The eint num will obtain the operation address through pins. Change the traversal method of irq handle from traversing a set of registers to traversing one by one. Signed-off-by: Hao Chang <ot_chhao.chang@xxxxxxxxxxxx> Signed-off-by: Qingliang Li <qingliang.li@xxxxxxxxxxxx> --- drivers/pinctrl/mediatek/mtk-eint.c | 219 +++++++++++++++++----------- drivers/pinctrl/mediatek/mtk-eint.h | 9 +- 2 files changed, 140 insertions(+), 88 deletions(-) diff --git a/drivers/pinctrl/mediatek/mtk-eint.c b/drivers/pinctrl/mediatek/mtk-eint.c index 27f0a54e12bf..540245c3128d 100644 --- a/drivers/pinctrl/mediatek/mtk-eint.c +++ b/drivers/pinctrl/mediatek/mtk-eint.c @@ -1,11 +1,13 @@ // SPDX-License-Identifier: GPL-2.0 -// Copyright (c) 2014-2018 MediaTek Inc. +// Copyright (c) 2014-2025 MediaTek Inc. /* * Library for MediaTek External Interrupt Support * * Author: Maoguang Meng <maoguang.meng@xxxxxxxxxxxx> * Sean Wang <sean.wang@xxxxxxxxxxxx> + * Hao Chang <ot_chhao.chang@xxxxxxxxxxxx> + * Qingliang Li <qingliang.li@xxxxxxxxxxxx> * */ @@ -20,6 +22,7 @@ #include <linux/platform_device.h> #include "mtk-eint.h" +#include "pinctrl-mtk-common-v2.h" #define MTK_EINT_EDGE_SENSITIVE 0 #define MTK_EINT_LEVEL_SENSITIVE 1 @@ -68,13 +71,11 @@ static void __iomem *mtk_eint_get_offset(struct mtk_eint *eint, unsigned int eint_num, unsigned int offset) { - unsigned int eint_base = 0; + unsigned int idx = eint->pins[eint_num].index; + unsigned int inst = eint->pins[eint_num].instance; void __iomem *reg; - if (eint_num >= eint->hw->ap_num) - eint_base = eint->hw->ap_num; - - reg = eint->base + offset + ((eint_num - eint_base) / 32) * 4; + reg = eint->base[inst] + offset + (idx / 32 * 4); return reg; } @@ -83,7 +84,7 @@ static unsigned int mtk_eint_can_en_debounce(struct mtk_eint *eint, unsigned int eint_num) { unsigned int sens; - unsigned int bit = BIT(eint_num % 32); + unsigned int bit = BIT(eint->pins[eint_num].index % 32); void __iomem *reg = mtk_eint_get_offset(eint, eint_num, eint->regs->sens); @@ -92,7 +93,7 @@ static unsigned int mtk_eint_can_en_debounce(struct mtk_eint *eint, else sens = MTK_EINT_EDGE_SENSITIVE; - if (eint_num < eint->hw->db_cnt && sens != MTK_EINT_EDGE_SENSITIVE) + if (eint->pins[eint_num].debounce && sens != MTK_EINT_EDGE_SENSITIVE) return 1; else return 0; @@ -102,9 +103,9 @@ static int mtk_eint_flip_edge(struct mtk_eint *eint, int hwirq) { int start_level, curr_level; unsigned int reg_offset; - u32 mask = BIT(hwirq & 0x1f); - u32 port = (hwirq >> 5) & eint->hw->port_mask; - void __iomem *reg = eint->base + (port << 2); + unsigned int mask = BIT(eint->pins[hwirq].index & 0x1f); + unsigned int port = (eint->pins[hwirq].index >> 5) & eint->hw->port_mask; + void __iomem *reg = eint->base[eint->pins[hwirq].instance] + (port << 2); curr_level = eint->gpio_xlate->get_gpio_state(eint->pctl, hwirq); @@ -126,11 +127,13 @@ static int mtk_eint_flip_edge(struct mtk_eint *eint, int hwirq) static void mtk_eint_mask(struct irq_data *d) { struct