Hi Miquèl, On 09/22/2018 03:48 PM, Miquel Raynal wrote:
Hi Christophe, I suppose you received the kbuildrobot issues already, please have a look at them.
Yes, kbuidroot issues will be solved in the v2 patchset.
The driver looks well, some comments below. <christophe.kerello@xxxxxx> wrote on Mon, 17 Sep 2018 17:47:39 +0200:From: Christophe Kerello <christophe.kerello@xxxxxx> The driver adds the support for the STMicroelectronics FMC2 NAND Controller found on STM32MP SOCs. This patch is based on FMC2 command sequencer. The purpose of the command sequencer is to facilitate the programming and the reading of NAND flash pages with the ECC and to free the CPU of sequencing tasks. It requires one DMA channel for write and two DMA channels for read operations. Only NAND_ECC_HW mode is actually supported. The driver supports a maximum 8k page size. The following ECC strength and step size are currently supported: - nand-ecc-strength = <8>, nand-ecc-step-size = <512> (BCH8) - nand-ecc-strength = <4>, nand-ecc-step-size = <512> (BCH4) - nand-ecc-strength = <1>, nand-ecc-step-size = <512> (Extended ecc based on HAMMING) This patch has been tested on Micron MT29F8G08ABACAH4 and MT29F8G16ABACAH4 Signed-off-by: Christophe Kerello <christophe.kerello@xxxxxx> --- drivers/mtd/nand/raw/Kconfig | 9 + drivers/mtd/nand/raw/Makefile | 1 + drivers/mtd/nand/raw/stm32_fmc2_nand.c | 1729 ++++++++++++++++++++++++++++++++ 3 files changed, 1739 insertions(+) create mode 100644 drivers/mtd/nand/raw/stm32_fmc2_nand.c diff --git a/drivers/mtd/nand/raw/Kconfig b/drivers/mtd/nand/raw/Kconfig index c7efc31..863662c 100644 --- a/drivers/mtd/nand/raw/Kconfig +++ b/drivers/mtd/nand/raw/Kconfig @@ -541,4 +541,13 @@ config MTD_NAND_TEGRA is supported. Extra OOB bytes when using HW ECC are currently not supported.+config MTD_NAND_STM32_FMC2+ tristate "Support for NAND controller on STM32MP SoCs" + depends on MACH_STM32MP157 || COMPILE_TEST + help + Enables support for NAND Flash chips on SoCs containing the FMC2 + NAND controller. This controller is found on STM32MP SoCs. + The driver supports a maximum 8k page size. HW ECC is enabled and + supports a maximum 8-bit correction error per sector of 512 bytes.HW ECC should not be enabled by default. 8-bit/512B of correctability is good but not that high and people might want to use software ECC in conjunction with raw accesses.
Yes, I agree. The driver only supports NAND_ECC_HW mode. NAND_ECC_SOFT mode was not requested by customers and was not implemented. The driver could be improved later to support mode like NAND_ECC_SOFT or NAND_ECC_ON_DIE. Should I remove "HW ECC is enabled" from Kconfig description?
+ endif # MTD_NAND diff --git a/drivers/mtd/nand/raw/Makefile b/drivers/mtd/nand/raw/Makefile index a6ef067..8c437e3 100644 --- a/drivers/mtd/nand/raw/Makefile +++ b/drivers/mtd/nand/raw/Makefile @@ -56,6 +56,7 @@ obj-$(CONFIG_MTD_NAND_BRCMNAND) += brcmnand/ obj-$(CONFIG_MTD_NAND_QCOM) += qcom_nandc.o obj-$(CONFIG_MTD_NAND_MTK) += mtk_ecc.o mtk_nand.o obj-$(CONFIG_MTD_NAND_TEGRA) += tegra_nand.o +obj-$(CONFIG_MTD_NAND_STM32_FMC2) += stm32_fmc2_nand.onand-objs := nand_base.o nand_bbt.o nand_timings.o nand_ids.onand-objs += nand_amd.o diff --git a/drivers/mtd/nand/raw/stm32_fmc2_nand.c b/drivers/mtd/nand/raw/stm32_fmc2_nand.c new file mode 100644 index 0000000..dd5762a --- /dev/null +++ b/drivers/mtd/nand/raw/stm32_fmc2_nand.c @@ -0,0 +1,1729 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright (C) STMicroelectronics 2018 - All Rights Reserved + * Author: Christophe Kerello <christophe.kerello@xxxxxx> for STMicroelectronicsI'm not sure the "All rights reserved" has a meaning here. But you definitely not have to add "for STMicroelectronics" because it is already in the copyright.
Ok.
+ */ + +#include <linux/clk.h> +#include <linux/dmaengine.h> +#include <linux/dma-mapping.h> +#include <linux/errno.h> +#include <linux/interrupt.h> +#include <linux/iopoll.h> +#include <linux/module.h> +#include <linux/mtd/rawnand.h> +#include <linux/pinctrl/consumer.h> +#include <linux/platform_device.h> +#include <linux/reset.h> + +/* Bad block marker length */ +#define FMC2_BBM_LEN 2 + +/* ECC step size */ +#define FMC2_ECC_STEP_SIZE 512 + +/* BCHDSRx registers length */ +#define FMC2_BCHDSRS_LEN 20 + +/* HECCR length */ +#define FMC2_HECCR_LEN 4 + +/* Max requests done for a 8k nand page size */ +#define FMC2_MAX_SG_COUNT 16 + +/* Max chip enable */ +#define FMC2_MAX_CE 2 + +/* Timings */ +#define FMC2_THIZ 1 +#define FMC2_TIO 8000 +#define FMC2_TSYNC 3000 +#define FMC2_PCR_TIMING_MASK 0xf +#define FMC2_PMEM_PATT_TIMING_MASK 0xff + +/* FMC2 Controller Registers */ +#define FMC2_BCR1 0x0 +#define FMC2_PCR 0x80 +#define FMC2_SR 0x84 +#define FMC2_PMEM 0x88 +#define FMC2_PATT 0x8c +#define FMC2_HECCR 0x94 +#define FMC2_CSQCR 0x200 +#define FMC2_CSQCFGR1 0x204 +#define FMC2_CSQCFGR2 0x208 +#define FMC2_CSQCFGR3 0x20c +#define FMC2_CSQAR1 0x210 +#define FMC2_CSQAR2 0x214 +#define FMC2_CSQIER 0x220 +#define FMC2_CSQISR 0x224 +#define FMC2_CSQICR 0x228 +#define FMC2_CSQEMSR 0x230 +#define FMC2_BCHIER 0x250 +#define FMC2_BCHISR 0x254 +#define FMC2_BCHICR 0x258 +#define FMC2_BCHPBR1 0x260 +#define FMC2_BCHPBR2 0x264 +#define FMC2_BCHPBR3 0x268 +#define FMC2_BCHPBR4 0x26c +#define FMC2_BCHDSR0 0x27c +#define FMC2_BCHDSR1 0x280 +#define FMC2_BCHDSR2 0x284 +#define FMC2_BCHDSR3 0x288 +#define FMC2_BCHDSR4 0x28c + +/* Register: FMC2_BCR1 */ +#define FMC2_BCR1_FMC2EN BIT(31) + +/* Register: FMC2_PCR */ +#define FMC2_PCR_PWAITEN BIT(1) +#define FMC2_PCR_PBKEN BIT(2) +#define FMC2_PCR_PWID_MASK GENMASK(5, 4) +#define FMC2_PCR_PWID(x) (((x) & 0x3) << 4) +#define FMC2_PCR_PWID_BUSWIDTH_8 0 +#define FMC2_PCR_PWID_BUSWIDTH_16 1 +#define FMC2_PCR_ECCEN BIT(6) +#define FMC2_PCR_ECCALG BIT(8) +#define FMC2_PCR_TCLR_MASK GENMASK(12, 9) +#define FMC2_PCR_TCLR(x) (((x) & 0xf) << 9) +#define FMC2_PCR_TCLR_DEFAULT 0xf +#define FMC2_PCR_TAR_MASK GENMASK(16, 13) +#define FMC2_PCR_TAR(x) (((x) & 0xf) << 13) +#define FMC2_PCR_TAR_DEFAULT 0xf +#define FMC2_PCR_ECCSS_MASK GENMASK(19, 17) +#define FMC2_PCR_ECCSS(x) (((x) & 0x7) << 17) +#define FMC2_PCR_ECCSS_512 1 +#define FMC2_PCR_ECCSS_2048 3 +#define FMC2_PCR_BCHECC BIT(24) +#define FMC2_PCR_WEN BIT(25) + +/* Register: FMC2_SR */ +#define FMC2_SR_NWRF BIT(6) + +/* Register: FMC2_PMEM */ +#define FMC2_PMEM_MEMSET(x) (((x) & 0xff) << 0) +#define FMC2_PMEM_MEMWAIT(x) (((x) & 0xff) << 8) +#define FMC2_PMEM_MEMHOLD(x) (((x) & 0xff) << 16) +#define