--- /dev/null
+From 1ecb38eabd90efe93957d0a822a167560c39308a Mon Sep 17 00:00:00 2001
+From: Xiangsheng Hou <xiangsheng.hou@mediatek.com>
+Date: Wed, 20 Mar 2019 16:19:51 +0800
+Subject: [PATCH 6/6] spi: spi-mem: MediaTek: Add SPI NAND Flash interface
+ driver for MediaTek MT7622
+
+Change-Id: I3e78406bb9b46b0049d3988a5c71c7069e4f809c
+Signed-off-by: Xiangsheng Hou <xiangsheng.hou@mediatek.com>
+---
+ drivers/spi/Kconfig | 9 +
+ drivers/spi/Makefile | 1 +
+ drivers/spi/spi-mtk-snfi.c | 1183 ++++++++++++++++++++++++++++++++++++
+ 3 files changed, 1193 insertions(+)
+ create mode 100644 drivers/spi/spi-mtk-snfi.c
+
+--- a/drivers/spi/Makefile 2020-03-02 15:33:05.223669793 +0800
++++ b/drivers/spi/Makefile 2020-03-02 15:33:44.146364315 +0800
+@@ -60,6 +60,7 @@
+ obj-$(CONFIG_SPI_MPC52xx_PSC) += spi-mpc52xx-psc.o
+ obj-$(CONFIG_SPI_MPC52xx) += spi-mpc52xx.o
+ obj-$(CONFIG_SPI_MT65XX) += spi-mt65xx.o
++obj-$(CONFIG_SPI_MTK_SNFI) += spi-mtk-snfi.o
+ obj-$(CONFIG_SPI_MT7621) += spi-mt7621.o
+ obj-$(CONFIG_SPI_MXIC) += spi-mxic.o
+ obj-$(CONFIG_SPI_MXS) += spi-mxs.o
+--- a/drivers/spi/Kconfig 2020-03-02 15:33:11.183468935 +0800
++++ b/drivers/spi/Kconfig 2020-03-02 15:34:53.304079101 +0800
+@@ -427,6 +427,15 @@
+ say Y or M here.If you are not sure, say N.
+ SPI drivers for Mediatek MT65XX and MT81XX series ARM SoCs.
+
++config SPI_MTK_SNFI
++ tristate "MediaTek SPI NAND interface"
++ select MTD_SPI_NAND
++ help
++ This selects the SPI NAND FLASH interface(SNFI),
++ which could be found on MediaTek Soc.
++ Say Y or M here.If you are not sure, say N.
++ Note Parallel Nand and SPI NAND is alternative on MediaTek SoCs.
++
+ config SPI_MT7621
+ tristate "MediaTek MT7621 SPI Controller"
+ depends on RALINK || COMPILE_TEST
+--- /dev/null
++++ b/drivers/spi/spi-mtk-snfi.c
+@@ -0,0 +1,1200 @@
++// SPDX-License-Identifier: GPL-2.0
++/*
++ * Driver for MediaTek SPI Nand interface
++ *
++ * Copyright (C) 2018 MediaTek Inc.
++ * Authors: Xiangsheng Hou <xiangsheng.hou@mediatek.com>
++ *
++ */
++
++#include <linux/clk.h>
++#include <linux/delay.h>
++#include <linux/dma-mapping.h>
++#include <linux/interrupt.h>
++#include <linux/iopoll.h>
++#include <linux/mtd/mtd.h>
++#include <linux/mtd/mtk_ecc.h>
++#include <linux/mtd/spinand.h>
++#include <linux/module.h>
++#include <linux/of.h>
++#include <linux/of_device.h>
++#include <linux/platform_device.h>
++#include <linux/spi/spi.h>
++#include <linux/spi/spi-mem.h>
++
++/* NAND controller register definition */
++/* NFI control */
++#define NFI_CNFG 0x00
++#define CNFG_DMA BIT(0)
++#define CNFG_READ_EN BIT(1)
++#define CNFG_DMA_BURST_EN BIT(2)
++#define CNFG_BYTE_RW BIT(6)
++#define CNFG_HW_ECC_EN BIT(8)
++#define CNFG_AUTO_FMT_EN BIT(9)
++#define CNFG_OP_PROGRAM (3UL << 12)
++#define CNFG_OP_CUST (6UL << 12)
++#define NFI_PAGEFMT 0x04
++#define PAGEFMT_512 0
++#define PAGEFMT_2K 1
++#define PAGEFMT_4K 2
++#define PAGEFMT_FDM_SHIFT 8
++#define PAGEFMT_FDM_ECC_SHIFT 12
++#define NFI_CON 0x08
++#define CON_FIFO_FLUSH BIT(0)
++#define CON_NFI_RST BIT(1)
++#define CON_BRD BIT(8)
++#define CON_BWR BIT(9)
++#define CON_SEC_SHIFT 12
++#define NFI_INTR_EN 0x10
++#define INTR_AHB_DONE_EN BIT(6)
++#define NFI_INTR_STA 0x14
++#define NFI_CMD 0x20
++#define NFI_STA 0x60
++#define STA_EMP_PAGE BIT(12)
++#define NAND_FSM_MASK (0x1f << 24)
++#define NFI_FSM_MASK (0xf << 16)
++#define NFI_ADDRCNTR 0x70
++#define CNTR_MASK GENMASK(16, 12)
++#define ADDRCNTR_SEC_SHIFT 12
++#define ADDRCNTR_SEC(val) \
++ (((val) & CNTR_MASK) >> ADDRCNTR_SEC_SHIFT)
++#define NFI_STRADDR 0x80
++#define NFI_BYTELEN 0x84
++#define NFI_CSEL 0x90
++#define NFI_FDML(x) (0xa0 + (x) * sizeof(u32) * 2)
++#define NFI_FDMM(x) (0xa4 + (x) * sizeof(u32) * 2)
++#define NFI_MASTER_STA 0x224
++#define MASTER_STA_MASK 0x0fff
++/* NFI_SPI control */
++#define SNFI_MAC_OUTL 0x504
++#define SNFI_MAC_INL 0x508
++#define SNFI_RD_CTL2 0x510
++#define RD_CMD_MASK 0x00ff
++#define RD_DUMMY_SHIFT 8
++#define SNFI_RD_CTL3 0x514
++#define RD_ADDR_MASK 0xffff
++#define SNFI_MISC_CTL 0x538
++#define RD_MODE_X2 BIT(16)
++#define RD_MODE_X4 (2UL << 16)
++#define RD_QDUAL_IO (4UL << 16)
++#define RD_MODE_MASK (7UL << 16)
++#define RD_CUSTOM_EN BIT(6)
++#define WR_CUSTOM_EN BIT(7)
++#define WR_X4_EN BIT(20)
++#define SW_RST BIT(28)
++#define SNFI_MISC_CTL2 0x53c
++#define WR_LEN_SHIFT 16
++#define SNFI_PG_CTL1 0x524
++#define WR_LOAD_CMD_SHIFT 8
++#define SNFI_PG_CTL2 0x528
++#define WR_LOAD_ADDR_MASK 0xffff
++#define SNFI_MAC_CTL 0x500
++#define MAC_WIP BIT(0)
++#define MAC_WIP_READY BIT(1)
++#define MAC_TRIG BIT(2)
++#define MAC_EN BIT(3)
++#define MAC_SIO_SEL BIT(4)
++#define SNFI_STA_CTL1 0x550
++#define SPI_STATE_IDLE 0xf
++#define SNFI_CNFG 0x55c
++#define SNFI_MODE_EN BIT(0)
++#define SNFI_GPRAM_DATA 0x800
++#define SNFI_GPRAM_MAX_LEN 16
++
++/* Dummy command trigger NFI to spi mode */
++#define NAND_CMD_DUMMYREAD 0x00
++#define NAND_CMD_DUMMYPROG 0x80
++
++#define MTK_TIMEOUT 500000
++#define MTK_RESET_TIMEOUT 1000000
++#define MTK_SNFC_MIN_SPARE 16
++#define KB(x) ((x) * 1024UL)
++
++/*
++ * supported spare size of each IP.