mtk_eint *eint = irq_data_get_irq_chip_data(d); - u32 mask = BIT(d->hwirq & 0x1f); + u32 idx = eint->pins[d->hwirq].index; + u32 inst = eint->pins[d->hwirq].instance; + u32 mask = BIT(idx & 0x1f); void __iomem *reg = mtk_eint_get_offset(eint, d->hwirq, eint->regs->mask_set); - eint->cur_mask[d->hwirq >> 5] &= ~mask; + eint->cur_mask[inst][idx >> 5] &= ~mask; writel(mask, reg); } @@ -138,32 +141,44 @@ static void mtk_eint_mask(struct irq_data *d) static void mtk_eint_unmask(struct irq_data *d) { struct mtk_eint *eint = irq_data_get_irq_chip_data(d); - u32 mask = BIT(d->hwirq & 0x1f); + u32 idx = eint->pins[d->hwirq].index; + u32 inst = eint->pins[d->hwirq].instance; + u32 mask = BIT(idx & 0x1f); void __iomem *reg = mtk_eint_get_offset(eint, d->hwirq, eint->regs->mask_clr); - eint->cur_mask[d->hwirq >> 5] |= mask; + eint->cur_mask[inst][idx >> 5] |= mask; writel(mask, reg); - if (eint->dual_edge[d->hwirq]) + if (eint->pins[d->hwirq].dual_edge) mtk_eint_flip_edge(eint, d->hwirq); } static unsigned int mtk_eint_get_mask(struct mtk_eint *eint, unsigned int eint_num) { - unsigned int bit = BIT(eint_num % 32); + unsigned int bit = BIT(eint->pins[eint_num].index % 32); void __iomem *reg = mtk_eint_get_offset(eint, eint_num, eint->regs->mask); return !!(readl(reg) & bit); } +static unsigned int mtk_eint_get_status(struct mtk_eint *eint, + unsigned int eint_num) +{ + unsigned int bit = BIT(eint->pins[eint_num].index % 32); + void __iomem *reg = mtk_eint_get_offset(eint, eint_num, + eint->regs->stat); + + return !!(readl(reg) & bit); +} + static void mtk_eint_ack(struct irq_data *d) { struct mtk_eint *eint = irq_data_get_irq_chip_data(d); - u32 mask = BIT(d->hwirq & 0x1f); + u32 mask = BIT(eint->pins[d->hwirq].index & 0x1f); void __iomem *reg = mtk_eint_get_offset(eint, d->hwirq, eint->regs->ack); @@ -174,7 +189,7 @@ static int mtk_eint_set_type(struct irq_data *d, unsigned int type) { struct mtk_eint *eint = irq_data_get_irq_chip_data(d); bool masked; - u32 mask = BIT(d->hwirq & 0x1f); + unsigned int mask = BIT(eint->pins[d->hwirq].index & 0x1f); void __iomem *reg; if (((type & IRQ_TYPE_EDGE_BOTH) && (type & IRQ_TYPE_LEVEL_MASK)) || @@ -186,9 +201,9 @@ static int mtk_eint_set_type(struct irq_data *d, unsigned int type) } if ((type & IRQ_TYPE_EDGE_BOTH) == IRQ_TYPE_EDGE_BOTH) - eint->dual_edge[d->hwirq] = 1; + eint->pins[d->hwirq].dual_edge = 1; else - eint->dual_edge[d->hwirq] = 0; + eint->pins[d->hwirq].dual_edge = 0; if (!mtk_eint_get_mask(eint, d->hwirq)) { mtk_eint_mask(d); @@ -223,27 +238,32 @@ static int mtk_eint_set_type(struct irq_data *d, unsigned int type) static int mtk_eint_irq_set_wake(struct irq_data *d, unsigned int on) { struct mtk_eint *eint = irq_data_get_irq_chip_data(d); - int shift = d->hwirq & 0x1f; - int reg = d->hwirq >> 5; + unsigned int idx = eint->pins[d->hwirq].index; + unsigned int inst = eint->pins[d->hwirq].instance; + unsigned int shift = idx & 0x1f; + unsigned int port = idx >> 5; if (on) - eint->wake_mask[reg] |= BIT(shift); + eint->wake_mask[inst][port] |= BIT(shift); else - eint->wake_mask[reg] &= ~BIT(shift); + eint->wake_mask[inst][port] &= ~BIT(shift); return 0; } static void mtk_eint_chip_write_mask(const struct mtk_eint *eint, - void __iomem *base, u32 *buf) + void __iomem *base, u32 **buf) { - int port; + int inst, port, port_num; void __iomem *reg; - for (port = 0; port < eint->hw->ports; port++) { - reg = base + (port << 2); - writel_relaxed(~buf[port], reg + eint->regs->mask_set); - writel_relaxed(buf[port], reg + eint->regs->mask_clr); + for (inst = 0; inst < eint->nbase; inst++) { + port_num = (eint->base_pin_num[inst] + 31) / 32; + for (port = 0; port < port_num; port++) { + reg = eint->base[inst] + (port << 2); + writel_relaxed(~buf[inst][port], reg + eint->regs->mask_set); + writel_relaxed(buf[inst][port], reg + eint->regs->mask_clr); + } } } @@ -303,15 +323,18 @@ static struct irq_chip mtk_eint_irq_chip = { static unsigned int mtk_eint_hw_init(struct mtk_eint *eint) { - void __iomem *dom_en = eint->base + eint->regs->dom_en; - void __iomem *mask_set = eint->base + eint->regs->mask_set; - unsigned int i; - - for (i = 0; i < eint->hw->ap_num; i += 32) { - writel(0xffffffff, dom_en); - writel(0xffffffff, mask_set); - dom_en += 4; - mask_set += 4; + void __iomem *dom_reg, *mask_reg; + unsigned int i, j; + + for (i = 0; i < eint->nbase; i++) { + dom_reg = eint->base[i] + eint->regs->dom_en; + mask_reg = eint->base[i] + eint->regs->mask_set; + for (j = 0; j < eint->base_pin_num[i]; j += 32) { + writel(0xffffffff, dom_reg); + writel(0xffffffff, mask_reg); + dom_reg += 4; + mask_reg += 4; + } } return 0; @@ -322,14 +345,16 @@ mtk_eint_debounce_process(struct mtk_eint *eint, int index) { unsigned int rst, ctrl_offset; unsigned int bit, dbnc; + unsigned int inst = eint->pins[index].instance; + unsigned int idx = eint->pins[index].index; - ctrl_offset = (index / 4) * 4 + eint->regs->dbnc_ctrl; - dbnc = readl(eint->base + ctrl_offset); - bit = MTK_EINT_DBNC_SET_EN << ((index % 4) * 8); + ctrl_offset = (idx / 4) * 4 + eint->regs->dbnc_ctrl; + dbnc = readl(eint->base[inst] + ctrl_offset); + bit = MTK_EINT_DBNC_SET_EN << ((idx % 4) * 8); if ((bit & dbnc) > 0) { - ctrl_offset = (index / 4) * 4 + eint->regs->dbnc_set; - rst = MTK_EINT_DBNC_RST_BIT << ((index % 4) * 8); - writel(rst, eint->base + ctrl_offset); + ctrl_offset = (idx / 4) * 4 + eint->regs->dbnc_set; + rst = MTK_EINT_DBNC_RST_BIT << ((idx % 4) * 8); + writel(rst, eint->base[inst] + ctrl_offset); } } @@ -337,20 +362,16 @@ static void mtk_eint_irq_handler(struct irq_desc *desc) { struct irq_chip *chip = irq_desc_get_chip(desc); struct mtk_eint *eint = irq_desc_get_handler_data(desc); - unsigned int status, eint_num; - int offset, mask_offset, index; - void __iomem *reg = mtk_eint_get_offset(eint, 0, eint->regs->stat); + unsigned int index, inst, idx, mask; + void __iomem *reg; int dual_edge, start_level, curr_level; chained_irq_enter(chip, desc); - for (eint_num = 0; eint_num < eint->hw->ap_num; eint_num += 32, - reg += 4) { - status = readl(reg); - while (status) { - offset = __ffs(status); - mask_offset = eint_num >> 5; - index = eint_num + offset; - status &= ~BIT(offset); + for (index = 0; index < eint->hw->ap_num; index++) { + if (mtk_eint_get_status(eint, index)) { + inst = eint->pins[index].instance; + idx = eint->pins[index].index; + mask = BIT(idx & 0x1f); /* * If we get an interrupt on pin that was only required @@ -358,21 +379,22 @@ static void mtk_eint_irq_handler(struct irq_desc *desc) * interrupt (as would mtk_eint_resume do anyway later * in the resume sequence). */ - if (eint->wake_mask[mask_offset] & BIT(offset) && - !(eint->cur_mask[mask_offset] & BIT(offset))) { - writel_relaxed(BIT(offset), reg - - eint->regs->stat + - eint->regs->mask_set); + if (eint->wake_mask[inst][idx >> 5] & mask && + !(eint->cur_mask[inst][idx >> 5] & mask)) { + reg = mtk_eint_get_offset(eint, index, + eint->regs->mask_set); + writel_relaxed(mask, reg); } - dual_edge = eint->dual_edge[index]; + dual_edge = eint->pins[index].dual_edge; if (dual_edge) { /* * Clear soft-irq in case we raised it last * time. */ - writel(BIT(offset), reg - eint->regs->stat + - eint->regs->soft_clr); + reg = mtk_eint_get_offset(eint, index, + eint->regs->soft_clr); + writel(mask, reg); start_level = eint->gpio_xlate->get_gpio_state(eint->pctl, @@ -388,13 +410,14 @@ static void mtk_eint_irq_handler(struct irq_desc *desc) * If level changed, we might lost one edge * interrupt, raised it through soft-irq. */ - if (start_level != curr_level) - writel(BIT(offset), reg - - eint->regs->stat + - eint->regs->soft_set); + if (start_level != curr_level) { + reg = mtk_eint_get_offset(eint, index, + eint->regs->soft_set); + writel(mask, reg); + } } - if (index < eint->hw->db_cnt) + if (eint->pins[index].debounce) mtk_eint_debounce_process(eint, index); } } @@ -423,6 +446,8 @@ int mtk_eint_set_debounce(struct mtk_eint *eint, unsigned long eint_num, int virq, eint_offset; unsigned int set_offset, bit, clr_bit, clr_offset, rst, i, unmask, dbnc; + unsigned int inst = eint->pins[eint_num].instance; + unsigned int idx = eint->pins[eint_num].index; struct irq_data *d; if (!eint->hw->db_time) @@ -432,8 +457,8 @@ int mtk_eint_set_debounce(struct mtk_eint *eint, unsigned long eint_num, eint_offset = (eint_num % 4) * 8; d = irq_get_irq_data(virq); - set_offset = (eint_num / 4) * 4 + eint->regs->dbnc_set; - clr_offset = (eint_num / 4) * 4 + eint->regs->dbnc_clr; + set_offset = (idx / 4) * 4 + eint->regs->dbnc_set; + clr_offset = (idx / 4) * 4 + eint->regs->dbnc_clr; if (!mtk_eint_can_en_debounce(eint, eint_num)) return -EINVAL; @@ -454,12 +479,12 @@ int mtk_eint_set_debounce(struct mtk_eint *eint, unsigned long eint_num, } clr_bit = 0xff << eint_offset; - writel(clr_bit, eint->base + clr_offset); + writel(clr_bit, eint->base[inst] + clr_offset); bit = ((dbnc << MTK_EINT_DBNC_SET_DBNC_BITS) | MTK_EINT_DBNC_SET_EN) << eint_offset; rst = MTK_EINT_DBNC_RST_BIT << eint_offset; - writel(rst | bit, eint->base + set_offset); + writel(rst | bit, eint->base[inst] + set_offset); /* * Delay a while (more than 2T) to wait for hw debounce counter reset @@ -487,27 +512,53 @@ EXPORT_SYMBOL_GPL(mtk_eint_find_irq); int mtk_eint_do_init(struct mtk_eint *eint) { - int i; + unsigned int size, i, port; /* If clients don't assign a specific regs, let's use generic one */ if (!eint->regs) eint->regs = &mtk_generic_eint_regs; - eint->wake_mask = devm_kcalloc(eint->dev, eint->hw->ports, - sizeof(*eint->wake_mask), GFP_KERNEL); - if (!eint->wake_mask) + eint->base_pin_num = devm_kmalloc_array(eint->dev, eint->nbase, sizeof(u16), + GFP_KERNEL); + if (!eint->base_pin_num) return -ENOMEM; - eint->cur_mask = devm_kcalloc(eint->dev, eint->hw->ports, - sizeof(*eint->cur_mask), GFP_KERNEL); - if (!eint->cur_mask) + if (!eint->pins) { + size = eint->hw->ap_num * sizeof(struct mtk_eint_pin); + eint->pins = devm_kmalloc(eint->dev, size, GFP_KERNEL); + if (!eint->pins) + return -ENOMEM; + + eint->base_pin_num[0] = eint->hw->ap_num; + for (i = 0; i < eint->hw->ap_num; i++) { + eint->pins[i].instance = 0; + eint->pins[i].index = i; + eint->pins[i].debounce = (i < eint->hw->db_cnt) ? 1 : 0; + } + } else { + for (i = 0; i < eint->hw->ap_num; i++) + eint->base_pin_num[eint->pins[i].instance]++; + } + + eint->wake_mask = devm_kmalloc(eint->dev, eint->nbase * sizeof(u32 *), GFP_KERNEL); + if (!eint->wake_mask) return -ENOMEM; - eint->dual_edge = devm_kcalloc(eint->dev, eint->hw->ap_num, - sizeof(int), GFP_KERNEL); - if (!eint->dual_edge) + eint->cur_mask = devm_kmalloc(eint->dev, eint->nbase * sizeof(u32 *), GFP_KERNEL); + if (!eint->wake_mask) return -ENOMEM; + for (i = 0; i < eint->nbase; i++) { + port = (eint->base_pin_num[i] + 31) / 32; + eint->wake_mask[i] = devm_kzalloc(eint->dev, port * sizeof(u32), GFP_KERNEL); + if (!eint->wake_mask[i]) + return -ENOMEM; + + eint->cur_mask[i] = devm_kzalloc(eint->dev, port * sizeof(u32), GFP_KERNEL); + if (!eint->cur_mask[i]) + return -ENOMEM; + } + eint->domain = irq_domain_add_linear(eint->dev->of_node, eint->hw->ap_num, &irq_domain_simple_ops, NULL); diff --git a/drivers/pinctrl/mediatek/mtk-eint.h b/drivers/pinctrl/mediatek/mtk-eint.h index c1efdf02fb49..959afc4b950b 100644 --- a/drivers/pinctrl/mediatek/mtk-eint.h +++ b/drivers/pinctrl/mediatek/mtk-eint.h @@ -1,10 +1,11 @@ /* SPDX-License-Identifier: GPL-2.0 */ /* - * Copyright (C) 2014-2018 MediaTek Inc. + * Copyright (C) 2014-2025 MediaTek Inc. * * Author: Maoguang Meng <maoguang.meng@xxxxxxxxxxxx> * Sean Wang <sean.wang@xxxxxxxxxxxx> - * + * Hao Chang <ot_chhao.chang@xxxxxxxxxxxx> + * Qingliang Li <qingliang.li@xxxxxxxxxxxx> */ #ifndef __MTK_EINT_H #define __MTK_EINT_H @@ -71,8 +72,8 @@ struct mtk_eint { int irq; int *dual_edge; - u32 *wake_mask; - u32 *cur_mask; + u32 **wake_mask; + u32 **cur_mask; /* Used to fit into various EINT device */ const struct mtk_eint_hw *hw; -- 2.34.1