FMC2_PMEM_MEMHIZ(x) (((x) & 0xff) << 24) +#define FMC2_PMEM_DEFAULT 0x0a0a0a0a + +/* Register: FMC2_PATT */ +#define FMC2_PATT_ATTSET(x) (((x) & 0xff) << 0) +#define FMC2_PATT_ATTWAIT(x) (((x) & 0xff) << 8) +#define FMC2_PATT_ATTHOLD(x) (((x) & 0xff) << 16) +#define FMC2_PATT_ATTHIZ(x) (((x) & 0xff) << 24) +#define FMC2_PATT_DEFAULT 0x0a0a0a0a + +/* Register: FMC2_CSQCR */ +#define FMC2_CSQCR_CSQSTART BIT(0) + +/* Register: FMC2_CSQCFGR1 */ +#define FMC2_CSQCFGR1_CMD2EN BIT(1) +#define FMC2_CSQCFGR1_DMADEN BIT(2) +#define FMC2_CSQCFGR1_ACYNBR(x) (((x) & 0x7) << 4) +#define FMC2_CSQCFGR1_CMD1(x) (((x) & 0xff) << 8) +#define FMC2_CSQCFGR1_CMD2(x) (((x) & 0xff) << 16) +#define FMC2_CSQCFGR1_CMD1T BIT(24) +#define FMC2_CSQCFGR1_CMD2T BIT(25) + +/* Register: FMC2_CSQCFGR2 */ +#define FMC2_CSQCFGR2_SQSDTEN BIT(0) +#define FMC2_CSQCFGR2_RCMD2EN BIT(1) +#define FMC2_CSQCFGR2_DMASEN BIT(2) +#define FMC2_CSQCFGR2_RCMD1(x) (((x) & 0xff) << 8) +#define FMC2_CSQCFGR2_RCMD2(x) (((x) & 0xff) << 16) +#define FMC2_CSQCFGR2_RCMD1T BIT(24) +#define FMC2_CSQCFGR2_RCMD2T BIT(25) + +/* Register: FMC2_CSQCFGR3 */ +#define FMC2_CSQCFGR3_SNBR(x) (((x) & 0x1f) << 8) +#define FMC2_CSQCFGR3_AC1T BIT(16) +#define FMC2_CSQCFGR3_AC2T BIT(17) +#define FMC2_CSQCFGR3_AC3T BIT(18) +#define FMC2_CSQCFGR3_AC4T BIT(19) +#define FMC2_CSQCFGR3_AC5T BIT(20) +#define FMC2_CSQCFGR3_SDT BIT(21) +#define FMC2_CSQCFGR3_RAC1T BIT(22) +#define FMC2_CSQCFGR3_RAC2T BIT(23) + +/* Register: FMC2_CSQCAR1 */ +#define FMC2_CSQCAR1_ADDC1(x) (((x) & 0xff) << 0) +#define FMC2_CSQCAR1_ADDC2(x) (((x) & 0xff) << 8) +#define FMC2_CSQCAR1_ADDC3(x) (((x) & 0xff) << 16) +#define FMC2_CSQCAR1_ADDC4(x) (((x) & 0xff) << 24) + +/* Register: FMC2_CSQCAR2 */ +#define FMC2_CSQCAR2_ADDC5(x) (((x) & 0xff) << 0) +#define FMC2_CSQCAR2_NANDCEN(x) (((x) & 0x3) << 10) +#define FMC2_CSQCAR2_SAO(x) (((x) & 0xffff) << 16) + +/* Register: FMC2_CSQIER */ +#define FMC2_CSQIER_TCIE BIT(0) + +/* Register: FMC2_CSQICR */ +#define FMC2_CSQICR_CLEAR_IRQ GENMASK(4, 0) + +/* Register: FMC2_CSQEMSR */ +#define FMC2_CSQEMSR_SEM GENMASK(15, 0) + +/* Register: FMC2_BCHIER */ +#define FMC2_BCHIER_DERIE BIT(1) +#define FMC2_BCHIER_EPBRIE BIT(4) + +/* Register: FMC2_BCHICR */ +#define FMC2_BCHICR_CLEAR_IRQ GENMASK(4, 0) + +/* Register: FMC2_BCHDSR0 */ +#define FMC2_BCHDSR0_DUE BIT(0) +#define FMC2_BCHDSR0_DEF BIT(1) +#define FMC2_BCHDSR0_DEN_MASK GENMASK(7, 4) +#define FMC2_BCHDSR0_DEN_SHIFT 4 + +/* Register: FMC2_BCHDSR1 */ +#define FMC2_BCHDSR1_EBP1_MASK GENMASK(12, 0) +#define FMC2_BCHDSR1_EBP2_MASK GENMASK(28, 16) +#define FMC2_BCHDSR1_EBP2_SHIFT 16 + +/* Register: FMC2_BCHDSR2 */ +#define FMC2_BCHDSR2_EBP3_MASK GENMASK(12, 0) +#define FMC2_BCHDSR2_EBP4_MASK GENMASK(28, 16) +#define FMC2_BCHDSR2_EBP4_SHIFT 16 + +/* Register: FMC2_BCHDSR3 */ +#define FMC2_BCHDSR3_EBP5_MASK GENMASK(12, 0) +#define FMC2_BCHDSR3_EBP6_MASK GENMASK(28, 16) +#define FMC2_BCHDSR3_EBP6_SHIFT 16 + +/* Register: FMC2_BCHDSR4 */ +#define FMC2_BCHDSR4_EBP7_MASK GENMASK(12, 0) +#define FMC2_BCHDSR4_EBP8_MASK GENMASK(28, 16) +#define FMC2_BCHDSR4_EBP8_SHIFT 16 + +enum stm32_fmc2_ecc { + FMC2_ECC_HAM = 1, + FMC2_ECC_BCH4 = 4, + FMC2_ECC_BCH8 = 8 +}; + +struct stm32_fmc2_timings { + u8 tclr; + u8 tar; + u8 thiz; + u8 twait; + u8 thold_mem; + u8 tset_mem; + u8 thold_att; + u8 tset_att; +}; + +struct stm32_fmc2 { + struct nand_chip chip; + struct device *dev; + void __iomem *io_base; + void __iomem *data_base[FMC2_MAX_CE]; + void __iomem *cmd_base[FMC2_MAX_CE]; + void __iomem *addr_base[FMC2_MAX_CE]; + phys_addr_t io_phys_addr; + phys_addr_t data_phys_addr[FMC2_MAX_CE]; + struct clk *clk; + + struct dma_chan *dma_tx_ch; + struct dma_chan *dma_rx_ch; + struct dma_chan *dma_ecc_ch; + struct sg_table dma_data_sg; + struct sg_table dma_ecc_sg; + u8 *ecc_buf; + int dma_ecc_len; + + struct completion complete; + struct completion dma_data_complete; + struct completion dma_ecc_complete; + + struct stm32_fmc2_timings timings; + u8 cs_assigned; + int cs_sel; + int ncs; + int cs_used[FMC2_MAX_CE]; +}; + +/* Enable irq sources in case of the sequencer is used */ +static inline void stm32_fmc2_enable_seq_irq(struct stm32_fmc2 *fmc2) +{ + u32 csqier = readl_relaxed(fmc2->io_base + FMC2_CSQIER); + + csqier |= FMC2_CSQIER_TCIE; + + writel_relaxed(csqier, fmc2->io_base + FMC2_CSQIER); +} + +/* Disable irq sources in case of the sequencer is used */ +static inline void stm32_fmc2_disable_seq_irq(struct stm32_fmc2 *fmc2) +{ + u32 csqier = readl_relaxed(fmc2->io_base + FMC2_CSQIER); + + csqier &= ~FMC2_CSQIER_TCIE; + + writel_relaxed(csqier, fmc2->io_base + FMC2_CSQIER); +} + +/* Clear irq sources in case of the sequencer is used */ +static inline void stm32_fmc2_clear_seq_irq(struct stm32_fmc2 *fmc2) +{ + writel_relaxed(FMC2_CSQICR_CLEAR_IRQ, fmc2->io_base + FMC2_CSQICR); +} + +/* Select function */ +static void stm32_fmc2_select_chip(struct nand_chip *chip, int chipnr) +{ + struct stm32_fmc2 *fmc2 = nand_get_controller_data(chip); + struct dma_slave_config dma_cfg; + + if (chipnr < 0 || chipnr >= fmc2->ncs) + return; + + if (fmc2->cs_used[chipnr] == fmc2->cs_sel) + return; + + fmc2->cs_sel = fmc2->cs_used[chipnr]; + + if (fmc2->dma_tx_ch && fmc2->dma_rx_ch) { + memset(&dma_cfg, 0, sizeof(dma_cfg)); + dma_cfg.src_addr = fmc2->data_phys_addr[fmc2->cs_sel]; + dma_cfg.dst_addr = fmc2->data_phys_addr[fmc2->cs_sel]; + dma_cfg.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES; + dma_cfg.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES; + dma_cfg.src_maxburst = 32; + dma_cfg.dst_maxburst = 32; + + if (dmaengine_slave_config(fmc2->dma_tx_ch, &dma_cfg)) + dev_warn(fmc2->dev, "tx DMA engine slave config failed\n"); + + if (dmaengine_slave_config(fmc2->dma_rx_ch, &dma_cfg)) + dev_warn(fmc2->dev, "rx DMA engine slave config failed\n"); + }What if there are two NAND chips using different timing modes? You should probably reconfigure the timings registers, unless there are already a set of timing registers per CS?
Yes, it's true. In case of 2 NAND chips, timings and pcr registers should have been reconfigured. But, the driver only supports one NAND chip connected to 1 or 2 CS. There was no requirement on our side to support 2 different NAND chips. I do not have a board to test such configuration, so i do not want to deliver this support without being able to test it on my side. The driver will be improved later to support 2 different NAND chips, in case this configuration is requested by customers.