++ * order should be the same with the spare size bitfiled defination of
++ * register NFI_PAGEFMT.
++ */
++static const u8 spare_size_mt7622[] = {
++ 16, 26, 27, 28
++};
++
++struct mtk_snfi_caps {
++ const u8 *spare_size;
++ u8 num_spare_size;
++ u32 nand_sec_size;
++ u8 nand_fdm_size;
++ u8 nand_fdm_ecc_size;
++ u8 ecc_parity_bits;
++ u8 pageformat_spare_shift;
++ u8 bad_mark_swap;
++};
++
++struct mtk_snfi_bad_mark_ctl {
++ void (*bm_swap)(struct spi_mem *mem, u8 *buf, int raw);
++ u32 sec;
++ u32 pos;
++};
++
++struct mtk_snfi_nand_chip {
++ struct mtk_snfi_bad_mark_ctl bad_mark;
++ u32 spare_per_sector;
++};
++
++struct mtk_snfi_clk {
++ struct clk *nfi_clk;
++ struct clk *spi_clk;
++};
++
++struct mtk_snfi {
++ const struct mtk_snfi_caps *caps;
++ struct mtk_snfi_nand_chip snfi_nand;
++ struct mtk_snfi_clk clk;
++ struct mtk_ecc_config ecc_cfg;
++ struct mtk_ecc *ecc;
++ struct completion done;
++ struct device *dev;
++
++ void __iomem *regs;
++
++ u8 *buffer;
++};
++
++static inline u8 *oob_ptr(struct spi_mem *mem, int i)
++{
++ struct spinand_device *spinand = spi_mem_get_drvdata(mem);
++ struct mtk_snfi *snfi = spi_controller_get_devdata(mem->spi->master);
++ struct mtk_snfi_nand_chip *snfi_nand = &snfi->snfi_nand;
++ u8 *poi;
++
++ /* map the sector's FDM data to free oob:
++ * the beginning of the oob area stores the FDM data of bad mark
++ */
++
++ if (i < snfi_nand->bad_mark.sec)
++ poi = spinand->oobbuf + (i + 1) * snfi->caps->nand_fdm_size;
++ else if (i == snfi_nand->bad_mark.sec)
++ poi = spinand->oobbuf;
++ else
++ poi = spinand->oobbuf + i * snfi->caps->nand_fdm_size;
++
++ return poi;
++}
++
++static inline int mtk_data_len(struct spi_mem *mem)
++{
++ struct mtk_snfi *snfi = spi_controller_get_devdata(mem->spi->master);
++ struct mtk_snfi_nand_chip *snfi_nand = &snfi->snfi_nand;
++
++ return snfi->caps->nand_sec_size + snfi_nand->spare_per_sector;
++}
++
++static inline u8 *mtk_oob_ptr(struct spi_mem *mem,
++ const u8 *p, int i)
++{
++ struct mtk_snfi *snfi = spi_controller_get_devdata(mem->spi->master);
++
++ return (u8 *)p + i * mtk_data_len(mem) + snfi->caps->nand_sec_size;
++}
++
++static void mtk_snfi_bad_mark_swap(struct spi_mem *mem,
++ u8 *buf, int raw)
++{
++ struct spinand_device *spinand = spi_mem_get_drvdata(mem);
++ struct mtk_snfi *snfi = spi_controller_get_devdata(mem->spi->master);
++ struct mtk_snfi_nand_chip *snfi_nand = &snfi->snfi_nand;
++ u32 bad_pos = snfi_nand->bad_mark.pos;
++
++ if (raw)
++ bad_pos += snfi_nand->bad_mark.sec * mtk_data_len(mem);
++ else
++ bad_pos += snfi_nand->bad_mark.sec * snfi->caps->nand_sec_size;
++
++ swap(spinand->oobbuf[0], buf[bad_pos]);
++}
++
++static void mtk_snfi_set_bad_mark_ctl(struct mtk_snfi_bad_mark_ctl *bm_ctl,
++ struct spi_mem *mem)
++{
++ struct spinand_device *spinand = spi_mem_get_drvdata(mem);
++ struct mtd_info *mtd = spinand_to_mtd(spinand);
++
++ bm_ctl->bm_swap = mtk_snfi_bad_mark_swap;
++ bm_ctl->sec = mtd->writesize / mtk_data_len(mem);
++ bm_ctl->pos = mtd->writesize % mtk_data_len(mem);
++}
++
++static void mtk_snfi_mac_enable(struct mtk_snfi *snfi)
++{
++ u32 mac;
++
++ mac = readl(snfi->regs + SNFI_MAC_CTL);
++ mac &= ~MAC_SIO_SEL;
++ mac |= MAC_EN;
++
++ writel(mac, snfi->regs + SNFI_MAC_CTL);
++}
++
++static int mtk_snfi_mac_trigger(struct mtk_snfi *snfi)
++{
++ u32 mac, reg;
++ int ret = 0;
++
++ mac = readl(snfi->regs + SNFI_MAC_CTL);
++ mac |= MAC_TRIG;
++ writel(mac, snfi->regs + SNFI_MAC_CTL);
++
++ ret = readl_poll_timeout_atomic(snfi->regs + SNFI_MAC_CTL, reg,
++ reg & MAC_WIP_READY, 10,
++ MTK_TIMEOUT);
++ if (ret < 0) {
++ dev_err(snfi->dev, "polling wip ready for read timeout\n");
++ return -EIO;
++ }
++
++ ret = readl_poll_timeout_atomic(snfi->regs + SNFI_MAC_CTL, reg,
++ !(reg & MAC_WIP), 10,
++ MTK_TIMEOUT);
++ if (ret < 0) {
++ dev_err(snfi->dev, "polling flash update timeout\n");
++ return -EIO;
++ }
++
++ return ret;
++}
++
++static void mtk_snfi_mac_leave(struct mtk_snfi *snfi)
++{
++ u32 mac;
++
++ mac = readl(snfi->regs + SNFI_MAC_CTL);
++ mac &= ~(MAC_TRIG | MAC_EN | MAC_SIO_SEL);
++ writel(mac, snfi->regs + SNFI_MAC_CTL);
