return 0;
> }
>
> ...and then call it like
> ret = mtk_smi_larb_clamp_protect_enable(dev, true);
>
> or
>
> ret = mtk_smi_larb_clamp_protect_enable(dev, false);
>
Thanks for your comments, I will merge these two functions
into one. This looks better.
> > +
> > +static int mtk_smi_genpd_callback(struct notifier_block *nb,
> > + unsigned long flags, void *data)
> > +{
> > + struct mtk_smi_larb *larb = container_of(nb, struct
> > mtk_smi_larb, nb);
> > + struct device *dev = larb->dev;
> > +
> > + if (flags == GENPD_NOTIFY_PRE_ON || flags ==
> > GENPD_NOTIFY_PRE_OFF) {
> > + /* disable related SMI sub-common port */
> > + mtk_smi_larb_clamp_protect_enable(dev);
> > + } else if (flags == GENPD_NOTIFY_ON) {
> > + /* enable related SMI sub-common port */
> > + reset_control_reset(larb->rst_con);
> > + mtk_smi_larb_clamp_protect_disble(dev);
> > + }
> > +
> > + return NOTIFY_OK;
> > +}
> > +
> > static int mtk_smi_device_link_common(struct device *dev, struct
> > device **com_dev)
> > {
> > struct platform_device *smi_com_pdev;
> > @@ -528,6 +615,66 @@ static int mtk_smi_dts_clk_init(struct device
> > *dev, struct mtk_smi *smi,
> > return ret;
> > }
> >
> > +static int mtk_smi_larb_parse_clamp_info(struct mtk_smi_larb
> > *larb)
> > +{
> > + struct device *dev = larb->dev;
> > + const struct mtk_smi_larb_gen *larb_gen = larb->larb_gen;
> > + struct device_node *smi_node;
> > + struct of_phandle_args args;
> > + int ret, index;
> > +
> > + /* Only SMI LARBs located in camera and image subsys need to
> > + * apply clamp and reset operation, others can be skipped.
> > + */
> > + ret = of_parse_phandle_with_fixed_args(dev->of_node,
> > + "resets", 1, 0,
> > &args);
> > + if (ret)
> > + return 0;
> > +
> > + smi_node = of_parse_phandle(dev->of_node, "mediatek,smi", 0);
> > + if (!smi_node)
> > + return -EINVAL;
> > +
> > + index = args.args[0];
> > + larb->sub_comm_inport = larb_gen->clamp_port[index];
> > + larb->sub_comm_syscon = device_node_to_regmap(smi_node);
> > + of_node_put(smi_node);
>
> If you declare BIT(x) as clamp_ports, or if your clamp_ports start
> from 1,
> anything that is not more than 0 is something that was not declared,
> and
> this means that you can then error check:
>
> if (!larb->sub_comm_inport)
> return dev_err_probe(dev, -EINVAL, "Unknown clamp port for
> larb %d\n", index);
>
Yes, you are right. I will fix it as this way.
> > +
> > + if (IS_ERR(larb->sub_comm_syscon) ||
> > + larb->sub_comm_inport >= SMI_SUB_COMM_INPORT_NR) {
> > + larb->sub_comm_syscon = NULL;
> > + return -EINVAL;
> > + }
> > +
> > + return 0;
> > +}
> > +
> > +static int mtk_smi_larb_parse_reset_info(struct mtk_smi_larb
> > *larb)
> > +{
> > + struct device *dev = larb->dev;
> > + int ret;
> > +
> > + /* Only SMI LARBs located in camera and image subsys need to
> > + * apply clamp and reset operation, others can be skipped.
> > + */
> > + if (!of_find_property(dev->of_node, "resets", NULL))
> > + return 0;
>
> You don't have to manually check whether 'resets' exists or not -
> that's already
> done (devm_)reset_control_get in a way, as that will return -ENOENT
> if there is
> no 'reset-names' property. Check the comments down there....
>
> > +
> > + larb->rst_con = devm_reset_control_get(dev, "larb");
> > + if (IS_ERR(larb->rst_con))
> > + return dev_err_probe(dev, PTR_ERR(larb->rst_con),
> > + "Can not get larb reset
> > controller\n");
> > +
> > + larb->nb.notifier_call = mtk_smi_genpd_callback;
> > + ret = dev_pm_genpd_add_notifier(dev, &larb->nb);
> > + if (ret) {
> > + dev_err(dev, "Failed to add genpd callback %d\n",
> > ret);
> > + return -EINVAL;
> > + }
> > +
> > + return 0;
> > +}
> > +
OK, I will just check the return value of devm_reset_control_get
instead of parsing 'resets' here. This looks better.