+} + +/* + * ECC HAMMING calculation + * ECC is 3 bytes for 512 bytes of data (supports error correction up to + * max of 1-bit) + */ +static inline void stm32_fmc2_ham_set_ecc(const u32 ecc_sta, u8 *ecc) +{ + ecc[0] = ecc_sta; + ecc[1] = ecc_sta >> 8; + ecc[2] = ecc_sta >> 16; +} + +static int stm32_fmc2_ham_correct(struct nand_chip *chip, uint8_t *dat, + uint8_t *read_ecc, uint8_t *calc_ecc) +{ + u8 bit_position = 0, b0, b1, b2; + u32 byte_addr = 0, b; + u32 i, shifting = 1; + + /* Indicate which bit and byte is faulty (if any) */ + b0 = read_ecc[0] ^ calc_ecc[0]; + b1 = read_ecc[1] ^ calc_ecc[1]; + b2 = read_ecc[2] ^ calc_ecc[2]; + b = b0 | (b1 << 8) | (b2 << 16); + + /* No errors */ + if (likely(!b)) + return 0; + + /* Calculate bit position */ + for (i = 0; i < 3; i++) { + switch (b % 4) { + case 2: + bit_position += shifting; + case 1: + break; + default: + return -EBADMSG; + } + shifting <<= 1; + b >>= 2; + } + + /* Calculate byte position */ + shifting = 1; + for (i = 0; i < 9; i++) { + switch (b % 4) { + case 2: + byte_addr += shifting; + case 1: + break; + default: + return -EBADMSG; + } + shifting <<= 1; + b >>= 2; + } + + /* Flip the bit */ + dat[byte_addr] ^= (1 << bit_position); + + return 1; +} + +/* BCH algorithm correction */ +static int stm32_fmc2_bch_decode(int eccsize, u8 *dat, u32 *ecc_sta) +{ + u32 bchdsr0 = ecc_sta[0]; + u32 bchdsr1 = ecc_sta[1]; + u32 bchdsr2 = ecc_sta[2]; + u32 bchdsr3 = ecc_sta[3]; + u32 bchdsr4 = ecc_sta[4]; + u16 pos[8]; + int i, den; + unsigned int nb_errs = 0; + + /* No errors found */ + if (likely(!(bchdsr0 & FMC2_BCHDSR0_DEF))) + return 0; + + /* Too many errors detected */ + if (unlikely(bchdsr0 & FMC2_BCHDSR0_DUE)) + return -EBADMSG; + + pos[0] = bchdsr1 & FMC2_BCHDSR1_EBP1_MASK; + pos[1] = (bchdsr1 & FMC2_BCHDSR1_EBP2_MASK) >> FMC2_BCHDSR1_EBP2_SHIFT; + pos[2] = bchdsr2 & FMC2_BCHDSR2_EBP3_MASK; + pos[3] = (bchdsr2 & FMC2_BCHDSR2_EBP4_MASK) >> FMC2_BCHDSR2_EBP4_SHIFT; + pos[4] = bchdsr3 & FMC2_BCHDSR3_EBP5_MASK; + pos[5] = (bchdsr3 & FMC2_BCHDSR3_EBP6_MASK) >> FMC2_BCHDSR3_EBP6_SHIFT; + pos[6] = bchdsr4 & FMC2_BCHDSR4_EBP7_MASK; + pos[7] = (bchdsr4 & FMC2_BCHDSR4_EBP8_MASK) >> FMC2_BCHDSR4_EBP8_SHIFT; + + den = (bchdsr0 & FMC2_BCHDSR0_DEN_MASK) >> FMC2_BCHDSR0_DEN_SHIFT; + for (i = 0; i < den; i++) { + if (pos[i] < eccsize * 8) { + change_bit(pos[i], (unsigned long *)dat); + nb_errs++; + } + } + + return nb_errs; +} + +/* Sequencer read/write configuration */ +static void stm32_fmc2_rw_page_init(struct stm32_fmc2 *fmc2, int page, + int raw, bool write_data) +{ + struct nand_chip *chip = &fmc2->chip; + struct mtd_info *mtd = nand_to_mtd(chip); + u32 csqcfgr1, csqcfgr2, csqcfgr3; + u32 csqar1, csqar2; + u32 ecc_offset = mtd->writesize + FMC2_BBM_LEN; + u32 pcr = readl_relaxed(fmc2->io_base + FMC2_PCR); + + if (write_data) + pcr |= FMC2_PCR_WEN; + else + pcr &= ~FMC2_PCR_WEN; + writel_relaxed(pcr, fmc2->io_base + FMC2_PCR); + + /* + * - Set Program Page/Page Read command + * - Enable DMA request data + * - Set timings + */ + csqcfgr1 = FMC2_CSQCFGR1_DMADEN | FMC2_CSQCFGR1_CMD1T; + if (write_data) + csqcfgr1 |= FMC2_CSQCFGR1_CMD1(NAND_CMD_SEQIN); + else + csqcfgr1 |= FMC2_CSQCFGR1_CMD1(NAND_CMD_READ0) | + FMC2_CSQCFGR1_CMD2EN | + FMC2_CSQCFGR1_CMD2(NAND_CMD_READSTART) | + FMC2_CSQCFGR1_CMD2T; + + /* + * - Set Random Data Input/Random Data Read command + * - Enable the sequencer to access the Spare data area + * - Enable DMA request status decoding for read + * - Set timings + */ + if (write_data) + csqcfgr2 = FMC2_CSQCFGR2_RCMD1(NAND_CMD_RNDIN); + else + csqcfgr2 = FMC2_CSQCFGR2_RCMD1(NAND_CMD_RNDOUT) | + FMC2_CSQCFGR2_RCMD2EN | + FMC2_CSQCFGR2_RCMD2(NAND_CMD_RNDOUTSTART) | + FMC2_CSQCFGR2_RCMD1T | + FMC2_CSQCFGR2_RCMD2T; + if (!raw) { + csqcfgr2 |= write_data ? 0 : FMC2_CSQCFGR2_DMASEN; + csqcfgr2 |= FMC2_CSQCFGR2_SQSDTEN; + } + + /* + * - Set the number of sectors to be written + * - Set timings + */ + csqcfgr3 = FMC2_CSQCFGR3_SNBR(chip->ecc.steps - 1); + if (write_data) { + csqcfgr3 |= FMC2_CSQCFGR3_RAC2T; + if (chip->chipsize > SZ_128M) + csqcfgr3 |= FMC2_CSQCFGR3_AC5T; + else + csqcfgr3 |= FMC2_CSQCFGR3_AC4T; + } + + /* + * Set the fourth first address cycles + * Byte 1 and byte 2 => column, we start at 0x0 + * Byte 3 and byte 4 => page + */ + csqar1 = FMC2_CSQCAR1_ADDC3(page); + csqar1 |= FMC2_CSQCAR1_ADDC4(page >> 8); + + /* + * - Set chip enable number + * - Set ecc byte offset in the spare area + * - Calculate the number of address cycles to be issued + * - Set byte 5 of address cycle if needed + */ + csqar2 = FMC2_CSQCAR2_NANDCEN(fmc2->cs_sel); + if (chip->options & NAND_BUSWIDTH_16) + csqar2 |= FMC2_CSQCAR2_SAO(ecc_offset >> 1); + else + csqar2 |= FMC2_CSQCAR2_SAO(ecc_offset); + if (chip->chipsize > SZ_128M) { + csqcfgr1 |= FMC2_CSQCFGR1_ACYNBR(5); + csqar2 |= FMC2_CSQCAR2_ADDC5(page >> 16); + } else { + csqcfgr1 |= FMC2_CSQCFGR1_ACYNBR(4); + } + + writel_relaxed(csqcfgr1, fmc2->io_base + FMC2_CSQCFGR1); + writel_relaxed(csqcfgr2, fmc2->io_base + FMC2_CSQCFGR2); + writel_relaxed(csqcfgr3, fmc2->io_base + FMC2_CSQCFGR3); + writel_relaxed(csqar1, fmc2->io_base + FMC2_CSQAR1); + writel_relaxed(csqar2, fmc2->io_base + FMC2_CSQAR2); +} + +static void stm32_fmc2_dma_callback(void *arg) +{ + complete((struct completion *)arg); +} + +/* Read/write data from/to a page */ +static int stm32_fmc2_xfer(struct stm32_fmc2 *fmc2, const u8 *buf, + int raw, bool write_data) +{ + struct nand_chip *chip = &fmc2->chip; + struct dma_async_tx_descriptor *desc_data, *desc_ecc; + struct scatterlist *sg; + struct dma_chan *dma_ch = fmc2->dma_rx_ch; + enum dma_data_direction dma_data_dir = DMA_FROM_DEVICE; + enum dma_transfer_direction dma_transfer_dir = DMA_DEV_TO_MEM; + u32 csqcr = readl_relaxed(fmc2->io_base + FMC2_CSQCR); + int eccsteps = chip->ecc.steps; + int eccsize = chip->ecc.size; + const u8 *p = buf; + int s, ret; + + /* Configure DMA data */ + if (write_data) { + dma_data_dir = DMA_TO_DEVICE; + dma_transfer_dir = DMA_MEM_TO_DEV; + dma_ch = fmc2->dma_tx_ch; + } + + for_each_sg(fmc2->dma_data_sg.sgl, sg, eccsteps, s) { + sg_set_buf(sg, p, eccsize); + p += eccsize; + } + + ret = dma_map_sg(fmc2->dev, fmc2->dma_data_sg.sgl, + eccsteps, dma_data_dir); + if (ret < 0) + return ret; + + desc_data = dmaengine_prep_slave_sg(dma_ch, fmc2->dma_data_sg.sgl, + eccsteps, dma_transfer_dir, + DMA_PREP_INTERRUPT); + if (!desc_data) { + ret = -ENOMEM; + goto err_unmap_data; + } + + reinit_completion(&fmc2->dma_data_complete); + reinit_completion(&fmc2->complete); + desc_data->callback = stm32_fmc2_dma_callback; + desc_data->callback_param = &fmc2->dma_data_complete; + ret = dma_submit_error(dmaengine_submit(desc_data)); + if (ret) + goto err_unmap_data; + + dma_async_issue_pending(dma_ch); + + if (!write_data && !raw) { + /* Configure DMA ecc status */ + p = fmc2->ecc_buf; + for_each_sg(fmc2->dma_ecc_sg.sgl, sg, eccsteps, s) { + sg_set_buf(sg, p, fmc2->dma_ecc_len); + p += fmc2->dma_ecc_len; + } + + ret = dma_map_sg(fmc2->dev, fmc2->dma_ecc_sg.sgl, + eccsteps, dma_data_dir); + if (ret < 0) + goto err_unmap_data; + + desc_ecc = dmaengine_prep_slave_sg(fmc2->dma_ecc_ch, + fmc2->dma_ecc_sg.sgl, + eccsteps, dma_transfer_dir, + DMA_PREP_INTERRUPT); + if (!desc_ecc) { + ret = -ENOMEM; + goto err_unmap_ecc; + } + + reinit_completion(&fmc2->dma_ecc_complete); + desc_ecc->callback = stm32_fmc2_dma_callback; + desc_ecc->callback_param = &fmc2->dma_ecc_complete; + ret = dma_submit_error(dmaengine_submit(desc_ecc)); + if (ret) + goto err_unmap_ecc; + + dma_async_issue_pending(fmc2->dma_ecc_ch); + } + + stm32_fmc2_clear_seq_irq(fmc2); + stm32_fmc2_enable_seq_irq(fmc2); + + /* Start the transfer */ + csqcr |= FMC2_CSQCR_CSQSTART; + writel_relaxed(csqcr, fmc2->io_base + FMC2_CSQCR); + + /* Wait end of sequencer transfer */ + if (!wait_for_completion_timeout(&fmc2->complete, + msecs_to_jiffies(1000))) { + dev_err(fmc2->dev, "seq timeout\n"); + stm32_fmc2_disable_seq_irq(fmc2); + dmaengine_terminate_all(dma_ch); + if (!write_data && !raw) + dmaengine_terminate_all(fmc2->dma_ecc_ch); + ret = -ETIMEDOUT; + goto err_unmap_ecc; + } + + /* Wait DMA data transfer completion */ + if (!wait_for_completion_timeout(&fmc2->dma_data_complete, + msecs_to_jiffies(100))) { + dev_err(fmc2->dev, "data DMA timeout\n"); + dmaengine_terminate_all(dma_ch); + ret = -ETIMEDOUT; + } + + /* Wait DMA ecc transfer completion */ + if (!write_data && !raw) { + if (!wait_for_completion_timeout(&fmc2->dma_ecc_complete, + msecs_to_jiffies(100))) { + dev_err(fmc2->dev, "ecc DMA timeout\n"); + dmaengine_terminate_all(fmc2->dma_ecc_ch); + ret = -ETIMEDOUT; + } + } + +err_unmap_ecc: + if (!write_data && !raw) + dma_unmap_sg(fmc2->dev, fmc2->dma_ecc_sg.sgl, + eccsteps, dma_data_dir); + +err_unmap_data: + dma_unmap_sg(fmc2->dev, fmc2->dma_data_sg.sgl, eccsteps, dma_data_dir); + + return ret; +} + +static int stm32_fmc2_sequencer_write(struct nand_chip *chip, + const u8 *buf, int oob_required, + int page, int raw) +{ + struct mtd_info *mtd = nand_to_mtd(chip); + struct stm32_fmc2 *fmc2 = nand_get_controller_data(chip); + int ret; + + /* Configure the sequencer */ + stm32_fmc2_rw_page_init(fmc2, page, raw, true); + + /* Write the page */ + ret = stm32_fmc2_xfer(fmc2, buf, raw, true); + if (ret) + return ret; + + /* Write oob */ + if (oob_required) { + ret = nand_change_write_column_op(chip, mtd->writesize, + chip->oob_poi, mtd->oobsize, + false); + if (ret) + return ret; + } + + return nand_prog_page_end_op(chip); +} + +static int stm32_fmc2_sequencer_write_page(struct nand_chip *chip, + const uint8_t *buf, + int oob_required, + int page) +{ + return stm32_fmc2_sequencer_write(chip, buf, oob_required, page, false); +} + +static int stm32_fmc2_sequencer_write_page_raw(struct nand_chip *chip, + const uint8_t *buf, + int oob_required, + int page) +{ + return stm32_fmc2_sequencer_write(chip, buf, oob_required, page, true); +} + +/* + * Get a status indicating which sectors have errors + * Only available when the sequencer is used (BCH only) + */ +static inline u16 stm32_fmc2_get_mapping_status(struct stm32_fmc2 *fmc2) +{ + u32 csqemsr = readl_relaxed(fmc2->io_base + FMC2_CSQEMSR); + + return csqemsr & FMC2_CSQEMSR_SEM; +} + +static int stm32_fmc2_sequencer_read_page(struct nand_chip *chip, + uint8_t *buf, + int oob_required, + int page) +{ + struct mtd_info *mtd = nand_to_mtd(chip); + struct stm32_fmc2 *fmc2 = nand_get_controller_data(chip); + int i, s, ret, eccsize = chip->ecc.size; + int eccbytes = chip->ecc.bytes; + int eccsteps = chip->ecc.steps; + int eccstrength = chip->ecc.strength; + u8 *p = buf; + u8 *ecc_calc = chip->ecc.calc_buf; + u8 *ecc_code = chip->ecc.code_buf; + u32 *ecc_sta = (u32 *)fmc2->ecc_buf; + u16 sta_map; + unsigned int max_bitflips = 0; + + /* Configure the sequencer */ + stm32_fmc2_rw_page_init(fmc2, page, 0, false); + + /* Read the page */ + ret = stm32_fmc2_xfer(fmc2, buf, 0, false); + if (ret) + return ret; + + sta_map = stm32_fmc2_get_mapping_status(fmc2); + + /* Check if errors happen */ + if (likely(!sta_map)) { + if (oob_required) + return nand_change_read_column_op(chip, mtd->writesize, + chip->oob_poi, + mtd->oobsize, false); + + return 0; + } + + /* Read oob */ + ret = nand_change_read_column_op(chip, mtd->writesize, + chip->oob_poi, mtd->oobsize, false); + if (ret) + return ret; + + ret = mtd_ooblayout_get_eccbytes(mtd, ecc_code, chip->oob_poi, 0, + chip->ecc.total); + if (ret) + return ret; + + /* Correct data */ + for (i = 0, s = 0; s < eccsteps; s++, i += eccbytes, p += eccsize) { + int stat = 0; + + if (eccstrength == FMC2_ECC_HAM) { + /* Ecc_sta = FMC2_HECCR */ + if (sta_map & BIT(s)) { + stm32_fmc2_ham_set_ecc(*ecc_sta, &ecc_calc[i]); + stat = chip->ecc.correct(chip, p, &ecc_code[i], + &ecc_calc[i]); + } + ecc_sta++; + } else { + /* + * Ecc_sta[0] = FMC2_BCHDSR0 + * Ecc_sta[1] = FMC2_BCHDSR1 + * Ecc_sta[2] = FMC2_BCHDSR2 + * Ecc_sta[3] = FMC2_BCHDSR3 + * Ecc_sta[4] = FMC2_BCHDSR4 + */ + if (sta_map & BIT(s)) + stat = stm32_fmc2_bch_decode(chip->ecc.size, + p, ecc_sta); + ecc_sta += 5; + } + + if (stat == -EBADMSG) + /* Check for empty pages with bitflips */ + stat = nand_check_erased_ecc_chunk(p, eccsize, + &ecc_code[i], + eccbytes, + NULL, 0, + eccstrength); + + if (stat < 0) { + mtd->ecc_stats.failed++; + } else { + mtd->ecc_stats.corrected += stat; + max_bitflips = max_t(unsigned int, max_bitflips, stat); + } + } + + return max_bitflips; +} + +static int stm32_fmc2_sequencer_read_page_raw(struct nand_chip *chip, + uint8_t *buf, + int oob_required, + int page) +{ + struct mtd_info *mtd = nand_to_mtd(chip); + struct stm32_fmc2 *fmc2 = nand_get_controller_data(chip); + int ret; + + /* Configure the sequencer */ + stm32_fmc2_rw_page_init(fmc2, page, 1, false); + + /* Read the page */ + ret = stm32_fmc2_xfer(fmc2, buf, 1, false); + if (ret) + return ret; + + /* Read oob */ + if (oob_required) + return nand_change_read_column_op(chip, mtd->writesize, + chip->oob_poi, mtd->oobsize, + false); + + return 0; +} + +static irqreturn_t stm32_fmc2_irq(int irq, void *dev_id) +{ + struct stm32_fmc2 *fmc2 = (struct stm32_fmc2 *)dev_id; + + stm32_fmc2_disable_seq_irq(fmc2); + + complete(&fmc2->complete); + + return IRQ_HANDLED;You don't have any bit to check/ack/read?