++}
++
++static int mtk_snfi_mac_op(struct mtk_snfi *snfi)
++{
++ int ret = 0;
++
++ mtk_snfi_mac_enable(snfi);
++
++ ret = mtk_snfi_mac_trigger(snfi);
++ if (ret)
++ return ret;
++
++ mtk_snfi_mac_leave(snfi);
++
++ return ret;
++}
++
++static irqreturn_t mtk_snfi_irq(int irq, void *id)
++{
++ struct mtk_snfi *snfi = id;
++ u16 sta, ien;
++
++ sta = readw(snfi->regs + NFI_INTR_STA);
++ ien = readw(snfi->regs + NFI_INTR_EN);
++
++ if (!(sta & ien))
++ return IRQ_NONE;
++
++ writew(~sta & ien, snfi->regs + NFI_INTR_EN);
++ complete(&snfi->done);
++
++ return IRQ_HANDLED;
++}
++
++static int mtk_snfi_enable_clk(struct device *dev, struct mtk_snfi_clk *clk)
++{
++ int ret;
++
++ ret = clk_prepare_enable(clk->nfi_clk);
++ if (ret) {
++ dev_err(dev, "failed to enable nfi clk\n");
++ return ret;
++ }
++
++ ret = clk_prepare_enable(clk->spi_clk);
++ if (ret) {
++ dev_err(dev, "failed to enable spi clk\n");
++ clk_disable_unprepare(clk->nfi_clk);
++ return ret;
++ }
++
++ return 0;
++}
++
++static void mtk_snfi_disable_clk(struct mtk_snfi_clk *clk)
++{
++ clk_disable_unprepare(clk->nfi_clk);
++ clk_disable_unprepare(clk->spi_clk);
++}
++
++static int mtk_snfi_reset(struct mtk_snfi *snfi)
++{
++ u32 val;
++ int ret;
++
++ /* SW reset controller */
++ val = readl(snfi->regs + SNFI_MISC_CTL) | SW_RST;
++ writel(val, snfi->regs + SNFI_MISC_CTL);
++
++ ret = readw_poll_timeout(snfi->regs + SNFI_STA_CTL1, val,
++ !(val & SPI_STATE_IDLE), 50,
++ MTK_RESET_TIMEOUT);
++ if (ret) {
++ dev_warn(snfi->dev, "spi state active in reset [0x%x] = 0x%x\n",
++ SNFI_STA_CTL1, val);
++ return ret;
++ }
++
++ val = readl(snfi->regs + SNFI_MISC_CTL);
++ val &= ~SW_RST;
++ writel(val, snfi->regs + SNFI_MISC_CTL);
++
++ /* reset all registers and force the NFI master to terminate */
++ writew(CON_FIFO_FLUSH | CON_NFI_RST, snfi->regs + NFI_CON);
++ ret = readw_poll_timeout(snfi->regs + NFI_STA, val,
++ !(val & (NFI_FSM_MASK | NAND_FSM_MASK)), 50,
++ MTK_RESET_TIMEOUT);
++ if (ret) {
++ dev_warn(snfi->dev, "nfi active in reset [0x%x] = 0x%x\n",
++ NFI_STA, val);
++ return ret;
++ }
++
++ return 0;
++}
++
++static int mtk_snfi_set_spare_per_sector(struct spinand_device *spinand,
++ const struct mtk_snfi_caps *caps,
++ u32 *sps)
++{
++ struct mtd_info *mtd = spinand_to_mtd(spinand);
++ const u8 *spare = caps->spare_size;
++ u32 sectors, i, closest_spare = 0;
++
++ sectors = mtd->writesize / caps->nand_sec_size;
++ *sps = mtd->oobsize / sectors;
++
++ if (*sps < MTK_SNFC_MIN_SPARE)
++ return -EINVAL;
++
++ for (i = 0; i < caps->num_spare_size; i++) {
++ if (*sps >= spare[i] && spare[i] >= spare[closest_spare]) {
++ closest_spare = i;
++ if (*sps == spare[i])
++ break;
++ }
++ }
++
++ *sps = spare[closest_spare];
++
++ return 0;
++}
++
++static void mtk_snfi_read_fdm_data(struct spi_mem *mem,
++ u32 sectors)
++{
++ struct mtk_snfi *snfi = spi_controller_get_devdata(mem->spi->master);
++ const struct mtk_snfi_caps *caps = snfi->caps;
++ u32 vall, valm;
++ int i, j;
++ u8 *oobptr;
++
++ for (i = 0; i < sectors; i++) {
++ oobptr = oob_ptr(mem, i);
++ vall = readl(snfi->regs + NFI_FDML(i));
++ valm = readl(snfi->regs + NFI_FDMM(i));
++
++ for (j = 0; j < caps->nand_fdm_size; j++)
++ oobptr[j] = (j >= 4 ? valm : vall) >> ((j % 4) * 8);
++ }
++}
++
++static void mtk_snfi_write_fdm_data(struct spi_mem *mem,
++ u32 sectors)
++{
++ struct mtk_snfi *snfi = spi_controller_get_devdata(mem->spi->master);
++ const struct mtk_snfi_caps *caps = snfi->caps;
++ u32 vall, valm;
++ int i, j;
++ u8 *oobptr;
++
++ for (i = 0; i < sectors; i++) {
++ oobptr = oob_ptr(mem, i);
++ vall = 0;
++ valm = 0;
++ for (j = 0; j < 8; j++) {
++ if (j < 4)
++ vall |= (j < caps->nand_fdm_size ? oobptr[j] :
++ 0xff) << (j * 8);
++ else
++ valm |= (j < caps->nand_fdm_size ? oobptr[j] :
++ 0xff) << ((j - 4) * 8);
++ }
++ writel(vall, snfi->regs + NFI_FDML(i));
++ writel(valm, snfi->regs + NFI_FDMM(i));
++ }
++}
++
++static int mtk_snfi_update_ecc_stats(struct spi_mem *mem,
++ u8 *buf, u32 sectors)
++{
++ struct spinand_device *spinand = spi_mem_get_drvdata(mem);
++ struct mtd_info *mtd = spinand_to_mtd(spinand);