> > static int mtk_smi_larb_probe(struct platform_device *pdev)
> > {
> > struct mtk_smi_larb *larb;
> > @@ -538,6 +685,7 @@ static int mtk_smi_larb_probe(struct
> > platform_device *pdev)
> > if (!larb)
> > return -ENOMEM;
> >
> > + larb->dev = dev;
> > larb->larb_gen = of_device_get_match_data(dev);
> > larb->base = devm_platform_ioremap_resource(pdev, 0);
> > if (IS_ERR(larb->base))
> > @@ -554,15 +702,29 @@ static int mtk_smi_larb_probe(struct
> > platform_device *pdev)
> > if (ret < 0)
> > return ret;
> >
> > - pm_runtime_enable(dev);
>
> ret = mtk_smi_larb_parse_reset_optional(larb);
>
> Choose between:
>
> if (ret == 0) {
> ret = mtk_smi_larb_parse_smi_clamp(larb);
> if (ret) {
> dev_err_probe(dev, ret, "Failed to get SMI
> clamp\n");
> goto err_link_remove;
> }
> } else if (ret != -ENOENT) {
> dev_err_probe(dev, ret, "Cannot get larb resets\n");
> goto err_link_remove;
> } else {
> /*
> * Only SMI LARBs located in camera and image subsys
> need to apply
> * clamp and reset operation. For the others, resets
> are optional.
> */
> ret = 0;
> }
>
> and...
>
> if (ret && ret != -ENOENT) {
> dev_err_probe .....
> } else if (ret) {
> /* only smi larbs .... */
> ret = 0;
> } else {
> ret = mtk_smi_larb_parse_smi_clamp .....
> }
>
> whatever you like the most.
Thanks for your comments, I will check the return value as this way.
>
> > + /* find sub common to clamp larb for ISP software reset */
> > + ret = mtk_smi_larb_parse_clamp_info(larb);
> > + if (ret) {
> > + dev_err(dev, "Failed to get clamp setting for
> > larb\n");
> > + goto err_link_remove;
> > + }
> > +
> > + ret = mtk_smi_larb_parse_reset_info(larb);
> > + if (ret) {
> > + dev_err(dev, "Failed to get power setting for
> > larb\n");
> > + goto err_link_remove;
> > + }
> > +
> > platform_set_drvdata(pdev, larb);
> > ret = component_add(dev, &mtk_smi_larb_component_ops);
> > if (ret)
> > - goto err_pm_disable;
> > + goto err_link_remove;
> > +
> > + pm_runtime_enable(dev);
> > +
> > return 0;
> >
> > -err_pm_disable:
> > - pm_runtime_disable(dev);
> > +err_link_remove:
> > device_link_remove(dev, larb->smi_common_dev);
> > return ret;
> > }
> > @@ -686,6 +848,10 @@ static const struct mtk_smi_common_plat
> > mtk_smi_common_mt8188_vpp = {
> > .init = mtk_smi_common_mt8195_init,
> > };
> >
> > +static const struct mtk_smi_common_plat mtk_smi_sub_common_mt8188
> > = {
> > + .type = MTK_SMI_GEN2_SUB_COMM,
>
> ...no gals in MT8188?!
>
Yes, there is gals in MT8188 hardware design. And the 'gals' clock
may be the same as the 'apb' and 'smi' clock.
I will add this in the next version, thanks for your comments.
> Cheers,
> Angelo
>
> > +};
> > +
> > static const struct mtk_smi_common_plat mtk_smi_common_mt8192 = {
> > .type = MTK_SMI_GEN2,
> > .has_gals = true,
> > @@ -729,6 +895,7 @@ static const struct of_device_id
> > mtk_smi_common_of_ids[] = {
> > {.compatible = "mediatek,mt8186-smi-common", .data =
> > &mtk_smi_common_mt8186},
> > {.compatible = "mediatek,mt8188-smi-common-vdo", .data =
> > &mtk_smi_common_mt8188_vdo},
> > {.compatible = "mediatek,mt8188-smi-common-vpp", .data =
> > &mtk_smi_common_mt8188_vpp},
> > + {.compatible = "mediatek,mt8188-smi-sub-common", .data =
> > &mtk_smi_sub_common_mt8188},
> > {.compatible = "mediatek,mt8192-smi-common", .data =
> > &mtk_smi_common_mt8192},
> > {.compatible = "mediatek,mt8195-smi-common-vdo", .data =
> > &mtk_smi_common_mt8195_vdo},
> > {.compatible = "mediatek,mt8195-smi-common-vpp", .data =
> > &mtk_smi_common_mt8195_vpp},
>
>