No, nothing to check.
+} + +void stm32_fmc2_read_data(struct nand_chip *chip, void *buf, + unsigned int len, bool force_8bit) +{ + struct stm32_fmc2 *fmc2 = nand_get_controller_data(chip); + void __iomem *io_addr_r = fmc2->data_base[fmc2->cs_sel]; + u8 *p = buf; + unsigned int i; + + if (force_8bit) + goto read_8bit; + + if (IS_ALIGNED((u32)buf, sizeof(u32)) && IS_ALIGNED(len, sizeof(u32))) {If you selected BOUNCE_BUFFER in the options, buf is supposedly aligned, or am I missing something?
2 FMC2 internal modes can be used:- Sequencer mode (Patch 2/3): dmas are used and NAND_USE_BOUNCE_BUFFER option is selected. - Manual mode (Patch 3/3): no dma channel is used and NAND_USE_BOUNCE_BUFFER is not selected. Should i select NAND_USE_BOUNCE_BUFFER for sequencer and manual mode, and remove IS_ALIGNED test on buf?
+ u32 *p = buf; + + len >>= 2;Please don't do such optimization by hand. Compilers are doing it for it, so let's keep the code readable. Just do a regular "/= sizeof(unsigned long)".
Ok, i will replace this line with: len /= sizeof(u32)
+ for (i = 0; i < len; i++) + p[i] = readl_relaxed(io_addr_r); + return; + } + + if (chip->options & NAND_BUSWIDTH_16) { + u16 *p = buf; + + len >>= 1;Ditto
Ok, i will replace this line with: len /= sizeof(u16) Same modifications will be done in stm32_fmc2_write_data function.
+ for (i = 0; i < len; i++) + p[i] = readw_relaxed(io_addr_r); + return; + } + +read_8bit: + for (i = 0; i < len; i++) + p[i] = readb_relaxed(io_addr_r); +} + +void stm32_fmc2_write_data(struct nand_chip *chip, const void *buf, + unsigned int len, bool force_8bit) +{ + struct stm32_fmc2 *fmc2 = nand_get_controller_data(chip); + void __iomem *io_addr_w = fmc2->data_base[fmc2->cs_sel]; + const u8 *p = buf; + unsigned int i; + + if (force_8bit) + goto write_8bit; + + if (IS_ALIGNED((u32)buf, sizeof(u32)) && IS_ALIGNED(len, sizeof(u32))) { + const u32 *p = buf; + + len >>= 2; + for (i = 0; i < len; i++) + writel_relaxed(p[i], io_addr_w); + return; + } + + if (chip->options & NAND_BUSWIDTH_16) { + const u16 *p = buf; + + len >>= 1; + for (i = 0; i < len; i++) + writew_relaxed(p[i], io_addr_w); + return; + } + +write_8bit: + for (i = 0; i < len; i++) + writeb_relaxed(p[i], io_addr_w); +} + +static int stm32_fmc2_exec_op(struct nand_chip *chip, + const struct nand_operation *op, + bool check_only)You should probably return 0 if check_only == true, otherwise you could do twice each operation.
Ok.
+{ + struct stm32_fmc2 *fmc2 = nand_get_controller_data(chip); + const struct nand_op_instr *instr = NULL; + unsigned int op_id, i; + int ret = 0; + + for (op_id = 0; op_id < op->ninstrs; op_id++) { + instr = &op->instrs[op_id]; + + switch (instr->type) { + case NAND_OP_CMD_INSTR: + writeb_relaxed(instr->ctx.cmd.opcode, + fmc2->cmd_base[fmc2->cs_sel]); + break; + + case NAND_OP_ADDR_INSTR: + for (i = 0; i < instr->ctx.addr.naddrs; i++) + writeb_relaxed(instr->ctx.addr.addrs[i], + fmc2->addr_base[fmc2->cs_sel]); + break; + + case NAND_OP_DATA_IN_INSTR: + stm32_fmc2_read_data(chip, instr->ctx.data.buf.in, + instr->ctx.data.len, + instr->ctx.data.force_8bit); + break; + + case NAND_OP_DATA_OUT_INSTR: + stm32_fmc2_write_data(chip, instr->ctx.data.buf.out, + instr->ctx.data.len, + instr->ctx.data.force_8bit); + break; + + case NAND_OP_WAITRDY_INSTR: + ret = nand_soft_waitrdy(chip, + instr->ctx.waitrdy.timeout_ms); + break; + } + } + + return ret; +} + +/* Timings configuration */ +static void stm32_fmc2_timings_init(struct stm32_fmc2 *fmc2) +{ + struct stm32_fmc2_timings *timings = &fmc2->timings; + u32 pcr = readl_relaxed(fmc2->io_base + FMC2_PCR); + u32 pmem, patt; + + /* Set tclr/tar timings */ + pcr &= ~FMC2_PCR_TCLR_MASK; + pcr |= FMC2_PCR_TCLR(timings->tclr); + pcr &= ~FMC2_PCR_TAR_MASK; + pcr |= FMC2_PCR_TAR(timings->tar); + + /* Set tset/twait/thold/thiz timings in common bank */ + pmem = FMC2_PMEM_MEMSET(timings->tset_mem); + pmem |= FMC2_PMEM_MEMWAIT(timings->twait); + pmem |= FMC2_PMEM_MEMHOLD(timings->thold_mem); + pmem |= FMC2_PMEM_MEMHIZ(timings->thiz); + + /* Set tset/twait/thold/thiz timings in attribut bank */ + patt = FMC2_PATT_ATTSET(timings->tset_att); + patt |= FMC2_PATT_ATTWAIT(timings->twait); + patt |= FMC2_PATT_ATTHOLD(timings->thold_att); + patt |= FMC2_PATT_ATTHIZ(timings->thiz); + + writel_relaxed(pcr, fmc2->io_base + FMC2_PCR); + writel_relaxed(pmem, fmc2->io_base + FMC2_PMEM); + writel_relaxed(patt, fmc2->io_base + FMC2_PATT); +} + +/* Controller initialization */ +static void stm32_fmc2_init(struct stm32_fmc2 *fmc2) +{ + u32 pcr = readl_relaxed(fmc2->io_base + FMC2_PCR); + u32 bcr1 = readl_relaxed(fmc2->io_base + FMC2_BCR1); + + /* Enable wait feature and nand flash memory bank */ + pcr |= FMC2_PCR_PWAITEN; + pcr |= FMC2_PCR_PBKEN; + + /* Set buswidth to 8 bits mode for identification */ + pcr &= ~FMC2_PCR_PWID_MASK; + + /* Ecc logic is disabled */ + pcr &= ~FMC2_PCR_ECCEN; + + /* Default mode */ + pcr &= ~FMC2_PCR_ECCALG; + pcr &= ~FMC2_PCR_BCHECC; + pcr &= ~FMC2_PCR_WEN; + + /* Set default ecc sector size */ + pcr &= ~FMC2_PCR_ECCSS_MASK; + pcr |= FMC2_PCR_ECCSS(FMC2_PCR_ECCSS_2048); + + /* Set default tclr/tar timings */ + pcr &= ~FMC2_PCR_TCLR_MASK; + pcr |= FMC2_PCR_TCLR(FMC2_PCR_TCLR_DEFAULT); + pcr &= ~FMC2_PCR_TAR_MASK; + pcr |= FMC2_PCR_TAR(FMC2_PCR_TAR_DEFAULT); + + /* Enable FMC2 controller */ + bcr1 |= FMC2_BCR1_FMC2EN; + + writel_relaxed(bcr1, fmc2->io_base + FMC2_BCR1); + writel_relaxed(pcr, fmc2->io_base + FMC2_PCR); + writel_relaxed(FMC2_PMEM_DEFAULT, fmc2->io_base + FMC2_PMEM); + writel_relaxed(FMC2_PATT_DEFAULT, fmc2->io_base + FMC2_PATT); +} + +/* Controller configuration */ +static void stm32_fmc2_setup(struct stm32_fmc2 *fmc2) +{ + struct nand_chip *chip = &fmc2->chip; + u32 pcr = readl_relaxed(fmc2->io_base + FMC2_PCR); + + /* Configure in HAMMING by default */In other comments you use "Hamming", not "HAMMING". And this comment is weird, you don't configure for Hamming below. Maybe you meant that the controller is by default configured to work with Hamming error detection?
Yes, it's true. Hamming error detection is set by default. Below implementation configures the controller in BCH4 or BCH8. I will modify the comment.