++ struct mtk_snfi *snfi = spi_controller_get_devdata(mem->spi->master);
++ struct mtk_ecc_stats stats;
++ int rc, i;
++
++ rc = readl(snfi->regs + NFI_STA) & STA_EMP_PAGE;
++ if (rc) {
++ memset(buf, 0xff, sectors * snfi->caps->nand_sec_size);
++ for (i = 0; i < sectors; i++)
++ memset(spinand->oobbuf, 0xff,
++ snfi->caps->nand_fdm_size);
++ return 0;
++ }
++
++ mtk_ecc_get_stats(snfi->ecc, &stats, sectors);
++ mtd->ecc_stats.corrected += stats.corrected;
++ mtd->ecc_stats.failed += stats.failed;
++
++ return 0;
++}
++
++static int mtk_snfi_hw_runtime_config(struct spi_mem *mem)
++{
++ struct spinand_device *spinand = spi_mem_get_drvdata(mem);
++ struct mtd_info *mtd = spinand_to_mtd(spinand);
++ struct nand_device *nand = mtd_to_nanddev(mtd);
++ struct mtk_snfi *snfi = spi_controller_get_devdata(mem->spi->master);
++ const struct mtk_snfi_caps *caps = snfi->caps;
++ struct mtk_snfi_nand_chip *snfi_nand = &snfi->snfi_nand;
++ u32 fmt, spare, i = 0;
++ int ret;
++
++ ret = mtk_snfi_set_spare_per_sector(spinand, caps, &spare);
++ if (ret)
++ return ret;
++
++ /* calculate usable oob bytes for ecc parity data */
++ snfi_nand->spare_per_sector = spare;
++ spare -= caps->nand_fdm_size;
++
++ nand->memorg.oobsize = snfi_nand->spare_per_sector
++ * (mtd->writesize / caps->nand_sec_size);
++ mtd->oobsize = nanddev_per_page_oobsize(nand);
++
++ snfi->ecc_cfg.strength = (spare << 3) / caps->ecc_parity_bits;
++ mtk_ecc_adjust_strength(snfi->ecc, &snfi->ecc_cfg.strength);
++
++ switch (mtd->writesize) {
++ case 512:
++ fmt = PAGEFMT_512;
++ break;
++ case KB(2):
++ fmt = PAGEFMT_2K;
++ break;
++ case KB(4):
++ fmt = PAGEFMT_4K;
++ break;
++ default:
++ dev_err(snfi->dev, "invalid page len: %d\n", mtd->writesize);
++ return -EINVAL;
++ }
++
++ /* Setup PageFormat */
++ while (caps->spare_size[i] != snfi_nand->spare_per_sector) {
++ i++;
++ if (i == (caps->num_spare_size - 1)) {
++ dev_err(snfi->dev, "invalid spare size %d\n",
++ snfi_nand->spare_per_sector);
++ return -EINVAL;
++ }
++ }
++
++ fmt |= i << caps->pageformat_spare_shift;
++ fmt |= caps->nand_fdm_size << PAGEFMT_FDM_SHIFT;
++ fmt |= caps->nand_fdm_ecc_size << PAGEFMT_FDM_ECC_SHIFT;
++ writel(fmt, snfi->regs + NFI_PAGEFMT);
++
++ snfi->ecc_cfg.len = caps->nand_sec_size + caps->nand_fdm_ecc_size;
++
++ mtk_snfi_set_bad_mark_ctl(&snfi_nand->bad_mark, mem);
++
++ return 0;
++}
++
++static int mtk_snfi_read_from_cache(struct spi_mem *mem,
++ const struct spi_mem_op *op, int oob_on)
++{
++ struct mtk_snfi *snfi = spi_controller_get_devdata(mem->spi->master);
++ struct spinand_device *spinand = spi_mem_get_drvdata(mem);
++ struct mtd_info *mtd = spinand_to_mtd(spinand);
++ u32 sectors = mtd->writesize / snfi->caps->nand_sec_size;
++ struct mtk_snfi_nand_chip *snfi_nand = &snfi->snfi_nand;
++ u32 reg, len, col_addr = 0;
++ int dummy_cycle, ret;
++ dma_addr_t dma_addr;
++
++ len = sectors * (snfi->caps->nand_sec_size
++ + snfi_nand->spare_per_sector);
++
++ dma_addr = dma_map_single(snfi->dev, snfi->buffer,
++ len, DMA_FROM_DEVICE);
++ ret = dma_mapping_error(snfi->dev, dma_addr);
++ if (ret) {
++ dev_err(snfi->dev, "dma mapping error\n");
++ return -EINVAL;
++ }
++
++ /* set Read cache command and dummy cycle */
++ dummy_cycle = (op->dummy.nbytes << 3) >> (ffs(op->dummy.buswidth) - 1);
++ reg = ((op->cmd.opcode & RD_CMD_MASK) |
++ (dummy_cycle << RD_DUMMY_SHIFT));
++ writel(reg, snfi->regs + SNFI_RD_CTL2);
++
++ writel((col_addr & RD_ADDR_MASK), snfi->regs + SNFI_RD_CTL3);
++
++ reg = readl(snfi->regs + SNFI_MISC_CTL);
++ reg |= RD_CUSTOM_EN;
++ reg &= ~(RD_MODE_MASK | WR_X4_EN);
++
++ /* set data and addr buswidth */
++ if (op->data.buswidth == 4)
++ reg |= RD_MODE_X4;
++ else if (op->data.buswidth == 2)
++ reg |= RD_MODE_X2;
++
++ if (op->addr.buswidth == 4 || op->addr.buswidth == 2)
++ reg |= RD_QDUAL_IO;
++ writel(reg, snfi->regs + SNFI_MISC_CTL);
++
++ writel(len, snfi->regs + SNFI_MISC_CTL2);
++ writew(sectors << CON_SEC_SHIFT, snfi->regs + NFI_CON);
++ reg = readw(snfi->regs + NFI_CNFG);
++ reg |= CNFG_READ_EN | CNFG_DMA_BURST_EN | CNFG_DMA | CNFG_OP_CUST;
++
++ if (!oob_on) {