+ if (chip->ecc.strength == FMC2_ECC_BCH8) { + pcr |= FMC2_PCR_ECCALG; + pcr |= FMC2_PCR_BCHECC; + } else if (chip->ecc.strength == FMC2_ECC_BCH4) { + pcr |= FMC2_PCR_ECCALG; + } + + /* Set buswidth */ + if (chip->options & NAND_BUSWIDTH_16) + pcr |= FMC2_PCR_PWID(FMC2_PCR_PWID_BUSWIDTH_16); + + /* Set ecc sector size */ + pcr &= ~FMC2_PCR_ECCSS_MASK; + pcr |= FMC2_PCR_ECCSS(FMC2_PCR_ECCSS_512); + + writel_relaxed(pcr, fmc2->io_base + FMC2_PCR); +} + +/* Controller timings */ +static void stm32_fmc2_calc_timings(struct stm32_fmc2 *fmc2, + const struct nand_sdr_timings *sdrt, + struct stm32_fmc2_timings *tims) +{ + unsigned long hclk = clk_get_rate(fmc2->clk); + unsigned long hclkp = NSEC_PER_SEC / (hclk / 1000); + int tar, tclr, thiz, twait, tset_mem, tset_att, thold_mem, thold_att; + + tar = hclkp; + if (tar < sdrt->tAR_min) + tar = sdrt->tAR_min; + tims->tar = DIV_ROUND_UP(tar, hclkp) - 1; + if (tims->tar > FMC2_PCR_TIMING_MASK) + tims->tar = FMC2_PCR_TIMING_MASK; + + tclr = hclkp; + if (tclr < sdrt->tCLR_min) + tclr = sdrt->tCLR_min; + tims->tclr = DIV_ROUND_UP(tclr, hclkp) - 1; + if (tims->tclr > FMC2_PCR_TIMING_MASK) + tims->tclr = FMC2_PCR_TIMING_MASK; + + tims->thiz = FMC2_THIZ; + thiz = (tims->thiz + 1) * hclkp; + + /* + * tWAIT > tRP + * tWAIT > tWP + * tWAIT > tREA + tIO + */ + twait = hclkp; + if (twait < sdrt->tRP_min) + twait = sdrt->tRP_min; + if (twait < sdrt->tWP_min) + twait = sdrt->tWP_min; + if (twait < sdrt->tREA_max + FMC2_TIO) + twait = sdrt->tREA_max + FMC2_TIO; + tims->twait = DIV_ROUND_UP(twait, hclkp); + if (tims->twait == 0) + tims->twait = 1; + else if (tims->twait > FMC2_PMEM_PATT_TIMING_MASK) + tims->twait = FMC2_PMEM_PATT_TIMING_MASK; + + /* + * tSETUP_MEM > tCS - tWAIT + * tSETUP_MEM > tALS - tWAIT + * tSETUP_MEM > tDS - (tWAIT - tHIZ) + */ + tset_mem = hclkp; + if (sdrt->tCS_min > twait && (tset_mem < sdrt->tCS_min - twait)) + tset_mem = sdrt->tCS_min - twait; + if (sdrt->tALS_min > twait && (tset_mem < sdrt->tALS_min - twait)) + tset_mem = sdrt->tALS_min - twait; + if (twait > thiz && (sdrt->tDS_min > twait - thiz) && + (tset_mem < sdrt->tDS_min - (twait - thiz))) + tset_mem = sdrt->tDS_min - (twait - thiz); + tims->tset_mem = DIV_ROUND_UP(tset_mem, hclkp); + if (tims->tset_mem == 0) + tims->tset_mem = 1; + else if (tims->tset_mem > FMC2_PMEM_PATT_TIMING_MASK) + tims->tset_mem = FMC2_PMEM_PATT_TIMING_MASK; + + /* + * tHOLD_MEM > tCH + * tHOLD_MEM > tREH - tSETUP_MEM + * tHOLD_MEM > max(tRC, tWC) - (tSETUP_MEM + tWAIT) + */ + thold_mem = hclkp; + if (thold_mem < sdrt->tCH_min) + thold_mem = sdrt->tCH_min; + if (sdrt->tREH_min > tset_mem && + (thold_mem < sdrt->tREH_min - tset_mem)) + thold_mem = sdrt->tREH_min - tset_mem; + if ((sdrt->tRC_min > tset_mem + twait) && + (thold_mem < sdrt->tRC_min - (tset_mem + twait))) + thold_mem = sdrt->tRC_min - (tset_mem + twait); + if ((sdrt->tWC_min > tset_mem + twait) && + (thold_mem < sdrt->tWC_min - (tset_mem + twait))) + thold_mem = sdrt->tWC_min - (tset_mem + twait); + tims->thold_mem = DIV_ROUND_UP(thold_mem, hclkp); + if (tims->thold_mem == 0) + tims->thold_mem = 1; + else if (tims->thold_mem > FMC2_PMEM_PATT_TIMING_MASK) + tims->thold_mem = FMC2_PMEM_PATT_TIMING_MASK; + + /* + * tSETUP_ATT > tCS - tWAIT + * tSETUP_ATT > tCLS - tWAIT + * tSETUP_ATT > tALS - tWAIT + * tSETUP_ATT > tRHW - tHOLD_MEM + * tSETUP_ATT > tDS - (tWAIT - tHIZ) + */ + tset_att = hclkp; + if (sdrt->tCS_min > twait && (tset_att < sdrt->tCS_min - twait)) + tset_att = sdrt->tCS_min - twait; + if (sdrt->tCLS_min > twait && (tset_att < sdrt->tCLS_min - twait)) + tset_att = sdrt->tCLS_min - twait; + if (sdrt->tALS_min > twait && (tset_att < sdrt->tALS_min - twait)) + tset_att = sdrt->tALS_min - twait; + if (sdrt->tRHW_min > thold_mem && + (tset_att < sdrt->tRHW_min - thold_mem)) + tset_att = sdrt->tRHW_min - thold_mem; + if (twait > thiz && (sdrt->tDS_min > twait - thiz) && + (tset_att < sdrt->tDS_min - (twait - thiz))) + tset_att = sdrt->tDS_min - (twait - thiz); + tims->tset_att = DIV_ROUND_UP(tset_att, hclkp); + if (tims->tset_att == 0) + tims->tset_att = 1; + else if (tims->tset_att > FMC2_PMEM_PATT_TIMING_MASK) + tims->tset_att = FMC2_PMEM_PATT_TIMING_MASK; + + /* + * tHOLD_ATT > tALH + * tHOLD_ATT > tCH + * tHOLD_ATT > tCLH + * tHOLD_ATT > tCOH + * tHOLD_ATT > tDH + * tHOLD_ATT > tWB + tIO + tSYNC - tSETUP_MEM + * tHOLD_ATT > tADL - tSETUP_MEM + * tHOLD_ATT > tWH - tSETUP_MEM + * tHOLD_ATT > tWHR - tSETUP_MEM + * tHOLD_ATT > tRC - (tSETUP_ATT + tWAIT) + * tHOLD_ATT > tWC - (tSETUP_ATT + tWAIT) + */ + thold_att = hclkp; + if (thold_att < sdrt->tALH_min) + thold_att = sdrt->tALH_min; + if (thold_att < sdrt->tCH_min) + thold_att = sdrt->tCH_min; + if (thold_att < sdrt->tCLH_min) + thold_att = sdrt->tCLH_min; + if (thold_att < sdrt->tCOH_min) + thold_att = sdrt->tCOH_min; + if (thold_att < sdrt->tDH_min) + thold_att = sdrt->tDH_min; + if ((sdrt->tWB_max + FMC2_TIO + FMC2_TSYNC > tset_mem) && + (thold_att < sdrt->tWB_max + FMC2_TIO + FMC2_TSYNC - tset_mem)) + thold_att = sdrt->tWB_max + FMC2_TIO + FMC2_TSYNC - tset_mem; + if (sdrt->tADL_min > tset_mem && + (thold_att < sdrt->tADL_min - tset_mem)) + thold_att = sdrt->tADL_min - tset_mem; + if (sdrt->tWH_min > tset_mem && + (thold_att < sdrt->tWH_min - tset_mem)) + thold_att = sdrt->tWH_min - tset_mem; + if (sdrt->tWHR_min > tset_mem && + (thold_att < sdrt->tWHR_min - tset_mem)) + thold_att = sdrt->tWHR_min - tset_mem; + if ((sdrt->tRC_min > tset_att + twait) && + (thold_att < sdrt->tRC_min - (tset_att + twait))) + thold_att = sdrt->tRC_min - (tset_att + twait); + if ((sdrt->tWC_min > tset_att + twait) && + (thold_att < sdrt->tWC_min - (tset_att + twait))) + thold_att = sdrt->tWC_min - (tset_att + twait); + tims->thold_att = DIV_ROUND_UP(thold_att, hclkp); + if (tims->thold_att == 0) + tims->thold_att = 1; + else if (tims->thold_att > FMC2_PMEM_PATT_TIMING_MASK) + tims->thold_att = FMC2_PMEM_PATT_TIMING_MASK; +} + +static int stm32_fmc2_setup_interface(struct nand_chip *chip, int chipnr, + const struct nand_data_interface *conf) +{ + struct stm32_fmc2 *fmc2 = nand_get_controller_data(chip); + struct stm32_fmc2_timings tims; + const struct nand_sdr_timings *sdrt; + + sdrt = nand_get_sdr_timings(conf); + if (IS_ERR(sdrt)) + return PTR_ERR(sdrt); + + stm32_fmc2_calc_timings(fmc2, sdrt, &tims);Why do you need 'tims' if the only use is to memcpy it in fmc2->timings? I suppose you can get rid of it.
Yes, it is not needed to use a local variable.
+ + if (chipnr == NAND_DATA_IFACE_CHECK_ONLY) + return 0;If the controller may support all the timing modes, then you can certainly return on CHECK_ONLY before *_calc_timings().