++ reg |= CNFG_AUTO_FMT_EN | CNFG_HW_ECC_EN;
++ writew(reg, snfi->regs + NFI_CNFG);
++
++ snfi->ecc_cfg.mode = ECC_NFI_MODE;
++ snfi->ecc_cfg.sectors = sectors;
++ snfi->ecc_cfg.op = ECC_DECODE;
++ ret = mtk_ecc_enable(snfi->ecc, &snfi->ecc_cfg);
++ if (ret) {
++ dev_err(snfi->dev, "ecc enable failed\n");
++ /* clear NFI_CNFG */
++ reg &= ~(CNFG_READ_EN | CNFG_DMA_BURST_EN | CNFG_DMA |
++ CNFG_AUTO_FMT_EN | CNFG_HW_ECC_EN);
++ writew(reg, snfi->regs + NFI_CNFG);
++ goto out;
++ }
++ } else {
++ writew(reg, snfi->regs + NFI_CNFG);
++ }
++
++ writel(lower_32_bits(dma_addr), snfi->regs + NFI_STRADDR);
++ readw(snfi->regs + NFI_INTR_STA);
++ writew(INTR_AHB_DONE_EN, snfi->regs + NFI_INTR_EN);
++
++ init_completion(&snfi->done);
++
++ /* set dummy command to trigger NFI enter SPI mode */
++ writew(NAND_CMD_DUMMYREAD, snfi->regs + NFI_CMD);
++ reg = readl(snfi->regs + NFI_CON) | CON_BRD;
++ writew(reg, snfi->regs + NFI_CON);
++
++ ret = wait_for_completion_timeout(&snfi->done, msecs_to_jiffies(500));
++ if (!ret) {
++ dev_err(snfi->dev, "read ahb done timeout\n");
++ writew(0, snfi->regs + NFI_INTR_EN);
++ ret = -ETIMEDOUT;
++ goto out;
++ }
++
++ ret = readl_poll_timeout_atomic(snfi->regs + NFI_BYTELEN, reg,
++ ADDRCNTR_SEC(reg) >= sectors, 10,
++ MTK_TIMEOUT);
++ if (ret < 0) {
++ dev_err(snfi->dev, "polling read byte len timeout\n");
++ ret = -EIO;
++ } else {
++ if (!oob_on) {
++ ret = mtk_ecc_wait_done(snfi->ecc, ECC_DECODE);
++ if (ret) {
++ dev_warn(snfi->dev, "wait ecc done timeout\n");
++ } else {
++ mtk_snfi_update_ecc_stats(mem, snfi->buffer,
++ sectors);
++ mtk_snfi_read_fdm_data(mem, sectors);
++ }
++ }
++ }
++
++ if (oob_on)
++ goto out;
++
++ mtk_ecc_disable(snfi->ecc);
++out:
++ dma_unmap_single(snfi->dev, dma_addr, len, DMA_FROM_DEVICE);
++ writel(0, snfi->regs + NFI_CON);
++ writel(0, snfi->regs + NFI_CNFG);
++ reg = readl(snfi->regs + SNFI_MISC_CTL);
++ reg &= ~RD_CUSTOM_EN;
++ writel(reg, snfi->regs + SNFI_MISC_CTL);
++
++ return ret;
++}
++
++static int mtk_snfi_write_to_cache(struct spi_mem *mem,
++ const struct spi_mem_op *op,
++ int oob_on)
++{
++ struct mtk_snfi *snfi = spi_controller_get_devdata(mem->spi->master);
++ struct spinand_device *spinand = spi_mem_get_drvdata(mem);
++ struct mtd_info *mtd = spinand_to_mtd(spinand);
++ u32 sectors = mtd->writesize / snfi->caps->nand_sec_size;
++ struct mtk_snfi_nand_chip *snfi_nand = &snfi->snfi_nand;
++ u32 reg, len, col_addr = 0;
++ dma_addr_t dma_addr;
++ int ret;
++
++ len = sectors * (snfi->caps->nand_sec_size
++ + snfi_nand->spare_per_sector);
++
++ dma_addr = dma_map_single(snfi->dev, snfi->buffer, len,
++ DMA_TO_DEVICE);
++ ret = dma_mapping_error(snfi->dev, dma_addr);
++ if (ret) {
++ dev_err(snfi->dev, "dma mapping error\n");
++ return -EINVAL;
++ }
++
++ /* set program load cmd and address */
++ reg = (op->cmd.opcode << WR_LOAD_CMD_SHIFT);
++ writel(reg, snfi->regs + SNFI_PG_CTL1);
++ writel(col_addr & WR_LOAD_ADDR_MASK, snfi->regs + SNFI_PG_CTL2);
++
++ reg = readl(snfi->regs + SNFI_MISC_CTL);
++ reg |= WR_CUSTOM_EN;
++ reg &= ~(RD_MODE_MASK | WR_X4_EN);
++
++ if (op->data.buswidth == 4)
++ reg |= WR_X4_EN;
++ writel(reg, snfi->regs + SNFI_MISC_CTL);
++
++ writel(len << WR_LEN_SHIFT, snfi->regs + SNFI_MISC_CTL2);
++ writew(sectors << CON_SEC_SHIFT, snfi->regs + NFI_CON);
++
++ reg = readw(snfi->regs + NFI_CNFG);
++ reg &= ~(CNFG_READ_EN | CNFG_BYTE_RW);
++ reg |= CNFG_DMA | CNFG_DMA_BURST_EN | CNFG_OP_PROGRAM;
++
++ if (!oob_on) {
++ reg |= CNFG_AUTO_FMT_EN | CNFG_HW_ECC_EN;
++ writew(reg, snfi->regs + NFI_CNFG);
++
++ snfi->ecc_cfg.mode = ECC_NFI_MODE;
++ snfi->ecc_cfg.op = ECC_ENCODE;
++ ret = mtk_ecc_enable(snfi->ecc, &snfi->ecc_cfg);
++ if (ret) {
++ dev_err(snfi->dev, "ecc enable failed\n");
++ /* clear NFI_CNFG */
++ reg &= ~(CNFG_DMA_BURST_EN | CNFG_DMA |
++ CNFG_AUTO_FMT_EN | CNFG_HW_ECC_EN);
++ writew(reg, snfi->regs + NFI_CNFG);
++ dma_unmap_single(snfi->dev, dma_addr, len,
++ DMA_FROM_DEVICE);
++ goto out;
++ }
++ /* write OOB into the FDM registers (OOB area in MTK NAND) */
++ mtk_snfi_write_fdm_data(mem, sectors);
++ } else {
++ writew(reg, snfi->regs + NFI_CNFG);
++ }
++ writel(lower_32_bits(dma_addr), snfi->regs + NFI_STRADDR);
++ readw(snfi->regs + NFI_INTR_STA);
++ writew(INTR_AHB_DONE_EN, snfi->regs + NFI_INTR_EN);
++
++ init_completion(&snfi->done);
++
++ /* set dummy command to trigger NFI enter SPI mode */
++ writew(NAND_CMD_DUMMYPROG, snfi->regs + NFI_CMD);
++ reg = readl(snfi->regs + NFI_CON) | CON_BWR;
++ writew(reg, snfi->regs + NFI_CON);
++
++ ret = wait_for_completion_timeout(&snfi->done, msecs_to_jiffies(500));
++ if (!ret) {
++ dev_err(snfi->dev, "custom program done timeout\n");
++ writew(0, snfi->regs + NFI_INTR_EN);
++ ret = -ETIMEDOUT;
++ goto ecc_disable;
++ }
++
++ ret = readl_poll_timeout_atomic(snfi->regs + NFI_ADDRCNTR, reg,
++ ADDRCNTR_SEC(reg) >= sectors,
++ 10, MTK_TIMEOUT);
++ if (ret)
++ dev_err(snfi->dev, "hwecc write timeout\n");
++
++ecc_disable:
++ mtk_ecc_disable(snfi->ecc);
++
++out:
++ dma_unmap_single(snfi->dev, dma_addr, len, DMA_TO_DEVICE);
++ writel(0, snfi->regs + NFI_CON);
++ writel(0, snfi->regs + NFI_CNFG);
++ reg = readl(snfi->regs + SNFI_MISC_CTL);
++ reg &= ~WR_CUSTOM_EN;
++ writel(reg, snfi->regs + SNFI_MISC_CTL);
++
++ return ret;
++}
++
++static int mtk_snfi_read(struct spi_mem *mem,
++ const struct spi_mem_op *op)
++{
++ struct spinand_device *spinand = spi_mem_get_drvdata(mem);
++ struct mtk_snfi *snfi = spi_controller_get_devdata(mem->spi->master);
++ struct mtd_info *mtd = spinand_to_mtd(spinand);
++ struct mtk_snfi_nand_chip *snfi_nand = &snfi->snfi_nand;
++ u32 col_addr = op->addr.val;
++ int i, ret, sectors, oob_on = false;
++
++ if (col_addr == mtd->writesize)
++ oob_on = true;
++
++ ret = mtk_snfi_read_from_cache(mem, op, oob_on);
++ if (ret) {
++ dev_warn(snfi->dev, "read from cache fail\n");
++ return ret;
++ }
++
++ sectors = mtd->writesize / snfi->caps->nand_sec_size;
++ for (i = 0; i < sectors; i++) {
++ if (oob_on)
++ memcpy(oob_ptr(mem, i),
++ mtk_oob_ptr(mem, snfi->buffer, i),
++ snfi->caps->nand_fdm_size);
++
++ if (i == snfi_nand->bad_mark.sec && snfi->caps->bad_mark_swap)
++ snfi_nand->bad_mark.bm_swap(mem, snfi->buffer,
++ oob_on);
++ }
++
++ if (!oob_on)
++ memcpy(spinand->databuf, snfi->buffer, mtd->writesize);
++
++ return ret;
++}
++
++static int mtk_snfi_write(struct spi_mem *mem,
++ const struct spi_mem_op *op)
++{
++ struct spinand_device *spinand = spi_mem_get_drvdata(mem);
++ struct mtk_snfi *snfi = spi_controller_get_devdata(mem->spi->master);
++ struct mtd_info *mtd = spinand_to_mtd(spinand);
++ struct mtk_snfi_nand_chip *snfi_nand = &snfi->snfi_nand;
++ u32 ret, i, sectors, col_addr = op->addr.val;
++ int oob_on = false;
++
++ if (col_addr == mtd->writesize)
++ oob_on = true;
++
++ sectors = mtd->writesize / snfi->caps->nand_sec_size;
++ memset(snfi->buffer, 0xff, mtd->writesize + mtd->oobsize);
++
++ if (!oob_on)
++ memcpy(snfi->buffer, spinand->databuf, mtd->writesize);
++
++ for (i = 0; i < sectors; i++) {
++ if (i == snfi_nand->bad_mark.sec && snfi->caps->bad_mark_swap)
++ snfi_nand->bad_mark.bm_swap(mem, snfi->buffer, oob_on);
++
++ if (oob_on)
++ memcpy(mtk_oob_ptr(mem, snfi->buffer, i),
++ oob_ptr(mem, i),
++ snfi->caps->nand_fdm_size);
++ }
++
++ ret = mtk_snfi_write_to_cache(mem, op, oob_on);
++ if (ret)
++ dev_warn(snfi->dev, "write to cache fail\n");
++
++ return ret;
++}
++
++static int mtk_snfi_command_exec(struct mtk_snfi *snfi,
++ const u8 *txbuf, u8 *rxbuf,
++ const u32 txlen, const u32 rxlen)
++{
++ u32 tmp, i, j, reg, m;
++ u8 *p_tmp = (u8 *)(&tmp);
++ int ret = 0;
++
++ /* Moving tx data to NFI_SPI GPRAM */
++ for (i = 0, m = 0; i < txlen; ) {
++ for (j = 0, tmp = 0; i < txlen && j < 4; i++, j++)
++ p_tmp[j] = txbuf[i];
++
++ writel(tmp, snfi->regs + SNFI_GPRAM_DATA + m);
++ m += 4;
++ }
++
++ writel(txlen, snfi->regs + SNFI_MAC_OUTL);
++ writel(rxlen, snfi->regs + SNFI_MAC_INL);
++ ret = mtk_snfi_mac_op(snfi);
++ if (ret)
++ return ret;
++
++ /* For NULL input data, this loop will be skipped */
++ if (rxlen)
++ for (i = 0, m = 0; i < rxlen; ) {
++ reg = readl(snfi->regs +
++ SNFI_GPRAM_DATA + m);
++ for (j = 0; i < rxlen && j < 4; i++, j++, rxbuf++) {
++ if (m == 0 && i == 0)