Yes, i will move this check before stm32_fmc2_calc_timings().
+ + /* Save and apply timings */ + memcpy(&fmc2->timings, &tims, sizeof(tims)); + stm32_fmc2_timings_init(fmc2); + + return 0; +} + +/* DMA configuration */ +static int stm32_fmc2_dma_setup(struct stm32_fmc2 *fmc2, u8 nbsect) +{ + struct nand_chip *chip = &fmc2->chip; + struct dma_slave_config dma_cfg; + int ret; + + memset(&dma_cfg, 0, sizeof(dma_cfg)); + dma_cfg.src_addr = fmc2->data_phys_addr[fmc2->cs_sel]; + dma_cfg.dst_addr = fmc2->data_phys_addr[fmc2->cs_sel]; + dma_cfg.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES; + dma_cfg.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES; + dma_cfg.src_maxburst = 32; + dma_cfg.dst_maxburst = 32; + + fmc2->dma_tx_ch = dma_request_slave_channel(fmc2->dev, "tx"); + if (fmc2->dma_tx_ch) { + ret = dmaengine_slave_config(fmc2->dma_tx_ch, &dma_cfg); + if (ret) { + dev_err(fmc2->dev, "data tx DMA engine slave config failed\n"); + return ret; + } + } + + fmc2->dma_rx_ch = dma_request_slave_channel(fmc2->dev, "rx"); + if (fmc2->dma_rx_ch) { + ret = dmaengine_slave_config(fmc2->dma_rx_ch, &dma_cfg); + if (ret) { + dev_err(fmc2->dev, "data rx DMA engine slave config failed\n"); + return ret; + } + + fmc2->dma_ecc_ch = dma_request_slave_channel(fmc2->dev, "ecc"); + if (fmc2->dma_ecc_ch) { + /* + * HAMMING: we read HECCR register + * BCH4/BCH8: we read BCHDSRSx registers + */ + memset(&dma_cfg, 0, sizeof(dma_cfg)); + dma_cfg.src_addr = fmc2->io_phys_addr; + dma_cfg.src_addr += chip->ecc.strength == FMC2_ECC_HAM ? + FMC2_HECCR : FMC2_BCHDSR0; + dma_cfg.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES; + + ret = dmaengine_slave_config(fmc2->dma_ecc_ch, + &dma_cfg); + if (ret) { + dev_err(fmc2->dev, "ecc DMA engine slave config failed\n"); + return ret; + } + + ret = sg_alloc_table(&fmc2->dma_ecc_sg, + nbsect, GFP_KERNEL); + if (ret) + return ret; + + /* Calculate ecc length needed for one sector */ + fmc2->dma_ecc_len = chip->ecc.strength == FMC2_ECC_HAM ? + FMC2_HECCR_LEN : FMC2_BCHDSRS_LEN; + + /* Allocate a buffer to store ecc status registers */ + fmc2->ecc_buf = devm_kzalloc(fmc2->dev, + fmc2->dma_ecc_len * nbsect, + GFP_KERNEL); + if (!fmc2->ecc_buf) + return -ENOMEM; + } else { + dev_err(fmc2->dev, "ecc DMA not defined in the device tree\n"); + return -ENOENT; + } + } + + if (fmc2->dma_tx_ch && fmc2->dma_rx_ch) { + ret = sg_alloc_table(&fmc2->dma_data_sg, nbsect, GFP_KERNEL); + if (ret) + return ret; + + init_completion(&fmc2->dma_data_complete); + init_completion(&fmc2->dma_ecc_complete); + } else { + dev_err(fmc2->dev, "rx/tx DMA not defined in the device tree\n"); + return -ENOENT; + } + + return 0; +} + +/* NAND callbacks setup */ +static void stm32_fmc2_nand_callbacks_setup(struct stm32_fmc2 *fmc2) +{ + struct nand_chip *chip = &fmc2->chip; + + /* Specific callbacks to read/write a page */ + chip->ecc.correct = stm32_fmc2_ham_correct; + chip->ecc.write_page = stm32_fmc2_sequencer_write_page; + chip->ecc.read_page = stm32_fmc2_sequencer_read_page; + chip->ecc.write_page_raw = stm32_fmc2_sequencer_write_page_raw; + chip->ecc.read_page_raw = stm32_fmc2_sequencer_read_page_raw; + chip->options |= NAND_USE_BOUNCE_BUFFER; + + /* Specific configurations depending on the algo used */ + if (chip->ecc.strength == FMC2_ECC_HAM) + chip->ecc.bytes = chip->options & NAND_BUSWIDTH_16 ? 4 : 3; + else if (chip->ecc.strength == FMC2_ECC_BCH8) + chip->ecc.bytes = chip->options & NAND_BUSWIDTH_16 ? 14 : 13; + else + chip->ecc.bytes = chip->options & NAND_BUSWIDTH_16 ? 8 : 7; +} + +/* FMC2 layout */ +static int stm32_fmc2_nand_ooblayout_ecc(struct mtd_info *mtd, int section, + struct mtd_oob_region *oobregion) +{ + struct nand_chip *chip = mtd_to_nand(mtd); + struct nand_ecc_ctrl *ecc = &chip->ecc; + + if (section) + return -ERANGE; + + oobregion->length = ecc->total; + oobregion->offset = FMC2_BBM_LEN; + + return 0; +} + +static int stm32_fmc2_nand_ooblayout_free(struct mtd_info *mtd, int section, + struct mtd_oob_region *oobregion) +{ + struct nand_chip *chip = mtd_to_nand(mtd); + struct nand_ecc_ctrl *ecc = &chip->ecc; + + if (section) + return -ERANGE; + + oobregion->length = mtd->oobsize - ecc->total - FMC2_BBM_LEN; + oobregion->offset = ecc->total + FMC2_BBM_LEN; + + return 0; +} + +const struct mtd_ooblayout_ops stm32_fmc2_nand_ooblayout_ops = { + .ecc = stm32_fmc2_nand_ooblayout_ecc, + .free = stm32_fmc2_nand_ooblayout_free, +}; + +/* FMC2 caps */ +static int stm32_fmc2_calc_ecc_bytes(int step_size, int strength) +{ + /* Hamming */ + if (strength == FMC2_ECC_HAM) + return 4; + + /* BCH8 */ + if (strength == FMC2_ECC_BCH8) + return 14; + + /* BCH4 */ + return 8; +} + +NAND_ECC_CAPS_SINGLE(stm32_fmc2_ecc_caps, stm32_fmc2_calc_ecc_bytes, + FMC2_ECC_STEP_SIZE, + FMC2_ECC_HAM, FMC2_ECC_BCH4, FMC2_ECC_BCH8); + +/* FMC2 controller ops */ +static int stm32_fmc2_attach_chip(struct nand_chip *chip) +{ + struct stm32_fmc2 *fmc2 = nand_get_controller_data(chip); + struct mtd_info *mtd = nand_to_mtd(chip); + u8 nbsect; + int ret; + + /* + * Only NAND_ECC_HW mode is actually supported + * HAMMING => ecc.strength = 1 + * BCH4 => ecc.strength = 4 + * BCH8 => ecc.strength = 8 + * ecc sector size = 512 + */ + if (chip->ecc.mode != NAND_ECC_HW) { + dev_err(fmc2->dev, "nand_ecc_mode is not well defined in the DT\n"); + return -EINVAL; + } + + ret = nand_ecc_choose_conf(chip, &stm32_fmc2_ecc_caps, + mtd->oobsize - FMC2_BBM_LEN); + if (ret) { + dev_err(fmc2->dev, "no valid ECC settings set\n"); + return ret; + } + + nbsect = mtd->writesize / chip->ecc.size; + if (nbsect > FMC2_MAX_SG_COUNT) { + dev_err(fmc2->dev, "nand page size is not supported\n"); + return -EINVAL; + } + + if (chip->bbt_options & NAND_BBT_USE_FLASH) + chip->bbt_options |= NAND_BBT_NO_OOB; + + /* FMC2 setup routine */ + stm32_fmc2_setup(fmc2); + + /* DMA setup */ + ret = stm32_fmc2_dma_setup(fmc2, nbsect); + if (ret) + return ret; + + /* NAND callbacks setup */ + stm32_fmc2_nand_callbacks_setup(fmc2); + + /* Define ECC layout */ + mtd_set_ooblayout(mtd, &stm32_fmc2_nand_ooblayout_ops); + + return 0; +} + +static const struct nand_controller_ops stm32_fmc2_nand_controller_ops = { + .attach_chip = stm32_fmc2_attach_chip, +}; + +/* FMC2 probe */ +static int stm32_fmc2_parse_child(struct device *dev, + struct stm32_fmc2 *fmc2, + struct device_node *dn) +{ + u32 cs; + int ret, i; + + if (!of_get_property(dn, "reg", &fmc2->ncs)) + return -EINVAL; + + fmc2->ncs /= sizeof(u32); + if (!fmc2->ncs) { + dev_err(dev, "invalid reg property size\n"); + return -EINVAL; + } + + for (i = 0; i < fmc2->ncs; i++) { + ret = of_property_read_u32_index(dn, "reg", i, &cs); + if (ret) { + dev_err(dev, "could not retrieve reg property: %d\n", + ret); + return ret; + } + + if (cs > FMC2_MAX_CE) { + dev_err(dev, "invalid reg value: %d\n", cs); + return -EINVAL; + } + + if (fmc2->cs_assigned & BIT(cs)) { + dev_err(dev, "cs already assigned: %d\n", cs); + return -EINVAL; + } + + fmc2->cs_assigned |= BIT(cs); + fmc2->cs_used[i] = cs; + } + + /* Default cs used */s/cs/CS/
OK
+ fmc2->cs_sel = fmc2->cs_used[0]; + + /* Timings */ + ret = of_property_read_u8_array(dn, "st,fmc2_timings", + (u8 *)&fmc2->timings, + sizeof(fmc2->timings)); + return ret ? 0 : 1; +} + +static int stm32_fmc2_parse_dt(struct device *dev, + struct stm32_fmc2 *fmc2) +{ + struct device_node *dn = dev->of_node; + struct device_node *child; + int nchips = of_get_child_count(dn); + int ret = 0; + + if (!nchips) { + dev_err(dev, "NAND chip not defined\n"); + return -EINVAL; + } + + if (nchips > 1) {If you have two CS, can't you have two NAND chips connected?