++ j = i + txlen;
++ *rxbuf = (reg >> (j * 8)) & 0xFF;
++ }
++ m += 4;
++ }
++
++ return ret;
++}
++
++/*
++ * mtk_snfi_exec_op - to process command/data to send to the
++ * SPI NAND by mtk controller
++ */
++static int mtk_snfi_exec_op(struct spi_mem *mem,
++ const struct spi_mem_op *op)
++
++{
++ struct mtk_snfi *snfi = spi_controller_get_devdata(mem->spi->master);
++ struct spinand_device *spinand = spi_mem_get_drvdata(mem);
++ struct mtd_info *mtd = spinand_to_mtd(spinand);
++ struct nand_device *nand = mtd_to_nanddev(mtd);
++ const struct spi_mem_op *read_cache;
++ const struct spi_mem_op *write_cache;
++ const struct spi_mem_op *update_cache;
++ u32 tmpbufsize, txlen = 0, rxlen = 0;
++ u8 *txbuf, *rxbuf = NULL, *buf;
++ int i, ret = 0;
++
++ ret = mtk_snfi_reset(snfi);
++ if (ret) {
++ dev_warn(snfi->dev, "reset spi memory controller fail\n");
++ return ret;
++ }
++
++ /*if bbt initial, framework have detect nand information */
++ if (nand->bbt.cache) {
++ read_cache = spinand->op_templates.read_cache;
++ write_cache = spinand->op_templates.write_cache;
++ update_cache = spinand->op_templates.update_cache;
++
++ ret = mtk_snfi_hw_runtime_config(mem);
++ if (ret)
++ return ret;
++
++ /* For Read/Write with cache, Erase use framework flow */
++ if (op->cmd.opcode == read_cache->cmd.opcode) {
++ ret = mtk_snfi_read(mem, op);
++ if (ret)
++ dev_warn(snfi->dev, "snfi read fail\n");
++
++ return ret;
++ } else if ((op->cmd.opcode == write_cache->cmd.opcode)
++ || (op->cmd.opcode == update_cache->cmd.opcode)) {
++ ret = mtk_snfi_write(mem, op);
++ if (ret)
++ dev_warn(snfi->dev, "snfi write fail\n");
++
++ return ret;
++ }
++ }
++
++ tmpbufsize = sizeof(op->cmd.opcode) + op->addr.nbytes +
++ op->dummy.nbytes + op->data.nbytes;
++
++ txbuf = kzalloc(tmpbufsize, GFP_KERNEL);
++ if (!txbuf)
++ return -ENOMEM;
++
++ txbuf[txlen++] = op->cmd.opcode;
++
++ if (op->addr.nbytes)
++ for (i = 0; i < op->addr.nbytes; i++)
++ txbuf[txlen++] = op->addr.val >>
++ (8 * (op->addr.nbytes - i - 1));
++
++ txlen += op->dummy.nbytes;
++
++ if (op->data.dir == SPI_MEM_DATA_OUT)
++ for (i = 0; i < op->data.nbytes; i++) {
++ buf = (u8 *)op->data.buf.out;
++ txbuf[txlen++] = buf[i];
++ }
++
++ if (op->data.dir == SPI_MEM_DATA_IN) {
++ rxbuf = (u8 *)op->data.buf.in;
++ rxlen += op->data.nbytes;
++ }
++
++ ret = mtk_snfi_command_exec(snfi, txbuf, rxbuf, txlen, rxlen);
++ kfree(txbuf);
++
++ return ret;
++}
++
++static int mtk_snfi_init(struct mtk_snfi *snfi)
++{
++ int ret;
++
++ /* Reset the state machine and data FIFO */
++ ret = mtk_snfi_reset(snfi);
++ if (ret) {
++ dev_warn(snfi->dev, "MTK reset controller fail\n");
++ return ret;
++ }
++
++ snfi->buffer = devm_kzalloc(snfi->dev, 4096 + 256, GFP_KERNEL);
++ if (!snfi->buffer)
++ return -ENOMEM;
++
++ /* Clear interrupt, read clear. */
++ readw(snfi->regs + NFI_INTR_STA);
++ writew(0, snfi->regs + NFI_INTR_EN);
++
++ writel(0, snfi->regs + NFI_CON);
++ writel(0, snfi->regs + NFI_CNFG);
++
++ /* Change to NFI_SPI mode. */
++ writel(SNFI_MODE_EN, snfi->regs + SNFI_CNFG);
++
++ return 0;
++}
++
++static int mtk_snfi_check_buswidth(u8 width)
++{
++ switch (width) {
++ case 1:
++ case 2:
++ case 4:
++ return 0;
++
++ default:
++ break;
++ }
++
++ return -ENOTSUPP;
++}
++
++static bool mtk_snfi_supports_op(struct spi_mem *mem,
++ const struct spi_mem_op *op)
++{
++ int ret = 0;
++
++ /* For MTK Spi Nand controller, cmd buswidth just support 1 bit*/
++ if (op->cmd.buswidth != 1)
++ ret = -ENOTSUPP;
++
++ if (op->addr.nbytes)
++ ret |= mtk_snfi_check_buswidth(op->addr.buswidth);
++
++ if (op->dummy.nbytes)
++ ret |= mtk_snfi_check_buswidth(op->dummy.buswidth);
++
++ if (op->data.nbytes)
++ ret |= mtk_snfi_check_buswidth(op->data.buswidth);
++
++ if (ret)
++ return false;
++
++ return true;
++}
++
++static const struct spi_controller_mem_ops mtk_snfi_ops = {
++ .supports_op = mtk_snfi_supports_op,
++ .exec_op = mtk_snfi_exec_op,
++};
++
++static const struct mtk_snfi_caps snfi_mt7622 = {
++ .spare_size = spare_size_mt7622,
++ .num_spare_size = 4,
++ .nand_sec_size = 512,
++ .nand_fdm_size = 8,
++ .nand_fdm_ecc_size = 1,
++ .ecc_parity_bits = 13,
++ .pageformat_spare_shift = 4,
++ .bad_mark_swap = 0,
++};
++
++static const struct mtk_snfi_caps snfi_mt7629 = {
++ .spare_size = spare_size_mt7622,
++ .num_spare_size = 4,
++ .nand_sec_size = 512,
++ .nand_fdm_size = 8,
++ .nand_fdm_ecc_size = 1,
++ .ecc_parity_bits = 13,
++ .pageformat_spare_shift = 4,
++ .bad_mark_swap = 1,
++};
++
++static const struct of_device_id mtk_snfi_id_table[] = {
++ { .compatible = "mediatek,mt7622-snfi", .data = &snfi_mt7622, },
++ { .compatible = "mediatek,mt7629-snfi", .data = &snfi_mt7629, },
++ { /* sentinel */ }
++};
++
++static int mtk_snfi_probe(struct platform_device *pdev)
++{
++ struct device *dev = &pdev->dev;
++ struct device_node *np = dev->of_node;
++ struct spi_controller *ctlr;
++ struct mtk_snfi *snfi;
++ struct resource *res;
++ int ret = 0, irq;
++
++ ctlr = spi_alloc_master(&pdev->dev, sizeof(*snfi));
++ if (!ctlr)
++ return -ENOMEM;
++
++ snfi = spi_controller_get_devdata(ctlr);
++ snfi->caps = of_device_get_match_data(dev);
++ snfi->dev = dev;
++
++ snfi->ecc = of_mtk_ecc_get(np);
++ if (IS_ERR_OR_NULL(snfi->ecc))
++ goto err_put_master;
++
++ res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
++ snfi->regs = devm_ioremap_resource(dev, res);
++ if (IS_ERR(snfi->regs)) {
++ ret = PTR_ERR(snfi->regs);
++ goto release_ecc;
++ }
++
++ /* find the clocks */
++ snfi->clk.nfi_clk = devm_clk_get(dev, "nfi_clk");
++ if (IS_ERR(snfi->clk.nfi_clk)) {
++ dev_err(dev, "no nfi clk\n");
++ ret = PTR_ERR(snfi->clk.nfi_clk);
++ goto release_ecc;
++ }
++
++ snfi->clk.spi_clk = devm_clk_get(dev, "spi_clk");
++ if (IS_ERR(snfi->clk.spi_clk)) {
++ dev_err(dev, "no spi clk\n");
++ ret = PTR_ERR(snfi->clk.spi_clk);
++ goto release_ecc;
++ }
++
++ ret = mtk_snfi_enable_clk(dev, &snfi->clk);
++ if (ret)
++ goto release_ecc;
++
++ /* find the irq */
++ irq = platform_get_irq(pdev, 0);
++ if (irq < 0) {
++ dev_err(dev, "no snfi irq resource\n");
++ ret = -EINVAL;
++ goto clk_disable;
++ }
++
++ ret = devm_request_irq(dev, irq, mtk_snfi_irq, 0, "mtk-snfi", snfi);
++ if (ret) {
++ dev_err(dev, "failed to request snfi irq\n");
++ goto clk_disable;
++ }
++
++ ret = dma_set_mask(dev, DMA_BIT_MASK(32));
++ if (ret) {
++ dev_err(dev, "failed to set dma mask\n");
++ goto clk_disable;
++ }
++
++ ctlr->dev.of_node = np;
++ ctlr->mem_ops = &mtk_snfi_ops;
++
++ platform_set_drvdata(pdev, snfi);
++ ret = mtk_snfi_init(snfi);
++ if (ret) {
++ dev_err(dev, "failed to init snfi\n");
++ goto clk_disable;
++ }
++
++ ret = devm_spi_register_master(dev, ctlr);
++ if (ret)
++ goto clk_disable;
++
++ return 0;
++
++clk_disable:
++ mtk_snfi_disable_clk(&snfi->clk);
++
++release_ecc:
++ mtk_ecc_release(snfi->ecc);
++
++err_put_master:
++ spi_master_put(ctlr);
++
++ dev_err(dev, "MediaTek SPI NAND interface probe failed %d\n", ret);
++ return ret;
++}
++
++static int mtk_snfi_remove(struct platform_device *pdev)
++{
++ struct mtk_snfi *snfi = platform_get_drvdata(pdev);
++
++ mtk_snfi_disable_clk(&snfi->clk);
++
++ return 0;
++}
++
++static int mtk_snfi_suspend(struct platform_device *pdev, pm_message_t state)
++{
++ struct mtk_snfi *snfi = platform_get_drvdata(pdev);
++
++ mtk_snfi_disable_clk(&snfi->clk);
++
++ return 0;
++}
++
++static int mtk_snfi_resume(struct platform_device *pdev)
++{
++ struct device *dev = &pdev->dev;
++ struct mtk_snfi *snfi = dev_get_drvdata(dev);
++ int ret;
++
++ ret = mtk_snfi_enable_clk(dev, &snfi->clk);
++ if (ret)
++ return ret;
++
++ ret = mtk_snfi_init(snfi);
++ if (ret)
++ dev_err(dev, "failed to init snfi controller\n");
++
++ return ret;
++}
++
++static struct platform_driver mtk_snfi_driver = {
++ .driver = {
++ .name = "mtk-snfi",
++ .of_match_table = mtk_snfi_id_table,
++ },
++ .probe = mtk_snfi_probe,
++ .remove = mtk_snfi_remove,
++ .suspend = mtk_snfi_suspend,
++ .resume = mtk_snfi_resume,
++};
++
++module_platform_driver(mtk_snfi_driver);
++
++MODULE_LICENSE("GPL v2");
++MODULE_AUTHOR("Xiangsheng Hou <xiangsheng.hou@mediatek.com>");
++MODULE_DESCRIPTION("Mediatek SPI Memory Interface Driver");
projects / openwrt / staging / zorun.git / commitdiff