No HW board has been designed with 2 NAND chips connected. I am not able to test this configuration. The driver will be improved when i will be able to test such configuration.
+ dev_err(dev, "too many NAND chips defined\n"); + return -EINVAL; + } + + for_each_child_of_node(dn, child) { + ret = stm32_fmc2_parse_child(dev, fmc2, child); + if (ret < 0) { + of_node_put(child); + return ret; + } + + nand_set_flash_node(&fmc2->chip, child); + } + + return ret; +} + +static int stm32_fmc2_probe(struct platform_device *pdev) +{ + struct device *dev = &pdev->dev; + struct reset_control *rstc; + struct stm32_fmc2 *fmc2; + struct resource *res; + struct mtd_info *mtd; + struct nand_chip *chip; + int i, j, ret, irq, timings_def; + + fmc2 = devm_kzalloc(dev, sizeof(*fmc2), GFP_KERNEL); + if (!fmc2) + return -ENOMEM; + + timings_def = stm32_fmc2_parse_dt(dev, fmc2); + if (timings_def < 0) + return timings_def; + + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + fmc2->io_base = devm_ioremap_resource(dev, res); + if (IS_ERR(fmc2->io_base)) + return PTR_ERR(fmc2->io_base); + + fmc2->io_phys_addr = res->start; + + for (i = 0, j = 1; i < FMC2_MAX_CE; i++, j += 3) {Maybe naming i 'chip_id' and j 'mem_region' or something else, more meaningful than just i/j would be more clear?
Ok.
+ if (!(fmc2->cs_assigned & BIT(i))) + continue; + + res = platform_get_resource(pdev, IORESOURCE_MEM, j); + fmc2->data_base[i] = devm_ioremap_resource(dev, res); + if (IS_ERR(fmc2->data_base[i])) + return PTR_ERR(fmc2->data_base[i]); + + fmc2->data_phys_addr[i] = res->start; + + res = platform_get_resource(pdev, IORESOURCE_MEM, j + 1); + fmc2->cmd_base[i] = devm_ioremap_resource(dev, res); + if (IS_ERR(fmc2->cmd_base[i])) + return PTR_ERR(fmc2->cmd_base[i]); + + res = platform_get_resource(pdev, IORESOURCE_MEM, j + 2); + fmc2->addr_base[i] = devm_ioremap_resource(dev, res); + if (IS_ERR(fmc2->addr_base[i])) + return PTR_ERR(fmc2->addr_base[i]); + } + + irq = platform_get_irq(pdev, 0); + ret = devm_request_irq(dev, irq, stm32_fmc2_irq, 0, + dev_name(dev), fmc2); + if (ret) { + dev_err(dev, "failed to request irq\n"); + return ret; + } + + init_completion(&fmc2->complete); + + fmc2->clk = devm_clk_get(dev, NULL); + if (IS_ERR(fmc2->clk)) + return PTR_ERR(fmc2->clk); + + ret = clk_prepare_enable(fmc2->clk); + if (ret) { + dev_err(dev, "can not enable the clock\n"); + return ret; + } + + rstc = devm_reset_control_get(dev, NULL); + if (!IS_ERR(rstc)) { + reset_control_assert(rstc); + reset_control_deassert(rstc); + } + + fmc2->dev = dev; + mtd = nand_to_mtd(&fmc2->chip); + chip = &fmc2->chip; + nand_set_controller_data(chip, fmc2); + mtd->dev.parent = dev; + + chip->exec_op = stm32_fmc2_exec_op; + chip->select_chip = stm32_fmc2_select_chip; + chip->options |= NAND_BUSWIDTH_AUTO | NAND_NO_SUBPAGE_WRITE; + + /* FMC2 init routine */ + stm32_fmc2_init(fmc2); + if (timings_def) + stm32_fmc2_timings_init(fmc2);You already implement ->setup_data_interface, why do you need this stm32_fmc2_timings_init()?
"st,fmc2-timings" is an optional property that allow the end user to overwrite the timings calculated by setup_data_interface callback. In case this property is defined in the device tree, timings are directly set and stm32_fmc2_setup_interface() will be never called.
+ else + chip->setup_data_interface = stm32_fmc2_setup_interface; + + /* Default settings */ + chip->ecc.mode = NAND_ECC_HW; + chip->ecc.size = FMC2_ECC_STEP_SIZE; + chip->ecc.strength = FMC2_ECC_BCH8; + + /* Scan to find existence of the device */ + chip->dummy_controller.ops = &stm32_fmc2_nand_controller_ops; + ret = nand_scan(chip, fmc2->ncs); + if (ret) + goto err_scan; + + ret = mtd_device_register(mtd, NULL, 0); + if (ret) + goto err_device_register; + + platform_set_drvdata(pdev, fmc2); + + return 0; + +err_device_register: + nand_cleanup(chip); + +err_scan: + if (fmc2->dma_ecc_ch) + dma_release_channel(fmc2->dma_ecc_ch); + if (fmc2->dma_tx_ch) + dma_release_channel(fmc2->dma_tx_ch); + if (fmc2->dma_rx_ch) + dma_release_channel(fmc2->dma_rx_ch); + + sg_free_table(&fmc2->dma_data_sg); + sg_free_table(&fmc2->dma_ecc_sg); + + clk_disable_unprepare(fmc2->clk); + + return ret; +} + +static int stm32_fmc2_remove(struct platform_device *pdev) +{ + struct stm32_fmc2 *fmc2 = platform_get_drvdata(pdev); + + nand_release(&fmc2->chip); + + if (fmc2->dma_ecc_ch) + dma_release_channel(fmc2->dma_ecc_ch); + if (fmc2->dma_tx_ch) + dma_release_channel(fmc2->dma_tx_ch); + if (fmc2->dma_rx_ch) + dma_release_channel(fmc2->dma_rx_ch); + + sg_free_table(&fmc2->dma_data_sg); + sg_free_table(&fmc2->dma_ecc_sg); + + clk_disable_unprepare(fmc2->clk); + + return 0; +} + +#ifdef CONFIG_PM_SLEEP +static int stm32_fmc2_suspend(struct device *dev)Please remove the pre-processor conditionals and add __maybe_unused in the suspend/resume functions definitions.
Ok.
+{ + struct stm32_fmc2 *fmc2 = dev_get_drvdata(dev); + + clk_disable_unprepare(fmc2->clk); + + pinctrl_pm_select_sleep_state(dev); + + return 0; +} + +static int stm32_fmc2_resume(struct device *dev) +{ + struct stm32_fmc2 *fmc2 = dev_get_drvdata(dev); + int i, ret; + + pinctrl_pm_select_default_state(dev); + + ret = clk_prepare_enable(fmc2->clk); + if (ret) { + dev_err(dev, "can not enable the clock\n"); + return ret; + } + + stm32_fmc2_init(fmc2); + stm32_fmc2_timings_init(fmc2); + stm32_fmc2_setup(fmc2); + + for (i = 0; i < fmc2->ncs; i++) + nand_reset(&fmc2->chip, i);This means you have one different NAND chip wired on each CS. We could have two CS wired to the same NAND chip. Calling nand_reset twice would be harmless but a lost of time.
Ok, so i will call nand_reset() only once as this driver only supports one or two CS wired connected to the same NAND chip.
+ + return 0; +} +#endif + +static SIMPLE_DEV_PM_OPS(stm32_fmc2_pm_ops, stm32_fmc2_suspend, + stm32_fmc2_resume); + +static const struct of_device_id stm32_fmc2_match[] = { + {.compatible = "st,stm32mp15-fmc2"}, + {} +}; +MODULE_DEVICE_TABLE(of, stm32_fmc2_match); + +static struct platform_driver stm32_fmc2_driver = { + .probe = stm32_fmc2_probe, + .remove = stm32_fmc2_remove, + .driver = { + .name = "stm32-fmc2", + .of_match_table = stm32_fmc2_match, + .pm = &stm32_fmc2_pm_ops, + }, +}; +module_platform_driver(stm32_fmc2_driver); + +MODULE_ALIAS("platform:" DRIVER_NAME); +MODULE_AUTHOR("Christophe Kerello <christophe.kerello@xxxxxx>"); +MODULE_DESCRIPTION("STMicroelectronics STM32 FMC2 nand driver"); +MODULE_LICENSE("GPL v2");Thanks, Miquèl
Thanks, Christophe Kerello.
